Open Access
Add to BibSonomy
Abstract
Optically controlled RF switches with a novel non-contact device architecture that achieves high performance in the millimeterwave-to-terahertz (mmW-THz) region are proposed and investigated through simulation. The significant change in conductivity in semiconductors caused by photogenerated carriers is used to develop RF switches having very high performance. By including a thin layer of insulator between the active semiconductor material and the metal contacts, the carrier concentration can be enhanced over that of conventional devices. For a prototype demonstration, G-band coplanar waveguide-based optical switches (using Si and Ge as active materials) with different contact geometries have been modeled and simulated. The proposed switches outperform both conventional solid-state switches and phase-change material-based switches in the switch figure-of-merit, and are promising for developing a novel class of tunable and reconfigurable mmW-THz circuits for advanced sensing, imaging, and communication.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
In recent years, electromagnetic (EM) waves in the millimeterwave-to-terahertz (mmW-THz) region have attracted extensive attention because of their promise for applications in radio astronomy, chemical detection, biometric imaging, wireless cognition, and precise positioning [1–4]. In these applications, tunable and/or reconfigurable mmW-THz circuits with high performance, flexibility, and reconfigurability are needed [5–8]. High performance RF switches are one of the most important building blocks for realizing such reconfigurable mmW-THz circuits [9]. However, current switch technologies do not deliver sufficient performance. Conventional solid-state switches are limited by their relatively low figure-of-merit (FOM=1/(2πRonCoff) [10]), significant insertion loss, and modest isolation at high frequencies [11]. Modern micro-electro-mechanical systems (MEMS) based switches offer superior RF performance (i.e., lower insertion loss and higher isolation), but the potential applications are limited by their limited reliability, complicated fabrication processes, and challenges for large-scale integration, especially in the mmW-THz region [12]. Emerging phase-change materials (PCM) such as VO2 have also been demonstrated in RF switches [11]. Although these switches have exhibited higher switch FOM than conventional solid-state switches, they often rely on thermally driven phase transitions, which leads to implementation challenges in practical circuits and systems.
To circumvent the limitations of existing devices, we propose a novel optically controlled integrated RF switch technology using a non-contact device architecture to achieve high performance in the mmW-THz region. The optical switching is based on the significant change in conductivity that can be achieved in semiconductors due to photogeneration of free carriers (e.g., ∼5-6 orders change in conductivity). In conventional photoconductive devices [13,14], however, the achievable RF performance (i.e., insertion loss and isolation) is not promising because of the limited photoconductivity. This is in part due to the lack of integrated optical source with high light intensity. But more importantly, this is also because conventional photoconductive switches use metal-semiconductor ohmic contacts; the photogenerated minority carriers rapidly recombine at the contacts leading to a much-reduced achievable carrier concentration and hence photoconductivity in these devices.
In order to improve the achievable performance for mmW-THz frequencies, a thin layer of insulator (e.g., SiO2) is introduced between the metal electrodes and the semiconductor active layer. This prevents contact-related recombination and leads to a higher photogenerated carrier concentration. By selecting the insulator thickness appropriately, the contact capacitance can be made high enough to efficiently couple mmW-THz signals in and out of the switch. In addition, the overall off-state capacitance of the proposed device is low due to the fully depleted active layer and simple device geometry. As a consequence, we find that optically controlled RF switches with superior performance (e.g., higher FOM than that of conventional solid-state and VO2 switches) in the mmW-THz region can be achieved.
In this paper, mmW-THz frequency switches based on these principles are modeled and simulated numerically. Physics-based calculation of the photoconductivity of Si and Ge for a range of illumination conditions is performed and used as the basis for the design and simulation of coplanar waveguide (CPW) based G-band (140-220 GHz) RF switches. We find that the on/off ratio of the Ge switches can be as high as 34.2 dB at 200 GHz with an insertion loss of only 0.62 dB. The simulations also show that an intrinsic FOM of 203.4 THz can be obtained for Ge active regions, while 49 THz can be achieved for Si active regions. This exceeds the performance of all other solid-state and PCM switches, and the Ge-based version approaches that of MEMS-based counterparts in the mmW-THz region. The results show that the proposed high performance RF switches are promising for tunable and reconfigurable circuits in the mmW-THz region for a wide range of applications, including sensing, imaging, and telecommunication.
2. Physics of Optical Switching
To achieve low on-state resistance in optically controlled switches, high carrier concentrations, and hence high light intensity, are needed. While a linear model for calculating the photogenerated carrier concentration and photoconductivity in semiconductors under low intensity illumination has been established and analyzed in Ref. [15], under strong illumination nonlinear Auger recombination may limit device performance. The effective carrier lifetime (τeff) of a semiconductor including Auger recombination can be expressed by [16]:
where τbulk is the bulk semiconductor carrier lifetime, γ is the Auger recombination coefficient of the material (τSi = 3×10−31 cm6/s, τGe = 1×10−31 cm6/s) [17,18], and n is the carrier concentration (electrons or holes). Assuming uniform illumination perpendicular to the wafer surface (i.e., along the z axis) and intrinsic semiconductor material (i.e., negligible carrier concentrations in the absence of illumination) the carrier distribution at steady state (i.e., the optical illumination is always on) can be determined by the 1-D continuity equation [15]:We further assume that surface recombination can be neglected since it is assumed that the top and bottom surfaces of the semiconductor are well passivated [15], leading to boundary conditions at z = 0 and z = H (semiconductor thickness):
Si and Ge are chosen here due to their large availability and easy access, as well as their relatively long intrinsic carrier lifetime compared with the compound semiconductor. Using Eq. (1)-(3), the carrier concentrations in Si and Ge thin films can be obtained as a function of semiconductor thickness (H) and light intensity (P). Typical results are shown in Fig. 1(a); for these examples, H was chosen to be 1 μm, 2 μm, and 5 μm, respectively, as Ref. [15] suggests that the carrier concentration increases as the semiconductor thickness is reduced. As can be seen from the plot, the carrier concentrations increase rapidly with optical intensity initially, and then saturate as the illumination intensity increases. This is a consequence of the onset of significant Auger recombination when the concentration is relatively high (∼1018 cm-3).
Fig. 1. (a) Photogenerated carrier concentration and (b) photoconductivity of Si and Ge as a function of incident light intensity.
The photoconductivity σ of the semiconductor can be calculated using σ = q(μn + μp)n, where μn and μp are the electron and hole mobility, respectively, and q is the elementary charge. Bipolar conduction is assumed because the material is intrinsic, and no external field is applied [19]. The conductivity results are shown in Fig. 1(b). As can be seen in this figure, thinner semiconductor films present higher photoconductivity, in agreement with Ref. [15], as a result of the higher achieved carrier concentration. It can also be seen that the conductivity of Ge is higher than that of Si under same illumination conditions due to its longer carrier lifetime, higher mobility, and smaller Auger coefficient. The conductivity of 1-μm-thick Ge can reach 260,000 S/m under 100 W/cm2 illumination, whereas that of Si could be as high as 48,000 S/m.
It should be noted that the carrier mobility, and hence the conductivity, in this device is projected to be higher than that achieved in MOSFETs. These optically controlled switches do not have high doping or inversion charge layers, leading to much lower scattering from ionized impurities or surface roughness. This effect has been experimentally demonstrated, with the photoconductivity of Si under illumination reaching 50,000 S/m [20], in close agreement with the calculated value above. In addition to the high conductivity under illumination, the optical switching material also presents high resistivity in the off-state. The material is intrinsic (σ < 5 S/m) in the absence of illumination. The significant change in conductivity (e.g., ∼5-6 orders) between the on- and off-state can be used in developing high performance RF switches. Finally, the carrier lifetime of the semiconductor also impacts the speed of the optical modulation [15]. The switching speed can be enhanced by choosing an active material with very short carrier lifetime (e.g., low temperature growth GaAs, or LT-GaAs). It has been experimentally demonstrated that the electrical transient of a photoconductive switch can be as short as 360 fs using LT-GaAs [21].
3. Novel RF switching architecture and modeling
The results in Fig. 1 show that high photoconductivity can be achieved in Ge and Si. However, in conventional photoconductive switch configurations the electrodes are connected directly to the active semiconductor, typically forming metal-semiconductor ohmic contacts [13]. As a result, photogenerated free carriers that reach the contacts rapidly recombine at the contacts. This dramatically lowers the practical achievable carrier concentration and photoconductivity, and hence degrades switch performance.
In order to circumvent this issue, a non-contact device architecture is proposed, as illustrated in Fig. 2. In this architecture, instead of forming direct metal-semiconductor contacts, a thin layer of insulator is inserted between the metal electrodes and the semiconductor, creating a non-contact, capacitively coupled configuration for the propagation of RF signals in the mmW-THz region.
Fig. 2. (a) Illustration of the optically controlled switch using Si or Ge as the active material with SiO2 isolation layer; (b) detailed device architecture (dimensions not to scale).
To explore the potential of this approach, a G-band CPW-based optically controlled RF switch has been designed and is illustrated in Fig. 2. The input and output lines of the series-configured switch are realized using 50 Ω (at 200 GHz) CPW transmission lines on a 100-μm-thick quartz substrate, corresponding to a 12-μm-wide center gold conductor and 2 μm gap between the signal and ground. The active region of the switch is a Si or Ge thin film located at the center conductor gap area with an exposed area having a length of L and width of W. The thickness of the film is initially chosen to be 1 μm on the basis of the results in Fig. 1. As shown in Fig. 2(b), the insulator layer is SiO2 with a thickness of 5 nm and covers the active material. As the thickness of the insulation layer is small, it is assumed that all optical power can reach the Si/Ge active region through the 5 nm-thick oxide. The metal contact connects the sidewall of the insulator layer in the “L” direction and additionally overlaps the top of the active region by 0.2 µm, over the full width (W) of the semiconductor active area. The active material (Si or Ge) on quartz substrate can be fabricated through direct wafer bonding [22] and chemical mechanical polishing (CMP) [23], followed by photolithography and reactive-ion-etching (RIE) to define the active region. The thin oxide layer can be deposited onto the semiconductor by using atomic layer deposition. For optical switching, the active area can be illuminated by a transfer-printed micro-LED, whose irradiance has been experimentally demonstrated to be as high as 822.4 W/cm2 [24,25]. It should also be noted that the modulation bandwidth of a GaN micro-LED has been demonstrated to be as high as 752 MHz [26], which is higher than the optical modulation bandwidth of Si reported in Ref. [27]. Therefore, the operation speed of the proposed optically controlled switches will not be limited by the micro-LED.
In order to maximize switch performance, our modeling suggests that the contact insulation layer is critical. Initial simulations show that with ohmic contacts, the rapid recombination (surface recombination velocity > 104 cm/s) at the contacts plus carrier diffusion leads to a much-reduced photoconductivity (e.g., only 3,800 S/m for 1-µm-thick Si with an active area of 12 μm × 12 μm under 100 W/cm2 illumination) that is less than 10% of that with insulated contacts (48,000 S/m from the results in Section 2). Consequently, the on-state series impedance (263.8 Ω, estimated from 3,800 S/m) is nearly 10 times as large as that with insulated contacts (21.6 Ω). Of course, the capacitively coupled switch sacrifices DC and low frequency switching response, but signals in the mmW-THz frequency range propagate easily through the insulator with minimal attenuation by way of capacitive coupling. In fact, the series capacitance of a 5 nm-thick oxide (estimated to be around 100 fF for a contact area of 14.4 µm2 corresponding to W = 12 µm) results in a lower insertion loss at 200 GHz than that of the typical contact resistance in conventional switch configurations (∼0.05 Ω·mm per contact). The net result is a lower overall effective on-state resistance (Ron), and hence a higher achievable intrinsic FOM, compared with conventional solid-state switches.
The proposed switch architecture not only lowers the effective Ron, but also maintains a low overall off-state capacitance (Coff). Since the active material of the optically controlled switch does not include any pn junctions, Coff consists solely of the geometrically limited the dielectric capacitance, which is lower than the junction capacitance associated with most FET-based switches.
To investigate and demonstrate the potential RF performance of the proposed switches, physics-based semiconductor modeling coupled to full-wave electromagnetic simulation was performed following the process described in Ref. [11]. To evaluate switch performance, the S-parameters of the proposed RF switches were obtained by using the EM simulation software HFSS [28] and de-embedded to the reference planes (illustrated in Fig. 2(a)). As an example, the S-parameters for a switch with the geometry of Fig. 2, using Si as the active material and with L = 12 μm and W = 12 μm are shown in Fig. 3(a) as the solid and dotted lines.
Fig. 3. (a) S-parameters from HFSS (solid and dotted lines) and circuit model (discrete points); (b) lumped-element circuit model for the on-state; (c) lumped-element model for the off-state.
In order to extract the effective Ron and Coff values for calculating the switch FOM, lumped-element models for the on- and off-states were analyzed. The on-state switch model includes a resistor Ron, an inductor Lon, and two capacitors Cox in series (one each from the insulation layer at each contact), with two shunt pad capacitors as illustrated in Fig. 3(b). The off-state model (Fig. 3(c)) contains a resistor Roff and a parallel capacitor Coff, with the same shunt pad capacitors. Cox is not included in the off-state model because in the intrinsic semiconductor case the semiconductor region is depleted. The values of the Ron and Coff are extracted by using the corresponding Y- and Z-matrix elements, as shown in the following expressions [29]:
The value of Cox can also be extracted from the on-state Y- matrix elements and is calculated to be 89 fF, which is reasonably close to our 100 fF estimation. The S-parameters calculated by using the lumped-element model are shown in Fig. 3(a) as discrete symbols. It can be seen that the results agree well with the HFSS simulation, indicating that the lumped-element model captures the behavior of the device accurately.
In this FOM, the contribution to insertion loss from the contact series capacitance Cox is not included; the Ron as calculated is the resistive contribution from the semiconductor. However, the reactive impedance of the Cox is a negligible contributor to insertion loss due to the significant (∼ 100 fF) capacitance achieved. For example, the insertion loss from one 89 fF capacitor is 0.035 dB at 200 GHz, whereas that from the resistive semiconductor in the above case is 1.68 dB. This insertion loss from the capacitor can be further reduced by engineering the insulation layer (e.g., using Al2O3 as the insulation material for higher dielectric constant).
4. Switch simulation results
4.1 Switches with gap geometry
Using the non-contact device configuration described above, the basic switch structure with a “gap” connection geometry (Fig. 2) was analyzed. The dimensions of the active area were investigated to understand their impact on performance. The on-state light illumination was fixed at 100 W/cm2. As an example, the S-parameters of switches with an active area length L = 12 µm and width W = 12 µm are shown for Ge and Si in Fig. 4(a) and (b) in the on- and off-state, respectively. The typical on/off ratio of the switch using Ge with L = 12 μm and W = 12 μm is 31.6 dB with an insertion loss of 0.53 dB at 200 GHz, demonstrating the potential for developing high performance mmW-THz switches. For Si, the insertion loss at 200 GHz is 1.80 dB with the same gap dimensions. The higher insertion loss of Si switch is mainly due to its lower photoconductivity.
Fig. 4. Typical S-parameters for (a) Ge and Si switches with L = W=12 µm in the on-state; (b) in the off- state.
To understand these results, the lumped circuit elements Ron and Coff were extracted from simulations of switch having a range of different active area dimensions; the results are shown in Fig. 5. Gap lengths from 1-15 µm were evaluated, as were widths from 4-12 µm. As can be seen in Fig. 5, as the device length L increases, Ron also increases while Coff decreases, as expected. In addition, Ron decreases as W increases, while Coff increases with increasing W. The resulting trends in switch FOM are shown in Fig. 6. It can be seen that for both Si and Ge switches, the FOM increases with increasing width W and length L, though the dependence on L is much weaker for Si than for Ge. From the above modeling and analysis, it is seen that an intrinsic FOM as high as 199.8 THz and 47.5 THz can be achieved using Ge and Si, respectively, for an active area of 12 μm × 15 μm.
4.2 Switches with inter-digitated geometry
In order to further improve the switch FOM, two additional switch geometries, namely an inter-digitated geometry and an “island” geometry, have been investigated. As shown in the inset of Fig. 7(a), the inter-digitated geometry consists of five gold fingers on top of the insulation and active semiconductor layers. The inter-digitated connection geometry reduces the on-state resistance of the switches (by reducing the effective path length between contacts), at the expense of an increased off-state capacitance.
Fig. 7. (a) Illustration of the inter-digit structure (inset, dimensions not to scale), and its corresponding S-parameters; (b) Calculated FOM with different finger length g.
For this study, the dimension of the active region is fixed to be 12 μm × 15 μm, consistent with the highest FOM for switches with the gap geometry discussed in Section 4.1. The light intensity is also kept the same as in Section 4.1, at 100 W/cm2. Switches with varying finger length (g), width (t), and inter-finger gap (u) were investigated. As an example, the S-parameters corresponding to g = 10 μm, t = 1 μm, and u = 1 μm are presented in Fig. 7(a). With this inter-digitated structure, the insertion loss of the Ge-based switch at 200 GHz drops to 0.265 dB, and that of Si switch also drops to 0.89 dB. However, the isolation of the switch in the off-state also decreases, to 17.1 dB and 18.6 dB for Ge and Si, respectively. By following the same procedures discussed above, the effective Ron and Coff of the devices were extracted and the FOM was calculated. The resulting FOMs for an interelectrode gap u = 1 μm and finger widths t = 0.5 and 1 µm as a function of finger length g are plotted in Fig. 7(b). It can be seen that a peak intrinsic FOM of 201 THz can be realized when the finger length g is 1 μm and t is 0.5 μm for the Ge-based switch. The maximum FOM for Si switch is 48.2 THz with the same finger dimensions. As g is increased beyond approximately 1 µm, the overall FOM drops, because the increase in Coff exceeds the decrease in Ron. It can also be seen from the figure that the FOM decreases slightly when the finger width t increases from 0.5 μm and to 1 μm, also as a consequence of the increased Coff.
4.3 Switches with “island” geometry
Another device geometry of interest is the “island” geometry as shown in the inset of Fig. 8(a). This geometry consists of a gold square at the center of the active area, on top of the contact insulating layer. Compared to the switches with the simple gap geometry, the insertion loss in the on-state is lower since the Ron is lower (due to the lower effective path length). For off-state, on the other hand, the off-state capacitance is modestly increased due to the presence of the island. The impact of geometry on the switch FOM for this design option was evaluated in simulation. In this analysis, the dimensions of the active region (dashed rectangle) were selected to be 12 µm × 15 µm, consistent with Sections 4.1 and 4.2 above. An example of switch S-parameters, for the case of an island length (p) of 1 µm and width (q) of 4 µm, is shown in Fig. 8(a). The insertion loss of the Ge switch under illumination becomes 0.61 dB at 200 GHz, whereas the isolation for off-state is 34.85 dB.
Fig. 8. (a) Illustration of the gold island structure (inset, dimensions not to scale), and its corresponding S-parameters; (b) Calculated FOM with different “island” width q and length p.
The FOMs for a range of p and q dimensions are shown in Fig. 8(b). It can be seen that for smaller island length p (p ≤ 1 µm), the calculated FOMs for both Si and Ge switches keep almost constant with varying island width q. This can be explained by noting that the additional Coff introduced by the island structure is compensated by a corresponding decrease in Ron. When the island length becomes large (p ≥ 2 µm), however, the FOMs drop significantly with the increasing island width q. This is because the increase in Coff becomes appreciably larger than the decrease in Ron in this situation. In terms of the island’s longitudinal dimension p, the FOMs first increase (from p = 0 to 0.5 µm) and then decrease as the island length p increases further for both Si and Ge switches. The peak value can be achieved when the island length is approximately 0.5 µm. This is because the Ron of the switch with 0.5-µm-long island is noticeably lower than that of the switch with the simple gap geometry. On the other hand, the increase in Coff is not significant when adding this small island structure. Therefore, the resulting FOM is larger than that of switches with the simple gap geometry. As the island length is increased further, the increment in Coff becomes more significant, leading to a lower FOM. The highest FOM of 203.4 THz was obtained using Ge for an island dimension (p × q) of 0.5 µm × 4 µm. Similarly, the FOM for a Si active area can reach 49 THz, with the same gold island dimensions.
In these simulations, the photoconductivity of the active material is assumed to be uniform; for dark regions with no direct illumination, the photogenerated free carriers diffuse into these areas. For the geometries explored here, the optically occluded regions are small relative to the diffusion length (61 µm for Si and 505 µm for Ge) [19,30]. In addition, the occluded area is < 10% of the active area in all cases, so little of the incident light is lost from reflection. As a result, the blockage of the light illuminating from the interdigitated or island surface structures has negligible impact on either the photo carrier concentration or the photoconductivity of the material. We should note, however, that even this small effect can be eliminated by using backside illumination, since the transmittance of the quartz substrate is larger than 99% at visible wavelengths [31].
5. Switch benchmarking
In order to benchmark the optically controlled RF switches, their performance is compared to that of other existing technologies (e.g., conventional solid-state switches, VO2-based switches, etc.). Figure 9(a) shows a comparison of the S12, on/S12, off ratio of the proposed RF switches vs. state-of-the-art FET and PCM-based switches vs. frequency [11,32–35]. As can be seen, the performance of the non-contact switches studied here are inferior to other technologies at low frequencies, due to the capacitively coupled contacts. However, as the frequency increases to a few tens of gigahertz, the performance of the proposed switches exceeds that of other technologies. At 20 GHz, the on/off ratio of the optically controlled Si switch is 4 dB higher than a VO2 switch reported in [11], and 16 dB higher than the GaN switch reported in [36]. The ratio is even higher when using Ge as the active material due to its higher photoconductivity. In addition, while the on/off ratio for all of the evaluated switches decreases with frequency, the switches proposed here maintain a higher on/off ratio well into the mmW-THz region (i.e., 34.2 dB for Ge and 25.4 dB for Si at 200 GHz). It is worth noting that the CPW transmission line designed in this report works properly for the broad range of frequencies (1 to 200 GHz as in Fig. 9(a)), with its simulated attenuation constant ranging from 0.3 to 1.6 dB/mm. Since the length of the CPW is usually very small in the mmW-THz region, the potential applications will not be limited by the CPW transmission line. In addition, the dimensions of the CPW transmission line, as well as the active region of the proposed switch, may be further optimized for the applications in other frequencies in order to achieve better performance.
Fig. 9. Comparison of (a) S12,on/S12,off ratio and (b) FOM between the optically controlled switches reported here and other technologies reported in Ref. [11,32–42]
In addition to on/off ratio, Fig. 9(b) illustrates the achievable Ron and Coff trade-space for a range of RF switch technologies [11,32–42]; the lower left corner represents the highest FOM. As discussed above, a switch FOM of 203.4 THz is projected for optically controlled switches using Ge, outperforming the conventional semiconductor-based switches as well as PCM-based switches. The FOM of the proposed switches using Ge also is competitive with, and in some cases exceeds, those of some MEMS devices [41,42]. The high FOM is primarily due to high conductivity of the active material under illumination, which is further facilitated by the insulating nature of the capacitively coupled contacts. In addition, Coff of the proposed switches consists of the geometrically limited dielectric capacitance, which is similar to the case of VO2 switches (but with a lower off-state permittivity), leading to an advantage in FOM for the proposed devices. Table 1 summarize the performance of the optically controlled switches reported here and other technologies.
Table 1. Comparison summary of the optically controlled switches reported here and other technologies
6. Conclusion
In this paper, high-performance optically controlled RF switches with a novel non-contact configuration for mmW-THz circuits were proposed and analyzed. The simulation results indicate that an intrinsic FOM of 203.4 THz can be potentially achieved while maintaining high on/off ratio of 34.2 dB using Ge as the active material at 200 GHz. Such performance is superior to that of conventional solid state and emerging PCM switches and approaches the performance of some MEMS switches. This indicates the promise of this approach for applications in advanced reconfigurable mmW-THz circuits for sensing and imaging, and telecommunication.
Funding
National Science Foundation (ECCS-1711052, ECCS-1711631).
Acknowledgments
The authors would like to thank the support from the Advanced Diagnostics and Therapeutics Initiative (AD&T), the Center for Nanoscience and Technology (NDnano), and the Center for Research Computing (CRC) at the University of Notre Dame.
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.
References
1. T. G. Phillips and J. Keene, “Submillmeter astronomy,” Proc. IEEE 80(11), 1662–1678 (1992). [CrossRef]
2. A. Markelz, A. Roitberg, and E. J. Heilweil, “Pulsed terahertz spectroscopy of DNA, bovine serum albumin and collagen between 0.1 and 2.0 THz,” Chem. Phys. Lett. 320(1-2), 42–48 (2000). [CrossRef]
3. I. S. Gregory, W. R. Tribe, B. E. Cole, C. Baker, M. J. Evans, I. V. Bradley, E. H. Linfield, A. G. Davies, and M. Missous, “Phase sensitive continuous-wave THz imaging using diode lasers,” Electron. Lett. 40(2), 143–145 (2004). [CrossRef]
4. T. S. Rappaport, Y. Xing, O. Kanhere, S. Ju, A. Madanayake, S. Mandal, A. Alkhateeb, and G. C. Trichopoulos, “Wireless Communications and Applications Above 100 GHz: Opportunities and Challenges for 6G and Beyond,” IEEE Access 7, 78729–78757 (2019). [CrossRef]
5. J. E. Muller, T. Bruder, P. Herrero, N. Norholm, P. Olesen, J. Rizk, and L. Schumacher, “Requirements for reconfigurable 4G front-ends,” IEEE MTT-S International Microwave Symposium Digest (MTT), 1–4 (2013).
6. F. Ma, Y. S. Lin, X. Zhang, and C. Lee, “Tunable multiband terahertz metamaterial using a reconfigurable electric split-ring resonator array,” Light: Sci. Appl. 3(5), e171 (2014). [CrossRef]
7. B. S. Rodriguez, R. Yan, L. Liu, D. Jena, and H. Xing, “Graphene for Reconfigurable Terahert Optoelectronics,” Proc. IEEE 101(7), 1705–1716 (2013). [CrossRef]
8. Y. Deng, J. Ren, Y. Shi, Y. Wang, L. Cheng, P. Fay, and L. Liu, “Advanced photo-induced substrate-integrated waveguides using pillar-array structures for tunable and reconfigurable THz circuits,” Opt. Express 28(5), 7259–7273 (2020). [CrossRef]
9. J. T. Bernhard, “Reconfigurable Antennas,” Synth. Lect. Antennas 2(1), 1–66 (2007). [CrossRef]
10. G. M. Rebeiz, C. D. Patel, S. K. Han, C. Ko, and K. M. J. Ho, “The Search for a Reliable MEMS Switch,” IEEE Microwave 14(1), 57–67 (2013). [CrossRef]
11. H. Madan, H. T. Zhang, M. Jerry, D. Mukherjee, N. Alem, R. Engel-Herbert, and S. Datta, “26.5 Terahertz electrically triggered RF switch on epitaxial VO2-on-Sapphire (VOS) wafer,” IEEE International Electron Devices Meeting (IEDM), 9.3.1-9.3.4 (2015).
12. R. R. Mansour, “RF MEMS-CMOS Device Integration: An Overview of the Potential for RF Researchers,” IEEE Microwave 14(1), 39–56 (2013). [CrossRef]
13. S. Lee, S. Lee, U. Ketprom, and Y. Kuga, “Optically CW-mode controlled microwave switches with carrier-confinement on a coplanar waveguide,” IEEE Antennas and Propagation Society International Symposium, 514–517 (2001).
14. C. Y. Fang, H. H. Lin, M. Alouini, Y. Fainman, and A. El Amili, “Microwave signal switching on a silicon photonic chip,” Sci. Rep. 9(1), 11166 (2019). [CrossRef]
15. A. Kannegulla, M. I. Bin Shams, L. Liu, and L. Cheng, “Photo-induced spatial modulation of THz waves: opportunities and limitations,” Opt. Express 23(25), 32098–32112 (2015). [CrossRef]
16. M. Rosling, H. Bleichner, M. Lundqvist, and E. Nordlander, “A novel technique for the simultaneous measurement of ambipolar carrier lifetime and diffusion coefficient in silicon,” Solid-State Electron. 35(9), 1223–1227 (1992). [CrossRef]
17. J. Dziewior and W. Schmid, “Auger coefficients for highly doped and highly excited silicon,” Appl. Phys. Lett. 31(5), 346–348 (1977). [CrossRef]
18. R. Conradt and J. Aengenheister, “Minority carrier lifetime in highly doped Ge,” Solid State Commun. 10(3), 321–323 (1972). [CrossRef]
19. J. Ren, Z. Jiang, M. I. B. Shams, P. Fay, and L. Liu, “Photo-Induced Electromagnetic Band Gap Structures for Optically Tunable Microwave Filters,” Prog. Electromagn. Res. 161, 101–111 (2018). [CrossRef]
20. H. Chen, J. F. O’hara, A. K. Azad, A. J. Taylor, R. D. Averitt, D. B. Shrekenhamer, and W. J. Padilla, “Experimental demonstration of frequency-agile terahertz metamaterials,” Nat. Photonics 2(5), 295–298 (2008). [CrossRef]
21. X. Zheng, Y. Xu, R. Sobolewski, R. Adam, M. Mikulics, M. Siegel, and P. Kordoš, “Femtosecond response of a free-standing LT-GaAs photoconductive switch,” Appl. Opt. 42(9), 1726–1731 (2003). [CrossRef]
22. C. Wang, Y. Wang, Y. Tian, C. Wang, and T. Suga, “Room-temperature direct bonding of silicon and quartz glass wafers,” Appl. Phys. Lett. 110(22), 221602 (2017). [CrossRef]
23. M. Krishnan, J. W. Nalaskowski, and L. M. Cook, “Chemical Mechanical Planarization: Slurry Chemistry, Materials, and Mechanisms,” Chem. Rev. 110(1), 178–204 (2010). [CrossRef]
24. T. Wu, C. Sher, Y. Lin, C. Lee, S. Liang, Y. Lu, S. H. Chen, W. Guo, H. Kuo, and Z. Chen, “Mini-LED and Micro-LED: Promising Candidates for the Next Generation Display Technology,” Appl. Sci. 8(9), 1557 (2018). [CrossRef]
25. I. Cho, Y. C. Sim, M. Cho, Y. Cho, and I. Park, “Monolithic Micro Light-Emitting Diode/Metal Oxide Nanowire Gas Sensor with Microwatt-Level Power Consumption,” ACS Sens. 5(2), 563–570 (2020). [CrossRef]
26. H. Lan, I. Tseng, H. Kao, Y. Lin, G. Lin, and C. Wu, “752-MHz Modulation Bandwidth of High-Speed Blue Micro Light-Emitting Diodes,” IEEE J. Quantum Electron. 54(5), 1 (2018). [CrossRef]
27. J. Ren, Z. Jiang, P. Fay, J. L. Hesler, C. E. Tong, and L. Liu, “High-Performance WR-4.3 Optically Controlled Variable Attenuator With 60-dB Range,” IEEE Microw. Wireless Compon. Lett. 28(6), 512–514 (2018). [CrossRef]
28. https://www.ansys.com/products/electronics/ansys-hfss
29. H. Madan, “Design and characterization of novel RF circuit based on emerging material systems,” Ph.D. Dissertation, The Pennsylvania State University, PS, (2016).
30. P. Würfela, T. Trupkeb, and T. Puzzer, “Diffusion lengths of silicon solar cells from luminescence images,” J. Appl. Phys. 101(12), 123110 (2007). [CrossRef]
31. L. Gao, F. Lemarchand, and M. Lequime, “Refractive index determination of SiO2 layer in the UV/Vis/NIR range: spectrophotometric reverse engineering on single and bi-layer designs,” J. Europ. Opt. Soc. Rap. Public. 8, 13010 (2013). [CrossRef]
32. N. El-Hinnawy, P. Borodulin, E. B. Jones, B. P. Wagner, M. R. King, J. S. Mason, J. Bain, J. Paramesh, T. E. Schlesinger, R. S. Howellet, M. J. Lee, and R. M. Young, “12.5 THz Fco GeTe Inline Phase-Change Switch Technology for Reconfigurable RF and Switching Applications,” 2014 IEEE Compound Semiconductor Integrated Circuit Symposium (CSICS), 1–3 (2014)
33. R. S. Howell, E. J. Stewart, R. Freitag, J. Parke, B. Nechay, H. Cramer, M. King, S. Gupta, J. Hartman, P. Borodulin, M. Snook, I. Wathuthanthri, P. Ralston, K. Renaldo, and H. G. Henry, “Low Loss, High Performance 1-18 GHz SPDT Based on the Novel Super-Lattice Castellated Field Effect Transistor (SLCFET),” 2014 IEEE Compound Semiconductor Integrated Circuit Symposium (CSICS), 1–5 (2014)
34. Q. Xiao, G. Samiotes, T. Galluccio, and B. Rizzi, “A high performance DC-20 GHz SPDT switch in a low cost plastic QFN package,” 2009 European Microwave Conference (EuMC), 1673–1676 (2009)
35. S. D. Ha, Y. Zhou, C. J. Fisher, S. Ramanathan, and J. P. Treadway, “Electrical switching dynamics and broadband microwave characteristics of VO2 radio frequency devices,” J. Appl. Phys. 113(18), 184501 (2013). [CrossRef]
36. R. S. Howell, E. J. Stewart, R. Freitag, J. Parke, B. Nechay, H. Cramer, M. King, S. Gupta, J. Hartman, P. Borodulin, M. Snook, I. Wathuthanthri, P. Ralston, K. Renaldo, H. G. Henry, and R. C. Clarke, “The Super-Lattice Castellated Field Effect Transistor (SLCFET): A novel high performance Transistor topology ideal for RF switching,” 2014 IEEE International Electron Devices Meeting, 11.5.1–11.5.4 (2014)
37. N. Wainstein, G. Ankonina, S. Kvatinsky, and E. Yalon, “Compact Modeling and Electrothermal Measurements of Indirectly Heated Phase-Change RF Switches,” IEEE Trans. Electron Devices 67(11), 5182–5187 (2020). [CrossRef]
38. H. Kamitsuna, Y. Yamane, M. Tokumitsu, H. Sugahara, and M. Muraguchi, “Low-power InP-HEMT switch ICs integrating miniaturized 2/spl times/2 switches for 10-Gb/s systems,” IEEE J. Solid-State Circuits 41(2), 452–460 (2006). [CrossRef]
39. A. Botula, A. Joseph, J. Slinkman, R. Wolf, Z.-X. He, D. Ioannou, L. Wagner, M. Gordon, M. Abou-Khalil, R. Phelps, M. Gautsch, W. Abadeer, D. Harmon, M. Levy, J. Benoit, and J. Dunn, “A Thin-Film SOI 180 nm CMOS RF Switch Technology,” 2009 IEEE Topical Meeting on Silicon Monolithic Integrated Circuits in RF Systems, 1–4 (2009)
40. C. Li, B. Ustundag, A. Kumar, M. Boenke, U. Kodak, and G. Rebeiz, “< 0.8 dB IL 46 dBm OIP3 Ka band SPDT for 5G communication,” 2018 IEEE 18th Topical Meeting on Silicon Monolithic Integrated Circuits in RF Systems (SiRF), 1–3 (2018)
41. R. R. Benoit, R. Q. Rudy, J. S. Pulskamp, and R. G. Polcawich, “Piezoelectric RF MEMS Switches on Si-on-Sapphire Substrates,” J. Microelectromech. Syst. 29(5), 1087–1090 (2020). [CrossRef]
42. P. D. Grant, M. W. Denhoff, and R. R. Mansour, “A Comparison between RF MEMS Switches and Semiconductor Switches,” 2004 International Conference on MEMS, NANO and Smart Systems (ICMENS'04), 515–521, (2004)
