Electronically Switchable Broadband Metamaterial Absorber

In this study, the novel electronically switchable broadband metamaterial absorber, using a PIN diode, is proposed. The unit cell of the absorber was designed with a Jerusalem-cross resonator and an additive ring structure, based on the FR-4 dielectric substrate. Chip resistors and PIN diodes were used to provide both a broadband characteristic and a switching capability. To satisfy the polarization insensitivity, the unit cell was designed as a symmetrical structure, including the DC bias network, electronic devices, and conductor patterns. The performance of the proposed absorber was verified using full-wave simulation and measurements. When the PIN diode was in the ON state, the proposed absorber had a 90% absorption bandwidth from 8.45–9.3 GHz. Moreover, when the PIN diode was in the OFF state, the 90% absorption bandwidth was 9.2–10.45 GHz. Therefore, the absorption band was successfully switched between the low-frequency band and the high-frequency band as the PIN diode was switched between the ON and OFF states. Furthermore, the unit cell of the proposed absorber was designed as a symmetrical structure, and its performance showed insensitivity with respect to the polarization angle.


Results
Absorber design. Figure 1 illustrates the geometry of the proposed absorber unit cell. On the top layer of the unit cell, conductor patterns are composed of the Jerusalem cross (JC) resonator, the additive ring structure, and the DC bias network. Moreover, there are four chip resistors, four PIN diode, and four chip inductors. Generally, LC resonance occurs in the JC resonator, and an electric field is generated. The additive ring structure can serve to gather the generated electric field inwards, toward the unit cell, and chip resistors are placed at this location to provide broadband operation. PIN diodes are used as switchable component, for the ON/OFF state. A separate DC bias network is required to operate the PIN diode. In this case, in order to minimize the performance limitation of the absorber due to the DC bias network, the cathode is designed as a conductor portion at the centre of the unit cell and the anode is designed as a conductor portion at the outermost part of the unit cell. The bottom layer of the unit cell is a fully covered conductor, to prevent the transmission wave. In addition, because the cathode conductor portions on the top layer and bottom layer are connected by via holes, the PIN diode can be powered through the bottom conductive layer. In the design process, to satisfy the polarization insensitivity, the unit cell is designed as both a horizontally and vertically symmetrical structure, including the DC bias network, electronic devices, and conductor patterns.
Fabricated absorber prototype. To verify the performance of the designed absorber, a prototype sample was fabricated, as shown in Fig. 2. The fabricated samples consisted of a total of 100 unit cells with a unit size of 10 × 10 and a total size of 162 × 162 mm. The FR-4 dielectric substrate was used and its relative permittivity and dielectric loss are 4.5 and 0.02, respectively. The conductor patterns on the top and bottom layers were implemented using a wet etching process. Electronic devices such as PIN diodes, chip resistors, and chip inductors were soldered on by SMT. In the fabricated prototype sample, a total of 400 chip resistors, 400 chip inductors, and 400 PIN diodes, were used. The value and the size of the chip resistors are 100 Ω and 0603 size (metric). In order to prevent the radio frequency (RF), chip inductors are used. The chip inductor used the 0402HP-3NSX_L and its value and size are 3.3 nH and 1005 size (metric). Moreover, because its self resonant frequency (SRF) is 12.8 GHz, this chip inductor is suitable for the proposed absorber, which operates in the X-band (8)(9)(10)(11)(12). The wires are connected to the outermost part of the top and bottom layers of the fabricated prototype, in order to supply the PIN diodes with power.
Simulated and experimental results. In this study, a SMP1340-079LF PIN diode provided by Skyworks Solutions, Inc. is used. The equivalent circuit of the PIN diode is represented by R (resistor), L (inductance), and C (capacitance) as its ON and OFF states 20 . When the PIN diode is in the OFF state, it is modelled as a series L, a parallel C 2 , and a high-value R P . When the diode is in the ON state, it is also modelled as an L and R 2 in series. The component values are L = 0.45 nH, R 2 = 5 Ω, C 2 = 0.03 pF, and R P = 5 MΩ. Consequently, the proposed absorber can be expressed as a transmission line model, consisting of three section, A, B, and C, as shown in Fig. 3(a). The first section, ' A' , demonstrates the top layer of the proposed absorber. The conductive patterns on the top layer are represented by C 1 , R 1 , and L in series, and are switched to C 2 and R 2 as the PIN diodes is switched between the ON and OFF state, respectively. In the section, 'B' , the finite length of the transmission line (L SUB ) represents the dielectric substrate, with an intrinsic impedance of η SUB . Finally, section 'C' represents the ground (GND) plane as a load impedance of Z L .
The reflection coefficient is given by the flowing equation:   Consequently, the zero reflection coefficient is achieved when Y in and Y 0 are equal. Figure 3(b) and (c) show the input impedance of the transmission line model from 8 to 12 GHz on Smith chart. L SUB and R 1 are critical parameters for broadband impedance matching. For instance, as L SUB is increased as shown in Fig. 3(b), the impedance trace is approaching to the center. In addition, when L SUB is 3.2 mm, the impedance trace is inside 2:1 VSWR (voltage standing wave ratio) circle. From Fig. 3(c), the impedance trace is approaching to the center as R 1 increases. Especially, when R 1 is 300 Ω, the impedance trace is inside 2:1 VSWR circle. Figure 4 shows the variation of the absorptivity with the change of the design parameters such as d, h, j, and the value of the chip resistor (r). In the design process, since the absorption frequency is switched depending on the state of the PIN diode, the design parameters satisfying both conditions are determined through full-wave simulation. Firstly, Fig. 4(a) and (b) show the variation of the absorptivity with the change of the width of JC, the width being represented by 'd' . Regardless of the state of the PIN diode, the absorption frequency band increases as the value of 'd' decreases from 2 mm to 1 mm. However, when the PIN diode is in the OFF state, the absorption bandwidth is the widest and most stable at a 'd' of 1.5 mm. Secondly, the change of the absorptivity is shown in Fig. 4(c) and (d), when the radius 'h' , of the additive ring structure, varies from 2.5 mm-4.5 mm. When the PIN diode is in the ON state, the absorptivity is best at an 'h' of 3.5 mm. Furthermore, when the PIN diode is in the OFF state, the absorption bandwidth is the widest and most stable at an 'h' of 3.5 mm. Next, Fig. 4(e) and (f) show the variation of the absorptivity when the width of stub 'j' varies from 1.5 mm-4.5 mm. In this case, when the PIN diode is in the OFF state, there are slight differences at various values of 'j' . However, when the PIN diode is in the ON state, the absorptivity is best, and the absorbing bandwidth is the widest at a 'j' of 1.5 mm. Finally, Fig. 4(g) and (h) show the change of the absorptivity when the value of the chip resistors (r), ranges from 100 Ω-500 Ω. When the PIN diode is in the ON state, the absorptivity is best, and the bandwidth is widest at an 'r' of 100 Ω. Likewise, when the PIN diode is in the OFF state, the absorptivity is best, and the absorption bandwidth decreases as 'r' becomes lager. As a result, the various design parameters of the proposed absorber are determined as follows: a = 16 mm, b = 0.5 mm, c = 0.5 mm, d = 1.5 mm, e = 9.8 mm, f = 0.2 mm, g = 1.74 mm, h = 3.5mm, i = 1 mm, j = 1 mm, and k = 3.2 mm. Figure 5(a) and (b) demonstrate the normalized complex intrinsic impedance (Z) at the top of the proposed absorber under normal incidence. The intrinsic complex impedance matrix (Z) can be calculated from the S-parameter (S) as follows (2) where U is an identity matrix. Normalized means that the calculated intrinsic complex impedance matrix is divided by the intrinsic impedance of free space (377 Ω). The frequency when the imaginary part value of the intrinsic impedance is 0, is the resonant frequency due to LC resonance. At this frequency, the real part value of the intrinsic impedance is closer to one, and the impedance matching between free space and proposed absorber is improved. Figure 5(c) and (d) show the simulated reflection and transmission coefficients and the absorptivity at each state. In this figure, even though there are via holes, the transmission wave is not in the proposed absorber. Figure 6 shows the measurement setup for the fabricated prototype. Wedge-tapered absorbers are placed around the prototype, to prevent signals reflected from the other parts. The reflection coefficient is measured using a horn gain antenna and an Anritsu MS2038C vector network analyzer (VNA). We used the WR-90 standard gain horn antenna with a norminal gain of 15 dB. Its 10-dB return loss bandwidth is 8.2-12.4 GHz. The absorptivity A(ω) of the proposed absorber is calculated using the measured reflection coefficient Γ(ω), as is shown the following equation (5): where T(ω) is the transmission coefficient, but in this case, the transmission coefficient is zero because of fully covered with conductor on bottom layer. Therefore, the absorptivity is calculated using only the measured reflection coefficient. To satisfy the far field condition, the horn gain antenna from the prototype sample, is located. Moreover, a time gating method is used to measure the just reflected wave from the prototype sample. A DC power supply is used to supply the PIN diode with power. Before measuring the prototype, the reflection coefficient of a conductor plate of the same size, is measured, to make reference |Γ| = 1 22 . The reflection coefficient of the prototype is the measured relative to the reference. Figure 7(a) shows the measurement results and simulation results under normal incidence. The proposed absorber has a 90% absorbing bandwidth from 8.45-9.3 GHz, when the PIN diode is in the ON state. When the PIN diode is in the OFF state, the 90% absorbing bandwidth is from 9.2-10.45 GHz. In Table 1, we compared the absorbing bandwidth of the proposed metamaterial absorber using the PIN diode, with the results of other researches. It is observed that the absorber proposed in this study, shows an absorption ratio and bandwidth that are similar to or better than those recorded in other researches.
Additional measurements are performed to verify that the performance of the proposed absorber is independent of the polarization and incident angles. Figure 8(a) and (b) show the absorptivity at different polarization angles (ϕ). The measurement setup is shown in Fig. 7(a). The absorptivity is measured by changing the polarization angle (ϕ) from 0°-90° by rotating the prototype sample, while keeping all the other conditions the same. From the measurement results, the proposed absorber shows no change in performance with respect to the change of polarization of both the ON and OFF state. This is due to the unit cell being designed as a horizontally and vertically symmetrical structure, including the DC bias network, electronic devices, and conductor patterns. Consequently, the proposed absorber gives the advantage of not only having a switchable frequency band, but also an insensitivity to changes in the polarization angles.
Next, the measurement is done to demonstrate the performance of the absorber with respect to a changing incident angle (θ), as shown in Fig. 7(b). Unlike the previous polarization insensitivity experiment, two horn antennas are used to measure absorptivity. The two horn antennas are placed at the same angle from 0°-30° according to the incident angle (θ), to measure the reflected signal. In addition, the reflection coefficient is measured for the two polarizations, including the transverse electric (TE) mode and the transverse magnetic (TM) mode, because the reflection coefficient differs between the perpendicular (Γ ⊥ ) and parallel polarization (Γ || ), as shown by the following equation: where θ i and θ t are incidence angle and transmitted angle, respectively. Figure 8(c) and (d) show the absorptivity at different incident angles (θ) for the TE mode from 0°-30°. When PIN the diode is in the ON state, the absorptivity remains the same up to 20°. However, when the incident angle is 30°, the absorbing bandwidth is blue shifted around 0.5 GHz. Even when the PIN diode is in the OFF state, it   Figure 8(e) and (f) show the absorptivity values at different incident angles (θ) for the TM mode from 0°-30°. Likewise, the measured results of the TE mode, irrespective of the PIN diode state, reveal that the absorptivity is maintained up to 20°, but it is also blue shifted around 0.5 GHz when the incident angle is 30°. The difference between the measurement results of the TE and the TM mode is the absorbing frequency bandwidth, which is wider in the TM mode than in the TE mode.

Discussion
In this paper, the novel electronically switchable broadband metamaterial absorber, using PIN diode, is proposed. The proposed absorber is designed as a horizontally and vertically symmetrical structure, including the DC bias network, electronic devices, and conductor patterns in order to satisfy the condition of polarization insensitivity. To verify the performance of the proposed absorber, it is designed using a full wave simulation tool, and a fabricated prototype sample. The total size of the prototype is 162 × 162 mm, consisting of 10 × 10 unit cells. Conductor patterns are implemented using wet etching process. The electronic devices such as chip resistors, chip inductors, and PIN diodes are soldered onto the substrate, using surface mount technology (SMT). After fabricating the prototype sample, the experimental measurement setup is created to measure the performance of the proposed absorber with respect the change of the polarization (ϕ) and incident angles (θ). As a result, the proposed absorber has absorbing frequency bandwidths from 8.45-9.3 GHz and 9.2-10.45 GHz, when the PIN diode is switched ON and OFF state, respectively. The proposed absorber has polarization insensitivity, irrespective of the state of the PIN diode. The performance of the proposed absorber is maintained when the incident angle is varied from 0°-20°, irrespective of the state of the PIN diode, for both the TE and the TM modes. The proposed absorber has broadband and frequency switching characteristics that are similar to or better than those recorded in other researches using the PIN diode.

Methods
Measurement. Wedge-tapered absorbers are placed around the prototype, to prevent signals reflected from the other parts. The reflection coefficient is measured using horn gain antennas and an Anritsu MS2038C vector network analyzer (VNA). To satisfy the far field condition, the horn gain antenna from the prototype sample, is located. Moreover, a time gating method is used to measure the just reflected wave from the prototype sample. A DC power supply is used to supply the PIN diode with power. The fabricated metamaterial absorber consists of 400 PIN diodes. In order to turn on all diodes, 0.108 W is consumed with 9 DC-V. Before measuring the  Physical Size of the unit cell. 2) Electri cal Size is calcualted at the center frequency of the absorbing bandwidth.
prototype, the reflection coefficient of a conductor plate of the same size, is measured, to make reference |Γ| = 1 22 . The reflection coefficient of the prototype is measured relative to the reference. Additional measurements are performed to verify that the performance of the proposed absorber is independent of the polarization and incident angles. The absorptivity is measured by changing the polarization angle (ϕ) from 0°-90° by rotating the prototype sample, while keeping all the other conditions the same. Next, the measurement is done to demonstrate the performance of the absorber with respect to a changing incident angle (θ). Unlike the previous polarization insensitivity experiment, two horn antennas are used to measure absorptivity. The two horn antennas are placed at the same angle from 0°-30° according to the incident angle (θ), to measure the reflected signal. In addition, the reflection coefficient is measured for the two polarizations, including the transverse electric (TE) mode and the transverse magnetic (TM) mode, because the reflection coefficient differs between the perpendicular (Γ ⊥ ) and parallel polarization (Γ || ).