Open Access
Add to BibSonomy
Abstract
Food safety is an important consideration for the food industry and for daily life, and food additives are essential in the modern food industry. Graphene-based metamaterial sensors are of great value and have potential applications in the detection of food additives, due to their ultra-sensitivity. This paper proposes a metasurface sensor consisting of graphene and dual elliptical ring resonators (Gr-DERRs) sensor for the detection of two common food additives. The limit of detection (LOD) for Sudan I solution is 581.43 fg/ml and, for taurine, 52.86 fg/ml. This ultra-sensitive detection is achieved by exploiting the unique electromagnetic properties of electromagnetically induced transparency (EIT) resonance, together with the Fermi energy level of graphene moving to the Dirac point, resulting in a dramatic change in the dielectric environment. The Gr-DERRs sensor has brings significant improvement in the detection of food additives with detection limits reduced to the femtogram level.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
There is a growing concern about food additives in the food supply because of the potential risks to human health from exposure to excessive amounts of food additives. Sudan I is a lipophilic chemical coloring agent that is illegally used as an additive in food products such as chili sauce and chili powder because of their bright colors that can improve the gloss of merchandise [1]. However, Sudan I is classified as a carcinogen by the International Agency for Research on Cancer (IARC), and its use in food production is banned all over the world [2]. Taurine promotes brain tissue and mental development in infants and is encouraged and approved as a food additive to be added to infant formula but with an upper limit in the amount. Taurine is specified in the Chinese National Standard for Food Safety of Infant Formulae as not less than 0.16 mg per 100 ml of milk powder in China, that is, the limit of detection (LOD) of taurine should be better than 0.16 mg/ml [3]. This is why researchers have placed great importance on the testing of its content. Near-infrared spectroscopy and hyperspectral imaging are common methods for the detection of foodadditives, with high sensitivity and selectivity, but with the disadvantage of long detection times and damage to the test object [4]. Hence, it is important to develop a new food-additive testing method that is both rapid and non-destructive.
Terahertz (THz) waves are becoming a mainstream technology for the detection of food additives, and THz spectroscopy offers a non-destructive technique for sensing, but the sensitivity of THz waves is limited when detecting trace analytes due to the mismatch between the THz wavelength and the size of the analyte [5–7]. Hence, methods have been developed to enhance the signal using metasurfaces to increase the interaction between the THz wave and the analyte, with the frequency and amplitude of the resonance peaks varying with the nature of the analyte, thus enabling the detection of low concentrations [8–10].
The coupling between metamaterials and THz waves produces a special electromagnetic response called the plasmonic analogue of electromagnetically induced transparency (EIT) [11]. EIT has a narrow transparency window and its absorption linewidth is affected only by radiation damping [12,13].
Graphene has excellent optoelectronic properties, such as ultra-low transmission losses and extremely high carrier mobility [14–16]. During its preparation, graphene is inevitably affected by impurities, defects and disorders, usually p-doping. The initial Fermi energy level (EF) of graphene is located in the valence band, slightly off the Dirac point [11]. To move EF from the valence band to the Dirac point requires a very small external stimulus and makes it possible to prepare ultra-sensitive sensors based on this principle [12]. In 2019, Xu et al [17] proposed THz biosensing based on a graphene heterostructure platform with a modulation depth of up to 35%. Analytes containing p -electrons, through p-p stacking with graphene interactions, have higher sensitivity than those without p-electrons. M. Amin et al [18] proposed a chiral biosensor to distinguish respiratory viruses based on the polarization state of reflected electromagnetic waves. The characteristics of biological samples can be fully characterized by detecting discrete-frequency-polarization (or chiral) states without the need for broad spectrum scanning, which can significantly reduce the cost of measurement equipment. Yao et al. [11] proposed a new THz biosensor composed of graphene-polyimide-perovskite. The LOD of the biosensor for detecting whey protein was 6.25 ng/ml.
Graphene sensors based on metamaterials also have important application value in the detection of food additives. This work proposes a metasurface sensor consisting of graphene and dual elliptical ring resonators (Gr-DERRs) for the detection of two common food additives, Sudan I and taurine. These can reach sensitivities of 581.43 fg/ml and 52.86 fg/ml, respectively. This ultra-sensitive detection is achieved by exploiting the unique electromagnetic properties of EIT resonance, together with moving the EF of graphene to the Dirac point, which results in a dramatic change in the dielectric environment.
2. Materials and methods
A schematic diagram of the Gr-DERRs sensor is shown in Fig. 1(a) From bottom substrate to top surface, it includes a 300-µm-thick quartz layer, a 5 µm-thick polyimide (PI) film, a 0.2 µm-thick aluminum (Al) structure, a 10-µm-thick PI film, and a layer of graphene. To fabricate the Gr-DERRs sensor, a PI film was first spin-coated on a quartz substrate, with the patterning of the Al structure being completed using standard photolithography and magnetron sputtering methods. Then the PI film was spin-coated, followed by the transfer of graphene. The detailed fabrication flow of the Gr-DERRs sensor is shown in Fig. S1 in the Supplement 1. Figures 1(b)(I and II) show the optical microscope image of the Gr-DERRs sensor, which generally meets the experimental requirements. The geometrical parameters of the Al structural unit are shown in Fig. 1(b)(III); it consists of a large elliptical ring resonator (LERR) and a small elliptical ring resonator (SERR). Figure 1(c) depicts the Raman spectrum of a single layer of graphene on a substrate, revealing a high-intensity 2D peak (∼2692 cm-1) and a low-intensity G peak (∼1581 cm-1), The full width at half maximum of the 2D peak is 52 cm-1. From these values it is concluded that the graphene used in this experiment is of high quality [19,20]. The transmission spectra of graphene the Gr-DERRs sensor was simulated with EF ranging from 0 eV to 0.04 eV, as shown in Fig. S2.
Fig. 1. The Gr-DERRs sensor. (a) A schematic diagram. (b) Enlarged diagram of the metamaterial structure under an optical microscope with geometric parameters P = 106 µm, Rx = 47 µm, Ry =42 µm, rx = 21 µm, ry = 16 µm, and W = 5 µm. (c) Raman spectra of the graphene monolayer.
Figure 2 shows the transmission spectra of the DERRs sensor and the Gr-DERRs sensor under experimental and simulated conditions. In order to protect and isolate graphene as much as possible — that is, to prevent the transmission coefficient of the sensor from being greatly attenuated — a layer of PI with a thickness of 10 µm is spin-coated between the Al structure and the graphene, so that the transmission curve is red-shifted. Graphene has a high carrier concentration and high conductivity, producing a large radiation loss, so that the amplitude of the transmission curve is attenuated. The transmission curves from the experiment and from both simulation software match well, except for slight intensity differences and frequency shifts. These slight differences can be attributed to errors in manufacturing the PI and metal structure.
3. Results and discussion
3.1 Simulation results
A numerical simulation of the DERRs sensor was carried out with a THz wave incident vertically on the upper surface of the sensor, with the electric field along the x-direction (Ex), and with the magnetic field along the y-direction (Hy). Figure 3(a) depicts the transmission curves of the LERR, SERR, and DERRs, contributing to understanding of the resonant coupling mechanism of this metamaterial. When the THz wave is perpendicular to the surface of the LERR or SERR, the amplitudes at 1.086 THz are close to 0 and 1, respectively. Since the LERR can be excited by electromagnetic waves, it (black line) can be regarded as a bright mode and the SERR (blue line) as a dark mode. The EIT resonance is formed when the two modes of light and dark produce destructive interference and a transparent window appears at 1.086 THz.
Fig. 3. (a) Transmission curves of the LERR, SERR and DERRs. (b)-(d) The electric field distributions of the LERR, SERR and DERRs at 1.086 THz. (e)-(g) The surface current distributions of DERRs at the resonance points of 0.85, 1.086 and 1.6 THz.
The electric field distributions of the LERR, SERR, and DERRS at 1.086 THz are simulated as shown in Figs. 3(b)-(d), respectively. The electric field intensity on the surface of the LERR is stronger, mainly concentrated on the left and right sides of the LERR (black circles in Fig. 3(b)), while the electric field intensity of the SERR is weaker. Figure 3(d) shows clearly that the electric field intensity of the LERR is completely suppressed due to phase-dissipative interference, with the field intensity concentrated mainly on the left half of the SERR (black circle), indicating the excitation of the dark mode. To further investigate the principle of the EIT resonance generated by this sensor, the surface current distributions at the resonance points of 0.85, 1.086, and 1.6 THz are simulated, as plotted in Figs. 3(e)-(g). At 0.85 THz, the current distribution is concentrated mainly in the LERR, whose direction is from left to right. A larger electric dipole (ED) moment produces a larger radiation loss, forming a dip in the transmission curve (f1) [6–25]. The LERR and SERR have the same direction of current at 1.6 THz, where the ED moment reaches its maximum, where the radiation loss is large, and where it is difficult for the incident electromagnetic wave to pass through the metamaterial, so that a dip in the transmission curve is formed (f3). The ED moments in the x-direction cancel each other at 1.086 THz when the currents of the LERR and SERR are in opposite directions, the ED moments are well suppressed, and the incident electromagnetic wave can pass through the metamaterial with low loss, forming a peak (f2) [26–29]. Hence, the EIT phenomenon in metamaterials has a narrow band and high transparency spectral characteristics similar to those of an EIT atomic system, so that it can be used to design high-sensitivity sensors and other devices.
3.2 Detection of Sudan I by sensors
The DERRs sensor and the Gr-DERRs sensor were used as detection platforms for two common food additives. Ten different concentrations of Sudan I solutions were tested, 581.43 fg/ml (C1), 25.01 pg/ml (C2), 269.76 pg/ml (C3), 514.52 pg/ml (C4), 10.79 ng/ml (C5), 2.50 µg/ml (C6), 52.37 µg/ml (C7), 91.24 µg/ml (C8), 0.89 mg/ml (C9), and 1.15 mg/ml (C10); also thirteen different concentrations of taurine solutions were tested, 52.86 fg/ml (C1), 79.29 fg/ml (C2), 660.72 fg/ml (C3), 36.18 pg/ml (C4), 47.84 pg/ml (C5), 292.60 pg/ml (C6), 782.12 pg/ml (C7), 227.06 ng/ml (C8), 4.96 µg/ml (C9), 9.71 µg/ml (C10), 62.1 µg/ml (C11), 0.16 mg/ml (C12), and 1.26 mg/ml (C13). To achieve greater reliability, all experimental data provided in this work are averaged over three tests. Figure 4(a) shows a schematic diagram of the Gr-DERRs sensor when detecting Sudan I. Figure 4(b) presents the transmission spectra of the DERRs sensor detecting an increase in the concentration of Sudan I solution from C1 to C10. The transmission spectra at all concentrations do not differ significantly from those with no Sudan I (bare). The results show that the use of the DERRs sensor for Sudan I detection at femtogram-level concentrations is challenging. Figure 4(c) shows the transmission spectrum of the Gr-DERRs sensor detecting Sudan I. To make comparison easier, the concentrations detected by both sensors are the same. When the Sudan I concentration was increased from C1 to C3, the transmittance was significantly enhanced, facilitating the qualitative sensing of Sudan I. As the concentration increased further, from C4 to C7, the transmittance decreases instead, eventually almost matching the transmittance of the bare Gr-DERRs sensor. This is because graphene is usually p-doped, so the initial EF of graphene is slightly off the Dirac point and in the valence band (see Fig. 5(a)). As the concentration of Sudan I increases, EF gradually shifts from the valence band to the Dirac point (see Fig. 5(b)). When the concentration of Sudan I increases to 269.76 pg/ml (C3), the Fermi energy level shifts to the Dirac point (see Fig. 5(c)). The closer EF of graphene is to the Dirac point, the lower the conductivity of graphene, and the smaller the loss, so the transmission coefficient reaches its maximum at f2. As the concentration of Sudan I increased from C3 to C7, EF moves away from the Dirac point to the conduction band (see Fig. 5(d)), and the conductivity continues to increase until it reaches a limit, so that the transmission coefficient of the sensor reaches a minimum when the concentration increases to C7. And, this does not mean that the amplitude change of the sensor is not obvious when the concentration of Sudan I is high. It is well known that to move EF from the conduction band to the Dirac point requires only a very small external stimulus. As the Sudan I concentration increases from C7 to C10 as shown in Fig. 6(c), the transmittance was significantly enhanced, EF gradually shifts from the valence band to the Dirac point, facilitating the quantitative sensing of Sudan I. In addition, the closer is EF of graphene to the Dirac point, the lower the conductivity of graphene and the stronger the current distribution and magnetic field strength, as shown in Figs. 5(e)-(h) and (i)-(l). The amplitude variation $\Delta \textrm{T}$ of the Gr-DERRs sensor at different concentrations, C1-C10, is defined as $\Delta \textrm{T =}\left| {\textrm{((}{\textrm{T}_{\textrm{Cc}}}\textrm{ - }{\textrm{T}_{\textrm{Bare}}}\textrm{)}{\textrm{T}_{\textrm{Bare}}}{)\% }} \right|$ [11,19], where TCc (TBare) is the transmittance at the resonance point with (without) analyte. As the concentration of Sudan I increases, the variations of ΔT at both f1 and f3 are small, and the variation of ΔT at the transmission window f2 increases, enabling ultra-sensitive detection, as shown in Fig. 4(d). The maximum ΔT was reached at a Sudan I concentration of 269.76 pg/ml (C3), $\Delta {\textrm{T}_{\textrm{max}}}\textrm{ = 31}{.34\% }$. At a minimum concentration of C1 (581.43 fg/ml), $\Delta \textrm{T = 21}{.41\%}$, which is still a high level of detection for the sensor.
Fig. 4. (a) The diagram of the Gr-DERRs sensor when detecting Sudan I (b) Transmission spectra of the DERRs sensor for various concentrations of Sudan I from C1 to C10. (c) Transmission spectra of the Gr-DERRs sensor for various concentrations from C1 to C10. (d) The amplitude variations of the Gr-DERRs sensor at f2 and at different solution concentrations.
Fig. 5. Mechanisms of sensing performance based on Gr-DERRs. (a)-(d) Variations of EF of graphene at different concentrations. (e)-(h) Current distributions. (i)-(l) Magnetic field distributions.
Fig. 6. Analysis of the sensing performance of this sensor. (a) Transmission spectra of the DERRs sensor for various concentrations of taurine from C1 to C13. (b) Transmission spectra of the Gr-DERRs sensor for various concentrations of taurine from C1 to C8. (c) Transmission spectra of the Gr-DERRs sensor for various concentrations of taurine from C9 to C13. (d) The amplitude variations of the Gr-DERRs sensor at f2 and at different solution concentrations.
3.3 Detection of taurine by sensors
The taurine solution was applied dropwise to the surface of the DERRs sensor and the Gr-DERRs sensor using the same method and was then tested, with concentration increasing from C1 to C13. Similar to the detection of Sudan I, the DERRs sensor is insensitive to taurine at different concentrations, as shown in Fig. 6(a). When taurine is detected using the Gr-DERRs sensor, EF gradually approaches the Dirac point from the valence band and as the concentration increases from C1 to C5, the conductivity decreases and the transmittance increases significantly, as shown in Fig. 6(b). EF reaches the Dirac point when the taurine concentration is C5, at which point the conductivity is at its lowest and the current and magnetic field strengths reach their maxima, as shown in Figs. 5(g) and (k). As the concentration of taurine gradually increases to C8, EF moves away from the Dirac point to the conduction band and the transmittance decreases, but it is still significantly higher than with the bare Gr-DERRs sensor (see Fig. 6(c)), extending the detection range of the sensor. As the concentration increases, the Gr-DERRs sensor's transmission curve again appeared to increase (C9-C11) and then decrease (C11-C13) as shown in Fig. 6(c). As the concentration increases from C9 to C11, EF gradually shifts from the valence band to the Dirac point. When the concentration of taurine increases to 62.1 µg/ml (C11), the Fermi energy level shifts to the Dirac point; As the concentration increases from C11 to C13, EF moves away from the Dirac point to the conduction band, and the conductivity continues to increase. Figure 6(d) shows that ΔT reaches a maximum at C5 = 47.84 pg/ml, where ΔTmax = 29.58% and the LOD of taurine can reach 52.86 fg/ml. Thus, the Gr-DERRs sensor has the potential to achieve femtogram-level detection of analytes; its sensitivity is higher than that previously reported, as shown in Table 1.
Table 1. Performance comparison with previous work using THz metasurface sensorsa
4. Conclusion
In this work, we have experimentally demonstrated a THz Gr-DERRs sensor that has the potential for ultra-sensitive detection of additives. First, in order to investigate the resonance mechanism by which this sensor achieves sensing, accurate numerical simulations and analysis of the model were carried out. The Gr-DERRs sensor was then prepared and validated for the ultra-sensitive detection of two additives, Sudan I and taurine. The detection limits for Sudan I and taurine can reach 581.43 fg/ml and 52.86 fg/ml, respectively. The depth of modulation of this Gr-DERRs sensor was 31.34% and 29.58% for the detection of Sudan I and taurine, respectively. In brief, the Gr-DERRs sensor has the potential to achieve femtogram-level detection of analytes, making it enormously significant for the detection of food additives.
Funding
Taishan Scholar Project of Shandong Province (tsqn201909150); Qingchuang Science and Technology Plan of Shandong Universities (2019KJN001); National Key Research and Development Program of China (2017YFA0700202, 2017YFB1401203); National Natural Science Foundation of China (61675147, 61701434, 61735010, 62201496, 61827818, 61975105, 62005013); Natural Science Foundation of Shandong Province (ZR2020FK008, ZR202102180769, ZR2021MF014); China Postdoctoral Science Foundation (2022M720395); National Key Research and Development Program of China (2021YFB2800900).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
References
1. L. Zhu, G. Zhang, and B. Hua, “Terahertz Spectrum of Sudan I,” Chinese Journal or Sensors and Actuators 21(1), 83–87 (2008).
2. L. Zong, Y. Yue, and R. Sun, “Evaluation of uncertainty in determination of Sudan I, II, III and IV in chili powder by HPLC,” Analysis and Testing 1(20), 189–192 (2020).
3. B. Wang, D. Cai, and X. Li, “Research progress of taurine detection in infant formula milk powder,” Food and Fermentation Industries 46(22), 287–291 (2020).
4. J. Zhang, T. Lang, Z. Hong, J. Liu, and P. Wang, “Sensitive Detection of Aspartame and Vanillin by Combining Terahertz Fingerprinting With a Metamaterial,” IEEE Sensors J. 22(16), 16513–16521 (2022). [CrossRef]
5. Z. Dai, M. Yang, T. Mou, et al., “Tracing pictogram-level chlorothalonil pesticide based on terahertz metal–graphene hybrid metasensors,” Opt. Commun. 529, 129025 (2023). [CrossRef]
6. T. Mu, Y. Ye, Z. Dai, et al., “Silver nanoparticles-integrated terahertz metasurface for enhancing sensor sensitivity,” Opt. Express 30(23), 41101–41109 (2022). [CrossRef]
7. R. Zhao, Y. Ye, Z. Dai, et al., “Research on specific identification method of substances through terahertz metamaterial sensors,” Results Phys. 43, 106055 (2022). [CrossRef]
8. S. Tabassum, S. K. Nayemuzzaman, M. Kala, A. Kumar Mishra, and S. K. Mishra, “Metasurfaces for Sensing Applications: Gas, Bio and Chemical,” Sensors 22(18), 6896 (2022). [CrossRef]
9. H. Hajian, I. D. Rukhlenko, A. L. Bradley, and E. Ozbay, “High-Figure-of-Merit Biosensing and Enhanced Excitonic Absorption in an MoS(2)-Integrated Dielectric Metasurface,” Micromachines 14(2), 370 (2023). [CrossRef]
10. X. Yan, Z. Zhang, L. Liang, M. Yang, D. Wei, X. Song, H. Zhang, Y. Lu, L. Liu, M. Zhang, T. Wang, and J. Yao, “A multiple mode integrated biosensor based on higher order Fano metamaterials,” Nanoscale 12(3), 1719–1727 (2020). [CrossRef]
11. H. Yao, Z. Sun, X. Yan, M. Yang, L. Liang, G. Ma, J. Gao, T. Li, X. Song, H. Zhang, Q. Yang, X. Hu, Z. Wang, Z. Li, and J. Yao, “Ultrasensitive, light-induced reversible multidimensional biosensing using THz metasurfaces hybridized with patterned graphene and perovskite,” Nanophotonics 11(6), 1219–1230 (2022). [CrossRef]
12. X. Guo, Z. Zhang, M. Yang, P. Bing, X. Yan, Q. Yang, D. Wei, L. Liu, L. Liang, and J. Yao, “Time-Frequency Double Domain Resolving by Electromagnetically Induced Transparency Metasensors for Rapid and Label-Free Detection of Cancer Biomarker Midkine,” Opt. Laser. Eng. 142, 106566 (2021). [CrossRef]
13. M. Chen and Z. Xiao, “Metal-graphene hybrid terahertz metamaterial based on dynamically switchable electromagnetically induced transparency effect and its sensing performance,” Diamond Relat. Mater. 124, 108935 (2022). [CrossRef]
14. N. L. Kazanskiy, S. N. Khonina, and M. A. Butt, “Recent Development in Metasurfaces: A Focus on Sensing Applications,” Nanomaterials 13(1), 118 (2022). [CrossRef]
15. X.-Q. Jiang, W.-H. Fan, X. Chen, and H. Yan, “Ultrahigh-Q terahertz sensor based on simple all-dielectric metasurface with toroidal dipole resonance,” Appl. Phys. Express 14(10), 102008 (2021). [CrossRef]
16. S. Alizadeh, E. Zareian-Jahromi, and V. Mashayekhi, “A dual-band simple graphene-based absorber for refractive index sensing applications,” Opt. Quantum Electron. 54(12), 863 (2022). [CrossRef]
17. W. Xu, L. Xie, J. Zhu, L. Tang, R. Singh, C. Wang, Y. Ma, H.-T. Chen, and Y. Ying, “Terahertz biosensing with a graphene-metamaterial heterostructure platform,” Carbon 141, 247–252 (2019). [CrossRef]
18. M. Amin, O. Siddiqui, H. Abutarboush, M. Farhat, and R. Ramzan, “A THz graphene metasurface for polarization selective virus sensing,” Carbon 176, 580–591 (2021). [CrossRef]
19. M. Yang, H. Yao, Y. Lu, et al., “Graphene-integrated toroidal resonance metasurfaces used for picogram-level detection of chlorothalonil in the terahertz region,” Opt. Express 30(19), 34034–34042 (2022). [CrossRef]
20. W. Xu, Y. Huang, R. Zhou, Q. Wang, J. Yin, J. Kono, J. Ping, L. Xie, and Y. Ying, “Metamaterial-free flexible graphene-enabled terahertz sensors for pesticide detection at bio-interface,” ACS Appl. Mater. Interfaces 12(39), 44281–44287 (2020). [CrossRef]
21. R. Cheng, L. Xu, X. Yu, L. Zou, Y. Shen, and X. Deng, “High-sensitivity biosensor for identification of protein based on terahertz Fano resonance metasurfaces,” Opt. Commun. 473, 125850 (2020). [CrossRef]
22. C. Zhang, T. Xue, J. Zhang, L. Liu, J. Xie, G. Wang, J. Yao, W. Zhu, and X. Ye, “Terahertz toroidal metasurface biosensor for sensitive distinction of lung cancer cells,” Nanophotonics 11(1), 101–109 (2021). [CrossRef]
23. M. A. Baqir and P. K. Choudhury, “Hyperbolic metamaterial-based optical biosensor for detecting cancer cells,” IEEE Photon. Technol. Lett. 35(4), 183–186 (2023). [CrossRef]
24. W. Chen, C. Li, D. Wang, W. An, S. Gao, C. Zhang, and S. Guo, “Tunable wideband-narrowband switchable absorber based on vanadium dioxide and graphene,” Opt. Express 30(23), 41328–41339 (2022). [CrossRef]
25. S. Zhang, M. Qin, B. Wu, and E. Wu, “All-dielectric Si metamaterials with electromagnetically induced transparency and strong gap-mode electric field enhancement,” Opt. Commun. 530, 129143 (2023). [CrossRef]
26. N. Yi, R. Zong, and R. Qian, “Multifunctional terahertz metasurface utilizing Dirac semi-metal and vanadium dioxide hybrid metamaterials,” Mater. Sci. Semicond. Process. 146, 106682 (2022). [CrossRef]
27. Z. Ding, W. Su, H. Wu, W. Li, Y. Zhou, L. Ye, and H. Yao, “Ultra-broadband tunable terahertz absorber based on graphene metasurface with multi-square rings,” Mater. Sci. Semicond. Process. 163, 107557 (2023). [CrossRef]
28. P. Wu, S. Qu, X. Zeng, N. Su, M. Chen, and Y. Yu, “High-Q refractive index sensors based on all-dielectric metasurfaces,” RSC Adv. 12(33), 21264–21269 (2022). [CrossRef]
29. J. Ruan, S. Ji, Z. Tao, and F. Lan, “Ultra-wideband metamaterial absorber doped GaAs in the infrared region,” J. Electromagnet. Wave. 35(8), 1088–1098 (2021). [CrossRef]
