Visualizing Electric Fields at Au(111) Step Edges via Tip-Enhanced

Oct 3, 2017 - This work was performed in the environmental and molecular sciences laboratory (EMSL), a DOE Office of Science User Facility sponsored b...
0 downloads 8 Views 3MB Size
Subscriber access provided by UNIVERSITY OF WISCONSIN-MILWAUKEE

Communication

Visualizing Electric Fields at Au(111) Step Edges via Tip-Enhanced Raman Scattering Ashish Bhattarai, Alan G. Joly, Wayne P Hess, and Patrick Z. El-Khoury Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.7b04027 • Publication Date (Web): 03 Oct 2017 Downloaded from http://pubs.acs.org on October 4, 2017

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.

Nano Letters 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 20

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

Nano Letters

Visualizing Electric Fields at Au(111) Step Edges via Tip-Enhanced Raman Scattering

Ashish Bhattarai, Alan G. Joly, Wayne P. Hess, and Patrick Z. El-Khoury*

Physical Sciences Division, Pacific Northwest National Laboratory, P.O. Box 999, Richland, WA 99352, USA

*

[email protected]

TOC Graphic

1 ACS Paragon Plus Environment

Nano Letters

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

ABSTRACT Tip-enhanced Raman scattering (TERS) can be used to image plasmon-enhanced local electric fields on the nanoscale. This is illustrated through ambient TERS measurements recorded using silver atomic force microscope tips coated with 4-mercaptobenzonitrile molecules and used to image step edges on an Au(111) surface. The observed 2D TERS images uniquely map electric fields localized at Au(111) step edges following 671-nm excitation. We establish that our measurements are not only sensitive to spatial variations in the enhanced electric fields but also to their vector components. We also experimentally demonstrate that (i) few nanometer precision is attainable in TERS nanoscopy using corrugated tips with nominally radii on the order of 100200 nm, and (ii) TERS signals do not necessarily exhibit the expected E4 dependence. Overall, we illustrate the concept of electric field imaging via TERS and establish the connections between our observations and conventional TERS chemical imaging measurements.

KEYWORDS Tip-Enhanced Raman, Raman Scattering, Nanoscale Chemical Imaging, Local Electric Fields, Plasmons

2 ACS Paragon Plus Environment

Page 2 of 20

Page 3 of 20

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

Nano Letters

Of the challenges associated with optical measurements aimed at the chemical characterization of an isolated nanoscopic analyte (e.g. a single molecule), the mismatch between the size of the analyte and the light source – conventionally diffracted-limited at best – is most prohibitive. Background scattering and signal-to-noise considerations aside, this disparity translates into an overall feeble interaction between light and a nanoscopic target, which limits the use of standard microscopic and/or spectroscopic approaches in this context.1 A viable approach to circumventing the aforementioned limitation comprises enhancing the optical crosssection of the nanoscopic system by coupling it to a plasmonic antenna. This is the underlying concept behind surface-enhanced Raman scattering (SERS),2,3,4 a technique that affords single molecule detection sensitivity,5,6 all while retaining the rich information of Raman scattering.2-6 Tip-enhanced Raman scattering (TERS)7,8,9,10 is a closely related technique that employs a plasmonic scanning probe to enhance molecular optical cross-sections. In the TERS scheme, it is thus also possible to achieve nanoscale chemical analysis and imaging by scanning the tip relative to the substrate. This realization resulted in numerous elegant demonstrations, as highlighted in recent TERS reviews.11,12,13,14 Although recent works call for the treatment of the coupled molecular-nanoplasmonic system as a whole,15,16 it is intuitive to decompose SERS (and TERS) signals into a combination of two cooperative effects.17 The first mechanism is chemical in nature,18 resulting from (i) changes in the polarizabilities of molecules dynamically interacting with a metal, and (ii) resonant enhancement, operative when plasmonic metal-to-molecule (or vice versa) charge transfer states are accessible at the incident photon energy.19 The second mechanism arises from the resonant interaction between light and plasmonic metal nanostructures, namely, the excitation of localized surface plasmon Eigenmodes.17,20 This results in the localization and

3 ACS Paragon Plus Environment

Nano Letters

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 20

enhancement of the incident and scattered radiation fields.20 In this framework, a comprehensive understanding of TERS spectra and images requires deconvoluting the problem and addressing each of the aforementioned mechanisms separately. This remains a challenge, requiring, e.g., correlating TERS nanographs with nanoscale-resolved ultrasensitive broad-band optical absorption images to capture hybrid molecular-metallic charge transfer states. These states, the energies and intensities of which are dictated by specific interactions between molecules and their heterogeneous local environments, govern the chemical enhancement mechanism.21 In the same vein, variations in the structures of nanoscopic plasmonic constructs (e.g. the metallic tip) may completely alter the structure22 and overall nature23 of the local fields, and hence, the plasmonic enhancement mechanism. This is particularly the case at plasmonic tip-sample nanojunctions, whereby hybrid plasmonic modes (e.g. bonding dipolar plasmon resonances) are exploited to achieve ultrasensitive and/or ultrahigh spatial resolution in TERS.24 In this regime, the commonly adopted practice of correlating structural/topographic images of plasmonic substrates (and/or tips) with molecular Raman spectra and images is often insufficient; the need for novel approaches that can be used to visualize plasmon-enhanced local electric fields on the nanoscale presents itself.25 Herein, we propose that molecular TERS is ideally suited for this purpose. Namely, assuming separable molecular and metallic polarizabities, TERS signals follow24 





  ∑   (Ω) 

(1)

 is the molecular polarizability derivative tensor of the nth vibrational in which  Eigenstate,  = , ,  is the Euler angle which dictate molecular orientation relative to the

vector components of the enhanced incident and scattered local electric fields,  , . In the realm

of ultrasensitive and/or ultra-high spatial resolution (a few nm or less) TERS, orientational 4 ACS Paragon Plus Environment

Page 5 of 20

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

Nano Letters

averaging is no longer appropriate;11-14,26 the optical response of a single scatterer (or a small number of scatterers) exposes the tensorial nature of Raman scattering. In this work, we take advantage of the tensorial nature of Raman scattering to map plasmon-enhanced local electric fields on the nanoscale. We employ silver atomic force microscope tips coated with 4mercaptobenzonitrile (MBN) molecules to visualize the vector components of plasmonic fields localized at Au(111) step edges. Our approach and choice of substrate are motivated by a recent report from our group,26 whereby TERS was used to image plasmon-enhanced local electric field variations on a sputtered silver film with extremely high spatial resolution (~1 nm). The attainable resolution was possible by virtue of the hybridization between the plasmon modes of the corrugated tip and the corrugated surface, i.e., the formation of a TERS gap mode. Nonetheless, both the orientation of the molecules on the tip and the electric fields on the surface were unknown, and as such, access to the vector components of the local electric fields was limited. In this regard, this work significantly expands the scope of our prior work, since the fields sustained on the step edges can be understood. Given molecular polarizability derivative tensors (from a density functional theory calculation) and numerically simulated electric fields (from finite-difference time-domain, or FDTD, simulations), we strive to forward simulate our TERS images following equation 1. This allows us to access the vector components of plasmonenhanced electric fields localized at Au(111) step-edges. We begin by describing diagnostic measurements of the tips and substrate. Figure 1A shows a representative atomic force microscopy (AFM) image of our Au(111) substrate. Cross-sectional line cuts tracing multiple terraces are shown in Figure 1B, and reveal an average height difference of ~2 nm between the terraces. For simplicity, we simply refer to the (multi-step) boundaries between the visualized Au(111) terraces as step edges. We also note

5 ACS Paragon Plus Environment

Nano Letters

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

that the subsequent step edges we consider in this study are comparable in height to the ones encountered in Figures 1A and 1B. Figures 1C and 1D show SERS and TERS point spectra recorded using two different MBN-coated AFM tips. The SERS spectra were recorded by focusing our incident laser on the MBN-coated tips, whereas the TERS spectra were recorded when the same tips were brought into contact with an Au(111) terrace. The apices of the first tip shown in the inset of Figure 1C (tip 1 from hereon) and its analogue in the inset of Figure 1D (tip 2 from hereon) are ~200 and 100 nm in diameter, respectively. Interestingly, similar TERS/SERS ratios were obtained using the two tips. This observation is attributed to a competition between the number of molecules (more on average for tip 1) and the plasmon resonance condition at 671-nm irradiation in the two cases. Namely, stronger Raman (and TERS) signals from tip 2 suggest that the plasmon resonance of this tip is better-aligned with the wavelength of the driving laser. Figures 2A and 2B show correlated AFM-TERS images of an Au(111) step edge. The TERS image (2B) was recorded with tip 1, and some of the information content therein is visualized in Figures 2C-E. The first plot (2C) shows 20 spectra recorded at an Au(111) terrace, all contained within the area highlighted by a dotted rectangle in Figure 2B; we observe nominally identical TERS spectra on terraces. This is in contrast with the second and third plot (2D and 2E), where we trace the change in TERS scattering activity at the ~1580 cm-1 (center of the vibrational resonance of the aromatic stretching vibration of MBN molecules on the tip) as the tip traverses a step edge. Both plots reveal that TERS activity at 1580 cm-1 is maximal in the immediate vicinity of the step edge, and weaker (to different extents) towards the left and right terraces. The effect is apparent in Figure 2E, which reveals a distinctive 1580 cm-1 TERS line profile, the origin of which is examined below. Notably, the TERS imaging measurements

6 ACS Paragon Plus Environment

Page 6 of 20

Page 7 of 20

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

Nano Letters

subsequently performed using tip 2 are dissimilar. These results are shown in Figure 3A-C and further analyzed in panels C-F of the same figure. In this case, the spectral variations (3D) and TERS intensity cross-section (3E) at 1580 cm-1 exhibit a pronounced dip centered at the Au(111) step edge position. More interestingly, the simultaneously recorded 2224 cm-1 image (2C) and profile (3F) are distinct from both their above-described 1580 cm-1 analogues. Not only are the recorded TERS images and profiles tip-dependent; slices belonging to the simultaneously recorded TERS spectral image at different vibrational resonances are also distinct. The above-described TERS images map the local electric fields in the vicinity of Au(111) step edges. To illustrate the concept, we started by performing FDTD simulations of our substrate. The two TERS cross-sections recorded with tips 1 and 2 are first compared to different electric field components in Figures 4A and 4B, respectively. Specifically, the TERS trace recorded with tip 1 is compared to ( ) and tip 2 is compared to ( ) , where n=0-4. In our laboratory frame of reference, the z-axis is defined as the axis perpendicular to the substrate and parallel to the tip, while the x axis is perpendicular to the edge axis. Besides some broadening of the experimental traces, the simulated electric field and TERS profiles are very similar in structure; our TERS images map variations in the vector components of the localized electric fields. Following our introductory remarks and our observation of different vector components with different tips (and molecular distributions), this strongly indicates that we are operative in the few molecule regime. Recall that our approach to electric field mapping relies on Raman scattering from small ensembles of MBN molecules residing on sputtered metallic AFM tips. That different electric field components on the substrate can be imaged using Raman scattering in the few-molecule regime is consistent with and supported by several prior reports from our group.24,27,28,29 At this stage, it is important to note that the chosen line profiles in Figures 4

7 ACS Paragon Plus Environment

Nano Letters

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

display limiting cases where the TERS profiles reflect predominantly either Ex or Ez field components. We do not ascribe any significance of the measured vectorial component to the actual tip radius, but simply make the case that the two tips have different nanoscale variations thereby enhancing different local electric field combinations of electric field components. The result is that most TERS step profiles display a combination of electric field components, see the supporting information section. In the few molecule regime where the tensorial nature of Raman scattering is operative, this results in TERS images that directly reflect the local field characteristics in both intensity and polarization. The results summarized in Figure 4A-B also seem to suggest that the commonly adopted   TERS enhancement law does not hold; the experimental traces follow ~  and ~  with tips 1 and 2, respectively. Classically, incident polarization along the tip axis is required to affect TERS enhancement.30 An incident polarization that is orthogonal to the tip axis does not allow for efficient coupling of the incident field to the metallic tip, in which case only the scattered radiation is enhanced (as   ) and the E4 law does not hold.31 Evidently, this treatment assumes idealized tips, which is far from our case of sputtered silver tips. Our observed deviation from the E4 law can be simply understood by considering the multiple possible projections of the incident polarization onto nanometric corrugations sustained on the sputtered silver tip. In this regard, our observed power law dependence further supports our picture of tensorial Raman scattering from small ensembles of MBN molecules residing in their unique local environments at the apices of the two tips. To further our understanding of our experimental observables, we numerically simulated our recorded TERS spectral images. The details of our simulations are documented in the supporting information section. Briefly, our model follows equation 1; we use molecular polarizability derivatives from density functional theory calculations of an isolated MBN

8 ACS Paragon Plus Environment

Page 8 of 20

Page 9 of 20

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

Nano Letters

molecule, and electric fields from FDTD simulations of our substrate. We then simulate the molecular orientation-dependent TERS spectral images as 

  



  ∑          

(2)

in which the rotation matrices ( ; " = #, $, %) and their transposes (  ; " = #, $, %) rotate (0 ) of MBN relative to the vector 360o in 4o increments) the molecular polarizability derivatives (

components of the FDTD local electric field vector components ( , ). The goal of these

simulations is to identify the effective molecular orientation on the tip that maps any experimentally encountered combination of local electric field components. To this end, we minimize the sum of absolute differences between the target experimental trace and all of our rotation angle-dependent TERS cross-sectional cuts. The best matches to the TERS profiles shown in Figures 4A and 4B are shown in panels C and D of the same figure, respectively. Interestingly, taking molecular orientation into account seems to recover most of the broadening of the electric field components that is observed in the experimental traces. In fact, deconvolving the matched simulated trace from its experimental analogue in Figure 4C yields a normal distribution, which when fitted to a Gaussian function yields a full-width-at-half maximum of only ~3.5 nm. That few nanometer precision in TERS is attainable using 100-200 nm tips is further supported by our below-described observations. To further support our inference of few nanometer precision in TERS imaging measurements performed using tips with nominally large radii, we imaged a corrugated gold substrate with tip 2, see Figure 5. In this case, the possibility of forming a plasmonic junction between the tip and substrate is increased, as local nanometric imperfections result in spatially varying and unique plasmonic properties on the corrugated film. The recorded spectral images are reminiscent of our prior report,26 and here again, the recorded TERS spectra exhibit pixel-to9 ACS Paragon Plus Environment

Nano Letters

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 20

pixel variations. This is indicative of sub-pixel resolution, which in this case translates into sub-5 nm precision. Several additional observations are noteworthy. First, the TERS images comprise flat (upper right areas in Figures 5A) and corrugated areas from the perspective of our MBNcoated tip. The spectra contained in the flat areas are comparable in intensity, see Figure 5B. Second, significantly different spectra are observed at pixels featuring high TERS activity, see Figure 5C. Again, these differences may be rationalized on the basis of equation 1; they are due to different molecular orientations relative to the local electric fields sustained at plasmonic tipsample nanojunctions. In this case, the vector components of the local electric fields and the molecular orientation are both unknown. We can nonetheless reproduce many of the spectral features by simulating the orientation-dependent TERS spectra as 



  ∑  (", &, ')  (", &, ')

(3)

where scanning the three indices of the normalized local electric field vectors (45 increments per rotation angle) yields different combinations of the fields relative to an assumed fixed molecular orientation. We first project the molecular orientation into the laboratory frame, such that it lies in the XY plane, with its long axis (to which the CS and CN moieties are aligned) along the X axis, and its aromatic framework orthogonal to the Z axis. We then simulate all the spectra and find the best matches to the two brightest TERS spectra in Figure 5C. The results are shown in Figure 5D. Besides minor differences between our simulated and experimental spectra, which are attributed to the limitations of the models and approximations in the of theories used, a general agreement is observed. This indicates that the model described herein captures the essential operative physics. This also further supports our picture of tensorial Raman scattering in ultrasensitive and/or ultrahigh spatial resolution TERS spectral imaging, which in the prior set of measurements that target step edges, broadcast the vector components of local electric fields. 10 ACS Paragon Plus Environment

Page 11 of 20

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

Nano Letters

In conclusion, this letter describes a non-standard application of TERS. It is driven by the need to image the vector components of plasmon-enhanced local electric fields, knowledge of which is prerequisite to rationalizing ultrasensitive and ultra-high resolution TERS nanoscopic imaging experiments. This is a logical extension to classical TERS experiments, which are chiefly geared towards nanoscale (bio)chemical imaging applications. To the best of our knowledge, these are the first measurements of the vector components of plasmon-enhanced local electric fields. We anticipate that ensuing correlated electric field mapping and chemical imaging measurements, both via TERS, will lead to a better understanding and broader applicability of this powerful technique. Besides our observations of ultrahigh spatial resolution using nominally coarse sputtered TERS tips, our results underscore the importance of considering molecular orientation and the tensorial nature of Raman scattering in interpreting ultrahigh spatial resolution and/or ultrasensitive TERS spectral imaging measurements. Our present work suggests that molecular orientation is another factor that needs to be considered to understand the spatial resolution in ultrasensitive and/or ultrahigh spatial resolution TERS imaging measurements.

Methods The measurements were performed using a previously described TERS imaging setup.27 Briefly, a 671 nm laser (100-200 µW) is incident onto the apex of the AFM probe (Nanosensors, ATEC) using a 100X air objective (Mitutoyo, NA=0.7) at a ~75o angle with respect to the

11 ACS Paragon Plus Environment

Nano Letters

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

surface normal. The polarization of the laser is controlled with a half waveplate and is set along the long axis of the AFM probe. The scattered radiation is collected through the same objective and further filtered through a long pass filter. The resulting light is detected by a CCD camera (Andor, iDus 416) coupled to a spectrometer (Andor, Shamrock 303i). The AFM probes were coated with varying amounts of silver by arc-discharge physical vapor deposition (target: Ted Pella Inc., 99.99% purity). The metallic AFM probes were then coated with 4-mercaptobenzonitrile (MBN, Aurum Pharmatech) by bringing the tip into contact with a substrate coated with a thin film of the molecules. Au(111) substrates were prepared by epitaxial growth of Au on freshly cleaved mica using physical vapor deposition (target: Ted Pella Inc., 99.99% purity). Subsequent AFM and TERS imaging measurements were performed under ambient conditions on hydrogen flame-annealed Au(111) substrates. All AFM measurements were performed in non-contact mode. FDTD simulations were performed using a commercially available software package (Lumerical Inc.) running on a local computer cluster. The computational models used replicate our experimental geometry by accounting for sample permittivity, laser wavelength, polarization, and angle of incidence. The calculations incorporate a 2 nm Au edge parsed in a three dimensional simulation volume. The calculations yield the spatially resolved relative intensities of the electric field components as a function of time. Standard Fourier transforms result in the corresponding spatial and frequency resolved relative field magnitudes. The field components shown in Figure 4 were taken at a constant height of 4 nm from the Au surface.

12 ACS Paragon Plus Environment

Page 12 of 20

Page 13 of 20

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

Nano Letters

Figure 1. A representative AFM image of the Au(111) substrate is shown in A. The scale bar in the same panel indicates 200 nm, and the color-coded dotted lines designate the 1D topographic cross-sections shown in B. Shown in C and D are SERS (tip-only, sample retracted by ~10 µm) and TERS (tip in contact with an Au(111) terrace) point spectra time-integrated for 1 s. The insets in C and D show helium ion micrographs of the two tips used to record the respective SERS/TERS measurements. The scale bars in insets of C and D indicate 100 nm.

13 ACS Paragon Plus Environment

Nano Letters

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. Correlated AFM (A) and TERS (B) images of a ~2 nm step edge on an Au(111) substrate. The scale bar in A indicates 50 nm, and the dashed rectangle designates the area imaged via TERS in B. The spectra plotted in C are contained within the rectangular area highlighted in B. In D, we zoom in on three TERS spectra in the 1560-1610 cm-1 spectral region. These spectra were recorded at Y = 10 nm, at the positions marked by ⊗ in B. In turn, the data plotted in E traces TERS intensity along Y = 10 nm (see B) at ~1580 cm-1. The Raman images and spectra shown in this panel were recorded with the tip shown in Figure 1C. TERS images are time-integrated for 1 s/pixel (10×10 nm2/pixel).

14 ACS Paragon Plus Environment

Page 14 of 20

Page 15 of 20

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

Nano Letters

Figure 3. Correlated AFM (A) and 1580 cm-1 (B) and 2224 cm-1 (C) TERS images of a ~2 nm step edge on an Au(111) substrate. The scale bar in A indicates 30 nm. Three TERS spectra in the 1000-2300 cm-1 spectral region are plotted in D. These spectra were recorded at the diagonal positions marked by ⊗ in B. The inset of panel D shows the same three spectra shown in D in the 1560-1610 cm-1 spectral region, to illustrate TERS signal variations at this vibrational resonance. The cross-section line profiles plotted in E and F trace TERS intensity along the diagonal at ~1580 cm-1 (dotted line in B) and 2224 cm-1 (dotted line in C). The Raman images and spectra shown in this panel were recorded using the tip shown in Figure 1D. The TERS images were time-integrated for 1 s/pixel (10×10 nm2/pixel).

15 ACS Paragon Plus Environment

Nano Letters

Normalized (arb. u.)

1.0

1.0

A

B 0.8

0.9 Exp EX

0.6

(EX)2

0.8

Exp EZ

(EX)3 (EX)4

(EZ)2

0.4

(EZ)3

0.7

(EZ)4 0

20

40

60

80

0

20

X (nm) 1.0

Normalized (arb. u.)

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 20

40

60

80

X (nm) 1.0

C

D

0.9 0.9

0.8

Exp Sim

0.7

Exp Sim

0.8

0.6 0.5

0.7

0

20

40

60

0

80

20

40

60

80

X (nm)

X (nm)

Figure 4. TERS cross-sections from Figures 2E and 3E are re-plotted in A and B, where they are compared to different vector components of plasmon-enhanced local electric fields predicted by FDTD simulations that closely emulate our experimental arrangement. Note that (i) we compare the experimental traces to () ) , where k=z,x and n=0-4, and (ii) the field components were taken at a constant height of 4 nm from the surface. In panels C and D, the same experimental TERS cross-sectional cuts are compared to their simulated analogues (thick red lines), obtained by solving equation (1) in the main text at different positions of the MBN-coated tip relative to the edge. The reader is referred to the main text and the supporting information for more details.

16 ACS Paragon Plus Environment

90

A

1580 cm-1

1320

200

80 70

Y (nm)

50

702.5

40

Counts / s

150

60

B

100

30 50

20 10

85.00 0 1000

0 0

10 20

30 40 50

60

70 80 90

X (nm)

C 1580 cm-1

2500 2000

1.0 0.8

Y = 65 X=5 X = 35 X = 45 X = 75

Normalized

3000

1250

2000

X,Y = (35,65) Exp Sim

0.6 0.4

1000

1200

1400

Raman Shift (cm-1)

1.0 0.8

Normalized

500

1750

D

0.0 800

1500 1000

1500

Raman Shift (cm-1)

0.2

2224 cm-1

Counts / s

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

Nano Letters

1580 cm-1

Page 17 of 20

0.6

1600

1800

2000

X,Y = (5,65) Exp Sim

0.4 0.2

0 1000

1250

1500

1750

2000

2250

0.0 800

1000

1200

1400

1600

1800

2000

Raman Shift (cm-1)

-1

Raman Shift (cm )

Figure 5. A TERS image (at 1580 cm-1) of a partially corrugated Au surface is shown in panel A. Panel B shows a collection of 63 spectra contained within a topographically flat area, which is highlighted using a white rectangle in A. In C, four TERS point spectra recorded at X = 5, 35, 45, and 75 at Y = 65 are shown; they are in turn marked by ⊗ in A. The TERS spectral images shown in this figure were time-integrated for 1 s and sampled at 5×5 nm2/pixel. Differences in the relative intensities of the observable vibrational states are evident in panel C. Molecular orientation-dependent spectra are simulated in D, and they are compared to two selected TERS point spectra from C. For more details on the simulations, the reader is referred to the main text and the supporting information.

17 ACS Paragon Plus Environment

Nano Letters

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 20

SUPPORTING INFORMATION Analysis of orientationally averaged vs oriented Raman scattering from MBN. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author *[email protected] The authors declare no competing financial interest.

ACKNOWLEDGMENTS AB is supported by the Department of Energy’s (DOE) Office of Biological and Environmental Research Bioimaging Technology project #69212. PZE acknowledges support from the Laboratory Directed Research and Development Program at Pacific Northwest National Laboratory (PNNL). AGJ and WPH are supported by the US DOE, Office of Science, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences & Biosciences. This work was performed in the environmental and molecular sciences laboratory (EMSL), a DOE Office of Science User Facility sponsored by BER and located at PNNL. PNNL is operated by Battelle Memorial Institute for the DOE under contract number DE-AC05-76RL1830. The authors are grateful to B. Arey for recording the HIM images of the AFM tips used in this study.

18 ACS Paragon Plus Environment

Page 19 of 20

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

Nano Letters

References

1

Moerner, W. E. Rev. Mod. Phys. 2015, 87, 1183-1212.

2

Fleischmann, M.; Hendra, P. J.; McQuillan, A. J. Chem. Phys. Lett. 1974, 26, 163-166.

3

Jeanmaire, D. L.; Van Duyne, R. P. J. Electroanal. Chem. 1977, 84, 1-20.

4

Moskovits, M. J. Chem. Phys. 1978, 69, 4159-4161.

5

Nie, S.; Emory, S. R. Science 1997, 275, 1102-1106.

6

Kneipp, K.; Wang, Y.; Kneipp, H.; Perelman, L. T.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Phys.

Rev. Lett. 1997, 78, 1667-1670. 7

Anderson, M. S. Appl. Phys. Lett. 2000, 76, 3130-3132.

8

Stockle, R. M.; Suh, Y. D.; Deckert, V.; Zenobi, R. Chem. Phys. Lett. 2000, 318, 131-136.

9

Pettinger, B.; Schambach, P.; Villagomez, C. J.; Scott N. Annu. Rev. Phys. Chem. 2012, 63,

379-399. 10

Stadler, J.; Schmid, T.; Zenobi, R. Nanoscale 2012, 4, 1856-1870.

11

Zrimsek, A. N.; Chiang, N.; Mattei, M.; Zaleski, S.; McAnally, M. O.; Chapman, C. T.; Henry,

A. –I.; Schatz, G. C.; Van Duyne, R. P. Chem. Rev. 2017, 117, 7583–7613. 12

Verma, P. Chem. Rev. 2017, 117, 6447–6466.

13

Wang, X.; Huang, S. –C; Huang, T. –X.; Su, H. –S.; Zhong, J. –H.; Zemd, Z. –C.; Li, M. –H.;

Ren, B. Chem. Soc. Rev. 2017, DOI: 10.1039/C7CS00206H. 14

Richard-Lacroix, M.; Zhang, Y.; Dong, Z.; Deckert, V. Chem. Soc. Rev. 2017, DOI:

10.1039/C7CS00203C. 15

El-Khoury, P. Z.; Hu, D.; Apkarian, V. A.; Hess, W. P. Nano Lett. 2013, 13, 1858-1861.

16

El-Khoury, P. Z.; Hess, W. P. Nano Lett. 2014, 14, 4114-4118.

19 ACS Paragon Plus Environment

Nano Letters

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

17

Zhao, L.; Jensen, L.; Schatz, G. C. J. Am. Chem. Soc. 2006, 128, 2911–2919.

18

Fromm, D. P.; Sundaramurthy, A.; Kinkhabwala, A.; Schuck, P. J.; Kino, G. S.; Moerner, W.

E. J Chem. Phys. 2006, 124, 061101-1-4. 19

Jäckel, F.; Kinkhabwala, A. A.; Moerner, W. E. Chem. Phys. Lett. 2007, 446, 339-343.

20

Hao, E.; Schatz, G. C. J. Chem. Phys. 2004, 120, 357-366.

21

Fischer, S. A.; Aprà, E.; Govind, N.; Hess, W. P.; El-Khoury, P. Z. J. Phys. Chem. A 2017,

121, 1344-1350. 22

Peppernick, S. J.; Joly, A. G.; Beck, K. M.; Hess, W. P. J. Chem. Phys. 2011, 134, 03457-

034514. 23

Halas, N. J.; Lal, S.; Chang, W. -S.; Link, S.; Nordlander, P. Chem. Rev. 2011, 111, 3913-3961.

24

El-Khoury, P. Z.; Hu, D.; Hess, W. P. J. Phys. Chem. Lett. 2013, 4, 3435-3439.

25

El-Khoury, P. Z.; Abellan, P.; Chantry, R. L.; Gong, Y.; Joly, A. G.; Novikova, I. V.; Evans, J.

E.; Aprà, E.; Hu, D.; Ramasse, Q. M.; Hess, W. P. Adv. Phys.: X 2016, 1, 35-54. 26

Bhattarai, A.; El-Khoury, P. Z. Chem. Commun. 2017, 53, 7310-7313.

27

El-Khoury, P. Z.; Hess, W. P. Chem. Phys. Lett. 2013, 581, 57-63.

28

El-Khoury, P. Z.; Ueltschi, T. W.; Mifflin, A. L.; Hu, D.; Hess, W. P. J. Phys. Chem. C 2014,

118, 27525-27530. 29

El-Khoury, P. Z.; Johnson, G. E.; Novikova, I. V.; Gong, Y.; Joly, A. G.; Evans, J. E.; Zamkov,

M.; Laskin, J.; Hess, W. P. Faraday Discuss. 2015, 184, 339-357. 30

Novotny, L.; Hecht, B. Principles of Nano-Optics, 1st ed.; Cambridge University Press: New

York, U.S.A.; 2011. 31

Cancado, L. G.; Hartschuh, A.; Novotny, L. J. Raman Spectrosc. 2009, 40, 1420–1426.

20 ACS Paragon Plus Environment

Page 20 of 20