Open Access
Add to BibSonomy
Abstract
In this paper, a highly integrated terahertz (THz) biosensor is proposed and implemented, which pioneered the preparation of low-temperature gallium arsenide (LT-GaAs) thin film photoconductive antenna (PCA) on the sensor for direct generation and detection of THz waves, simplifying complex terahertz time-domain spectroscopy (THz-TDS) systems. A latch type metasurface is deposited in the detection region to produce a resonance absorption peak at 0.6 THz that is independent of polarisation. Microfluidics is utilised and automatic injection is incorporated to mitigate the experimental effects of hydrogen bond absorption of THz waves in aqueous-based environment. Additionally, cell damage is minimised by regulating the cell flow rate. The biosensor was utilised to detect the concentration of three distinct sizes of bacteria with successful results. The assay was executed as a proof of concept to detect two distinct types of breast cancer cells. Based on the experimental findings, it has been observed that the amplitude and blueshift of the resonance absorption peaks have the ability to identify and differentiate various cancer cell types. The findings of this study introduce a novel approach for developing microfluidic THz metasurface biosensors that possess exceptional levels of integration, sensitivity, and rapid label-free detection capabilities.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
The Terahertz (THz) wave is an electromagnetic wave that falls within the frequency range of 0.1-10 THz [1–5]. This frequency band is situated between macroelectronics and microphotonics. The frequency range in question is considered highly promising within the electromagnetic spectrum. In recent years, THz waves have made label-free and non-invasive biosensing methods feasible due to their low photon energy and fingerprint spectral properties [6–11]. At the same time, combined with the rapid development of metasurface technology [12–18], the organic fusion of the two has opened up a new path for highly sensitive detection of biological cells. Silalahl et al proposed a floatable THz metasurface sensor with a great refractive index sensitivity of approximately 532 GHz/RIU at a floating height of 30 µm, which is three times that of a regular THz metasurface sensor. The biosensor has been successfully used for low concentration sensing of bovine serum albumin (BSA) [19]. The THz sensing method proposed by Zhang et al utilises a reflection time domain polarization spectroscopy (RTDPS) system and chiral metasurface. The method was tested on the chiral enantiomer of three amino acid aqueous solutions, resulting in improved sensing sensitivity and detection accuracy compared to the conventional THz resonance sensing method [20]. Chen et al proposed a three-dimensional graphene metasurface structure composed of a single-layer graphene coating, which can effectively excite multiple plasma resonance modes under THz wave incidence, that are highly sensitive to changes in the refractive index and thickness of the analyte [21].
The present study describes the development of an automatic injection THz microfluidic biosensor that integrates metasurface and low-temperature gallium arsenide (LT-GaAs) photoconductive antenna (PCA) [22–25]. The sensor was designed and implemented to generate and detect THz waves using two LT-GaAs thin-film PCAs prepared on cyclic olefin copolymer (COC) material through an unmasking process. This approach simplifies the complex terahertz time-domain spectroscopy (THz-TDS) systems. Next, a latch type metasurface is deposited within the sensor detection area, which can generate a polarization independent resonance absorption peak at approximately 0.6 THz, providing better sensing performance compared to traditional H type metasurface. At the same time, an automatic injection function is achieved using a flat micro-pump on the sensor, which allows control of the cell flow rate and reduces damage to the cells to a certain extent. The THz metasurface sensor used in this study demonstrates the potential for quantitative concentration detection and qualitative molecular type identification. Concentration sensing was performed on a solution of Candida albicans (CA) at varying detection area thicknesses. The experimental findings indicate that the biosensor attains maximum sensitivity at a detection area thickness of 50 µm. Various concentrations of Candida albicans (CA), Staphylococcus aureus (SA), and Escherichia coli (EC) solutions were tested at an analyte thickness of 50 µm. By measuring the THz transmission spectrum, the frequency and amplitude responses of cells with different concentrations were obtained, verifying the dependence of resonance peak shift and peak amplitude change on cell concentration. It was found that as the size of bacteria decreased, the degree of resonance frequency blue shift increased, which means that the sensor can detect and distinguish different types of bacteria. The assay was conducted to demonstrate the feasibility of identifying two distinct types of breast cancer cells, namely P53 mutant and P53 wild type. The results of the experiment demonstrate that the resonance absorption peak's amplitude and blueshift can be utilised to detect and differentiate various cancer cell types without the need for antibodies. The findings of this study indicate that the THz microfluidic sensor has a wide range of potential applications in the field of biological detection.
2. Experiment
2.1 Design and experiment of latch type metasurface
The designed latch type metasurface structure is depicted in Fig. 1(a). The substrate is composed of COC material and exhibits a transmission rate exceeding 90% for THz waves [26], as illustrated in Figure S1. The resonant metasurface units are arranged periodically, as depicted in the inset of Fig. 1(a). The units are fabricated using ultraviolet (UV) lithography, and further details of the process can be found in S2. The metasurface units are periodically arranged in two directions within the plane, and the shape of the units is similar to that of the latch, as shown in Fig. 1(b), using metallic gold with a thickness of 800 nm, each periodic cell is 80 µm long and 80 µm wide. Line width W1 = 16 µm. Thin line width W2 = 12 µm at the center of the metasurface. Latch shaped part length L2 = 33 µm. Seam width D = 9 µm.
Fig. 1. (a) Schematic diagram of latch type metasurface structure (inset shows SEM morphology characterization). (b) SEM morphology characterization of latch type metasurface units. (c) The simulation results of the electromagnetic enhancement effect of the H type metasurface are shown in the figure. (d) The simulation results of the electromagnetic enhancement effect of the latch type metasurface. (e) Simulated (dashed line) and experimental (solid line) results of resonance absorption peaks on H type metasurface (red line) and latch type metasurface (blue line).
The electromagnetic enhancement effects of latch type metasurface are stronger when compared to traditional H type metasurface. We simulated the electric field modes of two types of metasurface using COMSOL Multiphysics (as shown in Figure S3), and the simulation results are shown in Fig. 1(c) and 1(d). It has been observed that latch type metasurface exhibit more significant effects of electric field enhancement. This is because the surface current generated by the metasurface under the influence of the incident field excite the polarization field, thus inducing charges to arrange near the gap [27]. The gaps present in the metasurface units can be regarded as capacitors. Furthermore, it is worth noting that the latch type metasurface structure exhibits a more intricate configuration. The latch-like part at its central can be considered as three capacitors in parallel; in contrast, each unit of the conventional H type metasurface can only be equivalent to one capacitor, so the latch-type metasurface can excite a stronger electrodeposition field with the same charge density and thus has a stronger electromagnetic enhancement effect. Figure S4 displays the equivalent circuit of both metasurface.
The transmittance curves of H type metasurface and latch type metasurface with the same size were measured using the THz-TDS system. The experimental results are depicted by the solid lines in Fig. 1(e). Compared to the simulation results (dashed line in Fig. 1(e)), there is a slight difference in the amplitude of the resonance peak, which is due to some error between design and fabrication. The resonance absorption peaks at around 0.6 THz are generated by both H type metasurface and latch type metasurface, as observed from the experimental and simulation results. The resonance absorption intensity of H type metasurface is significantly weaker than that of latch type metasurface, and the measurement results are consistent with the simulation results. The manufacturing process adheres to the specified design.
2.2 Preparation of LT-GaAs thin film PCA and design of electrode structure
The LT-GaAs thin films were prepared using the epitaxial wafer stripping method. Figure 2(a) displays the structure of a four-layer epitaxial wafer, wherein a sacrificial layer of AlAs is removed to obtain a solitary LT-GaAs film in the uppermost layer. The process of growing the epitaxial wafer involves the utilisation of molecular beam epitaxy (MBE) equipment. For further information regarding the preparation process, please refer “The specific preparation process for LT-GaAs epitaxial wafer” in Supplement 1. This study employed the stripping process illustrated in Fig. 2(b) to adhere a single LT-GaAs film onto the COC material surface. For process details, refer to the section “The process of peeling off a single LT-GaAs film” in Supplement 1. Through numerous experiments, it was discovered that the AlAs sacrificial layer can react with the HCl solution. The epitaxial film underwent immersion in an HCl solution with a volume ratio of 13.5%. Subsequently, it was subjected to heating in a water bath at a temperature of 73 °C for a duration of 1 hour. Upon shaking the tube, it was observed that the epitaxial wafer underwent separation into two distinct portions. One portion was deposited at the bottom of the tube, while the other remained suspended in the HCl solution. This observation suggests that the AlAs sacrificial layer underwent complete reaction, while the LT-GaAs film was preserved. The SEM morphology of the single LT-GaAs film obtained by corrosion is shown in Fig. 2(c), which shows that the surface of the film is flat and without protrusions. Please refer to the S6-S8 for detailed fabrication procedures.
Fig. 2. (a) Schematic diagram of the structure of 4-layer LT-GaAs epitaxial wafer. (b) Schematic diagram of the peeling and bonding process for a single LT-GaAs film. (c) SEM morphology of a single LT-GaAs thin film. (d) SEM morphology of latch type electrode structure. (e) The simulation result diagram of the electromagnetic enhancement effect of the latch type electrode.
We also used UV lithography to prepare LT-GaAs thin-film PCA by depositing latch type electrode on LT-GaAs thin-films. The electrode morphology is characterized as shown in Fig. 2(d) to obtain stronger THz wave signals. The reason for this phenomenon is that upon exposure to a femtosecond laser beam, the latch type PCA surface undergoes excitation of photogenerated carriers, leading to the establishment of a potent electric field on its surface. This process is facilitated through a lens system. Under the action of a THz electric field on charge carriers, electrons and holes will continuously accelerate, stop, and twist between the lock and latch. After a certain transient process, a local electric field enhancement region will be formed between the lock and latch, and this enhancement effect is mainly attributed to the generation of surface plasmon polariton (SPP) [28]. The electromagnetic waves in this particular area are concentrated into a small volume, resulting in the creation of a powerful localised electric field. In contrast to conventional electrodes, which are depicted in Figure S9, the localised electric field in Fig. 2(e) exhibits a significantly enhanced electromagnetic effect. This effect results in the emission of stronger THz waves from the latch gap.
2.3 Design and experiment of latch type metasurface
Figure 3 illustrates the preparation procedure of the THz microfluidic biosensor devised in this investigation. For specific details, please refer to the “Sensor preparation process” section in the supporting information.
Fig. 3. Preparation process diagram of automatic injection THz microfluidic biosensor based on metasurface and LT-GaAs thin film PCA.
The microfluidic sensor facilitates the automated intake and outflow of the liquid specimen being analysed [29]. Figure S10 illustrates the principle. The system optical path system utilised in this study is depicted in Fig. 4(a). It mainly comprises an 800 nm femtosecond laser and a time-delay control system. The laser is divided into pump pulses and detection pulses after passing through a beam splitter (PBS). The first component is connected to the LT-GaAs thin film PCA located on one side of the sensor via a time-delay system, which is responsible for producing THz waves. Meanwhile, the second component is connected to the LT-GaAs thin film PCA on the opposite side of the sensor, which is responsible for detecting THz waves. The sensor is held in place on a three-dimensional table, and the THz-emitting PCA generates a THz wave that passes through the liquid-filled analyte sensor carrying the sample information, which is received by the detector film and fed into a lock-in amplifier for amplification. A computer is utilised for the purpose of collecting and processing data. The THz time-domain and frequency-domain spectra produced by the optical path and spatial sensor of the system are illustrated in Fig. 4(b) and 4(c), respectively.
Fig. 4. (a) Schematic diagram of the sensing system optical path, THz time-domain spectrum (b) and frequency-domain spectrum (c) generated by the sensor without injection of analyte.
3. Results and discussion
3.1 Influence of detection area thickness on sensing sensitivity
To optimise the sensitivity of a sensor towards an analyte, it is necessary to establish the ideal thickness of the detection region for achieving the most effective sensing outcome [30]. We selected the CA solution with the lowest refractive index among the three bacteria at the same concentration as the analyte, with a concentration gradient of PBS1 (process bacterial sample): 6 × 105 cell/mL, PBS2: 8 × 105 cell/mL, PBS3: 1 × 106 cell/mL and PBS4: 1.2 × 106 cell/mL. The detection area's thickness was selected to be 20 µm, 40 µm, 50 µm, 60 µm, and 80 µm. The outcomes of the experiment are illustrated in Fig. 5. As can be seen from the figure, the resonance absorption peaks show a certain degree of frequency shift Δf, which is defined as the absolute value of the resonance peak frequency distance between the analyte and the reference cell culture curve, as the concentration of CA solution increased under different detection area thickness conditions. When the thickness of the detection area measures 20 µm, as shown in Fig. 5(a), the maximum Δf limited to 19 GHz, resulting in a suboptimal sensing effect. As the thickness of the detection area increases from 20 µm to 50 µm (Fig. 5(a)-Fig. 5(c)), the maximum Δf increases significantly, and at 50 µm (Fig. 5(c)), the maximum Δf is 178 GHz, with a significant increase in sensing effect. In contrast, as the thickness of the detection area is increased from 50 µm to 80 µm (as shown in Fig. 5(c)-Fig. 5(e))), the maximum Δf experiences a sudden decrease, reaching 22 GHz at a detection area thickness of 80 µm (as shown in Fig. 5(e)). Additionally, the intensity of the sensor resonance absorption peak experiences a significant decrease during this process. We attribute this to the fact that as the thickness of the detection area increases, the H2O molecule content of the CA solution at the same PBS increases and some of the incident THz waves are absorbed by the hydrogen bonds in the water, resulting in a weaker resonance absorption effect from the latch type metasurface structure. The gradual approach of CA solution concentration sensing effect occurs in the absence of metasurface structure within the detection area, as depicted in Figure S11.
Fig. 5. THz transmittance curves of CA solutions for sensor detection area thicknesses of 20 µm (a), 40 µm (b), 50 µm (c), 60 µm (d) and 80 µm (e), respectively. (f) Plot of 1minus peak transmission (black line) and maximum Δf (red line) versus thickness of the detection area.
In Fig. 5(f), the 1 minus peak transmission (black line) and maximum Δf (red line) are plotted against the thickness of the detection area. The results indicate that the optimal values for both the resonance absorption peak amplitude and maximum Δf are obtained when the thickness of the detection area is around 50 µm. After determining the optimal thickness, it is possible to perform a quantitative calculation of the sensing sensitivity for the sensor. For CA solution, the biosensing sensitivity of the sensor is approximately 148 kHz·mL/cell. At the same time, we simulated the sensor's response ability to the refractive index of the analyte [31], and the experimental results are shown in Figure S13. Quantitative calculation shows that the refractive index sensing sensitivity of the THz sensor designed in this study is about 280 GHz/RIU.
3.2 Concentration sensing and type differentiation of three types of bacteria
Subsequent to the aforementioned research, we proceeded to carry out concentration sensing experiments on varying concentrations of CA, SA, and EC solutions, utilising a detection area thickness of 50 µm. The cell sizes of the three bacteria are CA: 4000-5000 nm, SA: 800-1000 nm and EC: 50 nm. The concentration gradients for each bacterium were the same as above, PBS1: 6 × 105 cell/mL, PBS2: 8 × 105 cell/mL, PBS3: 1 × 106 cell/mL and PBS4: 1.2 × 106 cell/mL, and the reference signal is a cell substrate signal without any bacteria. By measuring the THz transmission spectrum, we acquired the resonance frequency and resonance amplitude responses of three bacteria under different conditions. Figure 6 displays the experimental findings. Figures 6(a) and 6(b) demonstrate the detection of four concentrations of CA solutions in the detection area, both with and without the deposition of H type metasurface. From the experimental results, it can be seen that in the absence of metasurface, no resonance absorption peak can be generated and the CA solution has no intrinsic absorption peak in the range of 0-1.2 THz, so no concentration differentiation can be performed; after photolithography deposition of H type metasurface in the detection area, a weak resonance absorption peak can be generated at approximately 0.6 THz, which is consistent with the simulation results. We can use this absorption peak for concentration sensing of CA solution, but the sensing effect is poor and the biosensing sensitivity is low. Figures 6(c), 6(d), and 6(e) show the concentration sensing of CA, SA, and EC during the deposition of latch type metasurface in the detection area, and the experimental results show that the latch type metasurface can generate a significant resonance absorption peak at approximately 0.6 THz. During bacterial concentration sensing, an observable blue shift phenomenon occurs in the resonance absorption peak as the concentration of each bacterial type increases. This indicates a notable sensing effect.
Fig. 6. Plot of concentration sensing results for different concentrations of CA, SA and EC solutions with the sensor detection area thickness of 50 µm. (a) THz transmittance curve of CA solution with no metasurface in the detection area. (b) THz transmittance curve of CA solution with H type metasurface deposited in the detection area. THz transmittance curve of CA (c), SA (d), EC (e) solution with latch type metasurface deposited in the detection area. (f) Plot of three bacterial concentrations against the blue shift of the resonance absorption peak.
The fitting curve between bacterial concentration and resonance absorption peak blue shift is presented in Fig. 6(f). The results indicate a positive correlation between the concentration gradients of three distinct bacteria and their corresponding blue shift. This means that for a certain type of bacteria, we can quantitatively calculate its bacterial concentration based on its blue shift. The biosensing sensitivity of the THz microfluidic sensors prepared in this study was 148 kHz·mL/cell, 231 kHz·mL/cell and 301 kHz·mL/cell for CA, SA and EC bacteria respectively. Simultaneously, it was observed that a reduction in the size of bacteria leads to a gradual increase in the degree of resonance frequency blue shift. This is because as the size of bacteria decreases, more bacteria enter the gaps of the latch type metasurface, leading to an increase in capacitance and a more pronounced sensing effect. The blue shift amount can be used to qualitatively identify the type of bacteria among three different types, all present in equal concentrations.
In summary, the study presents a THz microfluidic biosensor capable of detecting and differentiating various types of bacteria by sensing three distinct bacterial concentrations.
3.3 Exploratory experiments: detection and differentiation of breast cancer cells
Breast cancer is a highly prevalent form of cancer worldwide. Its early clinical symptoms may not be readily apparent and can be easily missed. The detection of cancer cells often occurs after they have already metastasized, leading to the development of multi-organ lesions that pose a significant threat to the patient's life. Presently, the clinical identification of breast cancer relies on the utilisation of mammography, breast ultrasound, and breast MRI. Automated breast cancer detection systems lack the ability to make precise and conclusive assessments. As a result, they are frequently augmented by histopathological analyses of breast tissue sections. The successful detection of breast cancer cells on a specific scale holds significant medical importance [32–34]. As a conceptual verification, the THz microfluidic biosensor was utilised for conceptual verification purposes to detect and differentiate between two prevalent types of breast cancer cells, namely P53 mutant and P53 wild type. The cell models used were HCC1937 and HCC38, respectively. Further details regarding the culture procedure can be found the section “Materials” in the Supplement 1.
After multiple experiments, the resonance absorption peak of the two types of breast cancer cells did not exhibit a Δf in the hundreds of GHz when compared to the reference signal of the cell culture solution, as opposed to the bacterial concentration sensing results mentioned earlier. Figure 7(a) displays the experimental findings. The figure illustrates that the resonance absorption peak amplitude of the two breast cancer cells experiences a decline of approximately 6% compared to the reference signal, as a result of the blue shift of the peak frequency. The amplitude of the resonance absorption peak can be used to identify P53 breast cancer cells.
Fig. 7. Transmittance curves of cell cultures, P53 mutant (HCC1937) and P53 wild-type (HCC38) breast cancer cells on the biosensor (a) and concentrated in the range of 0.45-0.8 THz (b).
At the same time, the detection frequency is restricted to a range of 0.45-0.8 THz in order to exhibit specific details. The experimental results are shown in Fig. 7(b). For P53 mutant breast cancer cells (model: HCC1937) Δf is approximately 8 GHz, and the resonance absorption peak shifts from approximately 0.62 THz to 0.628 THz in red. For P53 wild-type breast cancer cells (model: HCC38), the resonance absorption peak is about 0.64 THz, Δf is approximately 20 GHz. The Δf parameter can be utilised to identify two distinct types of breast cancer cells.
The study presents a THz microfluidic biosensor capable of detecting breast cancer cells and cultures to a certain degree. The biosensor can also differentiate between the two prevalent types of breast cancer cells, namely P53 mutant and P53 wild type, at low concentrations. The biosensor is a fast and uncomplicated means of identifying breast cancer cells.
4. Conclusion
The present study involves the development and execution of a THz microfluidic biosensor that is highly integrated and based on metasurface and LT-GaAs. The LT-GaAs thin film PCA is utilised for the generation and detection of THz waves, which leads to a significant simplification of the complex THz-TDS system. Simultaneously, the integration of microfluidic and metasurface technologies was employed to fabricate a latch type metasurface. This metasurface is capable of producing a resonance absorption peak at 0.6 THz that is independent of polarisation. The metasurface sensor described herein showcases the capacity for detecting concentrations in a quantitative manner and recognising molecular types in a qualitative manner.
The optimal thickness of the detection area was calculated, and it was determined that the biosensor performs optimally when the thickness of the detection area is 50 µm. Subsequently, various concentrations of CA, SA, and EC solutions were evaluated with a detection area thickness of 50 µm. The dependence of resonance frequency shift and peak amplitude change on cell concentration was verified by measuring the THz transmission spectrum. It was observed that the degree of resonance frequency blue shift increases gradually as the size of the bacteria decreases. This suggests that the sensor is capable of detecting bacterial concentrations and distinguishing between different types of bacteria. As a proof of concept, two types of breast cancer cells were detected and differentiated. The detection of various cancer cell types can be achieved by measuring the amplitude and blue shift of the resonance peak. This method does not require the use of any antibodies. In brief, the THz microfluidic sensor exhibits extensive potential for use in biosensing applications.
Funding
Beijing Municipal Natural Science Foundation (4232066); National Natural Science Foundation of China (61575131); National Key Research and Development Program of China (No. 2021YFB3200100).
Acknowledgements
This work is funded by the National Key R&D Program of China (Grant No. 2021YFB3200100), the National Natural Science Foundation of China (61575131) and the Beijing Municipal Natural Science Foundation (4232066).
The authors would like to thank Enago for providing English proofreading.
Disclosures
There are no conflicts to declare.
Data availability
No data were generated or analyzed in the presented research.
Supplemental document
See Supplement 1 for supporting content.
References
1. X. Yang, X. Zhao, K. Yang, et al., “Biomedical Applications of Terahertz Spectroscopy and Imaging,” Trends Biotechnol. 34(10), 810–824 (2016). [CrossRef]
2. Y. Yang, Y. Yamagami, X. Yu, et al., “Terahertz topological photonics for on-chip communication,” Nat. Photonics 14(7), 446–451 (2020). [CrossRef]
3. B. Ferguson and X. C. Zhang, “Materials for terahertz science and technology,” Nat. Mater. 1(1), 26–33 (2002). [CrossRef]
4. R. T. Ako, A. Upadhyay, W. Withayachumnankul, et al., “Dielectrics for Terahertz Metasurfaces: Material Selection and Fabrication Techniques,” Adv. Opt. Mater. 8(3), 1900750 (2020). [CrossRef]
5. J. Li, C. Blake Wilson, R. Cheng, et al., “Spin current from sub-terahertz-generated antiferromagnetic magnons,” Nature 578(7793), 70–74 (2020). [CrossRef]
6. J. Lou, J. Liang, Y. Yu, et al., “Silicon-Based Terahertz Meta-Devices for Electrical Modulation of Fano Resonance and Transmission Amplitude,” Adv. Opt. Mater. 8(19), 2000449 (2020). [CrossRef]
7. M. Asgari, E. Riccardi, O. Balci, et al., “Chip-Scalable, Room-Temperature, Zero-Bias, Graphene-Based Terahertz Detectors with Nanosecond Response Time,” ACS Nano 15(11), 17966–17976 (2021). [CrossRef]
8. P. Pitchappa, A. Kumar, S. Prakash, et al., “Volatile Ultrafast Switching at Multilevel Nonvolatile States of Phase Change Material for Active Flexible Terahertz Metadevices,” Adv. Funct. Mater. 31, 2100200 (2021). [CrossRef]
9. G. Li, D. Xie, H. Zhong, et al., “Photo-induced non-volatile VO2 phase transition for neuromorphic ultraviolet sensors,” Nat. Commun. 13(1), 1729 (2022). [CrossRef]
10. T. L. Cocker, V. Jelic, R. Hillenbrand, et al., “Nanoscale terahertz scanning probe microscopy,” Nat. Photonics 15(8), 558–569 (2021). [CrossRef]
11. R. Zhou, C. Wang, Y. Huang, et al., “Label-free terahertz microfluidic biosensor for sensitive DNA detection using graphene-metasurface hybrid structures,” Biosens. Bioelectron. 188, 113336 (2021). [CrossRef]
12. C. X. Liu, F. Yang, X. J. Fu, et al., “Programmable Manipulations of Terahertz Beams by Transmissive Digital Coding Metasurfaces Based on Liquid Crystals,” Adv. Opt. Mater. 9, 2100932 (2021). [CrossRef]
13. T. C. Tan, Y. K. Srivastava, R. T. Ako, et al., “Active Control of Nanodielectric-Induced THz Quasi-BIC in Flexible Metasurfaces: A Platform for Modulation and Sensing,” Adv. Mater. (Weinheim, Ger.) 33, e2100836 (2021). [CrossRef]
14. S. Lee, S. Baek, T.-T. Kim, et al., “Metamaterials for Enhanced Optical Responses and their Application to Active Control of Terahertz Waves,” Adv. Mater. (Weinheim, Ger.) 32, e2000250 (2020). [CrossRef]
15. B. Chen, X. Wang, W. Li, et al., “Electrically addressable integrated intelligent terahertz metasurface,” Sci. Adv. 8(41), eadd1296 (2022). [CrossRef]
16. J. Xu, D. Liao, M. Gupta, et al., “Terahertz Microfluidic Sensing with Dual-Torus Toroidal Metasurfaces,” Adv. Opt. Mater. 9(15), 2100024 (2021). [CrossRef]
17. G. Choi, T. K. Chau, S. J. Hong, et al., “Augmented All-Optical Active Terahertz Device Using Graphene-Based Metasurface,” Adv. Opt. Mater. 9(16), 2100462 (2021). [CrossRef]
18. J. C. Deinert, D. Alcaraz Iranzo, R. Pérez, et al., “Grating-Graphene Metamaterial as a Platform for Terahertz Nonlinear Photonics,” ACS Nano 15(1), 1145–1154 (2021). [CrossRef]
19. H. M. Silalahi, Y.-P. Chen, Y.-H. Shih, et al., “Floating terahertz metamaterials with extremely large refractive index sensitivities,” Photonics Res. 9(10), 1970–1978 (2021). [CrossRef]
20. Z. Zhang, C. Zhong, F. Fan, et al., “Terahertz polarization and chirality sensing for amino acid solution based on chiral metasurface sensor,” Sens. Actuators, B 330, 129315 (2021). [CrossRef]
21. X. Chen, W. Fan, C. Song, et al., “Multiple plasmonic resonance excitations on graphene metamaterials for ultrasensitive terahertz sensing,” Carbon 133, 416–422 (2018). [CrossRef]
22. K. Lee, J. Park, B. J. Kang, et al., “Electrically Controllable Terahertz Second-Harmonic Generation in GaAs,” Adv. Opt. Mater. 8(18), 2000359 (2020). [CrossRef]
23. A. M. Buryakov, M.S. Ivanov, S.A. Nomoev, et al., “An advanced approach to control the electro-optical properties of LT-GaAs-based terahertz photoconductive antenna,” Mater. Res. Bull. 122, 110688 (2020). [CrossRef]
24. A. F. Bernardis, P. M. Sberna, J. Bueno, et al., “Time-Domain Modelling of Pulsed Photoconducting Sources-Part II: Characterization of an LT GaAs Bow-Tie Antenna,” IEEE Trans. Antennas Propag. 71(3), 2536–2545 (2023). [CrossRef]
25. K. M. Wang, Y. Gu, Y. Lu, et al., “Terahertz radiation enhancement in gallium arsenide nano-hole array under low power optical pump,” Appl. Phys. Lett. 121(7), 071703 (2022). [CrossRef]
26. H. Huang, S. Shao, G. Wang, et al., “Terahertz spectral properties of glucose and two disaccharides in solid and liquid states,” iScience 25(4), 104102 (2022). [CrossRef]
27. M. S. Islam, J. Sultana, M. Biabanifard, et al., “Tunable localized surface plasmon graphene metasurface for multiband superabsorption and terahertz sensing,” Carbon 158, 559–567 (2020). [CrossRef]
28. B. H. Fakhar, M. Ghazialsharif, M. S. Abrishamian, et al., “Graphene hybrid waveguide stimulation using a photoconductive terahertz generator,” Opt. Lett. 45(8), 2407–2410 (2020). [CrossRef]
29. B. Su and D. W. Wang, “The research of a novel electro-magnetically actuated valveless micropump,” Journal of Functional Materials & Devices (2011).
30. J. W. Wu, T. Yuan, J. Liu, et al., “Terahertz Metamaterial Sensor With Ultra-High Sensitivity and Tunability Based on Photosensitive Semiconductor GaAs,” IEEE Sensors J. 22(16), 15961–15966 (2022). [CrossRef]
31. Y. Z. Cheng, F. Chen, H. Luo, et al., “Triple-Band Perfect Light Absorber Based on Hybrid Metasurface for Sensing Application,” Nanoscale Res. Lett. 15(1), 103 (2020). [CrossRef]
32. F. Kazemi, “High Q-factor compact and reconfigurable THz aperture antenna based on graphene loads for detecting breast cancer cells,” Superlattices Microstruct. 153, 106865 (2021). [CrossRef]
33. A. Rezazadeh and M. R. Soheilifar, “THz absorber for breast cancer early detection based on graphene as multi-layer structure,” Opt. Quantum Electron. 53(10), 555 (2021). [CrossRef]
34. S. Alizadeh, E. Zareian-Jahrom, V. Mashayekhi, et al., “A tunable graphene-based refractive index sensor for THz bio-sensing applications,” Opt. Quantum Electron. 54 (2022).
