Dongxu
Wang
a,
Kai-Da
Xu
*abc,
Siyuan
Luo
d,
Yuqing
Cui
e,
Liuyang
Zhang
e and
Jianlei
Cui
e
aSchool of Information and Communications Engineering, Xi'an Jiaotong University, Xi'an 710049, China. E-mail: kaidaxu@ieee.org
bState Key Laboratory of Surface Physics and Department of Physics, Fudan University, Shanghai 200433, China
cState Key Laboratory of Precision Spectroscopy, East China Normal University, Shanghai 200241, China
dMicrosystem and Terahertz Research Center, China Academy of Engineering Physics, Chengdu 610200, China
eState Key Laboratory for Manufacturing Systems Engineering, Xi'an Jiaotong University, Xi'an 710049, China
First published on 9th January 2023
In this paper, a dual-band metamaterial absorber in the terahertz frequencies is proposed and its refractive index sensing characteristics is analyzed. The metamaterial structure is designed using a square metal ring with four T-shaped strips loaded outside of the ring, where the metal periodic array is on top of a silicon wafer backed with a metal ground plane. The resonant frequencies of the absorber are at 0.89 and 1.36 THz, whose absorption rates are both over 99% under normal TE and TM polarized incidences. The full widths at half maximum of them are 4.4 and 11.2 GHz, respectively, resulting in high quality factors (Q-factors) for these two frequency bands. The absorption rate of the absorber remains stable as the incident and polarized angles are changed. Several proposed metamaterial absorbers are experimentally fabricated and electron beam lithography (EBL) technology is employed. Good measurement results of the dual-band absorption performance are obtained using a terahertz time-domain spectroscopy system based on photoconductive antennas. Furthermore, the metamaterial absorber also shows sensing properties for analytes with different refractive indices or thicknesses. This work provides a new choice for the design of high-Q dual-band terahertz metamaterial absorbers and their application to refractive index sensing.
As one of the most popular applications for terahertz waves, terahertz absorbers have recently become a hot research topic. Since the absorber with two metamaterial resonators coupled separately was designed by Landy et al. in 2008,12 metamaterial absorbers have been extensively investigated. Diverse metamaterial structures have been discussed to improve the performance in the terahertz frequency band, including circular split rings,13–15 composite resonant structures,16 square rings,17 square split rings,18,19 polygonal rings20 and rotated twisted closed ring resonators.21 Meanwhile, metamaterial absorbers based on novel materials represented by graphene,22–24 strontium titanate (STO),25 InSb,26–28 and GaAs29 are also prominently prevalent subjects. In order to accommodate more complex application scenarios, broadband30–32 and multiband33,34 metamaterial absorbers have been intensively developed. As for dual-band terahertz metamaterial absorbers, a design based on two square metal rings in each unit cell showed perfect absorption peaks at 2.7 and 5 THz.35 In 2016, a periodically patterned elliptical nanodisk graphene structure was constructed for terahertz dual frequency absorption performance.36 Also, two resonant frequencies at 3.75 and 7.15 THz were excited by different resonant modes of an Au periodic array on a SiO2 substrate.37
It is very necessary to study the sensing characteristics of terahertz metamaterial absorbers for further applications in sensors or sensing systems. Simulations38–40 and measurements41 are carried out to determine the refractive index sensitivity. In some recently works, the sensing properties of sensors are studied by means of several solutions with different concentrations, such as glucose,42 chlorpyrifos (CPS), and 2,4-dichlorophenoxyacetic acid (2,4-D).43,44 Moreover, terahertz sensors using diverse novel materials have also been investigated in depth, including lead iodide (PbI2) for ultrafast all-optical switching,45 polymethyl methacrylate (PMMA) for molecular vibrational sensing,46 perovskites for flexible devices,47 and valley photonic crystals for ultrasensitive topological sensing.48 In addition to the familiar sensors that judged by the amplitude shift, some sensors rely on phase change characteristics, which also have potential applications in biosensing.49–51
In this paper, we propose a dual-band metamaterial absorber in terahertz frequencies with a periodically patterned metal structure on the top of a silicon wafer. It shows perfect absorption characteristics and high Q-factors at two frequency bands of 0.89 and 1.36 THz. Through simulations with varied incident angles and polarization angles, the absorption properties and polarization-independence of the absorber are studied. To investigate the structure experimentally, several samples are fabricated by micro-nano fabrication technology. Electron beam lithography (EBL) is employed to pattern the periodical structure and electron beam evaporation (EBE) is employed to deposit the metal. In addition, these samples are measured using a terahertz time-domain spectroscopy (TDS) system and the results are consistent with the simulations. Finally, the refractive index sensing characteristics of the metamaterial absorber is further investigated.
The structure of the metamaterial unit cell is shown in Fig. 1(b). A square ring is set at the center of each unit cell and four T-shaped strips are loaded at the midpoints of the four sides of the square ring. The dimensions of the T-shaped strips are designed as follows: L1 = 2 μm, L2 = 24.1 μm, W1 = 1 μm, and W2 = 2 μm, while the width and outer side length of the square ring are W = 1.6 μm and a = 21 μm, respectively. On the upper surface of the silicon wafer, every pattern is perfectly aligned with a lattice constant of P = 33 μm. The size of the periodically patterned metamaterial structure is optimized to avoid undesired coupling.52,53
Fig. 1(c) shows a photograph of the fabricated metamaterial absorber prototype. Its overall size is 10 mm × 10 mm, containing 303 × 303 unit cells. As shown in Fig. 1(d), the image of the fabricated metamaterial periodic array obtained using a microscope indicates that the fabrication process is satisfactory without structural adhesions between the metal structures.
A(ω) = 1 − R(ω) − T(ω) | (1) |
A(ω) = 1 − R(ω) = 1 − |S11(ω)|2. | (2) |
At the beginning, the absorption capabilities of the proposed metamaterial absorber were studied, including the performance of the conventional periodic array under normal transverse electric (TE) and transverse magnetic (TM) polarized incidences. The simulation results are shown in Fig. 2, which indicates that both the absorptions are over 99% at 0.89 and 1.36 THz under normal TE and TM polarized incident waves. The full widths at half maximum (FWHMs) of these two frequency bands are 4.4 and 11.2 GHz, which proves that a perfect dual-band absorber with low FWHMs is realized. Obviously, the two curves of absorption spectra are highly overlapping, which shows that the absorber has great polarization insensitivity. Another important index for absorbers is the quality factor (Q-factor), which can be expressed as
(3) |
Ref. | Resonant frequency (THz) | Absorption rate (%) | FWHM (THz) | Q-Factor | Structure |
---|---|---|---|---|---|
54 | 1.56 | >99 | 0.043 | 36.35 | Circular ring |
55 | 2.64 | >99 | 0.060 | 44 | Circular patch |
56 | 8.44 | 99 | 0.100 | 84.4 | Circular patch |
57 | 1.80 | 98 | 0.015 | 120 | Circular patch with square rings |
2.26 | 96 | 0.024 | 94 | ||
This paper | 0.89 | >99 | 0.004 | 203.2 | Square ring with T-shaped strips |
1.36 | >99 | 0.011 | 121.4 |
The unit cell of proposed metamaterial absorber can be considered as an inductive-capacitive circuit model. According to the LC circuit element, the resonant frequency of the metamaterial absorber can be expressed as:
(4) |
In order to clarify the impedance matching condition, the basic equivalent circuit model of the present metamaterial absorber is shown in Fig. 3, where Z0 is the free space wave impedance, ZA is the overall impedance of the absorber, ZM is the impedance of the metamaterial structure, and Z1 stands for the impedance of the substrate layer backed with a gold ground, which can be considered as a short-circuited transmission line whose length is the thickness of the substrate. When the impedances ZA and Z0 are well matched with each other at the two operating frequencies, dual-band perfect absorption performance can be achieved. Some other relevant analyses and discussions can be seen as follows.
The distributions of the z-component electric fields of the proposed metamaterial absorber under normal TM polarized incidence at 0.89 THz and 1.36 THz are shown in Fig. 4(c) and (d), respectively. These distributions are just rotated 90° from those under TE polarized incidence, illustrating that the absorber has outstanding polarization-independent characteristics.
As shown in Fig. 5(a), under normal TE polarization, when the parameter a increases from 20.2 to 21.8 μm, the first resonant frequency is blue-shifted slightly and the second one almost remains stable, which corresponds to the fact that the electric field concentrates around the square ring only at the lower resonant frequency. The two resonant frequencies are both blue-shifted with the parameter L1 changing from 1.6 to 2.4 μm in Fig. 5(b). As L2 increases from 22.6 to 24.6 μm, the first resonant frequency fluctuates slightly, while a blue-shift occurs at the second resonant frequency from 1.347 to 1.363 THz, which is shown in Fig. 5(c). Meanwhile, as shown in Fig. 5(d), the absorption spectra exhibit that the two resonant frequencies are both red-shifted with the width W1 increases from 0.2 to 1.8 μm. It can be seen in Fig. 5(e) that the two resonant frequencies are blue-shifted from 0.882 to 0.895 THz and from 1.349 to 1.363 THz, as the parameter W2 ranges from 1.5 to 2.5 μm.
Thus, the design process of the proposed dual-band metamaterial absorber can be divided into three steps. At the beginning, the side length of the square ring a was adjusted to determine the first absorption band at the anticipated resonant frequency. Then, the parameters of the T-shaped metal strips L1, L2, W1 and W2 were altered to shift the second absorption band to the required resonance frequency. Finally, a slight adjustment was made to the parameters in order to improve the absorption performance and modify the first frequency offset caused by the variations of T-shaped strips. Therefore, when a metamaterial structure needs to be constructed, it is necessary to figure out the relationship between the metal strip length and the operating wavelength. As for the specific metal structures, a few shapes could be considered, including but not limited to the square ring, the circular ring, the split ring resonator (SRR), the straight strip, and their combinations.
Through the analyses, it can be found that the variation trends of resonant frequencies are the same as the variation trend of parameter W1, which is different from the other parameters. The reason is that the increase of W1 creates a shorter current path on the T-shaped metal strips, leading to higher resonant frequencies. However, when the values of other parameters increase, the current path will be longer, which causes lower frequencies. As a result, the variation trends of resonant frequencies with shifting parameters are consistent with the principles of electromagnetism. Furthermore, the resonant frequency variations are maintained in a small range, which indicates that the proposed metamaterial absorber has good stability and the influence of fabrication accuracy on the absorption performance can be relatively low.
In addition, another index, polarization-independence, representing the stability of absorption rate with a varied incident polarization angle, is also of great importance for metamaterial absorbers. As shown in Fig. 6(e), the curves of the absorption spectra with the polarization angle φ varied from 0° to 90° are strongly overlapping, which demonstrates that the proposed metamaterial absorber is highly polarization-independent with the help of its outstanding central symmetry.
After measuring an absorber using the system introduced above, two groups of data are obtained, including measurement data from the experimented sample and reference data representing the results from the system with the platform vacant. The absorption rate can be calculated using the abovementioned eqn (2). Additionally, to exclude undesirable environmental absorption, the reflection coefficient |S11(ω)|2 is expressed as a function of the measurement results normalized to the reference results, namely,
(5) |
(6) |
A series of samples of the proposed metamaterial absorber have been fabricated with parameter W2 shifting from 1.9 to 2.2 μm. The measured absorption rates under normal TE polarized incidence are shown in Fig. 8(a), illustrating that the first resonant frequency fluctuates slightly while the second resonant frequency decreases with the increase of W2. This phenomenon is consistent with the simulation. Meanwhile, as shown in Fig. 8(b), the measured absorption rates under normal TM polarized incident waves exhibit the same properties. Consequently, the fabricated absorbers have high sensitivity to the value of the geometric parameter. The second resonant frequency is slightly moved when comparing the results between simulation and measurement. The Q-factors of the measured results are lower than those of the simulated ones. The discrepancies between simulation and measurement may be caused by the following reasons. First, the simulation cannot well emulate the material characteristics of the silicon wafer employed in the fabrication process, including the crystal face and semiconductor types. Second, the thickness of the fabricated silicon wafer is 250 ± 25 μm, which was set as 250 μm in simulation. The inaccuracy of the substrate thickness will lead to a frequency deviation. Third, the standing waves inside the substrate were not considered during the simulation, which might cause inaccuracies around the second resonant frequency.
Fig. 8 Measurement results of the fabricated dual-band metamaterial absorbers with different values of W2 under (a) normal TE polarized incidence and (b) normal TM polarized incidence. |
(7) |
Fig. 9 Sensing characteristics with the refractive index shifting from 1.0 to 2.0. (a) Absorption spectra and (b) further analysis of the second resonant frequency. |
Thus, the parameter Cn determined by the analyte plays a vital role in the sensing performance. In the simulation, the filling thickness of the analytes is designed as ta = 63 μm, which is capable of reflecting the sensing characteristics distinctly. In order to cover most biological and chemical specimens, including DNA, RNA, proteins, and amino acids, the refractive index of the analytes is set from 1.0 to 2.0.
The absorption spectra of the proposed metamaterial absorber with different refractive index analytes under normal TE polarized incidence are shown in Fig. 9(a). This illustrates that the first resonant frequency fluctuates slightly while a red-shift occurs for the second resonant frequency as the refractive index changes from 1.0 to 2.0. The details of the variations at the second frequency and their linear fits are shown in Fig. 9(b). On the one hand, the total red-shift of the second frequency is 37 GHz with an R2 value of 0.8868, and the fitting line function can be expressed as:
f = −0.025 × n + 1.385. | (8) |
On the other hand, since the existence of analytes leads to the reduction of the reflection power, the absorption rate decreases from 99% to 78%. In addition, the frequency sensitivity is defined by the frequency variation per refractive index unit (RIU), calculated as:
(9) |
(10) |
Through calculations, the frequency sensitivity at the second resonant frequency of the proposed metamaterial absorber is 37 GHz RIU−1 and the FoM is 3.3. Comparisons between previous refractive index sensors and the proposed work are given in Table 2 for a clear insight into the sensing performance.
Ref. | Resonant frequency (THz) | Sensitivity (GHz RIU−1) | FoM | Structure |
---|---|---|---|---|
61 | 0.33 | 0.15 | 65.2 | Rhombus with triangular holes |
62 | 0.6 | 18 | 0.3 | I-shaped |
0.55 | 39 | 1.95 | Fano design | |
63 | 0.9 | 24.5 | 0.31 | Split square ring |
64 | 0.69 | 128.3 | 2.3 | Cross shaped |
0.81 | 113.3 | 1.98 | Complementary cross shaped | |
This paper | 1.103 | 37 | 3.3 | Square ring with T-shaped strips |
As shown in Fig. 10, in order to study the effect of the analyte thickness, the frequency sensitivity and variation of the absorption rate (when the refractive index shifts from 1.0 to 2.0) with the thickness increasing from 5 to 80 μm are simulated. The frequency sensitivity increases from 24 to 38 GHz RIU−1 and 7 to 42 GHz RIU−1 with the thickness ranging from 5 to 20 μm and 35 to 80 μm, respectively, while a decrease appears in the middle part. However, as the thickness of the analyte increases from 30 to 55 μm, the absorption rates under different refractive indices will not change dramatically. As a result, both the tendencies of the frequency sensitivity and absorption rate variation show the characteristics of a sinusoidal curve, which indicates that the proposed metamaterial absorber is also sensitive to the analyte thickness. It is observed from the analysis carried out above that the proposed metamaterial absorber has satisfactory sensing characteristics for practical applications.
This journal is © The Royal Society of Chemistry 2023 |