Efficient Heterostructures for Combined Interference and Plasmon

Subscriber access provided by Fudan University

Article

Efficient heterostructures for combined interference and plasmon resonance Raman amplification Leo Alvarez-Fraga, Esteban Climent-Pascual, Montserrat Xochitl Aguilar-Pujol, Rafael Ramírez-Jiménez, Félix Jiménez-Villacorta, Carlos Prieto, and Alicia de Andres ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b12490 • Publication Date (Web): 05 Jan 2017 Downloaded from http://pubs.acs.org on January 9, 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.

ACS Applied Materials & Interfaces 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 26

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

ACS Applied Materials & Interfaces

Efficient heterostructures for combined interference and plasmon resonance Raman amplification

Leo Alvarez-Fraga1, Esteban Climent-Pascual1, Montserrat Aguilar-Pujol1, Rafael Ramírez-Jiménez2, Félix Jiménez-Villacorta1, Carlos Prieto1 and Alicia de Andrés1* 1

Instituto de Ciencia de Materiales de Madrid, Consejo Superior de Investigaciones

Científicas. Cantoblanco 28049 Madrid, Spain 2

Departamento de Física, Escuela Politécnica Superior, Universidad Carlos III de

Madrid, Avenida Universidad 30, Leganés, 28911 Madrid, Spain

Keywords: Raman enhancement, SERS, IERS, interference, sensing, imaging, graphene, nanoparticles.

ABSTRACT The detection, identification and quantification of different types of molecules and the optical imaging of, for example, cellular processes are important challenges. Here we present how interference enhanced Raman scattering (IERS) in adequately designed heterostructures can provide amplification factors relevant both for detection and imaging. Calculations demonstrate that the key factor is maximizing the absolute value of the refractive indices difference between dielectric and metal layers. Accordingly Si/Al/Al2O3/graphene heterostructures have been fabricated optimizing thickness and roughness and reaching enhancement values up to 700 for 488 nm excitation. The 1 ACS Paragon Plus Environment


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

deviation from the calculated enhancement, 1200, is mainly due to reflectivity losses and roughness of the Al layer. The IERS platforms are also demonstrated to improve significantly the quality of white light images of graphene and are foreseen to be adequate to reveal the morphology of 2D and biological materials. A graphene upper layer is adequate for most organic molecule deposition and often quenches possible fluorescence permitting Raman signal detection which, for a rhodamine 6G (R6G) monolayer presents 400 gain. Without graphene the non-quenched R6G fluorescence is similarly amplified. The wavelength dependence of the involved refractive indices predicts much higher amplification (around 104) for NIR excitation. Theses interference platforms can therefore be used to gain contrast and intensity in white light, Raman and fluorescence imaging. We also demonstrate that SERS and IERS amplifications can be efficiently combined, leading to gain > 105 (at 488 nm), by depositing an Ag nanostructured transparent film on the IERS platform. By optimizing the plasmonic structures deposited on the IERS platforms, single molecule detection can be actively envisaged.

2 ACS Paragon Plus Environment

Page 2 of 26

Page 3 of 26

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


INTRODUCTION Among the different mechanisms for Raman intensity amplification, the interference process has been scarcely investigated. Indeed, the design of amplification platforms for the detection and imaging of extremely diluted and/or complex materials still requires further research and development to get cheap, reliable, reproducible and stable over time systems that can be easily reused several times. The enhancement provided by electric field amplification due to localized plasmons of metallic nanoparticles, called SERS (surface enhanced Raman scattering),1,2, 3,4 is the most efficient process and to reach the single molecule detection complex structures and strategies have been proposed. 5,6,7 However, these enhancement substrates need to increase their robustness and stability against laser irradiation, corrosive media and degradation with time. Moreover the SERS platforms, most of the times, while extremely sensitive do not allow the reliable quantification of the concentration of the sensed molecule and their morphology impedes their use for white light imaging. The interference of light associated to multi-reflection processes at materials interfaces is well known and diverse heterostructures are commonly used for different applications, as selective filters or coatings. The interference enhanced Raman scattering (IERS) effect has been used since 1980 to amplify the Raman signal of ultrathin films as Ti and Ti2O3 8 or tellurium with a gain around 20. 9 A maximum amplification of 1000 for IERS was foreseen9 but in fact has never been demonstrated experimentally or with rigorous calculations. The application of this interference effect was limited to the detection of ultrathin films but, since 2008 the possibility to amplify the signal of graphene has been extensively exploited. The Raman characteristics of graphene have been studied using Si/SiO2/graphene heterostructures 10 ,

11 , 12

which

provide amplification factors around 30-40, as well as that of graphene nano-ribbons.13 3 ACS Paragon Plus Environment


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

Page 4 of 26

The same stacking was used to evaluate the graphene enhanced Raman scattering (GERS) process 14 and to investigate SERS effect.15 Gain up to 80 has been reported for F16CuPc thin films corresponding to an optimized SiO2 thickness of 90 nm.16 For (NiTi)/TiO2/graphene heterostructures gain values around 50 are found. 17 The use of silver as reflecting layer has been proposed, Si/Ag/Al2O3/graphene,

18

to promote

combined SERS and IERS amplification of graphene signal, reporting measured gain up to 20. On the other hand, graphene is an excellent bio-compatible platform19,20,21 which controls the metal/bio-material interactions, quenches molecular fluorescence and protects, stabilizes 22 and even de-oxidize metallic nanoparticles as Ag23 therefore its implementation in amplification platforms is interesting. This work reports on how interference enhancement substrates have to be designed to maximize their efficiency and how it is possible to combine SERS and IERS effects. The IERS platforms are also demonstrated to improve significantly the quality of white light images of graphene and are foreseen to be adequate to reveal the morphology of ultrathin films and of biological material. We use a transfer matrix method to calculate the propagation of light through the heterostructure (reflecting layer/ dielectric layer/ graphene) for a large set of materials in order to obtain the general trends of Raman interference process and to optimize the effect in view of its application. Graphene is used here as the ideal material to reveal the amplification power of the tested platform and as the appropriate substrate for the deposition of organic molecules. We have designed and fabricated optimized heterostructures for IERS which combined with nanostructured silver films demonstrate the combined IERS + SERS amplification.


Page 5 of 26

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


EXPERIMENTAL DETAILS The heterostructures were prepared at room temperature by sequential magnetron sputtering deposition on silicon substrates. Silicon (Si (001), 1×1 cm2) substrates with 350 ± 25 µm thickness were cleaned with acetone and dried with highly pure N2 gas. A strip of adhesive was placed along the Si substrate surface which is removed after the layer deposition (this step is repeated before/after deposition of each layer). The base pressure was around 1x10-6 mbar. The aluminum layer was deposited by direct current (DC) sputtering of Al target (99.0%) with an Ar pressure of 5x10-3 mbar and a target power of 0.04 A (deposition rate 3.4 nm/min). The aluminum oxide film was prepared by radio frequency (RF) magnetron sputtering (50 W) from an Al2O3 target with Ar gas at 7x10-3 mbar (deposition rate 2 nm/min). At the end of the process four areas can be observed in the sample: Si substrate, aluminum metal on Si substrate, aluminum oxide on Si substrate, and the complete stack: Al2O3/Al/Si. Samples with different thickness of Al and Al2O3 layers were prepared and their thickness was determined by a Veeco Dektak 150 profilometer. PMMA/Graphene on copper foil from Graphenea was transferred on the heterostructures. Cu was eliminated in a 2.1M FeCl3 and 1.3M HCl solution for 15min. The PMMA/Graphene stack was rinsed in a deionized water bath twice, immerged in a 10% HCl solution and then again into deionized water three times with the last bath for 20h to complete the removal of Cu. Subsequently, the Gr/PMMA film was fished onto the four-zone platforms and baked on a hot plate at 90ºC. PMMA was then removed by immersing in warm acetone at 50ºC followed by further vacuum thermal treatment at 250ºC. Rhodamine 6G (R6G) films were deposited by spin coating from a methanol solution on the samples and on a glass substrate for reference. The absorption spectrum of the sample on glass is used to estimate the molecular density of the film by 5 ACS Paragon Plus Environment


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

comparing with the spectrum of a 10-5 M solution of R6G. The molecular density is 2x1013 molecules/cm2. To check the sensitivity of the heterostructures, these have been immersed in different concentrations of R6G (10-8 – 10-6 M) solutions in methanol for two hours (dip-coating) in vertical position. Micro-Raman experiments were performed at room temperature with the 488nm line of an Ar+ laser, incident power in the 0.1–8 mW range, an Olympus microscope (x100 and x20 objectives), and a “super-notch-plus” filter from Kaiser. The scattered light was analyzed with a Jobin-Yvon HR-460 monochromator coupled to a Peltier cooled Synapse CCD. The estimated Raman spatial resolution is around 0.7 µm at 488 nm for the high NA (0.95) x100 objective. The morphology and roughness of the thin films (layers) were scrutinized using atomic force microscopy (equipment and software from NanotecTM).24 Topographic characterization was carried out in the tapping mode, using commercial Si tips (Nanosensors PPP-NCH-w) with a cantilever resonance frequency f0 ≈ 270 kHz and k ~ 30 Nm−1. Several regions were probed to confirm homogeneity of the layers to the micrometer scale. An estimation of the roughness of all the layers and heterostructures studied was carried out comparing the rms roughness and the full width at half-maximum (FWHM) of the height distribution analyzing the topographic images obtained.25 RESULTS AND DISCUSSION The simplest platform for Raman interference amplification requires a reflecting layer and a dielectric layer with a specific thickness. The analyte to be probed and intensified, in this case graphene and/or R6G, is deposited on top. Therefore, the first step to fabricate optimized heterostructures is to decide which are the most adequate materials and thickness for these layers.


Page 6 of 26

Page 7 of 26

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


Design of an IERS platform: predictions from calculations The propagation of light through the heterostructure (reflecting layer/ dielectric layer/ graphene) is calculated using a transfer matrix method. This methodology consists in calculating separately the absorbed and scattered lights at each point of the medium by using a matrix which relates neighboring layers by the product of two matrices. One contains the dependence on optical path length and determines the phase difference as light travels through the layer and the second one contains the reflection and transmission complex coefficients between adjacent layers, both matrices depend on the complex refractive indices of each layer. The whole stack is represented by a matrix which is the product of the individual matrices corresponding to each medium. 26 As a result, the effect on the Raman signal of successive interference processes depends on the thickness of each layer and on the wavelength of light through the wavelength dependence of the dielectric constants involved. For a metal/dielectric/graphene heterostructure we have analyzed the dependence of graphene G peak intensity amplification as a function of the different involved parameters. The details of the method are presented in the Supporting Information (S1 and Figure S1). The amplification depends on the thickness of the dielectric layer and the maximum amplification values are shown in Figure 1. This maximum amplification has been calculated for different metals (Cu, Ni, Al) and for Si and for dielectric layers with refractive index between 1 and 3 at several laser wavelengths commonly used in Raman spectroscopy (457, 488, 514 and 633 nm). A general trend is observed for the amplification factor for all calculated combinations of materials as shown in Figure 1. A systematic increase of the amplification factor with the modulus of the complex refractive indices difference ∆n = n(dielectric) – n(reflector) is observed.



400

Amplification Enhancement factor factor

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 26

4000

(a)

n=1 n=1.5

Al - 488 nm Al - 514 nm Al - 633 nm

3000

10000

Si - 457 nm Si - 488 Si - 515 Si - 633 Cu - 488 Ni - 488

300

200

n=3

2000

100

1000

0

0

0

12000

(b)

1

2

3

4

5

6

7

0

1

2

3

Refractive index difference

Figure 1: a) Amplification factor of G intensity for different graphene/dielectric/ reflecting layer (Al (triangles), Si (circles), Cu (squares) and Ni (stars)) systems at 633nm (red symbols), 514nm (green), 488nm (blue) and 457 nm (purple) laser excitation as a function of n(dielectric) – n(reflector). For each reflecting material several values of n(dielectric) ranging from 1 to 3 are calculated. b) Zoom of the selected region in a)

Similar amplification values are obtained for Cu, Ni and Si as reflecting layers, all of them with ∆n below around 3.5 (Figure 1b). The most important variation is due to the laser wavelength. Clearly aluminum is the best candidate as reflecting layer, combined with a dielectric with n(dielectric) =1 (Figure 1a). In this case the calculated amplification is close to 4000 for 633 nm excitation. But n=1 is the refractive index at visible wavelengths of air or any gas. For solids, the lowest refractive index values lies around 1.5. Fabrication of the heterostructures Following the results obtained in the calculations we have chosen aluminum as the best reflecting material, Al2O3 as dielectric layer (n = 1.775 at 488 nm), and silicon (001) as substrate because of its high flatness. To evaluate and compare exactly under the same 8 ACS Paragon Plus Environment

Page 9 of 26

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


conditions the Raman signal of graphene and of the reference molecule (rhodamine R6G) for the different stages of the heterostructure fabrication, the process described in Figure 2 (a - h) has been followed.

Figure 2: Fabrication process of the four-zone samples a) Si substrate b) adhesive strip, c) Al layer is deposited, d) strip is removed, e) perpendicular strip, f) Al2O3 layer is deposited, g) second strip is removed, h) final four-zones sample where 1: Si, 2: Si/Al, 3: Si/Al2O3 and 4: Si/Al/Al2O3. Finally, i) graphene/PMMA is transferred onto the sample, j) PMMA is eliminated and k) a thin film of R6G is spin-coated on the heterostructure.

In that way, four different situations are found at the same sample. Zone 1, with only Si, is used as the reference for Raman intensity of graphene and R6G. Zone 2 consists of a 90 nm aluminum layer on Si. Zone 3 is made of Al2O3 layers (70, 60 and 50 nm thicknesses were tested) on Si. Zone 4 represents the whole stacking: Si/Al/Al2O3. Within this protocol, three four-zone samples, with different Al2O3 layer thickness were fabricated which are labeled: 4Z-50, 4Z-60 and 4Z-70 for Al2O3 layer thickness of 50, 60 and 70 nm respectively. Finally CVD graphene is transferred on the 4-zone samples. Graphene is used here as the ideal material to reveal the amplification power of the tested platform. The quality of the transferred graphene layer is checked by Raman 9 ACS Paragon Plus Environment


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

spectroscopy as shown in Figure S3 (Supp. Inf.). Graphene is also an appropriate substrate for the deposition of most organic molecules and serves to ensure that the R6G concentration is identical in the four zones. Since the nature, hydrophobic character and morphology of the four zones are different, the final R6G concentration on each zone, either by depositing a droplet or by spinning a R6G solution, may differ. To ensure the homogeneity of the R6G molecular density a diluted suspension of R6G was spincoated on the 4-zone samples and also on a quartz/graphene sample from Graphenea. This sample is used to evaluate the molecular density using the measured optical absorption of the deposited R6G film by comparison with that of a 10-5 M solution of R6G in methanol.

Raman Amplification Performance The amplifications of 2D graphene intensity corresponding to different zones of the three samples (4Z-50, 4Z-60 and 4Z-70) have been obtained using the signals of graphene in zone 1 (Si/Gr) of each sample as reference. The obtained amplifications, with an excitation wavelength of 488 nm, corresponding to zones 4 (Si/Al/Al2O3/Gr) reach 550, 650 and 350 for the 50, 60 and 70nm Al2O3 layers, respectively, therefore the optimum Al2O3 thickness is 60 nm as predicted by our calculations for 488 nm excitation (Figure S1). For zones 3 (Si/ Al2O3/Gr), where the reflecting layer is silicon, the amplification factors are high (50, 73, 70) which are very much in agreement with the calculations (Figure S2) and for zones 2 (Si/Al/Gr) a significant amplification is detected (in the range 10 - 30). In the latter case no interference amplification is expected for a pristine Al layer on silicon. Considering that a passivation oxide layer, typically 3nm thick, is always formed on Aluminum, we have calculated the amplification factor using a 3 nm thick oxide layer on Al, the resulting amplification


Page 10 of 26

Page 11 of 26

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


factor is around 2 which is clearly below the experimental value. Therefore an additional weak enhancement probably associated to aluminum plasmon resonance, allowed by the morphology of the Al layer, accounts for the whole amplification. While spherical Al nanoparticles exhibit sharp plasmon resonances in the near and deep-UV, non-spherical particles present wide plasmons which span into the visible 27 and SERS effect is reported for excitation wavelengths even in the NIR (785 nm).

28

This

enhancement cannot contribute when the dielectric Al2O3 layer is present since the SERS effect is only effective at extremely short distances from the metallic nanoparticles. The next step is to evaluate the amplification of an analyte to test the platforms. We have used a rhodamine 6G solution in methanol which is spin coated onto the samples with and without graphene. The spin coating procedure ensures the homogeneity and reproducibility of the rhodamine films. The actual molecular density of the deposited films is 2x1013 molecules/cm2. The homogeneity of the molecular distribution is an important issue and cannot be obtained for example with the droplet procedure. In Figure 3a an optical image of the four-corner region of sample 4Z-70 is shown. The different reflectance of the four zones clearly reveals the different layer stacking.



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 3: a) Optical image (20 µm x 20µm) of the four zones of the sample with 70 nm Al2O3 layer (4Z-70) with transferred graphene and spin coated with R6G. Raman images of the same region of b) the 2D graphene peak and c) the rhodamine 6G 1645 cm-1 peak. Representative spectra of each of the four zones are plotted in d) with graphene and e) without graphene. In e) the R6G fluorescence also shows the interference amplification.

Images of the Raman intensities of the 2D graphene peak (Figure 3b) and of 1650 cm-1 peak of R6G (Figure 3c) clearly demonstrate the amplification power of the zones. Individual representative spectra of each zone are plotted for the sample with graphene and R6G (Figure 3d) and for the sample with only R6G (Figure 3e). R6G is strongly absorbing at the laser wavelength (488 nm) and shows a strong known fluorescence (∼ 552-556 nm) in the region of the Raman vibrations. However, graphene quenches almost totally this emission so that the Raman peaks can be easily detected (Figure 3d, for more details, see Figure S4). The amplification obtained for R6G Raman peaks is 12 ACS Paragon Plus Environment

Page 12 of 26

Page 13 of 26

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


about 20% lower than that of graphene. Because of the electronic structure of graphene such quenching is efficient for a large set of analytes29, 30 which may be important in many applications. When the R6G layer is deposited directly on the heterostructure, without graphene, the fluorescence completely hides the Raman peaks but a similar amplification as that observed for R6G vibrations is observed for the different zones (Figure 3e). Therefore, these interference systems are adequate for both Raman and fluorescence signals.

Figure 4. Left: Optical images of a transferred graphene monolayer on the 4 regions showing the obtained higher contrast image on the whole stack (zone 4). Graphene bilayers (darker regions of 2-3 µm diameter) and wrinkles are clearly revealed while on Si/Al2O3 (zone 3) only the bilayers are detected. The smallest black dots are dust particles. Right: simulations of the reflectance difference ∆R = R– R0, where R is the reflectance of single layer (SLG) and bilayer graphene (BLG) and R0 that without graphene, for Al2O3/Al and Al2O3/Si, versus the dielectric thickness.

These simple and flat platforms (Si/Al/ Al2O3) are extremely stable with time and robust against external agents and have a very straightforward application in enhancing the quality of the optical images obtained with a microscope. In Figure 4 the optical images of a graphene monolayer obtained in the same conditions for the 4 zones of sample 4Z60 evidence the possibility to visualize graphene defects. Graphene bilayers are seen in 13 ACS Paragon Plus Environment


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

zones 3 and 4 as darker regions of diameter around 1-2 µm, but the wrinkles of the transferred graphene are clearly seen only in zone 4 (the whole stack). Si/SiO2 substrates have been studied and used to enhance the visibility of graphene monolayers 31 and of thin graphite films. 32 The contrast is related to the different reflectance resulting from two regions; in this case we evaluate the contrast between single layer graphene and bilayer graphene regions on a substrate. A simplified case of the same transfer matrices defined before for Raman amplification are used to calculate the reflectance at 514 nm (the details are included in the Supporting Information S5). The calculations explain the observed increased contrast when the reflecting medium is Al compared to Si and is due to the high imaginary refractive index of aluminum which leads to the higher refractive index difference as in Figure 1.

Factors limiting interference amplification The obtained amplification factors, up to 650, are already very interesting values for applications however these experimental values are lower than the calculated ones. One of the crucial factors determining the quality of an interference process is the flatness of the layers involved; therefore, the morphology and roughness of Al and Al2O3 layers for different layer thickness and sputtering conditions were evaluated by Atomic Force Microscopy (AFM). The thickness of Al layer was limited to 90 nm to reduce roughness while optical transmission is maintained negligible. The AFM topography images and resulting rms roughness and full width (FWHM) of the height distributions of the four zones in Si/Al (90nm)/ Al2O3 (60 nm) sample are collected in Figure 5.


Page 14 of 26

Page 15 of 26

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


Figure 5 Optical image (center) and AFM topographic images and fitting results indicating the roughness of the uppermost layer surface (FWHM and rms) of the four zones of 4Z-60 heterostructure: 1- Si substrate, 2- Si/Al (t ~ 90 nm), 3- Si/Al2O3 (t ~ 60 nm) and 4- Si/Al/Al2O3. The drawings schematize the sections of the four zones.

Aluminum layer roughness (9.1 nm FWHM and 3.9 nm rms) is clearly larger than that of Al2O3 layer (2.4 nm FWHM, 1.1 nm rms) and dominates the overall roughness of the zone 4 stacking (10.4 nm FWHM, 4.4 nm rms). The roughness is directly related to the grain size which is significantly larger for Si/Al than for Si/Al2O3 (smaller than the tip size in this case). We can use the FWHM of the height distribution to estimate the impact of the morphology on the interference amplification of the Si/Al2O3/Gr and Si/Al/Al2O3/Gr regions, this is shown in Figure 6a, where we can observe how the reduction of the amplification factor is relevant at high FWHM values. For our actual samples it is < 10% for Si/Al/Al2O3/Gr stack and smaller for Si/Al2O3/Gr.



a)

b)

1200

70 nm 800

100

80

80

Si/Al2O3/G

70 nm

1200

Amplification factor

1000

Real part of n(Al)

Imaginary part of n(Al)

Al/Al2O3/G

Amplification factor

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 26

1200

+50% +20%

-1%

1000 800 600

-5%

1000

-10% -11%

800

-20% -50%

50

400

60

70

Gr/AlO O /Al Al/Al 2 2 33/Gr λ=488nm

200 80 nm 150

50

0 0 10 20 20 40 Half width of the distribution Height distribution FWHM (nm) Roughness HWHM (nm) (nm)

0

20

40

60

80

100

120

140

Al2O3 thickness (nm)

Figure 6 a) Reduction of Raman amplification as a function of the width (FWHM) of the heterostructure surface height distribution. The vertical dashed line corresponds to 4Z-60, b) modification of the amplification by the imaginary part (and real part in the inset) of the refractive index of aluminum. The grey dots correspond to the experimental amplifications for 4Z-50, 4Z-60 and 4Z-70 heterostructures.

The main reason for the expected large amplification factor of our layer arrangement is the refractive index value of Al which is related to the density of the layer, thus, any change in it has an impact on the amplification. In fact a small reduction of the imaginary part of the refractive index of aluminum, κ(Al), strongly hampers amplification (Figure 6b) while a change in the real part is almost irrelevant (inset of figure 6b). A reduction of 11% for κ(Al) explains the observed graphene amplification values for the different dielectric layer thickness (circles in figure 6b) and this reduction is probably related to a decrease of the density of Al layers compared to bulk. Increasing Al layer density, by reducing particle size using, for example, e-beam technique for Al layer deposition, would bring refractive index closer to bulk value. The absence of the Al layer for Si/Al2O3/G, Zone 3, results in a much closer agreement between measured


Page 17 of 26

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


and calculated amplification factors, confirming the validity of our calculations (Figure S2). Combined IERS + SERS amplifications An ultrathin (nominally 1 nm) Ag film was deposited on the 4Z-70 sample by dcsputtering (see schema inserted in Figure 7). These discontinuous films are formed by Ag islands with a bimodal size distribution, around lateral size ~40 nm and height ~10 nm and around 20nm width and 5 nm height (AFM analysis shown in Figure S5, Supporting Information) and yield a SERS amplification factor close to 3000 for R6G comparing the signal from zone 1 (Si/Ag) to that of the molecule on bare silicon.

Figure 7. Raman spectra in the 1600 cm-1 spectral range of R6G spin coated on the four zones of the sample with 70 nm Al2O3 layer (4Z-70). The inset shows a schema of the interference platform with the nanostructured Ag film. Spectra 1 to 4 correspond to the different zones: Si+Ag+R6G, Si/Al+Ag+R6G, Si/Al2O3+Ag+R6G and Si/Al/Al2O3+Ag+R6G, respectively.

In Figure 7 the spectrum 1 (zone 1) corresponds to SERS only amplification and spectrum 4 to SERS + IERS combined amplifications (zone 4). Therefore the combined Ag SERS film and the interference platform provides an amplification factor >105. 17 ACS Paragon Plus Environment


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

Raman amplification by IERS is due to electric field amplification caused by wave interference and one of the main causes behind SERS effect is also electric field amplification by plasmon excitations. Since electric field amplitudes are enhanced in both mechanisms in an independent way, the overall amplification factor should be the product of the two contributions. Two factors can very effectively improve these results: the wavelength used to probe the Raman signal and the optimization of the SERS component through size and shape variation. In the first place, IERS amplification increases with increasing wavelength and gains as high as 104 for 1000 nm are predicted by our calculations; this increase is due to the strong wavelength dependence of the refractive index of metals. In the second place, SERS can be optimized in different ways by changing the sizes and shapes of the metal particles since they impact not only on the electric field amplification but also the on the analyte adhesion to the particles. To test the sensitivity of the platform we have used the dip-coating process for different concentrations of R6G (Figure 8a). The platform was immerged vertically in 10-8 M, 10-7 M and 10-6 M solutions of R6G in methanol for 2 hours. The detection limit for this platform is found to be 10-8 M using 488 nm excitation and laser power of 0.3 mW. In Figure 8b the spectra for 10-6 R6G on different regions of the 4Z-50 sample are shown. The blue trace corresponds to the combined IERS + SERS signal, pink to the IERS amplification (observe that R6G fluorescence is not quenched since there is no Ag) and olive trace corresponds to the SERS enhancement due to the Ag film on Si. This figure shows the efficient combination of IERS and SERS enhancement processes in these platforms for the amplification of Raman signal.


Page 18 of 26

Page 19 of 26

a) Raman Intensity (cps)

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


b)

Si/Al/Al2O3/Ag

-6

10 M R6G

300

1000

Si/Al/Al2O3/Ag

R6G -6

10 M

200

Si/Al/Al2O3

500 -7

10 M -8

0

100

x10

Si/Ag

10 M 1200

1400

0

600

1600

Raman shift (cm-1)

700

800

Raman shift (cm-1)

Figure 8. Raman spectra of a) different concentrations of R6G dip-coated on the 4Z-50 platform with 1 nm Ag film (the spectrum for 10-8M is multiplied by 10) and b) 10-6 M R6G on different regions of the 4Z-50 sample: Zone 4 +Ag (blue) , zone 4 without Ag (pink) and zone 1 +Ag (olive).

CONCLUSIONS We have designed the optimum reflecting/dielectric/graphene heterostructures for Raman interference amplification (IERS). The calculations indicate that the effect is optimized by maximizing the imaginary part of the reflecting layer refractive index and minimizing the real part for the dielectric layer of the heterostructure. The optimum materials are found to be Aluminum as metal and a dielectric with n=1. The simplest options for the dielectric layer are low-n oxide films as Al2O3 or SiO2. The predicted maximum amplification for Si/Al/Al2O3/graphene systems (60 nm Al2O3) is close to 1200 for 488 nm excitation and increases up to 104 for 1000 nm. The fabricated heterostructures with Al thickness of 90 nm and 60 nm Al2O3 show an experimental enhancement of the Raman signal of graphene of 700 and close to 400 for a rhodamine 6G film with 1013 molecules/cm2 density. The theoretical amplification values are not reached mainly due to two factors related to the Aluminum film characteristics. A 19 ACS Paragon Plus Environment


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

reduction as small as 10% of the imaginary part of the Al refractive index depletes the amplification by 40%. The other factor is the roughness of all layers that compose the heterostructure which is dominated by that of the Al film. Nevertheless this effect has relatively low impact for moderate roughness. As a confirmation of the impact of Al film quality on the device performance, we observed that the amplification without Al layer (Si/Al2O3/Graphene) is in very close agreement with the calculations. A weak SERS effect (10-30 gain) in Si/Al/graphene zones is originated by the nanostructured morphology of the aluminum layer. The designed interference platforms also provide a significantly higher contrast for optical images allowing for example, a clear visualization of the defects of a transferred monolayer graphene (as bilayers and wrinkles). They may serve as an easy and rapid tool for inspection of graphene or any 2D or cuasi-2D materials. Similar amplification of R6G photoluminescence and Raman intensities is also demonstrated when graphene is absent. Combined SERS and IERS platforms have been fabricated by depositing a Ag nanostructured ultra-thin film on the heterostructures. The combined amplification has been demonstrated with gain > 105. The performance of the plasmonic layer of this proof of concept (gain in the 103 range) can be significantly improved so that the combined IERS + SERS amplification is foreseen to reach much higher values. Author Information: *E-mail: [email protected]. Acknowledgments L. A-F. acknowledges a FPI grant (BES-2013-062759) from Spanish Ministerio de Economía y Competitividad (MINECO). The research leading to these results has 20 ACS Paragon Plus Environment

Page 20 of 26

Page 21 of 26

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


received funding from MINECO (MAT2012-37276-C03-01 and MAT2015-65356-C31-R) and Comunidad de Madrid (S2013/MIT-2740, PHAMA2.0-CM).

Supporting Information: Simulation details for Raman amplification interference process for the general case and for the studied trilayer heterostructures. Figure S1 presents the calculated amplification for Al/ Al2O3 /Graphene as a function of the Al2O3 thickness and Figure S2 for Si/Al2O3/Graphene with the measured amplification factors. The quality of the transferred graphene is shown Figure S3. Figure S4 shows the Raman spectra of 4Z-60/Gr/R6G for zones 3 and 4 with graphene and R6G peaks. In S5, an AFM topography image of an Ag film shows the size of the grains.

REFERENCES 1

Stiles, P. L.; Dieringer, J. A.; Shah, N. C.; Van Duyne, R. P. Surface-Enhanced Raman Spectroscopy. Annu. Rev. Anal. Chem. (Palo Alto. Calif). 2008, 1, 601–626.

2

Sharma, B.; Frontiera, R. R.; Henry, A.-I.; Ringe, E.; Van Duyne, R. P. SERS: Materials, Applications, and the Future. Mater. Today 2012, 15, 16–25.

3

Xie,

W.;

Schlücker,

S.

Rationally

Designed

Multifunctional

Plasmonic

Nanostructures for Surface-Enhanced Raman Spectroscopy: A Review. Rep. Prog. Phys. 2014, 77, 116502. 4

Jimenez-Villacorta, F.; Climent-Pascual, E.; Ramirez-Jimenez, R.; Sanchez-Marcos, J.; Prieto, C.; de Andrés, A. Graphene–ultrasmall Silver Nanoparticle Interactions and Their Effect on Electronic Transport and Raman Enhancement. Carbon, 2016, 101, 305–314



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

5

Page 22 of 26

Qian, X.-M.; Nie, S. M. Single-Molecule and Single-Nanoparticle SERS: From Fundamental Mechanisms to Biomedical Applications. Chem. Soc. Rev. 2008, 37, 912–920.

6

Wang, P.; Liang, O.; Zhang, W.; Schroeder, T.; Xie, Y.-H. Ultra-Sensitive Graphene-Plasmonic Hybrid Platform for Label-Free Detection. Adv. Mater. 2013, 25, 4918–4924.

7

Xie,

W.;

Schlücker,

S.

Rationally

Designed

Multifunctional

Plasmonic

Nanostructures for Surface-Enhanced Raman Spectroscopy: A Review. Rep. Prog. Phys. 2014, 77, 116502. 8

Nemanich, R. J.; Tsai, C. C.; Connell, A. N. Interference-Enhanced Raman Scattering of Very Thin Titanium and Titanium Oxide Films. Phys. Rev. Lett. 1980, 44, 273–276.

9

Connell, G. A. N.; Nemanich, R. J.; Tsai, C. C. Interference Enhanced Raman Scattering from Very Thin Absorbing Films. Appl. Phys. Lett. 1980, 36, 31–33.

10

Wang, Y. Y.; Ni, Z. H.; Shen, Z. X.; Wang, H. M.; Wu, Y. H.; Wang, Y. Y.; Ni, Z. H.; Shen, Z. X. Interference Enhancement of Raman Signal of Graphene. Appl. Phys. Lett. 2008, 92, 043121.

11

Yoon, D.; Moon, H.; Son, Y.-W.; Choi, J. S.; Park, B. H.; Cha, Y. H.; Kim, Y. D.; Cheong, H. Interference Effect on Raman Spectrum of Graphene on SiO2/Si. Phys. Rev. B 2009, 80, 125422.

12

Ni, Z.; Wang, Y.; Yu, T.; Shen, Z. Raman Spectroscopy and Imaging of Graphene. Nano Res. 2008, 1, 273–291.


Page 23 of 26

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


13

Ren, W.; Saito, R.; Gao, L.; Zheng, F.; Wu, Z.; Liu, B.; Furukawa, M.; Zhao, J.; Chen, Z.; Cheng, H.-M. Edge Phonon State of Mono- and Few-Layer Graphene Nanoribbons Observed by Surface and Interference Co-Enhanced Raman Spectroscopy. Phys. Rev. B 2010, 81, 035412.

14

Ling, X.; Zhang, J. Interference Phenomenon in Graphene-Enhanced Raman Scattering. J. Phys. Chem. C 2011, 115, 2835–2840.

15

Schedin, F.; Lidorikis, E.; Lombardo, A.; Kravets, V. G.; Geim, A. K.; Grigorenko, A. N.; Novoselov, K. S.; Ferrari, A. C. Surface-Enhanced Raman Spectroscopy of Graphene. ACS Nano 2010, 4, 5617–5626.

16

Solonenko, D.; Gordan, O. D.; Milekhin, a; Panholzer, M.; Hingerl, K.; Zahn, D. R. T. Interference-Enhanced Raman Scattering of F

16

CuPc Thin Films. J. Phys. D.

Appl. Phys. 2016, 49, 115502. 17

Abidi, I. H.; Cagang, A. A.; Tyagi, A.; Riaz, M. A.; Wu, R.; Sun, Q.; Luo, Z. RSC Advances Oxidized Nitinol Substrate for Interference Enhanced Raman Scattering of Monolayer. RSC Adv. 2016, 6, 7093–7100.

18

Gao, L.; Ren, W.; Liu, B.; Saito, R.; Wu, Z.; Li, S. Surface and Interference Coenhanced Raman Scattering of Graphene. ACS Nano 2009, 3, 933–939.

19

Xu, W.; Mao, N.; Zhang, J. Graphene: A Platform for Surface-Enhanced Raman Spectroscopy. Small 2013, 9, 1206–1224.

20

Jesion, I.; Skibniewski, M.; Skibniewska, E.; Strupiński, W.; Szulc-Dąbrowska, L.; Krajewska, A.; Pasternak, I.; Kowalczyk, P.; Pinkowski, R.;. Graphene and Carbon



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

Nanocompounds: Biofunctionalization and Applications in Tissue Engineering. Biotechnol. Biotechnol. Equip. 2015, 29, 415–422. 21

Thompson, B. C.; Murray, E.; Wallace, G. G. Graphite Oxide to Graphene. Biomaterials to Bionics. Adv. Mater. 2015, 10.1002/adma.201500411.

22

Xu, W.; Ling, X.; Xiao, J.; Dresselhaus, M. S.; Kong, J.; Xu, H.; Liu, Z.; Zhang, J..Surface Enhanced Raman Spectroscopy on a Flat Graphene Surface. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 9281–9286.

23

Losurdo, M.; Bergmair, I.; Dastmalchi, B.; Kim, T.-H.; Giangregroio, M. M.; Jiao, W.; Bianco, G. V.; Brown, A. S.; Hingerl, K.; Bruno, G. Graphene as an Electron Shuttle for Silver Deoxidation: Removing a Key Barrier to Plasmonics and Metamaterials for Sers in the Visible. Adv. Funct. Mater. 2014, 24, 1864–1878.

24

Horcas, I.; Fernández, R.; Gómez-Rodríguez, J. M.; Colchero, J.; Gómez-Herrero, J.;

Baró, A. M. WSXM: A Software for Scanning Probe Microscopy and a Tool for Nanotechnology, Rev. Sci. Instrum., 2007, 78, 013705. 25

Jiménez Riobóo, R. J.; Climent-Pascual, E.; Díez-Betriu, X.; Jiménez-Villacorta, F.; Prieto, C.; De Andrés, A. Elastic Constants of Graphene Oxide Few-Layer Films: Correlations with Interlayer Stacking and Bonding. J. Mater. Chem. C 2015, 3, 4868–4875.

26

Ramírez-Jiménez, R.; Álvarez-Fraga, L.; Jimenez-Villacorta, F.; Climent-Pascual, E.;

Prieto, C.; De Andrés, A. Interference Enhanced Raman Effect in Graphene Bubbles. Carbon N. Y. 2016, 105, 556–565. 27

Ekinci, Y.; Solak, H. H.; Löffler, J. F. Plasmon Resonances of Aluminum Nanoparticles and Nanorods. J. Appl. Phys. 2008, 104. 24 ACS Paragon Plus Environment

Page 24 of 26

Page 25 of 26

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


28

Mogensen, K. B.; Gühlke, M.; Kneipp, J.; Kadkhodazadeh, S.; Wagner, J. B.; Espina Palanco, M.; Kneipp, H.; Kneipp, K. Surface-Enhanced Raman Scattering on Aluminum Using near Infrared and Visible Excitation. Chem. Commun. (Cambridge, U. K.). 2014, 50, 3744–3746.

29

Kasry, A.; Ardakani, A. A.; Tulevski, G. S.; Menges, B.; Copel, M.; Vyklicky, L. Highly Efficient Fluorescence Quenching with Graphene. J. Phys. Chem. C 2012, 116, 2858–2862.

30

Lin, W.; Tian, B.; Zhuang, P.; Yin, J.; Zhang, C.; Li, Q.; Shih, T.; Cai, W. GrapheneBased Fluorescence-Quenching-Related Fermi Level Elevation and ElectronConcentration Surge. Nano Lett. 2016, 16, 5737-5741.

31

Blake, P.; Hill, E. W.; Castro Neto, A. H.; Novoselov, K. S.; Jiang, D.; Yang, R.; Booth, T. J.; Geim, A. K. Making Graphene Visible. Appl. Phys. Lett. 2007, 91, 2007–2009.

32

Roddaro, S.; Pingue, P.; Piazza, V.; Pellegrini, V.; Beltram, F. The Optical Visibility of Graphene: Interference Colors of Ultrathin Graphite on SiO2. Nano Lett. 2007, 7, 2707–2710.



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

TOC


Page 26 of 26

Efficient Heterostructures for Combined Interference and Plasmon

Recommend Documents