Plasmonic Metal-Insulator-Metal Capped Polymer Nanopillars for

bottom Au layers and a 5 nm Ti adhesion layer was used for all top Au layers except for the top. Au layers greater than 90 nm where a 10 nm Ti adhesio...
0 downloads 7 Views 7MB Size
Subscriber access provided by - Access paid by the | University Library, UC Santa Cruz

Article

Plasmonic Metal-Insulator-Metal Capped Polymer Nanopillars for SERS Analysis of Protein-Protein Interactions Lisa Plucinski Hackett, Wenyue Li, Abid Ameen, Lynford Lawrence Goddard, and Gang Logan Liu J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b08656 • Publication Date (Web): 13 Feb 2018 Downloaded from http://pubs.acs.org on February 14, 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.

The Journal of Physical Chemistry C 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 38 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

The Journal of Physical Chemistry

Plasmonic Metal-Insulator-Metal Capped Polymer Nanopillars for SERS Analysis of Protein-Protein Interactions Lisa P. Hackett,†, § Wenyue Li,†,┴ Abid Ameen,† Lynford L. Goddard,†,§ and Gang Logan Liu†,§,* †

Micro and Nanotechnology Laboratory, §Department of Electrical and Computer Engineering,

and ┴Department of Agricultural and Biological Engineering, University of Illinois at UrbanaChampaign, Urbana, IL 61801, United States

ABSTRACT: The advancement of SERS as an analytical tool requires substrates that provide both sensitive and reproducible measurements. In this work, a gold-titanium dioxide-gold metalinsulator-metal capped polymer nanopillar array SERS substrate is presented and optimized for SERS-based biosensing applications. The optical properties of the multilayered nanoantenna array are investigated using a combination of simulation and experimental studies. It is found that hot spot engineering, the plasmon resonance of the array, and cavity structure optimization all contribute to fundamental SERS sensor properties such as enhancement factor and enhancement factor uniformity, which are critically studied using this highly tunable device. A spatially averaged enhancement factor of (2.4 ± 0.8) x 107 with sufficient error for quantitative studies is demonstrated at an excitation wavelength of 633 nm. The label-free detection of protein-protein interactions on the metal-insulator-metal nanopillar array surface is then demonstrated including for the cancer biomarker cancer antigen 125 at a concentration of 100 ng/mL.

ACS Paragon Plus Environment

1

The Journal of Physical 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

Page 2 of 38

INTRODUCTION The development of nanostructured plasmonic substrates for surface enhanced Raman spectroscopy (SERS) is promising for applications in areas such as biosensing,1-4 cell imaging and analysis,5-10 and materials science11-12 where the SERS substrate serves to enhance the Raman spectroscopy signal by several orders of magnitude. The amplification of the Raman scattering signal is primarily due to an electromagnetic enhancement effect at metallic nanostructures and is quantified by an enhancement factor (EF). Very low limits of detection, down to a single molecule, have already been demonstrated using SERS and surface enhanced resonance Raman spectroscopy (SERRS) where the SERS substrates are made by randomly organized gold (Au) or silver (Ag) metallic nanoparticles.13-18 In highly localized locations, the spacing between particles is 10 nm or less, which leads to a significant electromagnetic enhancement, referred to as a hot spot. In this case, the magnitude of the enhancement varies exponentially with the gap size. As target applications increase, the samples for Raman analysis vary greatly based on parameters such as size, uniformity, and baseline Raman signal. For example, samples are typically nonuniform on the microscale for applications in areas such as cell biology and materials science. To advance SERS as an analytical tool, current work on SERS substrates then seeks to develop reliable and reproducible sensors that can accommodate these sample variations with an EF that is sufficiently high and uniformly distributed with low error, enabling detection, but also quantitative analysis. Towards this objective, advanced nanofabrication methods have been applied to develop SERS substrates consisting of periodic nanostructures, such as metallic nanodome arrays, largely based on templates made from nanosphere lithography or other lithography methods.19-22 The fabrication parameters can be tuned such that the spacing between nanodomes is 10 nm or less,

ACS Paragon Plus Environment

2

Page 3 of 38 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

The Journal of Physical Chemistry

generating a high enhancement in the Raman scattering of a molecule located in this gap. Although it is a significant improvement over randomly distributed nanoparticles, wafer-scale reproducible samples with consistent, uniform gap spacing are difficult to obtain leading to high EF fluctuations. In addition, the enhancement to the Raman scattering signal will occur primarily only for the part of the sample at the hot spot location. Tuning the hot spot location is then crucial depending on the desired application. Promising work has been done in the development of periodic nanoantenna arrays based on Au or Ag coated nanopillars.23-32 The hot spot is engineered to occur primarily at the tops of the nanopillars. While the EF is typically less than the case of sub-10 nm gaps, a trade-off is accepted between EF, EF uniformity, and device reproducibility. Building upon these achievements, in this work, we explore engineering individual nanoantenna, on a polymer substrate, with a Au-titanium dioxide (TiO2)-Au metal-insulatormetal Fabry-Perot cavity to improve field localization at the tops of the nanopillars. We use this device to explore the relationship between plasmon resonance, hot spot location, EF, and EF uniformity of SERS substrates on the microscale. Fabry-Perot cavity structures have been incorporated

with

plasmonic

structures

previously

for

applications

including

light

manipulation,33-35 biosensing,36-39 SERS,40-43 surface enhanced infrared absorption (SEIRA),44 and enhanced-catalytic activity.45 For SERS, the Fabry-Perot cavity structure has not previously been incorporated in the way we approach here by directly tuning the properties of the nanoantennae. After analysis of the optical properties, device optimization, and assessment of the EF and uniformity, we apply this SERS sensor to the detection of protein-protein interactions. SERS sensors applied for biomolecular studies provide high spectral stability in addition to high

ACS Paragon Plus Environment

3

The Journal of Physical 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

Page 4 of 38

potential for multiplexing. The study of molecules binding on SERS sensors is of continuing high interest to study molecular structure and conformation in response to different environmental cues. For example, the Raman spectrum of 11-mercaptoundecanoic acid (11MUA) molecules varies widely depending on its pretreatment and local environment.46 Protein detection utilizing SERS in the label-free assay format4, 47 or with Raman reporter molecules48-52 has been previously developed where the majority require Raman reporter molecules for sensitive detection at the disadvantage of decreased potential applications and increased sample preparation requirements. Here we present label-free detection of protein-protein interactions including for the cancer biomarker cancer antigen 125 (CA-125). Another significant advantage to the approach presented here is that the SERS substrate is fabricated by a wafer-scale nanoreplica molding process onto a flexible polymer substrate. The wafer-scale fabrication onto a flexible substrate enables greater device reproducibility, larger-scale applications, and improved compatibility with existing Raman systems. METHODS Metal-Insulator-Metal SERS Substrate Fabrication: The polymer nanopillar array was fabricated by a nanoreplica molding process. Firstly, a polymer nanocup array was fabricated from a wafer-scale quartz tapered nanopillar mold. A pipette was used to deposit 2 mL of UVcurable polymer (NOA-61) on top of the mold followed by a PET sheet backing. Once evenly spread, the polymer was cured under UV exposure for 2 minutes. The polymer nanocups were peeled away from the quartz mold and then coated with 5 nm of silicon dioxide (SiO2) deposited by e-beam followed by immersion in a silane solution (PlusOne Repel-Silane). The polymer nanocups were then used as our mold for a second nanoreplica molding process, following the previously described steps, to fabricate the polymer nanopillar array. The metal-insulator-metal

ACS Paragon Plus Environment

4

Page 5 of 38 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

The Journal of Physical Chemistry

deposition was carried out using e-beam and sputtering techniques for the metal and insulator layers, respectively. E-beam was used for the Au deposition since the directional deposition will increase the chance to form Au nanostructures, which will enhance the localized surface plasmon resonance (LSPR) effect of the device. A 9 nm titanium (Ti) adhesion layer was used for all bottom Au layers and a 5 nm Ti adhesion layer was used for all top Au layers except for the top Au layers greater than 90 nm where a 10 nm Ti adhesion layer was used. The TiO2 deposition was done by sputtering deposition to form as conformal of a coating as possible for the cavity layer. 3D-FEM Simulation: The optical properties of the metal-insulator-metal capped polymer nanopillar SERS substrate were simulated with a 3D finite element method (3D-FEM) study using the COMSOL Multiphysics RF Module. The geometry of the device was built based on SEM images. A single tapered polymer nanopillar with a height of 400 nm, bottom radius of 80 nm, and top radius of 60 nm was modeled in the center of the simulation region on a polymer substrate. A Au-TiO2-Au multilayer, with Ti adhesion layers for each Au layer, was then placed surrounding the polymer nanopillar and on top of the polymer nanopillar. The Ti layer was 9 nm for the bottom Au layer and 5 nm for the top Au layer. The multilayer stack on top of the polymer nanopillar was set to be half of the thickness of the stack surrounding the nanopillar. Au hemispherical nanoparticles with 30 nm diameters were then placed along the polymer nanopillar as two rows with 8 nanoparticles in each row. To simulate the array effect of the nanopillars, periodic boundary conditions were placed along the x and y boundaries with a pitch of 320 nm. The z boundaries were set to be non-reflecting using periodic ports for reflection and transmission calculations. A plane wave source polarized in the x-direction was incident from the top of the simulation region and set to propagate in the -z-direction. The refractive index value

ACS Paragon Plus Environment

5

The Journal of Physical 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

Page 6 of 38

for the polymer substrate and nanopillar was set to 1.56, the refractive index value for the TiO2 cavity layer was set to 2.35, and the wavelength-dependent refractive index values for Ti and Au were interpolated from experimental data.53-54 Reflection and Raman Measurements: Diffuse reflectance measurements were carried out with a Cary 5000G UV-Vis Spectrophotometer from 400 nm to 850 nm with baseline correction and a 1 nm step size. All Raman measurements were carried out using a Renishaw inVia Raman Microscope with an excitation wavelength of 633 nm from a HeNe laser and 20X objective. For acquiring Raman spectra, a grating with 1800 l/mm was used while a grating with 600 l/mm was used for acquiring Raman maps. For all non-mapping data, the laser power was set to 0.7 mW with an integration time of 10 s. For the Raman mapping data, the laser power was set to 0.7 mW with an integration time of 1 s. The mapping window was 100 µm x 100 µm with a step size of 5 µm for a total of 400 spectra. BPE Characterization Sample Preparation and Data Analysis: The molecules trans-1,2bis(4-pyridyl)ethylene (BPE) (Sigma-Aldrich) was chosen for SERS substrate characterization because this molecule has distinct Raman peaks and forms a self-assembled monolayer on Au. The sensors were first cleaned in isopropyl alcohol (IPA), water, IPA, and dried with a nitrogen (N2) gun. A 10 mM BPE solution was prepared in ethanol (EtOH) and the Au SERS substrates were immersed in the solution for 24 hours. The devices were then cleaned in EtOH three times and dried with N2 to ensure that a single monolayer of the BPE molecules was left on the Au surfaces. To characterize the different devices with immobilized BPE in air and water environments, the peak signal intensity and the signal-to-noise ratio (SNR) were both utilized. To obtain the peak signal intensity, the average of the peak values at 1600 cm-1 was used for two consecutively taken Raman spectra at a single point. After each Raman spectrum was acquired

ACS Paragon Plus Environment

6

Page 7 of 38 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

The Journal of Physical Chemistry

for the BPE analysis, a sixth-order polynomial baseline correction was carried out to obtain the final Raman spectrum. The SNR for each peak was calculated as SNR = S/σ where S is the average value of the two peaks and σ is the standard deviation in the peak region divided by √2. To obtain σ accurately, the two spectra were subtracted leaving only the noise and the standard deviation was computed at the peak region.55 Protein-Protein Interaction Detection: After cleaning in IPA, water, IPA, and drying with a N2 gun, the multilayered nanopillar SERS samples were immersed in a 10 mM 11-MUA (SigmaAldrich) ethanolic solution for 24 hours. After incubation, each sensor was washed in pure EtOH three times and dried with a N2 gun. The devices were then immersed in a solution of 400 mM N-(3-Dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC) (Sigma-Aldrich) and 100 mM N-Hydroxysuccinimide (NHS) (Sigma-Aldrich) in Milli-Q water for 30 minutes to activate covalent binding between the MUA molecules and the capture proteins. The solution was then removed with a pipette and without drying the sensors were immersed in a 50 µg/mL avidin solution (Sigma-Aldrich) or 1 µg/mL cancer antigen 125 antibody (anti-CA-125) solution (Abcam), all prepared in phosphate buffered saline (PBS), for 1 hour. Each sensor was then washed in Milli-Q water three times and dried with N2. For the final incubation, the sensors were immersed in a 1 µg/mL biotin solution (Sigma-Aldrich) or 100 ng/mL CA-125 solution (Abcam), all prepared in PBS, for 2 hours. After, each sensor was washed three times with MilliQ water and dried with N2. RESULTS AND DISCUSSION A schematic of the multilayered nanoantenna SERS substrate presented in this work is shown in Figure 1a. The device consists of a periodic nanopillar array, fabricated by a wafer-scale

ACS Paragon Plus Environment

7

The Journal of Physical 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

Page 8 of 38

nanoreplica molding process into a UV-curable polymer. On top of the polymer nanopillar array, a multilayer of Au, TiO2, and Au is deposited, which forms a Fabry-Perot cavity on the top of each nanopillar. Au is chosen for the top and bottom layers because of its favorable plasmonic properties, low reactivity, and high reflectivity, which make it a suitable reflector material in a cavity structure. TiO2 is chosen for the insulator because it has a high refractive index and low loss, making it an appropriate thin film cavity layer material. The optical properties of this Fabry-Perot cavity can be tuned by controlling the deposited thickness of each layer. It is important to note that the cavity design, where the multilayered structure is deposited on top of a periodically nanotextured surface, leads to an intentionally leaky cavity. This enables the stored optical energy in the cavity layer to interact with the confined near-field electric field from the plasmonic effect to provide an enhanced optical interaction with the analyte of interest. If the cavity is designed with Au and TiO2 layers of the proper thicknesses, the result is an overall increase in the SERS EF. A wafer-scale camera image of the fabricated device is shown in Figure 1b. The periodic nanopillar array leads to diffraction of the reflected light, as can be seen by the polychrome color of the camera image. Tilted SEM images of the fabricated device with a bottom Au layer of 90 nm, TiO2 layer of 80 nm, and top Au layer of 90 nm, are shown in Figure 1c and Figure 1d. As can be seen, the result after deposition shows a nanopillar array with a large nanoparticle on top and smaller nanoparticles along the nanopillar sidewall.

ACS Paragon Plus Environment

8

Page 9 of 38 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

The Journal of Physical Chemistry

Figure 1: Plasmonic nanocavity SERS substrate overview. (a) Schematic showing the Au-TiO2Au metal-insulator-metal multilayer on the nanopillar array. (b) Camera image of the wafer-scale fabricated device. (c) Tilted SEM image of the nanopillar array. (d) Close-up tilted SEM image of the nanopillars to show the morphology of individual nanoantennae.

The optical properties of the multilayered SERS substrate were investigated using a 3DFEM simulation study. Snapshots of the developed 3D-FEM model are shown in Figure 2a. A single nanopillar was modeled with periodic boundary conditions along the x and y directions to simulate the array. The near-field electric field distribution for a single nanopillar with a modeled thin (10 nm) protein layer is shown in Figure 2b with a superstrate of air and water at an excitation wavelength of 633 nm. In both cases, there is a highly localized near-field at the Au nanoparticles, along the nanopillar sidewall, and at the Au-TiO2-Au cavity on top of the polymer nanopillar, leading to a large near-field experienced by the modeled 10 nm protein layer. The simulated absorption and reflectance spectra for the polymer nanopillar with a FabryPerot cavity were then compared to the case where the TiO2 layer is replaced with Au, such that

ACS Paragon Plus Environment

9

The Journal of Physical 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

Page 10 of 38

the polymer nanopillar is only capped with a Au layer. The simulated absorption spectra with a superstrate of air and water are shown in Figure 2c for the cavity nanopillar and the Au only nanopillar. As can be seen, the absorption for the cavity case exceeds that obtained for the Au only nanopillar. We anticipate that this is primarily due to reduced reflection and increased field localization for the cavity case. The simulated reflection spectra for the metal-insulator-metal capped nanopillar and Au only capped nanopillar are shown in Figure 2d for superstrates of air and water. As expected, the cavity case shows reduced reflection compared to the Au only case. The near-field electric fields, absorption spectra, and reflection spectra show that the overall optical interaction between the nanostructure and the molecular layer occurs due to a combination of the plasmonic effect and photonic effect. The near-field electric field distributions in Figure 2b show a highly confined near-field electric field at the multilayered cavity structure and at the Au nanoparticles along the nanopillar sidewall. Au nanoparticles are well-known to produce an electromagnetic SERS EF due to the plasmonic effect. The addition of the leaky cavity structure leads to an increase in optical energy at each nanopillar and therefore an increased optical interaction with the SERS sample of interest. An additional advantage of the metal-insulator-metal capped nanopillar device geometry is that the thicknesses of the cavity layers can be changed to optimize the plasmon resonance of the nanopillar array to match well with the excitation wavelength and the wavelength of the Stokes scattered photon. Figure 2e and Figure 2f show the reflection spectra for the multilayered nanopillar with different TiO2 cavity thickness values for fixed top and bottom Au layers of 90 nm with superstrates of air and water, respectively. As can be seen, even by changing the thickness of the TiO2 by 20 nm, there is a significant change in the reflection spectrum in air and

ACS Paragon Plus Environment

10

Page 11 of 38 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

The Journal of Physical Chemistry

water. Therefore, the cavity structure can be used for hot spot engineering and to optimize the plasmon resonance of the array.

Figure 2: 3D-FEM simulation of SERS substrate optical properties. (a) Snapshots of the 3DFEM model. (b) Near-field electric field (|E|) distribution for the multilayered nanopillar with air and water superstrates at a wavelength of 633 nm. (c) Simulated absorption spectra for the multilayered and Au only nanopillar with air and water superstrates. (d) Simulated reflectance spectra for the multilayered and Au only nanopillar array. Simulated reflectance spectra with a superstrate of air (e) and water (f) for different TiO2 cavity thickness values.

To assess the cavity parameters for device optimization, 36 different devices were fabricated with varying cavity parameters (bottom Au thickness of 50 nm, 75 nm, and 90 nm; TiO2 cavity thickness of 40 nm, 60 nm, 80 nm, and 100 nm; top Au thickness of 50 nm, 75 nm, and 90 nm). The reflection spectra in air are shown in Figure 3a (bottom Au layer of 90 nm), Figure 3b (bottom Au layer of 75 nm), and Figure 3c (bottom Au layer of 50 nm). The spectral

ACS Paragon Plus Environment

11

The Journal of Physical 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

Page 12 of 38

properties of the nanopillar array are affected by the thickness of each layer in the multilayer stack. The reflectance dip, corresponding to the plasmon resonance wavelength of the array, is plotted as a function of top Au thickness, cavity thickness, and bottom Au thickness in Figure 3d, Figure 3e, and Figure 3f, respectively. As can be seen, increasing the cavity thickness is a reliable way to redshift the plasmon resonance wavelength of the device. Interestingly, at a top Au thickness of 90 nm, the plasmon resonance tuning effect of changing the cavity layer thickness is significantly dampened, possibly due to dominating cavity effects. The experimental reflection spectra with a superstrate of water are shown in Figure S1 and likewise demonstrate how tuning the cavity layer thicknesses can be used for tuning the reflectance peak and dip for device optimization.

ACS Paragon Plus Environment

12

Page 13 of 38 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

The Journal of Physical Chemistry

Figure 3. Measured reflection spectra with varying cavity parameters. Reflection spectra in air for cavity structures with a Au bottom layer of (a) 90 nm, (b) 75 nm, and (c) 50 nm. The cavity thickness is as labeled in the legend. Each set of curves is vertically offset for clarity: the bottom row corresponds to a top Au thickness of 50 nm, the middle row corresponds to a top Au thickness of 75 nm, and the top row corresponds to a top Au thickness of 90 nm. A plot of the reflectance dip, corresponding to the plasmon resonance wavelength, is plotted as a function of (d) top Au thickness, (e) cavity thickness, and (f) bottom Au thickness.

It is well-known that to maximize the Raman scattering efficiency, the plasmon resonance wavelength of the device structure should be designed such that it lies between the excitation wavelength and the wavelengths corresponding to the Stokes scattered photons of interest. In this work, a laser wavelength of 633 nm is used and Raman spectra are collected in a range of 800-1800 cm-1, which corresponds to a wavelength range of 667-714 nm. Based on the

ACS Paragon Plus Environment

13

The Journal of Physical 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

Page 14 of 38

experimental reflectance spectra, 20 of the 36 measured cavity combinations were identified as being within the desired range. A complete list of these cavity structures is given in Table S1 in the Supporting Information with the device best corresponding to the laser wavelength listed first and the device best corresponding to the Stokes scattered photon listed last. Table 1 lists the cavity structures for the five devices that are at the center of Table S1 and therefore expected to have the best SERS performance in terms of plasmon resonance matching. In this work, the EF is calculated based on the Raman peak at 1600 cm-1 for the BPE molecules, which corresponds to a Stokes scattered photon wavelength of 704 nm. Therefore, from Table S1, with a cavity structure of 90 nm bottom Au layer, TiO2 insulator layer of 100 nm, and top Au layer of 50 nm, the device resonance lies almost exactly between the laser wavelength and the wavelength corresponding to the Stokes scattered photon at a wavenumber of 1600 cm-1. Table 1. Metal-insulator-metal capped polymer nanopillar array configurations with best match in air between plasmon resonance wavelength, 633 nm excitation wavelength, and Stokes scattered photon wavelength. Bottom Au Thickness (nm)

Cavity Thickness (nm)

Top Au Thickness (nm)

Plasmon Resonance Wavelength (nm)

90

80

90

643

75

40

90

645

50

60

90

646

50

80

75

646

75

60

90

648

To further assess the cavity parameters for device optimization, BPE was used to compare the SERS properties of the fabricated sensors. This molecule has strong Raman

ACS Paragon Plus Environment

14

Page 15 of 38 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

The Journal of Physical Chemistry

signatures at 1200 cm-1 and 1600 cm-1 and forms a self-assembled monolayer on Au. Example Raman spectra for three different cavity configurations (bottom Au 90 nm, TiO2 40 nm, top Au 50 nm; bottom Au 90 nm, TiO2 80 nm, top Au 75 nm; bottom Au 75 nm, TiO2 60 nm, top Au 90 nm) with a superstrate of air are shown in Figure 4a with corresponding schematics to show the changes to the cavity layer thicknesses. Example Raman spectra for three different cavity configurations (bottom Au 50 nm, TiO2 100 nm, top Au 50 nm; bottom Au 90 nm, TiO2 60 nm, top Au 75 nm; bottom Au 50 nm, TiO2 80 nm, top Au 90 nm) are shown in Figure 4b for a superstrate of water. The Raman peaks at 1200 cm-1 and 1600 cm-1, characteristic of the BPE molecule, are clearly visible for all the Raman spectra of the shown cavity combinations. However, there is also a clear difference in the Raman spectrum depending on the cavity structure. With an optimized cavity structure, the prominent peaks have a higher intensity and additional peaks become detectable. This level of optimization is crucial for the detection of samples with low Raman scattering efficiency, such as biological molecules. The intensity values from the Raman spectra at a wavenumber of 1600 cm-1 are plotted as a function of top Au thickness, cavity thickness, and bottom Au thickness in Figure 4c, Figure 4d, and Figure 4e, respectively, for a superstrate of air. The Raman intensity at a wavenumber of 1600 cm-1 from the BPE spectra is plotted as a function of top Au thickness, cavity thickness, and bottom Au thickness for a superstrate of water in Figure 4f, Figure 4g, and Figure 4h, respectively. These experimental results show that that a thicker top Au layer will lead to a larger Raman signal both in a superstrate of air and water. The cavity thickness and bottom Au thickness do not show as strong of a trend. Overall, the top three configurations to maximize SERS intensity for BPE at 1600 cm-1 (bottom Au thickness, TiO2 thickness, top Au thickness) in air are (75 nm, 60 nm, 90 nm), (75 nm, 40 nm, 90 nm), and (90 nm, 80 nm, 90 nm).

ACS Paragon Plus Environment

15

The Journal of Physical 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

Page 16 of 38

Figure 4. Tuning the cavity structure for maximum SERS intensity. Raman spectra of a monolayer of BPE on the Au surface are shown for different cavity combinations, indicated in the legend as bottom Au thickness/TiO2 cavity thickness/top Au thickness, with a superstrate of (a) air and (b) water with corresponding schematics of the changing cavity parameters. The intensity values at 1600 cm-1 from the Raman spectra of BPE immobilized on the nanopillar array SERS substrate are shown as a function of (c) top Au thickness, (d) cavity thickness, and (e) bottom Au thickness with a superstrate of air and as a function of (f) top Au thickness, (g) cavity thickness, and (h) bottom Au thickness with a superstrate of water.

We also analyzed the data in terms of the SNR. Biological molecules, such as those involved in the target protein-protein interactions we study in this work, typically do not have

ACS Paragon Plus Environment

16

Page 17 of 38 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

The Journal of Physical Chemistry

strong Raman signatures and we do not want to use high laser powers, which can damage biological samples and lead to high autofluorescence. Therefore, the SNR is an important parameter to assess the device performance under these constraints. Plots of the SNR as a function of the top Au thickness, cavity layer thickness, and bottom Au thickness are shown in Figure 5a, Figure 5b, and Figure 5c, respectively, with a superstrate of air. The SNR as a function of the top Au thickness, cavity layer thickness, and bottom Au thickness for a superstrate of water is shown in Figure 5d, Figure 5e, and Figure 5f, respectively. In air, the SNR is improved for a thicker top Au layer, a TiO2 cavity layer of 80 nm, and a thicker bottom Au layer. In water, the SNR is improved for a thicker top Au layer and 80 nm TiO2 cavity layer. For a superstrate of water, the thickness of the bottom Au layer does not show a reliable trend for the measured configurations. The top thee cavity structures (bottom Au thickness, TiO2 insulator thickness, and top Au thickness) for improving the SNR are given by (90 nm, 80 nm, 90 nm), (75 nm, 80 nm, 90 nm), and (50 nm, 80 nm, 75 nm). Therefore, to optimize both the Raman intensity and the SNR, a structure with a bottom Au layer of 90 nm, TiO2 cavity layer of 80 nm, and top Au layer of 90 nm should be used. The fact that the three optimized structures for SNR all have the same TiO2 cavity thickness suggests that forming an optimized cavity may have a significant impact on the SNR. Overall, these results suggest that the structure of the Fabry-Perot cavity has a measurable impact on the Raman intensity and on the SNR for SERS detection with this multilayered nanopillar substrate. The structures with the highest Raman intensity and SNR values do have plasmon resonance wavelengths that fall within the excitation wavelength and the Stokes scattered photon wavelength. However, the cavity design of 90 nm bottom Au layer, 100 nm TiO2 insulator layer, and 50 nm top Au layer, which was identified as being the best match to fall between the

ACS Paragon Plus Environment

17

The Journal of Physical 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

Page 18 of 38

excitation wavelength and the Stokes scattered photon wavelength for the Raman peak at 1600 cm-1 shows relatively low Raman intensity and SNR. Overall, the results suggest that the SERS substrate optimization relies on a combination of plasmon resonance matching, hot spot engineering, and cavity design.

Figure 5. Cavity tuning for maximum SNR. The SNR at 1600 cm-1 for BPE SERS spectra is shown as a function of (a) top Au thickness, (b) cavity thickness, and (c) bottom Au thickness with a superstrate of air. The SNR at 1600 cm-1 for a superstrate of water is shown as a function of (d) top Au thickness, (e) cavity thickness, and (f) bottom Au thickness.

To assess the relationship between Au thickness, gap size, EF, EF uniformity, and reproducibility, three different multilayered nanopillar SERS sensors were fabricated with bottom Au layers of 90 nm, TiO2 cavity layer thicknesses of 80 nm, and three different top Au layer thicknesses of 90 nm, 125 nm, and 150 nm. Although the Au deposition is done by e-beam evaporation, which is a highly directional deposition method, it is expected that during the deposition there will also be some sidewall coating along the sides of the nanopillars, which will

ACS Paragon Plus Environment

18

Page 19 of 38 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

The Journal of Physical Chemistry

decrease the gap size between neighboring nanopillars. Top-down SEM images of the fabricated multilayered devices with top Au thicknesses of 90 nm, 125 nm, and 150 nm, are shown in Figure 6a, Figure 6b, and Figure 6c, respectively. The morphological uniformity of the device is consistent with increasing Au thickness. Tilted SEM images, which show the structure of the individual nanopillars, are presented in Figure 6d, Figure 6e, and Figure 6f for top Au thicknesses of 90 nm, 125 nm, and 150 nm, respectively. The thicker top Au layer does lead to a reduced gap spacing. At a top Au thickness of 90 nm, the gap spacing is approximately 75 nm; at a top Au thickness of 125 nm, the gap spacing is approximately 60 nm; and at a top Au thickness of 150 nm, the gap spacing is approximately 35 nm. It is expected that the reduced gap spacing between the nanopillars will lead to a higher EF due to improved hot spot engineering. The gap size remains relatively large, even at a top Au thickness of 150 nm, and therefore molecules of interest in this work will easily be able to diffuse and attach in the gaps.

Figure 6. SEM images of the nanopillar array with increasing top Au thickness. Top-down SEM images for a top Au thickness of (a) 90 nm, (b) 125 nm, and (c) 150 nm. The insets show top-

ACS Paragon Plus Environment

19

The Journal of Physical 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

Page 20 of 38

down SEM images at a higher magnification. Tilted SEM images showing individual nanoantenna morphology for a top Au thickness of (a) 90 nm, (b) 125 nm, and (c) 150 nm. This relationship between gap size, EF, and EF uniformity was examined by Raman spectra acquisition and Raman mapping of a BPE monolayer immobilized on the nanopillar SERS substrates with top Au layers of 90 nm, 125 nm, and 150 nm. The SERS spectra were taken with a laser power of 0.7 mW and an integration time of 10 s. As a reference, a 100 mM BPE solution was prepared in a petri dish and the Raman spectra were acquired with a laser power of 1.4 mW and integration time of 60 s. The SERS spectra for the devices with a superstrate of air are shown in Figure 7a and the SERS spectra with a superstrate of water are shown in Figure 7b. As expected, a thicker top Au layer leads to a higher signal intensity. This can be seen in Figure 7c where the average SERS intensity at a wavenumber of 1600 cm-1 for four measurements on each sample is plotted as a function of the top Au thickness with the corresponding error taken as the standard deviation of the four measurements. The experimental EF was then calculated based on EF = (ISERS/NSERS) / (IRaman/NRaman) with corresponding error propagation and the result is shown in Figure 7d. Complete details on the EF calculation are found in the Supporting Information. With an increase in the top Au thickness, there is both an increase in the EF and an increase in the error in the EF. With a top Au layer of 90 nm, the EF is given by (7.3 ± 0.7) x 106 with a superstrate of air and (6 ± 1) x 106 with a superstrate of water. When the top Au layer is 125 nm, the EF is (1.7 ± 0.2) x 107 in air and (1.7 ± 0.6) x 107 in water. For a top Au layer of 150 nm, the EF is (3.2 ± 0.6) x 107 in air and (2.5 ± 0.6) x 107 in water. Raman mapping for BPE immobilized on the SERS substrate was also carried out to further investigate the EF uniformity. The Raman maps, with mapping sizes of 100 µm x 100 µm and step sizes of 5 µm, are shown in Figure 7e, Figure 7f, and Figure 7g with the scale set to

ACS Paragon Plus Environment

20

Page 21 of 38 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

The Journal of Physical Chemistry

cover plus or minus one standard deviation. For mapping, a laser power of 0.7 mW was used with an integration time of 1 s. The average BPE SERS spectra from the mapping data are shown in Figure S2. The spatially averaged EF from the mapping study with a superstrate of air is shown in Figure 7d. The spatially averaged EF for a top Au layer of 90 nm is (9 ± 1) x 106, for a top Au layer of 125 nm is (1.4 ± 0.5) x 107, and for a top Au layer of 150 nm is (2.4 ± 0.8) x 107. The largest deviation between the EF for averaging four spectra and the spatially averaged EF occurs for the device with a top Au thickness of 150 nm. Therefore, this result suggests that to fabricate reliable SERS substrates, accurate reports of the EF should be done using spatially averaged calculations especially when the gap size between neighboring nanostructures is less than 50 nm. The complete datasets for the three devices were also plotted against the theoretical values expected from a normal distribution in quantile-quantile (Q-Q) plots, as shown in Figure S3. It is found that the intensity values are normally distributed over most of the mapping area, but with some outliers at high intensity values, which are likely due to large background signals or nonuniform Au deposition. Therefore, for all devices the EF is sufficient and, even with a top Au layer of 150 nm, the EF error is sufficiently low. These devices can then all be used for quantitative analysis where the thinner top Au layer will lead to a lower EF, but with a lower error.

ACS Paragon Plus Environment

21

The Journal of Physical 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

Page 22 of 38

Figure 7. SERS EF and uniformity of the multilayered nanoantenna array. (a) The Raman spectra of BPE are shown with a superstrate of air and nanoantenna arrays with increasing top Au thicknesses. (b) BPE Raman spectra taken in water with increasing top Au thicknesses. (c) The SERS intensity with error at 1600 cm-1 is plotted as a function of top Au thickness. (d) Calculated EF with corresponding error at 1600 cm-1 as a function of top Au thickness. (e-g) Raman maps of BPE on the multilayered SERS substrate showing the uniformity over a single standard deviation. The mapping area is 100 µm by 100 µm and the scale bar is 25 µm.

ACS Paragon Plus Environment

22

Page 23 of 38 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

The Journal of Physical Chemistry

The multilayered nanopillar device with a 150 nm top Au layer was then applied to detect protein-protein interactions by measurement of SERS spectra for 11-MUA self-assembled monolayer formation followed by two different protein-protein systems. The first is biotin and avidin and the second is CA-125 and its antibody anti-CA-125. Biotin-avidin binding is wellknown as a protein complex standard for sensors due to its high binding affinity and CA-125 is a cancer biomarker for ovarian cancer. The two protein-protein systems were analyzed to assess spectral differences as molecules are added. The capture molecules (avidin and anti-CA-125) were immobilized onto the sensor surface using 11-MUA. The as-acquired SERS spectrum of the device itself is shown in Figure S4. There is a sharp background peak at 1002 cm-1 and a broader peak at 1603 cm-1. The as-acquired SERS spectra at different steps of the binding process are presented in Figure 8. Figure 8a shows the SERS spectrum after 11-MUA monolayer formation. There are several strong SERS peaks, which show the power of SERS sensors for small molecule detection. The locations of the peaks agree well with previously reported SERS spectra of 11-MUA.56-57 The peak at 1179 cm-1 corresponds to a C-H deformation, the peak at 1240 cm-1 corresponds to a C-O deformation, and the peak at 1477 cm-1 corresponds to a CH2 deformation. The peaks at 1527 cm-1 and 1576 cm-1 likely correspond to O-C-O.56, 58 The SERS spectra after immobilization of avidin and biotin on the 11-MUA coated device are shown in Figure 8b and Figure 8c, respectively. The spectra after immobilization of anti-CA-125 and CA-125 are shown in Figure 8d and Figure 8e, respectively. There are clear spectral changes in terms of Raman peak positions and intensities as subsequent molecules are immobilized onto the device surface. In addition, there are clear spectral differences between the two protein systems. In both cases, the addition of the capture molecule onto the MUA

ACS Paragon Plus Environment

23

The Journal of Physical 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

Page 24 of 38

monolayer leads to a significant decrease in signal intensity. This is possibly due to a modification of the MUA monolayer after incubation and binding of the capture proteins. Since MUA is in direct contact with the Au on the device surface, it is expected that spectral signatures from MUA will have a significant contribution to the Raman spectrum even after subsequent molecules are added. A list of the prominent Raman peaks for each molecule are given in the Supporting Information in Table S2. The results show that this SERS sensor is especially promising for small molecule measurements that may be not be detectable using other methods. The Raman signal for the protein-protein interactions can likely be increased by using thiolated capture proteins instead of a 11-MUA monolayer for covalent binding to the Au surface because the thiol molecule will be much smaller than 11-MUA and will keep the protein-protein interactions closer to the device surface.

ACS Paragon Plus Environment

24

Page 25 of 38 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

The Journal of Physical Chemistry

Figure 8. Protein-protein interactions on the multilayered SERS substrate. (a) The SERS spectrum after incubation in 11-MUA is shown. SERS spectra taking after immobilization of (b) avidin and (c) biotin on the MUA-coated device. The SERS spectra are shown after incubation in (d) anti-CA-125 and (e) the cancer biomarker CA-125. All SERS spectra are as-acquired and therefore do not have any background correction.

CONCLUSION A SERS sensor based on a Au-TiO2-Au Fabry-Perot cavity nanoantenna array was developed and the optical properties of the novel device were investigated using a combination of simulation and experiment. It was demonstrated that a combination of plasmon resonance tuning, hot spot engineering, and cavity optimization impact crucial parameters for the development of SERS substrates such as EF, EF uniformity, SNR, and reproducibility. A spatially averaged EF

ACS Paragon Plus Environment

25

The Journal of Physical 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

Page 26 of 38

of (2.4 ± 0.8) x 107 was demonstrated with sufficiently low error for quantitative studies. A tradeoff is also demonstrated between EF and EF uniformity, but with a gap size as small as 35 nm between neighboring nanopillars, a spatially averaged EF with sufficiently small error is obtained. The developed SERS substrate was then applied to detect protein-protein interactions, including the cancer biomarker CA-125 at a concentration of 100 ng/mL, and the power of the developed SERS sensor to detect small molecules was demonstrated.

AUTHOR INFORMATION Corresponding Author *[email protected] Author Contributions All authors have given approval to the final version of the manuscript. ACKNOWLEDGMENT This work was carried out in part in the Micro and Nanotechnology Laboratory and the Frederick Seitz Material Research Laboratory Central Research Facilities and the University of Illinois at Urbana-Champaign. The research was supported by the U.S. Department of Agriculture award number 2014-06959. In addition, funding for L.P.H. was provided by the Linda Su-Nan Chang Sah Doctoral Fellowship.

REFERENCES 1. Nath, N.; Chilkoti, A., Label-Free Biosensing by Surface Plasmon Resonance of Nanoparticles on Glass: Optimization of Nanoparticle Size. Anal Chem 2004, 76, 5370-5378.

ACS Paragon Plus Environment

26

Page 27 of 38 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

The Journal of Physical Chemistry

2. Fan, Z.; Kanchanapally, R.; Ray, P. C., Hybrid Graphene Oxide Based Ultrasensitive SERS Probe for Label-Free Biosensing. J Phys Chem Lett 2013, 4, 3813-3818. 3. Ngo, H. T.; Wang, H. N.; Fales, A. M.; Vo-Dinh, T., Label-Free DNA Biosensor Based on SERS Molecular Sentinel on Nanowave Chip. Anal Chem 2013, 85, 6378-6383. 4. Pavel, I.; McCarney, E.; Elkhaled, A.; Morrill, A.; Plaxco, K.; Moskovits, M., Label-Free SERS Detection of Small Proteins Modified to Act as Bifunctional Linkers. J Phys Chem C 2008, 112, 4880-4883. 5. Klein, K.; Gigler, A. M.; Aschenbrenne, T.; Monetti, R.; Bunk, W.; Jamitzky, F.; Morfill, G.; Stark, R. W.; Schlegel, J., Label-Free Live-Cell Imaging with Confocal Raman Microscopy. Biophys J 2012, 102, 360-368. 6. Chourpa, I.; Lei, F. H.; Dubois, P.; Manfait, M.; Sockalingum, G. D., Intracellular Applications of Analytical SERS Spectroscopy and Multispectral Imaging. Chem Soc Rev 2008, 37, 993-1000. 7. Kang, J. W.; So, P. T. C.; Dasari, R. R.; Lim, D. K., High Resolution Live Cell Raman Imaging Using Subcellular Organelle-Targeting SERS-Sensitive Gold Nanoparticles with Highly Narrow Intra-Nanogap. Nano Lett 2015, 15, 1766-1772. 8. Kneipp, J.; Kneipp, H.; Rajadurai, A.; Redmond, R. W.; Kneipp, K., Optical Probing and Imaging of Live Cells Using SERS Labels. J Raman Spectrosc 2009, 40, 1-5. 9. Lee, S.; Chon, H.; Lee, J.; Ko, J.; Chung, B. H.; Lim, D. W.; Choo, J., Rapid and Sensitive Phenotypic Marker Detection on Breast Cancer Cells Using Surface-Enhanced Raman Scattering (SERS) Imaging. Biosens Bioelectron 2014, 51, 238-243. 10. Palonpon, A. F.; Ando, J.; Yamakoshi, H.; Dodo, K.; Sodeoka, M.; Kawata, S.; Fujita, K., Raman and SERS Microscopy for Molecular Imaging of Live Cells. Nat Protoc 2013, 8, 677692. 11. Li, X. X.; Blinn, K.; Fang, Y. C.; Liu, M. F.; Mahmoud, M. A.; Cheng, S.; Bottomley, L. A.; El-Sayed, M.; Liu, M. L., Application of Surface Enhanced Raman Spectroscopy to the Study of SOFC Electrode Surfaces. Phys Chem Chem Phys 2012, 14, 5919-5923. 12. Kalish, R.; Reznik, A.; Nugent, K. W.; Prawer, S., The Nature of Damage in IonImplanted and Annealed Diamond. Nucl Instrum Meth B 1999, 148, 626-633. 13. Freeman, R. G., et al., Self-Assembled Metal Colloid Monolayers - an Approach to SERS Substrates. Science 1995, 267, 1629-1632. 14. Grabar, K. C.; Freeman, R. G.; Hommer, M. B.; Natan, M. J., Preparation and Characterization of Au Colloid Monolayers. Anal Chem 1995, 67, 735-743. 15. Leopold, N.; Lendl, B., A New Method for Fast Preparation of Highly Surface-Enhanced Raman Scattering (SERS) Active Silver Colloids at Room Temperature by Reduction of Silver Nitrate with Hydroxylamine Hydrochloride. J Phys Chem B 2003, 107, 5723-5727. 16. Kneipp, K.; Wang, Y.; Dasari, R. R.; Feld, M. S., Approach to Single-Molecule Detection Using Surface-Enhanced Resonance Raman-Scattering (SERRS) - a Study Using Rhodamine 6G on Colloidal Silver. Appl Spectrosc 1995, 49, 780-784. 17. Kneipp, K.; Wang, Y.; Kneipp, H.; Perelman, L. T.; Itzkan, I.; Dasari, R.; Feld, M. S., Single Molecule Detection Using Surface-Enhanced Raman Scattering (SERS). Phys Rev Lett 1997, 78, 1667-1670. 18. Nie, S. M.; Emery, S. R., Probing Single Molecules and Single Nanoparticles by SurfaceEnhanced Raman Scattering. Science 1997, 275, 1102-1106. 19. Dick, L. A.; McFarland, A. D.; Haynes, C. L.; Van Duyne, R. P., Metal Film over Nanosphere (Mfon) Electrodes for Surface-Enhanced Raman Spectroscopy (SERS):

ACS Paragon Plus Environment

27

The Journal of Physical 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

Page 28 of 38

Improvements in Surface Nanostructure Stability and Suppression of Irreversible Loss. J Phys Chem B 2002, 106, 853-860. 20. Hulteen, J. C.; Vanduyne, R. P., Nanosphere Lithography - a Materials General Fabrication Process for Periodic Particle Array Surfaces. J Vac Sci Technol A 1995, 13, 15531558. 21. Zhang, X. Y.; Yonzon, C. R.; Van Duyne, R. P., Nanosphere Lithography Fabricated Plasmonic Materials and Their Applications. J Mater Res 2006, 21, 1083-1092. 22. Merk, V.; Kneipp, J.; Leosson, K., Gap Size Reduction and Increased SERS Enhancement in Lithographically Patterned Nanoparticle Arrays by Templated Growth. Adv Opt Mater 2013, 1, 313-318. 23. Caldwell, J. D.; Glembocki, O.; Bezares, F. J.; Bassim, N. D.; Rendell, R. W.; Feygelson, M.; Ukaegbu, M.; Kasica, R.; Shirey, L.; Hosten, C., Plasmonic Nanopillar Arrays for LargeArea, High-Enhancement Surface-Enhanced Raman Scattering Sensors. ACS Nano 2011, 5, 4046-4055. 24. Hu, Y. S.; Jeon, J.; Seok, T. J.; Lee, S.; Hafner, J. H.; Drezek, R. A.; Choo, H., Enhanced Raman Scattering from Nanoparticle-Decorated Nanocone Substrates: A Practical Approach to Harness in-Plane Excitation. ACS Nano 2010, 4, 5721-5730. 25. Li, J. Q.; Chen, C.; Jans, H.; Xu, X. M.; Verellen, N.; Vos, I.; Okumura, Y.; Moshchalkov, V. V.; Lagae, L.; Van Dorpe, P., 300 mm Wafer-Level, Ultra-Dense Arrays of Au-Capped Nanopillars with Sub-10 Nm Gaps as Reliable SERS Substrates. Nanoscale 2014, 6, 12391-12396. 26. Huang, Z. L.; Meng, G. W.; Huang, Q.; Yang, Y. J.; Zhu, C. H.; Tang, C. L., Improved SERS Performance from Au Nanopillar Arrays by Abridging the Pillar Tip Spacing by Ag Sputtering. Adv Mater 2010, 22, 4136-+. 27. Li, X. H.; Chen, G. Y.; Yang, L. B.; Jin, Z.; Liu, J. H., Multifunctional Au-Coated Tio2 Nanotube Arrays as Recyclable SERS Substrates for Multifold Organic Pollutants Detection. Adv Funct Mater 2010, 20, 2815-2824. 28. Oh, Y. J.; Jeong, K. H., Glass Nanopillar Arrays with Nanogap-Rich Silver Nanoislands for Highly Intense Surface Enhanced Raman Scattering. Adv Mater 2012, 24, 2234-2237. 29. Ruan, C. M.; Eres, G.; Wang, W.; Zhang, Z. Y.; Gu, B. H., Controlled Fabrication of Nanopillar Arrays as Active Substrates for Surface-Enhanced Raman Spectroscopy. Langmuir 2007, 23, 5757-5760. 30. Hackett, L. P.; Goddard, L. L.; Liu, G. L., Plasmonic Nanocone Arrays for Rapid and Detailed Cell Lysate Surface Enhanced Raman Spectroscopy Analysis. Analyst 2017, 142, 44224430. 31. Huang, J. A.; Zhao, Y. Q.; Zhu, X. Y.; Zhang, W. J., Averaging Effect on Improving Signal Reproducibility of Gap-Based and Gap-Free SERS Substrates Based on Ordered Si Nanowire Arrays. Rsc Adv 2017, 7, 5297-5305. 32. Zhao, Y. Q.; Zhang, Y. L.; Huang, J. A.; Zhang, Z. Y.; Chen, X. F.; Zhang, W. J., Plasmonic Nanopillar Array Embedded Microfluidic Chips: An in Situ Sers Monitoring Platform. J Mater Chem A 2015, 3, 6408-6413. 33. Zhou, W.; Suh, J. Y.; Hua, Y.; Odom, T. W., Hybridization of Localized and Guided Modes in 2D Metal-Insulator-Metal Nanocavity Arrays. J Phys Chem C 2013, 117, 2541-2546. 34. Sorger, V. J.; Oulton, R. F.; Yao, J.; Bartal, G.; Zhang, X., Plasmonic Fabry-Perot Nanocavity. Nano Lett 2009, 9, 3489-3493.

ACS Paragon Plus Environment

28

Page 29 of 38 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

The Journal of Physical Chemistry

35. Yang, X. D.; Ishikawa, A.; Yin, X. B.; Zhang, X., Hybrid Photonic - Plasmonic Crystal Nanocavities. ACS Nano 2011, 5, 2831-2838. 36. Ameen, A.; Hackett, L. P.; Seo, S. J.; Dar, F. K.; Gartia, M. R.; Goddard, L. L.; Liu, G. L. G., Plasmonic Sensing of Oncoproteins without Resonance Shift Using 3D Periodic Nanocavity in Nanocup Arrays. Adv Opt Mater 2017, 5. 37. Artar, A.; Yanik, A. A.; Altug, H., Fabry-Peacuterot Nanocavities in Multilayered Plasmonic Crystals for Enhanced Biosensing. Appl Phys Lett 2009, 95. 38. Cattoni, A.; Ghenuche, P.; Haghiri-Gosnet, A. M.; Decanini, D.; Chen, J.; Pelouard, J. L.; Collin, S., Lambda(3)/1000 Plasmonic Nanocavities for Biosensing Fabricated by Soft UV Nanoimprint Lithography. Nano Lett 2011, 11, 3557-3563. 39. Hackett, L. P.; Ameen, A.; Li, W.; Dar, F. K.; Goddard, L. L.; Liu, G. L., SpectrometerFree Plasmonic Biosensing with Metal-Insulator-Metal Nanocup Arrays. ACS Sens 2018. 40. Chanda, D.; Shigeta, K.; Truong, T.; Lui, E.; Mihi, A.; Schulmerich, M.; Braun, P. V.; Bhargava, R.; Rogers, J. A., Coupling of Plasmonic and Optical Cavity Modes in Quasi-ThreeDimensional Plasmonic Crystals. Nat Commun 2011, 2. 41. Schmidt, M. A.; Lei, D. Y.; Wondraczek, L.; Nazabal, V.; Maier, S. A., Hybrid Nanoparticle-Microcavity-Based Plasmonic Nanosensors with Improved Detection Resolution and Extended Remote-Sensing Ability. Nat Commun 2012, 3. 42. Tabatabaei, M.; Najiminaini, M.; Davieau, K.; Kaminska, B.; Singh, M. R.; Carson, J. J. L.; Lagugne-Labarthet, F., Tunable 3d Plasmonic Cavity Nanosensors for Surface-Enhanced Raman Spectroscopy with Sub-Femtomolar Limit of Detection. ACS Photonics 2015, 2, 752759. 43. Xu, J. J.; Guan, P.; Kvasnicka, P.; Gong, H.; Homola, J.; Yu, Q. M., Light Transmission and Surface-Enhanced Raman Scattering of Quasi-3D Plasmonic Nanostructure Arrays with Deep and Shallow Fabry-Perot Nanocavities. J Phys Chem C 2011, 115, 10996-11002. 44. Brown, L. V.; Yang, X.; Zhao, K.; Zheng, B. Y.; Nordlander, P.; Halas, N. J., FanShaped Gold Nanoantennas above Reflective Substrates for Surface-Enhanced Infrared Absorption (SEIRA). Nano Lett 2015, 15, 1272-1280. 45. Wadell, C.; Antosiewicz, T. J.; Langhammer, C., Optical Absorption Engineering in Stacked Plasmonic Au-SiO2-Pd Nanoantennas. Nano Lett 2012, 12, 4784-4790. 46. Kudelski, A., Raman Studies on the Coverage Integrity of Monolayers Formed on Silver from Various Omega-Functionalised Alkanethiols. Vib Spectrosc 2006, 41, 83-89. 47. Han, X. X.; Jia, H. Y.; Wang, Y. F.; Lu, Z. C.; Wang, C. X.; Xu, W. Q.; Zhao, B.; Ozaki, Y., Analytical Technique for Label-Free Multi-Protein Detection-Based on Western Blot and Surface-Enhanced Raman Scattering. Anal Chem 2008, 80, 2799-2804. 48. Han, X. X.; Kitahama, Y.; Tanaka, Y.; Guo, J.; Xu, W. Q.; Zhao, B.; Ozaki, Y., Simplified Protocol for Detection of Protein-Ligand Interactions Via Surface-Enhanced Resonance Raman Scattering and Surface-Enhanced Fluorescence. Anal Chem 2008, 80, 65676572. 49. Kustner, B.; Gellner, M.; Schutz, M.; Schoppler, F.; Marx, A.; Strobel, P.; Adam, P.; Schmuck, C.; Schlucker, S., Sers Labels for Red Laser Excitation: Silica-Encapsulated SAMS on Tunable Gold/Silver Nanoshells. Angew Chem Int Edit 2009, 48, 1950-1953. 50. Wang, Y. L.; Lee, K.; Irudayaraj, J., Sers Aptasensor from Nanorod-Nanoparticle Junction for Protein Detection. Chem Commun 2010, 46, 613-615.

ACS Paragon Plus Environment

29

The Journal of Physical 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

Page 30 of 38

51. Yoon, K. J.; Seo, H. K.; Hwang, H.; Pyo, D.; Eom, I. Y.; Hahn, J. H.; Jung, Y. M., Bioanalytical Application of SERS Immunoassay for Detection of Prostate-Specific Antigen. B Korean Chem Soc 2010, 31, 1215-1218. 52. Zhao, B.; Shen, J. L.; Chen, S. X.; Wang, D. F.; Li, F.; Mathur, S.; Song, S. P.; Fan, C. H., Gold Nanostructures Encoded by Non-Fluorescent Small Molecules in Polya-Mediated Nanogaps as Universal SERS Nanotags for Recognizing Various Bioactive Molecules. Chem Sci 2014, 5, 4460-4466. 53. Johnson, P. B.; Christy, R. W., Optical Constants of Noble Metals. Phys Rev B 1972, 6, 4370-4379. 54. Johnson, P. B.; Christy, R. W., Optical-Constants of Transition-Metals - Ti, V, Cr, Mn, Fe, Co, Ni, and Pd. Phys Rev B 1974, 9, 5056-5070. 55. McCreery, R. L., Raman Spectroscopy for Chemical Analysis; John Wiley & Sons: New York, 2000, p xxiv, 420 p. 56. Ignat, T.; Munoz, R.; Irina, K.; Obieta, I.; Mihaela, M.; Simion, M.; Iovu, M., Nanostructured Au/Si Substrate for Organic Molecule Sers Detection. Superlattice Microst 2009, 46, 451-460. 57. Ignat, T.; Munoz, R.; Kleps, I.; Miu, M.; Obieta, I.; Simion, M., SERS-Active Substrate Based on Macroporous Silicon. Rom J Inf Sci Tech 2009, 12, 496-503. 58. Ignat, T.; Husanu, M. A.; Munoz, R.; Kusko, M.; Danila, M.; Teodorescu, C. M., Gold Nano-Island Arrays on Silicon as Sers Active Substrate for Organic Molecule Detection. Thin Solid Films 2014, 550, 354-360. TOC Graphic

ACS Paragon Plus Environment

30

Page 31 of 38 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

The Journal of Physical Chemistry

463x381mm (96 x 96 DPI)

ACS Paragon Plus Environment

The Journal of Physical 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

648x400mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 32 of 38

Page 33 of 38 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

The Journal of Physical Chemistry

556x484mm (96 x 96 DPI)

ACS Paragon Plus Environment

The Journal of Physical 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

444x406mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 34 of 38

Page 35 of 38 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

The Journal of Physical Chemistry

450x274mm (96 x 96 DPI)

ACS Paragon Plus Environment

The Journal of Physical 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

707x382mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 36 of 38

Page 37 of 38 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

The Journal of Physical Chemistry

461x467mm (96 x 96 DPI)

ACS Paragon Plus Environment

The Journal of Physical 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

365x291mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 38 of 38