We deposited Au-Cu-Si, an Au-based thin-film metallic glass (TFMG) of ∼50 nm thickness, as the activation layer for propagating surface plasmon resonance (PSPR)-based sensors on a BK7 glass substrate to substitute the commonly used gold layer. The film composition was tuned to yield the maximum Au content (∼65 at %), while the structure remained amorphous. The results showed that the Au-based TFMG could support surface plasmon resonance and gave rise to the extinction in the angle-resolved reflection spectrum. Using deionized water and ethyl alcohol with the refractive index difference of ∼0.03 as the analytes, the angle shift given by Au-based TFMG was 4° compared to 5° given by the Au film. Hence, Au-based TFMG is feasible to be used as the activation layer in PSPR-based sensors. Compared to the Au film, Au-based TFMG has the advantages of being less expensive, lacking grain boundary scattering, better adhesion to the substrate, and higher resistance to scratch and corrosion because of its amorphous structure with excellent mechanical properties.
We deposited Au-Cu-Si, an Au-based thin-film metallic glass (TFMG) of ∼50 nm thickness, as the activation layer for propagating surface plasmon resonance (PSPR)-based sensors on a BK7 glass substrate to substitute the commonly used gold layer. The film composition was tuned to yield the maximum Au content (∼65 at %), while the structure remained amorphous. The results showed that the Au-based TFMG could support surface plasmon resonance and gave rise to the extinction in the angle-resolved reflection spectrum. Using deionized water and ethyl alcohol with the refractive index difference of ∼0.03 as the analytes, the angle shift given by Au-based TFMG was 4° compared to 5° given by the Au film. Hence, Au-based TFMG is feasible to be used as the activation layer in PSPR-based sensors. Compared to the Au film, Au-based TFMG has the advantages of being less expensive, lacking grain boundary scattering, better adhesion to the substrate, and higher resistance to scratch and corrosion because of its amorphous structure with excellent mechanical properties.
Surface plasmon resonance
(SPR) is a concept of nonradiative electromagnetic
(EM) surface wave for macroscopic interfaces which is related to the
collective oscillation of conduction electrons on the metal surface
due to the excitation of the incident light at the metal–dielectric
interface.[1−3] Comprehensive reviews on SPR[4,5] and
its applications for optical biosensors[6−9] have been performed recently. The ordered
array of metal nanoparticles resulting in high-quality factors of
plasmonic surface lattice resonances has also been reviewed,[10,11] and the recent state-of-art achievements in the field of plasmonic
biosensing-based terahertz spectroscopy using toroidal metadevices
has been reviewed by Ahmadivand et al.[12,13] In addition,
comprehensive coverage of recent design and development, including
processing and fabrication, of 2D materials in the context of plasmonic-based
devices has also been provided.[14] The SPR
modes can be generally divided into two types, propagating surface
plasmon resonance (PSPR) and localized surface plasmon resonance (LSPR).[15] PSPR, also known as surface plasmon polariton,
is induced by evanescent EM waves at the planar metal–dielectric
interfaces. Different from PSPR, LSPR occurs at the surface of metallic
nanostructures due to the confinement of EM waves.[16−19] Both PSPR- and LSPR-based sensors
can detect the change in the refractive index of the dielectric medium.
Compared to the LSPR-based sensor, the PSPR-based sensor has a higher
refractive index sensitivity (∼2 × 106 vs ∼2
× 102 nm RIU–1) and a longer EM
field decay length (200–300 vs 5–15 nm).[20] However, combining the two abovementioned factors,
both the PSPR-based sensor and LSPR-based sensor are very competitive
in their sensitivities.[20]The performance
of the SPR sensor can be commonly determined in
two parts. On the one hand, the sensitivity is a concern and the shift
of resonance angle should be large even in the case of a small change
in the refractive index of the dielectric medium. On the other hand,
the accuracy is the concern and the smaller full width at half maximum
of the corresponding SPR spectrum represents the higher accuracy.
A comprehensive list of comparisons of the performance of biosensors
in terms of limit of detection and sensitivity among different optical
sensing schemes could be found elsewhere.[6] For both PSPR and LSPR, plasmonic material is the basic to excite
plasmons and it is required to have a negative real component and
a small, positive imaginary component of the dielectric constant.[1,21−23] Thus, noble metals, such as gold and silver have
been commonly used as the substrates of SPR-based sensors. Compared
with silver, gold has an inert nature with higher chemical stability
and direct functionalization[6] with excellent
sensitivity to the change of refractive index of the dielectric medium.Metallic glasses (MGs) are obtained by rapid quenching from the
melt to the solid amorphous state.[24,25] Different
from the ordered atomic arrangement of crystalline alloys, MGs have
a liquid-like random atomic arrangement due to the rapid quenching
process. Thus, crystalline defects, such as dislocations, grain boundaries,
and voids, would not appear. Due to the absence of crystalline defects,
MGs have excellent mechanical properties and have been studied extensively
for engineering applications. However, the studies of the optical
properties of MGs have been sparse. In the absence of grain boundaries
in amorphous MGs, damping due to scattering at grain boundaries would
not occur and would benefit the plasmonic applications.[26−29] Specifically, the effects of grain size of crystalline materials
on surface-enhanced Raman scattering (SERS) were first studied by
Dawson et al.,[26] and it was found that
the SERS signals obtained from the slow-deposited Ag films with the
larger grain sizes were higher than those obtained from the fast-deposited
Ag films with the smaller grain sizes. Since then, the concept of
elastic scattering of surface-plasmon polaritons at grain boundaries
was proposed, which would result in the increasing internal damping
and decreasing SERS intensity.[27,28]Thin-film metallic
glasses (TFMGs) are a new group of MGs which
have been commonly fabricated by physical vapor deposition (PVD).[30−35] With the process of vapor-to-solid by PVD, TFMGs are subjected to
a much faster cooling rate compared to the process of liquid-to-solid
by casting for bulk MGs. As a result, TFMGs not only possess the unique
properties of MGs but also have a better glass-forming ability with
a wider composition range.[36,37] Moreover, TFMGs can
exhibit higher fracture resistance compared with their bulk MG counterparts
because of the size effect.[37,38] Previously, we have
confirmed the Au-based TFMGs as a plasmonic material, used nanoimprint
to fabricate periodic nanostructures by taking advantage of the viscous
flow behavior of TFMGs in the supercooled liquid region, and proved
the feasibility of using Au-based TFMGs as LSPR-based sensor.[39] In addition, we have studied the effects of
crystallinity of Au-based TFMGs on SERS recently.[29] Specifically, the as-deposited Au-based TFMGs were imprinted
in the supercooled liquid region to develop the periodic nanostructures.
The imprinted amorphous films were then heat-treated at different
temperatures to obtain different crystallinities. Using crystal violet
as the analyte, the Raman spectra were measured and the intensity
showed an initial quick decrease and then a slow increase with the
increasing heat treatment temperature. In this case, the Raman intensity
decreased upon the initial crystallization of heat-treated amorphous
film due to scattering at grain boundaries, and the grain growth at
higher heat-treated temperatures resulted in the reduction of grain
boundary area and the Raman intensity increased.[29] In the present study, we utilized the Au-based TFMGs as
the activation layer in the PSPR-based sensor to substitute the traditional
gold layer to reduce the cost. Also, compared to pure Au, the absence
of crystalline defects and better adhesion to the substrate would
result in the absence of grain boundary scattering and excellent scratch/corrosion
resistance of Au-based TFMGs.
Experimental Procedure
The schematic illustration of the fabrication process of Au-based
TFMGs for PSPR-based sensor applications is shown in Scheme .
Scheme 1
Schematic Illustration
of the Fabrication of Au-Based TFMGs for PSPR-Based
Sensor Applications
Sputtering
The Au-based TFMGs were
deposited on BK7 glass substrates by magnetron co-sputtering from
an Au target and a Cu30Si70 target. While both
Au and Cu are plasmonic materials, Si with a small atomic size was
added to facilitate the formation of an amorphous structure. The purities
of both targets were over 99.99 wt %, and the distance between the
substrate and the target was fixed at 10 cm. The background pressure
of the system was lower than 5 × 10–5 Pa, and
the working pressure was maintained by Ar gas at 0.3 Pa. While the
radio frequency power (13.56 MHz) was set at 55 W for the Au target,
the direct current power was set at 25 W for the Cu30Si70 target. To keep the uniform distribution of elements in
the film, the substrate holder with a diameter of 10.16 cm was rotated
at a speed of 60 rpm. It is worth noting that the effects of Ar working
pressure, ranging from 0.4 to 10 Pa, on the growth mode of Au-based
TFMGs were studied by Denis et al.,[34] and
the surface morphology changed from homogeneous at low working pressures
to nanostructured morphology at high working pressures.
Characterization
To measure the compositions
of Au-based TFMG, an electron probe X-ray microanalyzer (EPMA, JEOL,
JXA-8200) with an accelerating voltage of 10 kV and an electron beam
size of 10 μm was used, and the data were collected randomly
from seven points. The structure of the sample was investigated by
X-ray diffraction (XRD, Rigaku, TTRAX 3) using Cu-Kα radiation
(λ = 0.14506 nm) with a grazing incident angle
of 0.7° and 2θ scanned from 20 to 60° at a scanning
rate of 4°/min. The real part, εr, and the imaginary
part, εi, of the dielectric constants and the reflectance
were measured by a spectroscopic ellipsometer (J.A. Woolam Co, M-2000
Ellipsometer) to check the feasibility of Au-based TFMG as PSPR-based
activation layer and to investigate the angle shift of PSPR when different
dielectric media were used as analytes.The different dielectric
properties and crystallinities with the different contents of Au in
Au-based films were systematically studied in our previous works.[39,40] As the content of Au increased in the Au-based film, the dielectric
constant would become closer to that of Au.[39] However, the degree of crystallinity would simultaneously increase
with the increasing Au content in the Au-based film, and the maximum
Au content could be achieved with the amorphous structure was ∼65
at %.[40] In this study, the compositions
were measured by EPMA and the sample was named by the atomic percentage
of Au in the film. To obtain the maximum Au content and simultaneously
control the structure in the amorphous state, Au65 TFMG, for which
Au, Cu, and Si contents were 65.5 ± 0.2, 17.3 ± 0.2, and
17.2 ± 0.1 at %, respectively, was chosen as the activation layer
in the PSPR-based sensor.
PSPR Measurements
The selection of
material for the glass prism used in PSPR measurements has been discussed
by Singh et al.[41] and it is usually based
on the resonance wavelength. Specifically, BK7, SF2, and SF10 have
been selected for visible light-PSPR; sapphire and quartz prisms are
used for UV–PSPR; and fluoride glass prism is suitable for
NIR–PSPR. Because the resonance wavelength in our work was
in the range of visible light, BK7 glass was selected for the present
study. Specifically, the BK7 glass prism purchased from Precision
Systems Industrial Ltd. with a width of 20 mm was used in the present
study. The surface of the prism in contact with the substrate was
fully covered by the substrate during the PSPR measurements. Prior
to each PSPR measurement, the sensor surface was cleaned by rinsing
in acetone, alcohol, and deionized water each for at least 1 min in
sequence and then heat-dried to remove the residual liquid. A comparative
analysis of using different prisms was not performed in the present
study. However, it is worth noting that the effects of using different
prisms (SiO2, BaF2, CsF, and BK7), different
plasmonic metals (Au, Ag, Cu, and Al), and doping in the intermediate
layer on the sensitivity improvement of the PSPR sensor was investigated
recently by Kumar et al.[42]
Results and Discussion
Structure
The
surface morphologies
of Au-based TFMGs scanned by the Berkovich nanoindenter with a scanned
area of 20 μm × 20 μm at various temperatures were
obtained in our previous work.[40] The surface
of the as-deposited film was homogeneous and smooth at room temperature,
which agreed with Denis et al.’s results of sputtering at low
working pressures,[34] and became rough and
granular when the temperature reached the crystallization temperature.[40] The cross-sectional SEM image of Au65 TFMG deposited
on the BK7 glass substrate is shown in Figure , and the thickness of Au65 TFMG was about
50 nm.
Figure 1
SEM image showing the cross-section of Au65 TFMG on the BK7 glass
substrate.
SEM image showing the cross-section of Au65 TFMG on the BK7 glass
substrate.The XRD spectra of Au and Au65
TFMG are shown in Figure . The spectrum for Au showed
the crystalline peaks at 38.2 and 44.4° in 2θ and were
identified to be that of the (111) and (200) lattice planes, respectively,
of the face-centered cubic Au structure (JCPDS 04-0784). For Au65
TFMG, the broad peak around 40.5° and a lack of crystalline peaks
supported the amorphous structure of the film. Upon annealing of Au65
TFMG at 200 °C for 10 min, a sharp crystalline peak appeared
at 40.5° (not shown), which could be identified as the (400)
lattice plane of simple cubic AuCu (JCPDS 38-0741).
Figure 2
XRD spectra of Au and
Au65 TFMG (scanning rate of 4°/min).
XRD spectra of Au and
Au65 TFMG (scanning rate of 4°/min).
Dielectric Properties
The dielectric
constants of Au65 TFMG were measured by an ellipsometer in the wavelength
range of 370–1200 nm, and the results are shown in Figure . For comparison,
the dielectric constants of pure Au and Pt (data taken from Palik’s
handbook) are also included in Figure . The value of εr of Au65 TFMG was
similar to that of pure Au in the visible light range (400–700
nm), as shown in Figure a. The value of εi of Au65 TFMG, shown in Figure b, was higher than
that of pure Au representing the higher energy loss and lower sensitivity;
however, it was sufficiently low in the visible light range. Specifically,
it was lower than Pt, one of the plasmonic materials. Hence, the dielectric
constants shown in Figure supported the feasibility of Au65 TFMG as a plasmonic material
in the visible light range. However, it was not suitable to be used
in the infrared region for the wavelength modulation spectroscopy.
Figure 3
(a) Real
part, εr, and (b) imaginary part, εi, of the dielectric constants of Au65 TFMG and pure Au film
measured by an ellipsometer. The data for Pt were taken from Palik’s
handbook.
(a) Real
part, εr, and (b) imaginary part, εi, of the dielectric constants of Au65 TFMG and pure Au film
measured by an ellipsometer. The data for Pt were taken from Palik’s
handbook.
PSPR
Measurements
The instrument
set-up for PSPR measurements is schematically shown in Figure . The Au film and Au65 TFMG
with a thickness of ∼50 nm, which has been commonly adopted
in PSPR measurements, were deposited, respectively, on a BK7 (refractive
index n = 1.5168) glass substrate. While the uniform
and continuous films could not be achieved for the thinner films,
too much adsorption would occur for the thicker films. The PSPR substrate
was attached to a BK7 prism with index-matching oil (n = 1.5150 ± 0.0002) to eliminate the reflection/refraction losses
associated with the interface and to keep intimate contact between
the BK7 prism and BK7 substrate in order to avoid the air gap in between.
Deionized water (H2O with polar covalent hydrogen-oxygen
bonds) and ethyl alcohol (CH3–CH2–OH
with nonpolar covalent carbon-hydrogen bonds and polar covalent carbon-oxygen
and hydrogen-oxygen bonds) with refractive indexes of 1.333 and 1.361,
respectively, were used as analytes in the PSPR measurements. Polarized
incident light with its wavelength in the range of 370 to 1690 nm
was illuminated through the prism onto the film to select the most
efficient wavelength for conducting PSPR measurements. The prism,
film, and container were fixed on the stage, and the reflectance was
measured by a spectroscopic ellipsometer with an angle of incidence
ranging from 65 to 85°.
Figure 4
Schematic diagram of instrument set-up for angle-resolved
PSPR-based
sensors.
Schematic diagram of instrument set-up for angle-resolved
PSPR-based
sensors.The phenomenon of PSPR is schematically
shown in Figure .
When the incident angle reaches
a critical value (θSPR), the condition of resonance
is fulfilled and a sharp dip of optical reflectance would appear.
This resonance angle (θSPR) depends on the refractive
index of the dielectric medium, and the change of refractive index
near the metal/dielectric interface would result in a shift of θSPR; that is, ΔθSPR, to be detected.[43] Thus, PSPR can be used to detect the change
of refractive index at the interface and further be applied as high
sensitivity sensor in the field of optics.[1,44,45]
Figure 5
Schematic diagrams showing no plasmon excitation
and plasmon excitation
phenomena of PSPR.
Schematic diagrams showing no plasmon excitation
and plasmon excitation
phenomena of PSPR.Using deionized water
and ethyl alcohol as the analytes in the
PSPR measurements, the reflectance intensity versus incident angle
curves are shown in Figure (in which each curve is the average of 50 measurements) for
Au film and Au65 TFMG, respectively, as the activation layers. With
a difference in the refractive index of ∼0.03 between deionized
water and ethyl alcohol and an incident wavelength of 850 nm, the
angle shifts, ΔθSPR, for Au thin film and Au65
TFMG were 5 and 4°, respectively, when the analyte was switched
from deionized water to ethyl alcohol. Therefore, the Au65 TFMG could
be regarded as a potential candidate for PSPR-based sensors. In addition,
while the gold layer was prone to damage and detachment after repeated
usage and cleaning, the Au-based TFMG showed good reusability because
of the improved adhesion to the substrate. Also, because of the amorphous
nature, the Au-based TFMG has excellent corrosion resistance and could
be applied to a wider range of analytes compared to the gold layer.
However, the service life of Au-based TFMG was not studied in the
present work, and we hope to study the durability issue in the future.
Also, while deionized water and ethyl alcohol with a refractive index
difference of ∼0.03 were used as the analytes in the present
work, it would be more meaningful if Au-based TFMG could be used for
more practical testing (such as biomass testing), and we hope to perform
this study in the future work. It is worth noting that the prism structure
was used in the present study for sensing. This structure has a number
of disadvantages, including a bulky structure and moving components.
On the other hand, fiber structure has been developed recently[6] and it has the advantages of low EM interference,
small sizes, multiplexing, and remote sensing capabilities, and it
could be preferable over the prism structure.[6]
Figure 6
Resonance
angles (θSPR) for deionized water and
ethyl alcohol with a small refractive index difference of ∼0.03
were measured to verify that Au65 TFMG is a potential candidate for
PSPR-based sensors.
Resonance
angles (θSPR) for deionized water and
ethyl alcohol with a small refractive index difference of ∼0.03
were measured to verify that Au65 TFMG is a potential candidate for
PSPR-based sensors.
Conclusions
In conclusion, the feasibility of using the amorphous Au65 TFMG
fabricated by magnetron co-sputtering as the activation layer in the
PSPR-based sensor was verified in the present study. Compared to the
traditional activation layer of Au film, Au65 TFMG had slightly inferior
dielectric properties (Figure ) to be used as the plasmonic material because of the lower
gold content in the film. As a result, the sensitivity of Au65 TFMG
as a PSPR-based sensor was slightly lower than that of the Au film.
Nevertheless, using water (n = 1.333) and ethyl alcohol
(n = 1.361) with the refractive index difference
of ∼0.03 as the analytes, the angle shift given by Au film
was 5°, while it was only one degree less for Au65 TFMG (Figure ). However, compared
to Au film, Au65 TFMG had the advantages of being less expensive,
lacking grain boundary scattering, better adhesion to the substrate,
and higher resistance to scratch and corrosion[30] because of its amorphous structure. The characteristics
of high scratch and corrosion resistances of Au65 TFMG made it suitable
for tip-enhanced Raman scattering (TERS) applications, while the traditional
TERS tips of Au and Ag had an extremely short lifetime because of
low hardness and poor adhesion.[46] Our work
is hoped to trigger more advanced studies on Au-based TFMGs for the
potential applications of sensors.
Authors: Tao Li; Kaiyu Wu; Tomas Rindzevicius; Zhongli Wang; Lars Schulte; Michael S Schmidt; Anja Boisen; Sokol Ndoni Journal: ACS Appl Mater Interfaces Date: 2016-06-09 Impact factor: 9.229