Applied Materials Today 35 (2023) 101929

Contents lists available at ScienceDirect

Applied Materials Today

journal homepage: www.elsevier.com/locate/apmt

Exploring the surface-enhanced Raman scattering (SERS) activity of gold

nanostructures embedded around nanogaps at wafer scale: Simulations
and experiments
Marta Lafuente a, b, *, Pablo Muñoz b, Erwin J.W. Berenschot a, Roald M. Tiggelaar c,
Arturo Susarrey-Arce a, Sergio G. Rodrigo d, f, Lucas J. Kooijman a, Sonia M. García-Blanco b,
Reyes Mallada e, f, g, María P. Pina e, f, g, Niels R. Tas a, *
a
Mesoscale Chemical Systems, MESA+ Institute, University of Twente, P.O. Box 217, AE Enschede 7500, the Netherlands
b
Integrated Optical Systems, MESA+ Institute, University of Twente, P.O. Box 217, AE Enschede 7500, the Netherlands
c
NanoLab cleanroom, MESA+ Institute, University of Twente, P.O. Box 217, AE Enschede 7500, the Netherlands
d
Departamento de Física Aplicada, Facultad de Ciencias, Universidad de Zaragoza, Zaragoza 50009, Spain
e
Department of Chemical & Environmental Engineering, Edificio I+D+i, Campus Rio Ebro, C/Mariano Esquillor s/n, Zaragoza 50018, Spain
f
Instituto de Nanociencia y Materiales de Aragón (INMA), CSIC-Universidad de Zaragoza, Zaragoza 50009, Spain
g
Networking Research Center on Bioengineering, Biomaterials and Nanomedicine, CIBER-BBN, Madrid 28029, Spain

A R T I C L E I N F O A B S T R A C T

Keywords: A unique way of converting free space light into a local electromagnetic field in small spaces is via metallic
Ordered nano-wedges nanostructuring. In this work fabrication, experimental characterization and simulation of surface-enhanced
Size-controllable nanogaps Raman scattering (SERS) active specimens based on Au nanostructures are discussed. We used displacement
FDTD simulations
Talbot lithography (DTL) to fabricate silicon nano-wedge substrates with Au nanostructures embedded around
Ordered symmetric and asymmetric
their apices. After the ion beam etching process, a nanogap is introduced between two Au nanostructures
nanostructures
Ion beam etching (IBE) templated over nano-wedges, yielding specimens with SERS characteristics. The Au nanostructures and the
Gold sputtering nanogaps have symmetric and asymmetric configurations with respect to the wedges. With this nanofabrication
method, various wafer-scale specimens were fabricated with highly controllable nanogaps with a size in the
order of 6 nm for symmetric gaps and 8 nm for asymmetric gaps. SERS characteristics of these specimens were
analyzed experimentally by calculating their analytical enhancement factor (AEF). According to finite-difference
time-domain (FDTD) simulations, the Raman enhancement arises at the narrow gap due to plasmonic resonances,
yielding a maximum AEF of 6.9 × 106. The results highlight the SERS activity of the nanostructures and ul­
timately comply with reliable substrates for practical applications.

1. Introduction disorders [2] and the presence of virus [3] or genetic diseases [4]. The
implementation of SERS in other fields has substantially expanded, e.g.,
Raman spectroscopy is a non-destructive technique to study the in­ for monitoring and quantification of environmental toxins [5,6], pesti­
elastic scattering of photons caused by matter and can provide rich in­ cides [7,8], forensic substances [9,10] or nerve agents [11,12].
formation about the molecular structure and composition of analytes SERS occurs mainly due to a plasmonic effect in metallic nano­
based on their vibrational fingerprints. However, light has a low prob­ structures [13]. Under light irradiation, a localized collective oscillation
ability of being inelastically scattered by a material, thereby making of electrons in metallic nanostructures produces a high local electric
Raman spectroscopy a technique with low sensitivity [1]. This drawback field, which leads to enhanced Raman scattering from molecules in close
can be overcome by using the Surface Enhanced Raman Scattering contact with the plasmonic nanostructure. SERS can also occur over
(SERS) effect, which increases the Raman signal greatly. SERS is used in adsorption sites nearby the electric field enhancement volume. Along­
multiple biomedical applications, such as diagnosis of degenerative side, localized SERS can be produced in nanometer-sized gaps in

* Corresponding authors.
E-mail addresses: [email protected], [email protected] (M. Lafuente), [email protected] (N.R. Tas).

https://doi.org/10.1016/j.apmt.2023.101929
Received 16 June 2023; Received in revised form 30 August 2023; Accepted 7 September 2023
Available online 15 September 2023
2352-9407/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
M. Lafuente et al. Applied Materials Today 35 (2023) 101929

between metallic nanostructures, known as SERS “hot-spots”, i.e. re­ et al. after a FIB treatment converted a stack of Au (100 nm) – TiO2 (40
gions with the highest local electric field. nm) – Au (50 nm) in two symmetric nano-antennas with a length of 35
Plasmonic substrates including metal nanoparticles and lithograph­ nm a gap width of 6 nm. Photolithography has been widely used for the
ically prepared nanostructures are ideal substrates to produce enhanced fabrication of structures ranging from 500 nm to several microns due to
electric fields. For chemically synthesized nanoparticles, self-assembly low costs, high uniformity and high throughput [40–42].
on solid supports is one of the most widely applied bottom-up fabrica­ Random-ordered silicon nanopillars with a large aspect ratio has been
tion techniques [14–17]. There, the hot-spots can be found in small free fabricated by means of reactive ion etching (RIE) by controlling the
spaces between metal nanoparticles [14,18,19], the assembly of which composition of RIE gas mixture [43].
typically ends randomized in the form of aggregates [20], lacking uni­ Herein, the SERS activity of Au nanostructures templated over silicon
formity and reproducibility that negatively affect the Raman signals nano-wedges is studied. The nanofabrication of symmetric and asym­
[21]. Irreproducibility of Raman signals might be a result of in­ metric Au nanostructures was achieved by utilizing ion beam etching of
homogeneities at the nanogap level. Self-assembly strategies have been Au films sputter-deposited over silicon nano-wedges. It is important to
adopted widely to alleviate such particle-to-particle nanogap in­ note that the gap between the symmetric Au nanostructures at the apex
homogeneities [22]. Nevertheless, reaching those SERS substrates with of the nano-wedge is also referred to as ‘symmetric nanogap’, whereas
highly reproducible and uniform distribution of “hot-spots” at centi­ the gap between the asymmetric Au nanostructures is named ‘asym­
metre scale is still a challenge [23–25]. metric nanogap’. Whether or not a symmetric or asymmetric Au nano­
Top-down nanofabrication offers a plethora of options for controlled structure is formed, depends on the initial thickness of the sputtered gold
nanogap fabrication with a wide selection of sizes and geometries, layer (see Fig. 1). The nano-wedges were realized by a combination of
excellent reproducibility, and large-scale uniformity [14,26]. For convex corner lithography and anisotropic wet etching. This fabrication
example, electron beam lithography (EBL) [27] or focus ion beam method provides size-controllable nanogaps which were used here to go
lithography (FIB) [28] provide maximum control over the nanogap size down to ~ 6 ± 2 nm between Au nanostructures. Various such nanogap
between nanostructures but suffer from limited scalability. Nano­ specimens were prepared to understand the effect of symmetry and
lithography is a flexible and powerful tool that can aid in accelerating asymmetry around nano-wedges. We identified that the etched silicon
the nanofabrication of optical and sensing devices, especially when depth (at the apices of nano-wedges) as well as the Au nanogap width
combined with corner lithography and ion beam etching. A key nano­ affect the Raman enhancement factor. The simulations conducted with
lithographic resource compatible with other nanofabrication methods is the Finite-Difference Time-Domain (FDTD) method allowed for a better
displacement Talbot lithography (DTL). This interference nanolitho­ understanding of the electromagnetic (EM) modes responsible for the
graphic method is an emerging patterning technique by means of which SERS signal measured over the various Au nanostructure configurations.
ordered nanostructures can be produced with a monochromatic beam For the evaluation of their SERS activity, the Analytical Enhancement
over large areas [23,26,29–31]. Likewise, corner lithography combined Factor (AEF) was estimated using rhodamine 6 G (R6G) as the molecular
with the selective anisotropic wet etching of silicon has resulted in the probe.
realization of different ordered structures in silicon [32]. Both tech­
nologies can ultimately be combined for the top-down nanofabrication 2. Experimental section
of silicon nano-wedges [33]. Such nano-wedges can then render as
templates for the formation and size/shape tuning of Au structures with 2.1. Fabrication of Au nanostructures embedded on nano-wedges
a well-controlled nanogap in-between them. In fact, these nanogaps can
be configured symmetrically and asymmetrically with respect to the The SERS-active specimens were fabricated in (100)-oriented silicon
supporting wedges. (Si) substrates. A simplified summary of the fabrication route followed
In literature different top-down nanotechnologies have been used for to produce the SERS-active samples can be found in this section. A
the fabrication of well-ordered, periodic and uniform plasmonic nano­ detailed explanation of the fabrication process is provided in the sup­
structures and nanogaps. EBL can be used to nanofabricate SERS porting information.
nanostructures with different geometries: spheres [27], nanostars [34], Nanolithography by means of DTL and continuation of the top-down
nanoholes [35] or bimetallic [36] with a resolution up to 10 nm. The fabrication process employing reactive ion etching (RIE) yielded a pe­
main drawback of EBL is it high time consuming and its high costs, for riodic line pattern defined in stoichiometric silicon nitride (Si3N4)
example, fabricating 1 cm2 with nanostructures of 20 nm requires 111 h (Fig. 2a) [30,31,44,45]. It is noted that this pattern was aligned with the
[37]. FIB has been used for fabricating nano-antennas [38,39]. Ahmed silicon (100) crystal orientation. Etching in potassium hydroxide (KOH)

Fig. 1. Schematic overview of the fabrication process for creating symmetric and asymmetric gold nanostructures and, therefore, symmetric and asymmetric
nanogaps, respectively.

2
M. Lafuente et al. Applied Materials Today 35 (2023) 101929

Fig. 2. Schematic representation of the fabrication process of sharpened nano-wedges: (a) Patterning of Si3N4 on (100)-Si by DTL and reactive ion etching (RIE), (b)
KOH etching of 1st generation of V-grooves using Si3N4 as hard mask, (c) LOCOS, (d) Selective removal of Si3N4 with H3PO4, (e) TMAH etching of 2nd generation of
V-grooves using SiO2 as hard mask, (f) Dry oxidation of nano-wedges, (g) Removal of SiO2 with HF results in sharpened nano-wedges, (h) Dry oxidation of sharpened
nano-wedges. (i) Cross-section SEM image of realized sharpened nano-wedge. The black box shows a magnification of the nano-wedges.

was then applied to generate the first generation of V-shaped 2.2. Materials characterization
nano-grooves (Fig. 2b). Nano-wedges can be realized by removing the
flat spacers in-between these nano-grooves. This was accomplished by High-Resolution scanning electron microscopy (HR-SEM) images of
local oxidation of silicon (LOCOS) (Fig. 2c), followed by selective the fabricated nanotextured SERS-active specimens were obtained using
removal of the Si3N4 hard mask with hot phosphoric acid (H3PO4) a FEI Sirion system. With ImageJ analysis software, five to ten SEM
(Fig. 2d). Subsequently, anisotropic etching of the exposed silicon in images of each specimen were analyzed to obtain dimensional infor­
tetramethyl ammoniumhydroxide (TMAH) was performed, which gave mation about the Au nanostructures around a nanogap, i.e., the gap size
the second generation of V-shaped nano-grooves and hence as well as the etched silicon depth at the top of a nano-wedge. The un­
nano-wedges (Fig. 2e) [46]. The obtained nano-wedges were sharpened certainty is expressed in ± 1σ. The process of Cr removal was evaluated
using dry oxidation (Fig. 2f) and subsequent removal of the formed by preparing a TEM lamella using Focus Ion Beam (FIB) etching (Dual
silicon dioxide (SiO2) film using hydrofluoric acid (HF) (Fig. 2g). Prior to Beam Helios): an asymmetric specimen was cut in a transversal direction
Au sputtering, the sharpened nano-wedges were coated with dry SiO2 and Scanning-transmission electron microscopy (STEM) images were
(Fig. 2h). recorded using an FEI Tecnai F30 (see SI Fig. S1).
Two different Au thicknesses were sputtered on nano-wedge arrays,
i.e., 50 nm and 190 nm (Fig. 3a). Before Au sputtering, a chromium (Cr) 2.3. Finite-difference time-domain (FDTD) simulations
layer of 2 nm was sputter-deposited. The Au thickness determines
whether size-symmetric or size-asymmetric Au nanostructures will The plasmonic behavior of nanogap based SERS substrates was
surround the tip of the nano-wedge: size-symmetric implies that the simulated using the FDTD method by means of a commercial software
nanostructures present at both sides of each nano-wedge are identical in package (Lumerical Solutions Inc, FDTD solutions). In the simulations, a
cross-sectional size (and shape), whereas size-asymmetric means that model was created for each of the fabricated configurations according to
differently sized Au nanostructures surround the nano-wedges. The Au the dimensions as obtained from HR-SEM image analysis. Conformal
thickness was reduced with ion beam etching (IBE) until the utmost tip mesh (mesh accuracy 4) and mesh refinement around the nanogaps were
of the nano-wedges were protruding (Fig. 3b). Post to IBE, wet chemical used to define the structure with a minimum mesh size of 0.1 nm. Pe­
removal of the Cr film from the tip of the nano-wedges was performed, riodic boundary conditions were considered along the x-direction, and
followed by selective removal of the SiO2 with buffered hydrofluoric the perfectly matched layer (PML) boundary conditions with a steep
acid (BHF) (Fig. 3c). The exposed silicon at the top of each nano-wedge angle profile were selected along the y-direction (it is noted that the x
was anisotropically etched in KOH, which yielded openings with a cross- and y-directions are graphically represented in Fig. 6a). Ideally, PML
section in the form of rhombic-like cages at the tip of the nano-wedges boundaries would absorb all incident light without creating any back
(Fig. 3d). By means of immersion in BHF and chromium etchant, the reflection. A plane wave source propagating in the y-direction with the
SiO2 and Cr could be removed from within the KOH-etched cage at the electric field parallel to the x-direction was employed to excite the
tip of the nano-wedge (Fig. 3e). This option was applied to both types of system. A convergence test was performed to determine the optimum
specimen, i.e. with symmetric and asymmetric Au nanostructures. The distance between the SERS nanostructure and the PML boundary con­
use of a second Au sputtering step with the aim to reduce the gap be­ ditions in the y-direction (2.5 µm). The analysis was performed by
tween the Au nanostructures was only applied to specimens with considering the “Palik” dispersion curve for Au, silicon and silicon di­
asymmetric Au-structures around the nano-wedges (Fig. 3f). oxide [47], whereas the refractive index of chromium was taken from
the CRC handbook [48].

3
M. Lafuente et al. Applied Materials Today 35 (2023) 101929

Fig. 3. Schematic representation of the sputtering of Au on oxidized sharpened nano-wedges and further processing into SERS-active specimen: (a) sputtering of Cr/
Au, (b) IBE of Au, (c) selective removal of SiO2 from protruding tips with BHF, (d) KOH etching, (e) selective removal of Cr and SiO2, (f) sputter deposition of Au to
downsize the gap.

2.4. Raman and surface-enhanced Raman scattering (SERS) 2.5. Determination of analytical enhancement factor (AEF)
measurements
The SERS properties of a plasmonic substrate can be characterized
An Alpha 300 Raman spectrometer (WITec, Ulm, Germany) was used following various methods, the results of which may vary even by a few
with a confocal optical microscope (spectral resolution 2 cm− 1). Raman- orders of magnitude [49]. In this work, we use the analytical enhance­
SERS spectra were collected in backscattering configuration. Excitation ment factor (AEF). The AEF was calculated according to Eq. (1):
of the samples was carried out with a 633 nm He-Ne laser, a scattering ( ∗ / )/( ∗ / )
(1)
area of 1.86 µm2, at room temperature, a laser power of 3 mW, and an
AEF = ISERS CSERS IRaman CRaman
acquisition time of 1 s.
Where: CRaman and CSERS are the R6G concentrations during the Raman
SERS substrates (4 mm x 6 mm) were incubated (overnight) in
measurements and SERS conditions, respectively. I*SERS is the normal­
Rhodamine (R6G, Sigma-Aldrich) 1 µM in a vertical position; then, they
ized intensity of the R6G molecules on the SERS substrates, and I*Raman
were rinsed in distilled water and air-dried. Subsequently, on each
corresponds to the normalized intensity of the R6G molecules measured
specimen, five different maps of 50 × 50 µm2 containing 100 mea­
in the liquid phase (1 mM concentration). In order to perform a fair
surement locations were measured. The relative standard deviation
comparison, both intensities were normalized to laser power and time
(RSD) was the average of these 100 measurements. The Raman intensity
(cts mW− 1 s− 1).
of the C–C stretching vibrational mode of R6G at 1510 cm− 1 was used
as a mapping signal. In addition, normal Raman spectra were measured
3. Results & discussion
in a 1 mM R6G solution by focusing the laser beam inside the liquid,
with 15 mW of power and 50 s collection time. The solution was placed
3.1. Fabrication of Au nanostructures on nano-wedges
in a polytetrafluoroethylene (PTFE) container of 5 mm × 5 mm × 5
mm (length, width and depth, respectively). In all SERS and Raman
The nanofabrication of SERS substrates is based on the sputtering of
spectra, automatic baseline correction, i.e., background subtraction, has
thin films of Au onto Si etched nano-wedges. The combination of DTL
been done with WITec Software 2.10.
and etching to create nano-wedges onto which Au is sputtered and

4
M. Lafuente et al. Applied Materials Today 35 (2023) 101929

further processed into nano-structures is shown in Figs. 2 and 3. Fig. 3 Table 1

depicts a three-step process that combines DTL, RIE, and anisotropic wet Dimensions and characteristics of nano-textured SERS-active specimens studied
etching of silicon. This process is applied for the top-down fabrication of in this work (it is noted that ‘sym’ and ‘asym’ in the sample I.D. refer to symmetry
sharpened nano-wedges. Named as the “nano-wedges process”, it pro­ or asymmetry (in size and shape) of the Au-patterns around a wedge). The un­
duces a template for realization of well-defined Au nanostructures with certainty is expressed in ± 1σ.
highly reproducible nanogaps. Fig. 3 shows the subsequent sputter Sample I. Gap Etched silicon depth Additional processing
deposition of Au onto these wedges and further nanomachining into two D. (G) (SiD)
[nm] [nm]
different types of gold nanostructures with defined nanogaps, i.e., size-
symmetric and size-asymmetric Au-shapes. This process is referred to as sym1 6±2 0 removal of SiO2 and Cr
sym2 6±2 13 ± 1 removal of SiO2 and Cr
“Au-shaping process”. It is important to remember that the symmetry or
sym3 6±2 40 ± 3 removal of SiO2 and Cr
asymmetry between the Au-nanostructures at the apex of a wedge de­ sym4 6±2 40 ± 3 –
pends on the initial thickness of the sputtered gold layer (as illustrated in asym1 16 ± 4 100 ± 5 removal of SiO2 and Cr
Fig. 3). asym2 12 ± 3 100 ± 5 removal of SiO2 and Cr, Au
We first address the nano-wedges process from Fig. 2. Three main sputtering
asym3 8±3 100 ± 5 removal of SiO2 and Cr, Au
steps constitute this process, which are applied for the nanofabrication
sputtering
of both symmetric as well as asymmetric geometries. The first step starts
with the fabrication of the 1st generation of V-grooves in the silicon
wafer. Here, a combination of three techniques was employed, i.e., (i) a different depth of the rhombic cages. The other two additional process
DTL, (ii) RIE, and (iii) anisotropic wet etching of silicon. The DTL-step sequences that were employed concern i) removal of the SiO2 and Cr
was used to define a periodic line pattern in a positive photoresist. present at the inside of the KOH-etched feature at the top of the wedges
Subsequently, selective and directional RIE was applied to pattern a 13 (Fig. 3e) and ii) downsizing of the gap between the Au-structures by
nm thick Si3N4 layer (Fig. 2a). Then, selective wet etching with KOH means of a 2nd sputtering step (Fig. 3f). Table 1 summarizes the typical
produced the first generation of V-shaped nano-grooves, with a pitch of dimensions and characteristics of the fabricated SERS substrates. Taking
250 nm and in-between the V-grooves flat spacers of 68 nm wide into account the cleanroom costs, one specimen costs approximately 45
(Fig. 2b). €.
The 2nd generation of V-grooves was created by the doubling of the Now we turn our attention to the fabricated specimen with sym­
periodicity of the V-grooves. This can be achieved by removing the flat metric Au nanostructures (named as ‘sym’) in Table 1 and SEM images in
spacers (indicated by the red bars in Fig. 2b) between the 1st generation Fig. 4. From the SEM images, the etched silicon depth was estimated and
of V-grooves. Hence, LOCOS was used in combination with an aniso­ showed small variations around 13 ± 1 nm or 40 ± 3 nm for a fixed gap
tropic wet etching of silicon with TMAH [46]. The aim of the of 6 ± 2 nm. For the same structures, the length of the Au flap was 67 ±
LOCOS-step was to protect the 1st generation of V-grooves during sub­ 7 nm, and the thickness was 16 ± 3 nm. It is worth noting that due to the
sequent fabrication steps. LOCOS yielded a SiO2 layer of 7.9 nm on the Si IBE process, the Au-film in the widest V-groove has become thinner than
{111}-walls of the V-grooves, whereas the Si3N4 was only slightly the gold layer on the sides of the wedges. As a consequence, during the
oxidized. Then, the Si3N4 layer was removed by immersion of the sub­ HF-step at which the SiO2 is removed from the tips of the wedges, HF can
strate in H3PO4. Lastly, the exposed silicon was etched in TMAH for 12 also penetrate through the locally thinned (or through defects/holes in
min, resulting in the 2nd generation of V-grooves and hence the) Au-film and etch the SiO2. Moreover, some points in the wide V-
nano-wedges (Fig. 2e). After this sequence, the widths of the V-grooves groove could lead to the defective etched silicon underneath the Au,
were 120 ± 2 nm for the 2nd generation and 130 ± 2 nm for the 1st yielding “artifacts” as indicated with a red arrow in Fig. 4. It is noted that
generation. Post to TMAH etching, the SiO2 layer was removed with HF. these artifacts are not considered in our numerical simulations.
The third step in Fig. 2, named Sharpening nano-wedges, was applied Figs. 5a-5f show HR-SEM images of the asymmetric Au nanogaps
to obtain smaller gaps between the Au nanostructures. To accomplish (named as ‘asym’). The nanogap size was tuned between 8 and 16 nm
this, the substrate was dry oxidized, during which the nano-wedges were with an additional Au sputtering step, as illustrated in Fig. 3f. The SEM
sharpened (Fig. 2f) [50]. Afterwards, the SiO2 film was removed with images in Fig. 5 reveal the nanogap reduction as a function of the
HF. Finally, in order to avoid the formation of Au-silicide, the sharpened sputtered-deposited additional Au layer. The structures shown in
nano-wedges were coated with 3 nm of SiO2 before Au sputtering Figs. 5a & 5d (sample ‘asym1’) are employed as control with the widest
(Fig. 2h). nanogap of 16 ± 4 nm, while the structure in Figs. 5b & 5e (sample
Next, two different Au thicknesses (i.e., 50 nm and 190 nm) were ‘asym2’) and Figs. 5c & 5f (sample ‘asym3’) confirm nanogap reduction
sputtered on nano-wedges to achieve symmetric and asymmetric Au down to 12 ± 3 and 8 ± 3 nm, respectively. For a clear comparison
nanogaps (Fig. 3a). The Au thickness was reduced with IBE until the tips between ‘sym’ and ‘asym’, consult Table 1 and the histograms in Figs. 5g-
of the nano-wedges protruded (Fig. 3b). Note that the thickness of the 5i. The widths of the largest parts of asymmetric nanostructures were
initially sputtered Au film determined whether the remaining Au- measured and found to be 126 ± 3 nm (asym1), 131 ± 5 nm (asym2)
pattern around a nano-wedge was symmetric or asymmetric and the and 136 ± 2 nm (asym3), respectively (Fig. 4). The widths of the
time length of IBE the width of Au nanogap together with the last smallest parts were 91 ± 3 nm, 95 ± 3 nm and 98 ± 3 nm, respectively.
sputtering step of Au for further finetuning. Hence, the IBE of 50 nm Au The etched silicon depth was 100 ± 5 nm for all asymmetric specimens.
yielded equally sized Au-shapes on the left-hand and right-hand side of Finally, we verified that the Cr layer was removed, as explained in
each wedge (Fig. 3b(i)), whereas IBE of 190 nm Au resulted in differ­ Fig. 2e, i.e., from the Au nanostructure, by STEM analysis (see
ently sized Au-structures at both sides of each nano-wedge (Fig. 3b(ii)). Figure S1). The chemical analysis confirmed that Cr was completely
The metallic adhesion layer (i.e., 2 nm Cr) and SiO2 were removed to removed from the Au “flaps”, whereas it remained at the interface be­
gain access to the silicon tip of the nano-wedge (Fig. 3c). The exposed tip tween Au and silicon. The STEM images also evidence the absence of Au
was etched in KOH, yielding to a cage with a rhombic-like cross-section inside the rhombic cages after the second (optional) gold sputtering step.
at the top of each wedge (Fig. 3d). Applying this KOH-step resulted in a To summarize, our nanofabrication process, which combines the
nano-textured SERS-active specimen in which Au adhered with Cr was fabrication of nano-wedges in silicon with Au sputter-deposition, allows
present on SiO2 flaps (Fig. 3d). for the fabrication of robust and highly uniform SERS substrates at
The described fabrication process can be easily adapted to create wafer-scale. In addition, upon performing slight modifications, various
multiple variations of SERS-active specimens as shown in Table 1. The SERS nanostructures can be fabricated with well-controlled gaps and,
slightly adapted process utilizes a shorter immersion in KOH, generating therefore, hot-spot distribution.

5
M. Lafuente et al. Applied Materials Today 35 (2023) 101929

Fig. 4. Cross-sectional SEM images of realized wedge-based samples with symmetric Au nanostructures: a) sym1; b) sym2; c) sym3 and d) sym4. For more details
about the sample, see Table 1. The scale bar is 100 nm. The black box shows the unit cells used in the simulations. Red arrows indicate the mentioned artifacts. e)
Corresponding size distribution histogram (N>150) of the gap as estimated from 3 different top-view HR-SEM images.

Fig. 5. (a-c) Top-view and (d-f) cross-sectional HR-SEM images of the asymmetric Au nanostructures around nano-wedges (asym1, asym2, asym3). The scale bar is
100 nm. The dashed black box shows the unit cells used in the simulations. (g-i) Corresponding size distribution histograms (N>500) of the gap as estimated from 10
different top-view HR-SEM images.

3.2. Effect of gap and etched silicon depth on electric field intensity (CM) origin(s) [18]. In this section we elucidate about the SERS
distribution enhancement observed in the experiments, which is likely due to the EM
field enhancement associated to plasmonic excitations. Focusing on the
The SERS enhancement has electromagnetic (EM) and/or chemical application of SERS-active symmetric and asymmetric Au nanogaps, we

6
M. Lafuente et al. Applied Materials Today 35 (2023) 101929

first show by FDTD simulations that these structures may exhibit sig­ arrays are highly sensitive to the polarization of incident light, with their
nificant SERS enhancement based on the high EM field enhancements most notable effect being that resonances are only observable when the
obtained from simulations. Furthermore, we discuss the possible types of electric field is perpendicular to the invariant direction (y-direction)
plasmonic resonances supported by these structures that could explain [52]. This has important consequences on the SERS effect formation, as
such high enhancements. we demonstrate in the next section.
The 3D-dimensions of the nanostructures and the dielectric constants Slit arrays support two different types of plasmonic resonances,
of the materials of which they are made have a significant impact on the originating either from Surface Plasmon Polaritons (SPPs) or Localized
SERS signal. To identify potential designs for effective SERS, FDTD Plasmon Polaritons (LSPs) [51]. The first have been related to excita­
simulations can be used to establish specific criteria. tions of resonances in the optical spectrum of optically thick films
Firstly, we created a base simulation with a parametric symmetric Au (thicknesses > skin-depth of the metal) at wavelengths slightly above
nanogap structure model (Fig. 6a), which then was used to perform a the condition λR ∼ n × p (where p is the lattice constant, n the refractive
parameter sweep with different nanogaps (G, G = 2–16 nm), and depths index of the substrate and R states for resonant). For films that are
of the rhombic cross-sectional feature etched at the top of the silicon optically thin, the surface resonances may experience significant red­
nano-wedges (SiD, SiD = 0–100 nm). A constant parameter in our Au shifts [53]. The second type of plasmonic resonances are Fabry-Perot
nanogap model was the thickness of the overhanging Au nanostructures modes, which depend on the optical path of light inside the slits [51].
(“Au-nanoflaps”), which was set to 17 nm for the complete study, in In plasmonic structures, these resonances are also dependent on the
accordance with the fabricated samples. After the nanogap model con­ phase change that occurs upon reflection at the cavity’s end, defined by
struction, the FDTD were configured as described in the method section. the slits. Our SERS nanostructures are optically thin, such that we can
Fig. 6b shows the enhancement of the electric field intensity in the safely neglect the first effect (SPPs resonances) [52].
gap (averaged in a small nearby region, black box of Fig. 6c) as a We elucidate that the main parameter controlling the position and
function of the wavelength, for different etched silicon depths (gap of 6 shape of the resonances in Fig. 6b is the phase change, which depend on
nm). The spectral window investigated shows a clear resonance for each the dielectric environment around the nanogaps. However, surface
silicon depth, which develops from a low broadband peak at lower plasmonic effects cannot be disregarded. As we will show later, without
etching depths to a strong and narrow resonance around SiD ~ 40 nm. removal the Cr layer strongly impairs the resonant response of the sys­
As the etched depth increases further, the resonance intensity starts to tem. Therefore, it is possible that hybridization could occur between
diminish again. localized and surface modes, as SPP resonances are expected to occur at
The nanostructured systems studied can be viewed as subwavelength wavelengths greater than n × p ∼ 3.83 × 120 = 460nm. The etched
metallic gratings composed of slits. The study of slit arrays on metals has silicon depth, SiD, affects both modes. When SiD increases, the reso­
been extensively researched in relation to Extraordinary Optical nance shifts to shorter wavelengths, which is caused by the metal
Transmission phenomena [51]. The slits are usually carved into flat “seeing” a lower refractive index, as illustrated in Fig. 6b.
metal surfaces, which makes them more convenient for theoretical Another shared feature between flat metal slits and our Au nano­
analysis. Although the SERS nanostructures fabricated here are much structures is the similar field distribution of light inside the gap [51].
more complex than flat metal films corrugated with slits, they share Fig. 6c displays the distribution of normalized electrical field intensity
most of the optical features displayed by them, as shown hereafter. Slit for a nanostructure with an etched silicon depth of SiD = 39 nm and a

Fig. 6. (a) Schematic side view of the nano-

wedge-based symmetric Au nanostructure
model for FDTD simulations. color code: gray,
Si; blue, SiO2; orange, Cr; yellow, Au. (b) In­
tensity of the electric field enhancement as a
function of the wavelength, for different etched
silicon depths (SiD) (the gap, G, fixed at 6 nm).
The system is illuminated from the air side or
topside from the positive y-direction at normal
incidence. The incident electric field is pointing
perpendicular to the gap (x-direction). (c)
Color-coded FDTD-simulated image of the
electric field intensity distribution for a gap of
6 nm and an etched SiD of 39 nm at the
wavelength with the maximum of intensity
enhancement shown in panel b) at λ = 680 nm;
(d) FDTD-simulated electromagnetic field
enhancement averaged over the gap region
(black box of c)) versus etched silicon depth for
six different gaps. The gray dashed line shows
the trend of the maximum field enhancement as
a function of G and SiD for λ = 633 nm.

7
M. Lafuente et al. Applied Materials Today 35 (2023) 101929

nanogap of G = 6 nm. The near-field map corresponds to the maximum homogeneity of the width of the nanogap, reproducibility during
point on the intensity enhancement curve shown in Fig. 6b (at a wave­ nanofabrication and uniform SERS signal intensity [15,55]. These pa­
length of 680 nm). The electric field points from side to side inside the rameters are assessed by means of the evaluation of SERS activity at
gap (x-direction), which in the study of slit arrays is attributed to the three different locations on the same SERS substrate. The recorded
excitation of TEM modes [51]. These modes have no cutoff, so light can measurements were performed over three areas of 50 µm × 50 µm
always propagate inside the slits, which at the Fabry-Perot resonance measuring 100 spectra (examples of these mappings are shown in Figs
condition makes the existence of strong peaks in the transmission 8a-8b). These kinds of measurements, i.e., SERS mappings, provide a
spectrum of a slit array possible. robust methodology for SERS analysis [55].
Fig. 6d shows the average electric field intensity enhancement in the Fig. 8c shows SERS spectra of R6G that were acquired, under iden­
gap region (black box of Fig. 6c), obtained for the region between the tical experimental conditions, for the different specimens with sym­
gap boundaries, as a function of SiD (calculated for six different gap metric Au nanogaps. In accordance with FDTD simulations,
values) at the laser wavelength used in this work, at 633 nm. This measurements on symmetrical nanogaps of the SERS signal of R6G
wavelength is indicated in Fig. 6b with a vertical dashed line. Given the exhibited a dependence on the etched silicon depth. In the case of the
evolution of the resonance as a function of SiD, we expect such a vari­ specimen with symmetric Au nanostructures with a gap of ~ 6 nm, the
ation in the averaged intensity enhancement. The results in Fig. 6d show SERS intensity increased for a larger etched silicon depth. It has to be
that the enhancement features a different maximum value per SiD, and noted that the RSD of the SERS intensities were 9%, 38%, and 11% for
strongly depends on the nanogap size G. Narrowing the gap results in symmetric Au nanogaps templated over silicon with depths of 0 nm
higher valued of the maximum local field enhancement factor ((E/E0)2), (sym1), ~ 13 nm (sym2) and ~ 40 nm (sym3), respectively. In addition,
as could be expected [23]. A parameter sweep of SiD shows the existence when the SiO2 support and the metallic Cr layer used as the adhesion
of an optimum SiD value that maximizes the local field enhancement layer remained underneath the Au nanoflaps (substrate sym4), the SERS
factor (E/E0)2. The found optimum SiD values depend on the G values intensity decreased. In this last case, the RSD was 15%. Fig. S2 shows the
and are in the range of 20 to 60 nm, for the simulations considered in this experimental and theoretical SERS enhancement: the presence of Cr
study (Fig. 6d). causes a reduction of the SERS intensity due to plasmon coupling losses
Finally, it is worth to discuss the effects of absorption in the Si sub­ introduced by the Cr [56–58]. Identical beneficial results have also been
strate and the presence of Cr in the SERS nanostructures. Fig. 7a shows reported for silver-capped silicon nanopillars upon removal of the
the EM field enhancement over the gap region as a function of etched chromium layer [59].
silicon depth for three different SERS nanostructures with a gap size of 6 In the case of asymmetric Au nanostructures, as shown in Fig. 8d,
nm. The three structures were characterized by the presence or absence reduction of the gap resulted in an increase of the SERS intensity: the
of a Cr layer and the use of the real or complex part of the dielectric asymmetric Au specimen with the smallest gap (~8 nm) exhibited a
constant of Si in the calculations. The dominant role of Si in the optical response with a significantly higher SERS intensity than the specimen
response of the system is due to its high real part value of dielectric with larger gaps. It has to be noted that the standard deviations of the
constant, as observed in Fig. 7a: Si without imaginary part of the SERS intensities recorded on asymmetric Au nanostructures with gaps of
dielectric constant (i.e., transparent material) produces a similar optical ~16 nm, ~12 nm, and ~8 nm, were 21%, 30%, and 30%, respectively.
response to that of Si with damping. The relatively low uniformity of the SERS activity for these small gaps
Absence of the Cr layer introduces an interesting dependence, a sort could be attributed to the variation of the size of the gap. For an average
of plateau for large values of the etched silicon depth. This optical gap of 8 nm, variations of 1 nm could greatly change the SERS intensity
response comes from the reduction of the absorption after removing the because the electric field amplitude shows an approximate inverse
Cr layer. Cr is considered as a poor metal in plasmonic gratings because dependence with the gap size [18].
it strongly absorbs the field components of localized and surface plas­ From averaging the mapping results, the analytical enhancement
mons [54]. Removing the Cr layer from the system in Fig. 6b results in factor (AEF) was calculated using Eq. (1) and the results are shown in
narrower and more intense resonances (Fig. 7b), particularly at lower Fig. 9. It is worth mentioning that the theoretical SERS electromagnetic
wavelengths where the absorption of Cr is higher. enhancement factor is approximately described by Eq. (2) [18]:
[ ]4
|E|
3.3. Experimental results of symmetric and asymmetric Au nanogaps on EF∝ (2)
|E0 |
nano-wedges
Where |E| is the amplitude of the enhanced electric field in the gap re
Essential requirements for SERS substrate are large-scale

Fig. 7. (a) FDTD-simulated electromagnetic

field enhancement averaged over the gap re­
gion versus etched silicon depth for G = 6 nm
for three different SERS nanostructures: (tri­
angles) the Cr layer is present and the Si layer
absorbs the light; (diamonds) the Cr layer is
removed (Si still absorbs); (hexagons); the Cr
layer is present but only the real part of the
dielectric constant of Si is considered in calcu­
lations (the substrate is transparent). (b) Elec­
tric field intensity enhancement as a function of
the wavelength, for different etched silicon
depths (SiD) (the gap, G, is set to 6 nm). The Cr
layer has been removed in the calculations. The
system is illuminated from the air side at
normal incidence. The incident electric field is
pointing perpendicular to the gap.

8
M. Lafuente et al. Applied Materials Today 35 (2023) 101929

Fig. 8. a)-b) SERS mappings (50 × 50 µm, 100 spectra) monitoring peak at 1510 cm− 1 of R6G for a) symmetric and b) asymmetric nanostructures. The scale bar
represents 10 μm. c)-d) The average SERS spectra of R6G 1 µM obtained on these SERS mappings as well as with the Raman spectrum of pure R6G in solid phase. Each
spectrum is the average of 100 spectra. For better interpretation of the results SERS spectra are displayed with an offset in the y-direction.

Fig. 9. (a) Effect of SiD on SERS enhancement: theoretical (black) and experimental (blue) SERS enhancement results of specimen with symmetric Au nanostructures
with different SiD-values: 0, 13 and 40 nm. (b) Effect of G on SERS enhancement: theoretical (black) and experimental (blue) SERS enhancement results of specimen
with asymmetric Au nanostructures with different G-values: 8, 12 and 16 nm. The error bars indicate standard deviation.

gion and |E0| is the amplitude of the incoming electromagnetic wave. known that this expression is able to predict qualitative trends of SERS
As can be seen in Fig. 9b (see also Figure S3 in SI), the asymmetric Au but not quantitative results.
nanostructures show a better SERS response than the symmetric spec­ Finally, the polarization effect has been verified experimentally from
imen (Fig. 9a), where asym3 achieved an AEF of 6.9 × 106. In addition, SERS signals in asymmetric structures (see SEM images in Fig. 10). In the
we can observe that the experimental AEF is about a factor ~102 - 103 experiments, the incoming light was polarized parallel and perpendic­
larger than the theoretical EM SERS enhancement calculated with Eq. ular to the width-direction of the Au nanostructures (Fig. 10a). The re­
(2), nevertheless, both the theoretical and experimental EM SERS sults showed that the Raman signals of R6G for two polarization
enhancement factors follow the same relative trend, which supports the directions differ by almost a factor of 5 for asym1 and 10 for asym2, by
validity of the FDTD simulations. The discrepancy between the being higher for the E0 parallel-orientation to the width of the Au
measured and calculated EM SERS enhancements arises from a varieties nanostructure (see Fig. 10b). In other words, in our experiments, the
aspect that include differences between experimental samples and the largest Raman enhancement was found for light perpendicularly ori­
theoretical models (e.g. dimensional imperfections of the samples, ented to the width gap.
slightly different dielectric constants of materials, etc.) that eventually From the experimental and theoretical approaches, we can now
result in the fact that resonances appear at different spectral locations. narrow the selection of the best performing SERS substrate. The asym­
Another important source of discrepancies is that we used an approach metric Au nanogaps are found to yield a better SERS response than the
based on a simple expression of the SERS signal (Eq. (2)). It is well symmetric nanogaps. The highest SERS enhancement of 6.9 × 106 is

9
M. Lafuente et al. Applied Materials Today 35 (2023) 101929

Fig. 10. a) Scheme indicating the direction of light polarized in parallel and perpendicular to the width-direction of the gold nanostructures or gap width. b) The
average SERS intensity of the peak at 1510 cm− 1 of R6G 1 μM recorded on samples ‘asym1’and ‘asym2’ with the incoming laser light polarized parallel and
perpendicular to the width-direction of the gold nanostructures.

achieved with specimen asym3. Funding sources

Conclusions S.G.R., R.M. and M.P.P are grateful for financial support from: the
European Union’s Horizon 2020 research and innovation program
We have fabricated symmetric and asymmetric Au gaps templated under grant agreement No. 883390 (H2020-SU-SECU-2019 SERSing
over silicon nano-wedges. The nano-wedges were realized by a combi­ Project) and R.M and M.P.P for financial support from AEI Spain
nation of displacement Talbot lithography (DTL), corner lithography (PID2019-108660RB-I00). P.M. and S.M.G.B thanks funding by the
and selective anisotropic etching of silicon. Subsequently, Au nanogaps Dutch Research Council (contract number 13328).
were formed on the nano-wedges by means of ion beam etching (IBE) of
sputter-deposited Au (Au) films. The thickness of the sputtered Au film
determined the nanogap type: Symmetric or asymmetric, and the time of Declaration of Competing Interest
IBE the width of the Au nanogap. The gap dimensions in-between the Au
nanostructures surrounding the tip of nano-wedges, and thus the hot- The authors declare that they have no known competing financial
spot formation, was highly tuneable by means of the silicon depth of interests or personal relationships that could have appeared to influence
the rhombic cage and by varying the thickness of the secondary sput­ the work reported in this paper.
tered Au layer. This was demonstrated by varying gap width from 2 to
16 nm while varying rhombic cage depth from 0 to 100 nm. This novel Data availability
fabrication protocol resulted in robust and uniform SERS specimen, with
very well controlled gaps in the range of 6 ± 2 nm to 16 nm ± 4 nm. We Data will be made available on request.
have firstly analyzed the SERS activity of these symmetric and asym­
metric Au nanostructures by FDTD simulations. In the case of symmetric
Au nanogaps, the SERS effects arise mainly at the gap between the Au Supplementary materials
nanoflaps. For both symmetric as well as asymmetric Au nanostructures
upon reduction of the gap, the enhanced electric field increases. In Supplementary material associated with this article can be found, in
addition, for asymmetric Au nanostructures the electric field enhance­ the online version, at doi:10.1016/j.apmt.2023.101929.
ment is the highest close to the Au nanostructure with the largest cross-
sectional. It shows that the Raman enhancement depends on the orien­ References
tation of the light polarization with respect to the width direction of the
Au nanogaps. Theoretically predicted designs for more active SERS [1] D.C. Harris, M.D. Bertolucci, Symmetry and spectroscopy: an Introduction to
nanostructures agree with the experimental tendencies shown by the Vibrational and Electronical Spectroscopy, Dover Publications, New York, 1978.
[2] R.A. Alvarez-Puebla, A. Agarwal, P. Manna, B.P. Khanal, P. Aldeanueva-Potel,
fabricated specimens with R6G. The exhibited AEF values are paving the E. Carbo-Argibay, N. Pazos-Perez, L. Vigderman, E.R. Zubarev, N.A. Kotov, L.
way for future developments where SERS is combined with sieving effect M. Liz-Marzan, Gold nanorods 3D-supercrystals as surface enhanced Raman
at the nanogaps. scattering spectroscopy substrates for the rapid detection of scrambled prions,
Proc. Natl Acad. Sci. 108 (2011) 8157–8161, https://doi.org/10.1073/
pnas.1016530108.
Author contributions [3] S. Shanmukh, L. Jones, J. Driskell, Y. Zhao, R. Dluhy, R.A. Tripp, Rapid and
sensitive detection of respiratory virus molecular signatures using a silver nanorod
array SERS substrate, Nano Lett. 6 (2006) 2630–2636, https://doi.org/10.1021/
M.L.: conceptualization, methodology, software, investigation, data nl061666f.
curation, writing the original draft and editing the manuscript. P.M.: [4] E. Garcia-Rico, R.A. Alvarez-Puebla, L. Guerrini, Direct surface-enhanced Raman
methodology and software. E.J.W.B.: methodology and validation. R.M. scattering (SERS) spectroscopy of nucleic acids: from fundamental studies to real-
life applications, Chem. Soc. Rev. (2018), https://doi.org/10.1039/C7CS00809K.
T.: writing – review and editing, supervision. A.S-A.: writing – review
[5] C.L. Wong, U.S. Dinish, M.S. Schmidt, M. Olivo, Non-labeling multiplex surface
and editing. S.G.R.: investigation, validation, data curation and writing. enhanced Raman scattering (SERS) detection of volatile organic compounds
L.J.K.: data curation and writing. S.M.G-B.: review and funding acqui­ (VOCs), Anal. Chim. Acta 844 (2014) 54–60, https://doi.org/10.1016/j.
aca.2014.06.043.
sition. R.M.: original idea, review, supervision and funding acquisition.
[6] Y. Zhu, H. Kuang, L. Xu, W. Ma, C. Peng, Y. Hua, L. Wang, C. Xu, Gold nanorod
M.P.P.: original idea, review, supervision and funding acquisition. N.R. assembly based approach to toxin detection by SERS, J. Mater. Chem. 22 (2012)
T.: original idea, review, supervision and funding acquisition. All au­ 2387–2391, https://doi.org/10.1039/C2JM15238J.
thors have given approval to the final version of the manuscript. [7] M. Lafuente, I. Pellejero, A. Clemente, M.A. Urbiztondo, R. Mallada, S. Reinoso, M.
P. Pina, L.M. Gandía, In situ synthesis of SERS-active Au@POM nanostructures in a
microfluidic device for real-time detection of water pollutants, ACS Appl. Mater.
Interfaces 12 (2020) 36458–36467, https://doi.org/10.1021/acsami.0c06725.

10
M. Lafuente et al. Applied Materials Today 35 (2023) 101929

[8] E. Segal, E. Haleva, A. Salomon, Ultrasensitive plasmonic sensor for detecting Sub- [32] E.J.W. Berenschot, H.V. Jansen, N.R. Tas, Fabrication of 3D fractal structures using
PPB levels of alachlor, ACS Appl. Nano Mater. 2 (2019) 1285–1293, https://doi. nanoscale anisotropic etching of single crystalline silicon, J. Micromech. Microeng.
org/10.1021/acsanm.8b02164. 23 (2013), https://doi.org/10.1088/0960-1317/23/5/055024.
[9] A. Hakonen, K. Wu, M. Stenbæk Schmidt, P.O. Andersson, A. Boisen, [33] E. Berenschot, R.M. Tiggelaar, B. Borgelink, C. van Kampen, C.S. Deenen,
T. Rindzevicius, Detecting forensic substances using commercially available SERS Y. Pordeli, H. Witteveen, H.J.G.E. Gardeniers, N.R. Tas, Self-aligned
substrates and handheld Raman spectrometers, Talanta 189 (2018) 649–652, crystallographic multiplication of nanoscale silicon wedges for high-density
https://doi.org/10.1016/j.talanta.2018.07.009. fabrication of 3D nanodevices, ACS Appl. Nano Mater. 5 (2022) 15847–15854,
[10] S.K. Islam, Y. Pak, R.L. Birke, M. Vega, C. Muehlethaler, J.R. Lombardi, An analysis https://doi.org/10.1021/acsanm.2c04079.
of tetrahydrocannabinol (THC) and its analogs using surface enhanced Raman [34] M. Chirumamilla, A. Toma, A. Gopalakrishnan, G. Das, R.P. Zaccaria, R. Krahne,
Scattering (SERS), Chem. Phys. 536 (2020), 110812, https://doi.org/10.1016/j. E. Rondanina, M. Leoncini, C. Liberale, F. de Angelis, E. di Fabrizio, 3D nanostar
chemphys.2020.110812. dimers with a sub-10-nm gap for single- /few-molecule surface-enhanced Raman
[11] M. Lafuente, D. Sanz, M.A. Urbiztondo, J. Santamaría, M.P. Pina, R. Mallada, Gas scattering, Adv. Mater. 26 (2014) 2353–2358, https://doi.org/10.1002/
phase detection of chemical warfare agents CWAs with portable Raman, J. Hazard. adma.201304553.
Mater. 384 (2020), 121279, https://doi.org/10.1016/j.jhazmat.2019.121279. [35] M. Tabatabaei, M. Najiminaini, K. Davieau, B. Kaminska, M.R. Singh, J.J.L. Carson,
[12] M. Lafuente, E.J.W. Berenschot, R.M. Tiggelaar, R. Mallada, N.R. Tas, M.P. Pina, F. Lagugné-Labarthet, Tunable 3D plasmonic cavity nanosensors for surface-
3D fractals as SERS active platforms: preparation and evaluation for gas phase enhanced Raman spectroscopy with sub-femtomolar limit of detection, ACS
detection of G-nerve agents, Micromachines (Basel) 9 (2018), https://doi.org/ Photonics 2 (2015) 752–759, https://doi.org/10.1021/acsphotonics.5b00104.
10.3390/mi9020060. [36] A. Gopalakrishnan, M. Chirumamilla, F. de Angelis, A. Toma, R. Proietti Zaccaria,
[13] E.Le Ru, P.G. Etchegoin, Principles of Surface-Enhanced Raman Spectroscopy and R. Krahne, Bimetallic 3D nanostar dimers in ring cavities: recyclable and robust
Related Plasmonic Effects, 2009th ed., Elsevier, 2009. surface-enhanced Raman scattering substrates for signal detection from few
[14] A. Shiohara, Y. Wang, L.M. Liz-marzán, Recent approaches toward creation of hot molecules, ACS Nano 8 (2014) 7986–7994, https://doi.org/10.1021/nn5020038.
spots for SERS detection, J. Photochem. Photobiol., C: Photochem. Rev. 21 (2014) [37] H.I. Smith, R. Menon, A. Patel, D. Chao, M. Walsh, G. Barbastathis, Zone-plate-
2–25, https://doi.org/10.1016/j.jphotochemrev.2014.09.001. array lithography: a low-cost complement or competitor to scanning-electron-beam
[15] M. Fan, G.F.S. Andrade, A.G. Brolo, A review on recent advances in the lithography, Microelectron. Eng. 83 (2006) 956–961, https://doi.org/10.1016/j.
applications of surface-enhanced Raman scattering in analytical chemistry, Anal. mee.2006.01.226.
Chim. Acta 1097 (2020) 1–29, https://doi.org/10.1016/j.aca.2019.11.049. [38] X. Zhang, W.J. Salcedo, M.M. Rahman, A.G. Brolo, Surface-enhanced Raman
[16] A. Susarrey-Arce, K.M. Czajkowski, I. Darmadi, S. Nilsson, I. Tanyeli, S. Alekseeva, scattering from bowtie nanoaperture arrays, Surf. Sci. 676 (2018) 39–45, https://
T.J. Antosiewicz, C. Langhammer, A nanofabricated plasmonic core-shell- doi.org/10.1016/j.susc.2018.02.003.
nanoparticle library, Nanoscale 11 (2019) 21207–21217, https://doi.org/ [39] A. Ahmed, R. Gordon, Single molecule directivity enhanced Raman scattering
10.1039/c9nr08097j. using nanoantennas, Nano Lett. 12 (2012) 2625–2630, https://doi.org/10.1021/
[17] M. Lafuente, I. Pellejero, V. Sebastián, M.A. Urbiztondo, R. Mallada, M.P. Pina, nl301029e.
J. Santamaría, Highly sensitive SERS quantification of organophosphorous [40] Q. Tao, S. Li, C. Ma, K. Liu, Q.Y. Zhang, A highly sensitive and recyclable SERS
chemical warfare agents: a major step towards the real time sensing in the gas substrate based on Ag-nanoparticle-decorated ZnO nanoflowers in ordered arrays,
phase, Sens. Actuat. B Chem. 267 (2018) 457–466, https://doi.org/10.1016/j. Dalton Trans. 44 (2015) 3447–3453, https://doi.org/10.1039/C4DT03596H.
snb.2018.04.058. [41] Q. Zhang, Y.H. Lee, I.Y. Phang, C.K. Lee, X.Y. Ling, Hierarchical 3D SERS substrates
[18] P.L. Stiles, J.A. Dieringer, N.C. Shah, R.P. Van Duyne, Surface-enhanced Raman fabricated by integrating photolithographic microstructures and self-assembly of
spectroscopy, Annu. Rev. Anal. Chem. 1 (2008) 601–626, https://doi.org/ silver nanoparticles, Small 10 (2014) 2703–2711, https://doi.org/10.1002/
10.1146/annurev.anchem.1.031207.112814. smll.201303773.
[19] M. Lafuente, S. Ruiz-Rincón, R. Mallada, P. Cea, M.Pilar Pina, Towards the [42] S. Yan, F. Chu, H. Zhang, Y. Yuan, Y. Huang, A. Liu, S. Wang, W. Li, S. Li, W. Wen,
reproducible fabrication of homogeneous SERS substrates by Langmuir-Schaefer Rapid, one-step preparation of SERS substrate in microfluidic channel for detection
technique: a low cost and scalable approach for practical SERS based sensing of molecules and heavy metal ions, Spectrochim. Acta A Mol. Biomol. Spectrosc.
applications, Appl. Surf. Sci. (2019), 144663, https://doi.org/10.1016/j. 220 (2019), 117113, https://doi.org/10.1016/j.saa.2019.05.018.
apsusc.2019.144663. [43] M.S. Schmidt, J. Hübner, A. Boisen, Large area fabrication of leaning silicon
[20] J. Lee, B. Hua, S. Park, M. Ha, Y. Lee, Z. Fan, H. Ko, Tailoring surface plasmons of nanopillars for surface enhanced Raman spectroscopy, Adv. Mater. 24 (2012)
high-density gold nanostar assemblies on metal films for surface-enhanced Raman 11–18, https://doi.org/10.1002/adma.201103496.
spectroscopy, Nanoscale 6 (2014) 616–623, https://doi.org/10.1039/c3nr04752k. [44] H. Le-The, E. Berenschot, R.M. Tiggelaar, N.R. Tas, A. van den Berg, J.C.T. Eijkel,
[21] R. Tantra, R.J.C. Brown, M.J. Milton, Strategy to improve the reproducibility of Shrinkage control of photoresist for large-area fabrication of sub-30nm periodic
colloidal SERS, J. Raman Spectrosc. 38 (2007) 1469–1479, https://doi.org/ nanocolumns, Adv. Mater. Technol. 2 (2017), https://doi.org/10.1002/
10.1002/jrs.1797. admt.201600238.
[22] J. Chen, L. Guo, B. Qiu, Z. Lin, T. Wang, Application of ordered nanoparticle self- [45] H. Le-The, R.M. Tiggelaar, E. Berenschot, A. Van Den Berg, N. Tas, J.C.T. Eijkel,
assemblies in surface-enhanced spectroscopy, Mater. Chem. Front. (2018) Postdeposition UV-ozone treatment: an enabling technique to enhance the direct
835–860, https://doi.org/10.1039/C7QM00557A. adhesion of gold thin films to oxidized silicon, ACS Nano 13 (2019) 6782–6789,
[23] H. Le-The, J.J.A. Lozeman, M. Lafuente, P. Muñoz, J.G. Bomer, H. Duy-Tong, https://doi.org/10.1021/acsnano.9b01403.
E. Berenschot, A. van den Berg, N.R. Tas, M. Odijk, J.C.T. Eijkel, Wafer-scale [46] J.G.E. Wilbers, J.W. Berenschot, R.M. Tiggelaar, T. Dogan, K. Sugimura, W.G. Van
fabrication of high-quality tunable gold nanogap arrays for surface-enhanced Der Wiel, J.G.E. Gardeniers, N.R. Tas, 3D-fabrication of tunable and high-density
Raman scattering, Nanoscale 11 (2019) 12152–12160, https://doi.org/10.1039/ arrays of crystalline silicon nanostructures, J. Micromech. Microeng. 28 (2018),
C9NR02215E. https://doi.org/10.1088/1361-6439/aaab2d.
[24] A. Habermehl, P. Brenner, R. Huber, A. Mertens, F. Winkler, L. Hahn, [47] E.D. Palik, Handbook of Optical Constants of Solids, Academic press, 1998. Volume
M. Guttmann, C. Eschenbaum, U. Lemmer, Roll-to-roll hot embossing of 1D and 2D III.
photonic nanostructures, Adv. Eng. Mater. 21 (2019), 1900110, https://doi.org/ [48] W.M. Haynes, CRC handbook of chemistry and physics, 95th Editi, 2014.
10.1002/adem.201900110. [49] E.C. Le Ru, E. Blackie, M. Meyer, P.G. Etchegoint, Surface enhanced Raman
[25] T. Jiang, P. Goel, H. Zhao, R. Ma, L. Zhu, X. Liu, Internal structure tailoring in 3D scattering enhancement factors: a comprehensive study, J. Phys. Chem. C 111
nanoplasmonic metasurface for surface-enhanced Raman spectroscopy, Particle (2007) 13794–13803, https://doi.org/10.1021/jp0687908.
Particle Syst. Characteriz. 37 (2020) 1900345. doi:10.1002/ppsc.201900345. [50] R.B. Marcus, T.T. Sheng, The oxidation of shaped silicon surfaces, J. Electrochem.
[26] D. Jonker, Z. Jafari, J.P. Winczewski, C. Eyovge, J.W. Berenschot, N.R. Tas, J.G. Soc. 129 (1982) 1278–1282, https://doi.org/10.1149/1.2124118.
E. Gardeniers, I. De Leon, A. Susarrey-Arce, A wafer-scale fabrication method for [51] F.J. Garcia-Vidal, L. Martin-Moreno, T.W. Ebbesen, L. Kuipers, Light passing
three-dimensional plasmonic hollow nanopillars, Nanoscale Adv. 3 (2021) through subwavelength apertures, Rev. Mod. Phys. 82 (2010) 789, https://doi.org/
4926–4939, https://doi.org/10.1039/d1na00316j. 10.1103/RevModPhys.82.729.
[27] M. Kahl, E. Voges, S. Kostrewa, C. V, W. Hill, Periodically structured metallic [52] Y. Pang, C. Genet, T.W. Ebbesen, Optical transmission through subwavelength slit
substrates for SERS, Sens. Actuat. B Chem. 51 (1998) 285–291, https://doi.org/ apertures in metallic films, Opt. Commun. 280 (2007) 10–15, https://doi.org/
10.1016/S0925-4005(98)00219-6. 10.1016/j.optcom.2007.07.063.
[28] C.-W. Chang, J.-D. Liao, A.-L. Shiau, C.-K. Yao, Non-labeled virus detection using [53] S.G. Rodrigo, L. Martín-Moreno, A.Yu. Nikitin, A.V. Kats, I.S. Spevak, F.J. Garcia-
inverted triangular Au nano-cavities arrayed as SERS-active substrate, Sens. Vidal, Extraordinary optical transmission through hole arrays in optically thin
Actuat. B Chem. 156 (2011) 471–478, https://doi.org/10.1016/j.snb.2011.04.006. metal films, Opt. Lett. 34 (2009) 4–6.
[29] H.H. Solak, C. Dais, F. Clube, Displacement Talbot lithography: a new method for [54] S.G. Rodrigo, F.J. García-Vidal, L. Martín-Moreno, Influence of material properties
high-resolution patterning of large areas, Opt. Express 19 (2011) 10686, https:// on extraordinary optical transmission through hole arrays, Phys. Rev. B Condens.
doi.org/10.1364/oe.19.010686. Matter Mater. Phys. 77 (2008) 1–8, https://doi.org/10.1103/
[30] D. Jonker, E.J.W. Berenschot, N.R. Tas, R.M. Tiggelaar, A. Van Houselt, H.J.G. PhysRevB.77.075401.
E. Gardeniers, Large dense periodic arrays of vertically aligned sharp silicon [55] J. Langer, D. Jimenez de Aberasturi, J. Aizpurua, R.A. Alvarez-Puebla, B. Auguié, J.
nanocones, Nanoscale Res. Lett. 17 (2022) 100, https://doi.org/10.1186/s11671- J. Baumberg, G.C. Bazan, S.E.J. Bell, A. Boisen, A.G. Brolo, J. Choo, D. Cialla-May,
022-03735-y. V. Deckert, L. Fabris, K. Faulds, F.J. García de Abajo, R. Goodacre, D. Graham, A.
[31] D. Jonker, L. Kooijman, Y. Pordeli, B. van der Wel, E. Berenschot, B. Borgelink, J. Haes, C.L. Haynes, C. Huck, T. Itoh, M. Käll, J. Kneipp, N.A. Kotov, H. Kuang, E.
H. Le-The, M. de Boer, J. Eijkel, R. Hueting, R. Tiggelaar, A. van Houselt, C. Le Ru, H.K. Lee, J.F. Li, X.Y. Ling, S.A. Maier, T. Mayerhöfer, M. Moskovits,
H. Gardeniers, N. Tas, Wafer-scale fabrication and modification of silicon nano- K. Murakoshi, J.M. Nam, S. Nie, Y. Ozaki, I. Pastoriza-Santos, J. Perez-Juste,
pillar arrays for nanoelectronics, nanofluidics and beyond, Int. J. Nanotechnol. 17 J. Popp, A. Pucci, S. Reich, B. Ren, G.C. Schatz, T. Shegai, S. Schlücker, L.L. Tay, K.
(2020) 583–606, https://doi.org/10.1504/IJNT.2020.111322. G. Thomas, Z.Q. Tian, R.P. Van Duyne, T. Vo-Dinh, Y. Wang, K.A. Willets, C. Xu,

11
M. Lafuente et al. Applied Materials Today 35 (2023) 101929

H. Xu, Y. Xu, Y.S. Yamamoto, B. Zhao, L.M. Liz-Marzán, Present and future of [58] D.T. Debu, P.K. Ghosh, D. French, J.B. Herzog, Surface plasmon damping effects
surface-enhanced Raman scattering, ACS Nano 14 (2019) 28–117, https://doi.org/ due to Ti adhesion layer in individual gold nanodisks, Opt Mater. Express 7 (2017)
10.1021/acsnano.9b04224. 73–84, https://doi.org/10.1364/OME.7.000073.
[56] J. Liu, H. Cai, L. Kong, X. Zhu, Effect of chromium interlayer thickness on optical [59] K. Wu, T. Rindzevicius, M. Stenbæk Schmidt, K.B. Mogensen, A. Hakonen,
properties of Au-Ag nanoparticle array, J. Nanomater. 2014 (2014), 650359, A. Boisen, Wafer-scale leaning silver nanopillars for molecular detection at ultra-
https://doi.org/10.1155/2014/650359. low concentrations, J. Phys. Chem. C 119 (2015) 2053–2062, https://doi.org/
[57] C. Chow, J.A. Bain, Effect of thin Cr and Cu adhesion layers on surface plasmon 10.1021/jp510073y.
resonance at Au/SiO2 interfaces, IEEE Trans. Magn. 52 (2016) 1–4, https://doi.
org/10.1109/TMAG.2015.2509252.

12