IEEE Xplore At-A-Glance
  • Abstract

Effect of FeCoB-SiO2-Film-Based Fractal Frequency Selective Surface on the Absorption Properties of Microwave Absorbers

The effects of FeCoB-SiO2-film-based fractal frequency selective surface (FSS) on the absorption properties of multilayer microwave absorbers (MMA) were investigated in detail. These magnetic films were prepared in argon pressure by radio frequency magnetron sputtering targets of SiO2 and Fe40Co40B20. Laser etching was used to fabricate FSS samples using the prepared magnetic films (SFSS) and comparative nonmagnetic aluminum film (AFSS). The reflection properties of the MMAs were measured in the range 2–18 GHz. The combined samples decreased the reflectivity by 50% and increased the bandwidths under −5dB by 100% by substituting SFSS for AFSS, with the absorbing peak frequencies almost kept unchanged. The 0.67-mm-thick MMA samples that SFSS and AFSS both used could further enhance the minimum peak vale to −9.84 dB and broaden the bandwidth under −5 dB to 5.36 GHz. As a consequence, magnetic-film-based fractal FSSs are able to strongly improve absorption characteristics in microwave absorbers.



Frequency selective surfaces (FSSs) [Munk 2000], a kind of 2-D periodic arrays of conducting metal patches or aperture elements, were usually used in the fields of antennas, radomes, e.g., based on their passband or stopband properties according to different circumstances, especially for their characteristic shapes, sizes, and metal materials of the FSSs unit cell, and also the thicknesses and electromagnetic parameters of the substrates, etc. However, the recent work has showed that the multilayer microwave absorbers (MMAs) [Vinoy and Jha, 1996] with introducing FSSs can adjust and improve the microwave absorbing properties of MMAs [Sha 2002, Xie 2009] for their electric and magnetic resonant features [Zou 2008]. These traditional FSS layers are usually fabricated out by printing nonmagnetic metal (μ = 1): such as aluminum and copper, on the dielectric substrates proposing standard lithographic technique, commonly with thick thicknesses and high densities. The conductivities of metal aluminum and copper are extremely large, usually at the magnitude of 107, and the permittivity could be approaching to infinite (ε∼∞), which is much larger than the permeability (∊≫ μ) for traditional FSS layers. This brings in the bad performances on the impedance matching condition. Meanwhile, these traditional kinds of FSS layers nearly have no metallic loss within the gigahertz frequency band [Raynolds 2003] due to the large conductivities of the metals. Based on the reports about developing the microwave absorbing properties by adding magnetic materials into traditional nonmagnetic-material-based FSSs [Archer2008, Buris 1993, Liu 2004], there exist the possibilities of further enhancing the absorption properties by changing the metal materials of FSS layers.

Recently, the sputtered soft magnetic films (μ ≫ 1) have attracted a lot of attentions because of their unique features, especially for the controllable resistivity under different technological conditions and the high permeability (larger than 1000) within the GHz frequency band [Chen 2000; Klemmer 2000; Munakata2005]. Via increasing the resistivity of Fe40Co40B20-SiO2 based magnetic film by proper technological conditions, the conductivity could be decreased to the magnitude of 105, and thus the permittivity is shifted down [Liao 2008]. When the magnetic film is manufactured into periodic FSS structure, the separated metal cells could further decrease the permittivity ε effectively, bringing in the possibility of εapproaching to the increased permeability μ(ε∼μ), which is good for satisfying the impedance matching condition. Meanwhile, the magnetic film has awfully large imaginary part of permeability compared with nonmagnetic metal, which could produce magnetic loss. As a consequence, using magnetic film to fabricate FSS structure probably provides another way to enhance the absorption properties for MMAs.

In this letter, we first prepared the FeCoB-SiO2-based magnetic film using radio frequency magnetron sputtering, and then the magnetic film is fabricated into FSS by laser etching technique originating from its advantages of high-precision and short-manufacturing time. The Minikowski fractal loop was chosen as the FSS unit cell because of its tunable and controllable frequency responses just according to its geometrical feature. By comparing the reflectivity of combined MMAs containing FeCoB-SiO2-film and aluminum-metal based FSS screens with different arrangements, we investigated the effect of this novel FSS on the absorbing properties of MMAs. Moreover, employing both kinds of FSSs in MMAs were discussed to find out the best absorbing performances.

Figure 1
Fig. 1. (a) 3-D schematic diagram of MMA layer structure;(b) unit cell of proposed FSSs pasted in MMA samples.


Fig. 1(a) shows the 3-D schematic diagram of the MMA layer structure. After the periodic FSS layers being pasted:1) on the surface; 2) in the middle; and 3) at the bottom of the two microwave absorbing material (MAM) layers, respectively, MMA samples with different kinds of arrangements were fabricated out. The Minkowski loop was employed as the unit cell of proposed fractal FSS, as shown in Fig. 1(b). The periodic constant of the unit cell is P = 20 mm and the loop width is W = 0.2 mm. The geometric dimensions of Minkowski loop are demonstrated as follows: L1 = 18 mm, L2 = 6 mm, L3 = L4 = L6 = L7 = H2 = 2 mm, L5 = H1 = 1 mm.

The soft magnetic FeCoB-SiO2-based thin films with thickness of 200–400 nm were prepared in argon pressure by radio frequency magnetron sputtering, with the sample dimensions of 180 × 180 mm2. The magnetic films employed the targets with nominal compositions of SiO2 and Fe40Co40B20 to cosputter at the substrate of 0.05-mm-thick polytetrafluoro ethylene (PTFE) cloth. The background pressure was below 6 × 10−5 Pa and the distance from target to substrate was 100 mm. Other sputtering conditions, including sputter pressure, supplied power, and deposition rate were 0.3 Pa, 160 W, and 4.8 nm/min, respectively.

After preparing the sputtering FeCoB-SiO2-based soft magnetic films, the laser etching technique was proposed to fabricate the fractal periodic FSS samples(SFSS). Meanwhile, the fractal FSS samples laser etched by nonmagnetic metal aluminum film (AFSS) with the same dimensions and thicknesses were also fabricated out. The flaky metal magnetic powders were chosen as absorbents and the hydrogenation acrylonitrile-butadiene rubbles (HNBR) as agglomerants. After the procedures of blending, roll mixing, and vulcanizing, two traditional MAM samples NO.1 and 2 were prepared with the dimensions of 180 × 180 mm2. The characteristics of these FSSs and MAM sheets are shown in Table 1.

The prepared plates of NO. 1, NO. 2, AFSS, and SFSS were properly used in certain arrangements illustrated in Table 2 to produce combined MMAs. In order to find out the effect of SFSS on the absorption properties of microwave absorbers, the reflectivity of samples was measured using an 8722ES network analyzer within the swept frequency range from 2 to18 GHz.

Table 1
Table 1 Characteristics of microwave absorber plates and FSSs.
Table 2
Table 2 Arrangements of the samples.


The reflection results of these samples without FSSs are presented in Fig. 2(a), with the inset showing the impedances of these samples. The single-layer sample S1 exhibits very weak absorption properties. Contrarily, the single-layer sample S2 has an absorbing peak at about 12.48 GHz, with peak value of −3.7dB. The impedance of S2 is much larger to that of S1. When S1 and S2 were combined to produce the double-layer-composite sample S3, much of the incident microwave was reflected because of the impedance mismatching. However, after switching the places of S1 and S2, the sample S4 increases the impedance and presents much better absorbing performances, with the absorbing peak around 10.64 GHz and the peak value −5 dB. Consequently, the sample S4 was chosen as the original multilayer composites needed to be developed.

Figure 2
Fig. 2. Reflectivity results (a) samples without FSSs and (b) MMAs with AFSS. The inset shows the impedances of corresponding samples.

Fig. 2(b) shows the reflectivity results of MMAs with AFSS plates, with the inset showing the impedances of these samples. As we can see, the absorption properties could be developed by introducing ultrathin AFSS plate into traditional microwave absorbers in proper arrangements. The introducing extrathin AFSS nearly has no effect on the absorbing peak frequencies, which is kept around 10.5 GHz. With different locations of AFSS embedded in the composites, the multilayer samples exhibit different absorption peak values. The corresponding peak values of S5, S6, and S7 are −5.68 dB, −5.41dB, and −4.23 dB, respectively. Clearly, there exists the change in the peak values by placing the AFSS at the bottom compared to placing it on the surface. The FSS plates are fabricated out using metal materials. Therefore, the MMA samples with AFSS plate placed at the top reflect more incident microwave and have less peak values compared to the samples without AFSS. As we know, the place nearest the metal plate in MMAs has the strongest magnetic field, derived from the boundary conditions of the electromagnetic theories [17]. Then the introduced AFSS placed at the bottom may strengthen this field by the magnetic resonance of FSS and bring in more magnetic loss in return. As a result, the peak values of the sample with AFSS arranged at the bottom are larger than the samples without AFSS. In addition, the bandwidths under −3 dB are 9.04, 10.24, and 7.44GHz, respectively. In our study, both of the peak value and the required bandwidth are considered to judge the absorption performances of the samples. The difference of the peak value between S5 and S6 is only about 0.27 dB, and we could consider both of the samples almost have the same peak values. The differences of the required bandwidth between S5 and S6 could approach to 1.2 GHz. Therefore, the reflectivity curve of S6 sample is wider than that of S5, with both of the peak values almost the same. From the aforementioned discussions, S6 with AFSS locating in the middle of microwave absorbers behaviors the best multiabsorbing performances among these samples. Additionally, the bandwidths under −5 dB of S5, S6, and S7 are 2.88, 2.32, and 0 GHz, respectively. However, the results are not good enough to satisfy our real demands, of which the peak value at least larger than −7 dB and the bandwidth under −5 dB at least wider than 4 GHz.

Figure 3
Fig. 3. (a) Reflectivity results by using the SFSS plate to replace the AFSS plate in MMAs. The inset shows the complex permeability of the magnetic film. (b) Impedances of the corresponding samples.

In order to further develop the absorbing properties, the novel prepared FeCoB-SiO2-film-based fractal FSS sample (SFSS) was employed to take the place of AFSS, with results shown in Fig. 3(a). The inset of Fig. 3(a) shows the complex permeability (μ) of the FeCoB-SiO2 magnetic film and Fig. 3(b) exhibits the impedances of the corresponding samples. The maximum values of the real part (μ′) and imaginary part (μ′′) are 700 and 1300, respectively. Remarkably, μ of the magnetic film is much larger than that of nonmagnetic aluminum metal and μ′′, standing for the magnetic loss factor is extremely huge. Meanwhile, the resistivity of the magnetic film measured by the four-probe method is 642 μΩ·cm. This means that the conductivity of the film is 1.56 × 105 siemens/m, smaller by two orders of magnitude compared to that of aluminum (normally 3.72 × 107 siemens/m). In view of the reasons mentioned in the introduction, our prepared magnetic film not only could satisfy the impedance matching condition better, but also has huger magnetic loss factor, which are both fascinating for developing the absorption properties of MMAs. Comparing to the samples with AFSS (S5, S6, S7), the samples containing SFSS with the same arrangements were shown in Table 1, namely, S8, S9, and S10, respectively. Similarly to the absorbers containing AFSS plates, the absorbing peak frequencies of the samples with SFSSs almost maintained at 10.5 GHz, too. The testing samples S8, S9, and S10 have the peak values of −8.03, −7.81, and −4.77 dB; meanwhile, the bandwidths under −5 dB of 3.92, 4.96, and 0 GHz, respectively. Obviously, the multilayer absorber composites with SFSSs of S8, S9, and S10 have better absorption properties compared with the corresponding samples of S5, S6, and S7, respectively. In addition, caused by the same reasons discussed for AFSS case, the MMA samples with SFSS placed on the surface still has the least peak values and the ones with SFSS located at the bottom also has the largest peak values. The sample S9 with SFSS located in the middle presents the best multiabsorption performance according to the judgments mentioned earlier in Fig. 2(b) and especially S9 could nearly increase the peak value by 50% and the bandwidth under −5 dB by 100%compared to those of S6. As a consequence, the absorbing characteristics can be enhanced in effect after introducing the SFSS instead of the traditional AFSS plate locating in the middle of the MMAs.

The extra-thin SFSS plate was introduced into the AFSS-based MAMs in different kinds of placing orders, producing the combined samples from S11 to S14. The reflectivity results are shown in Fig. 4(a), with Fig. 4(b) showing the impedances of these samples. The peak values of the samples after proposing both AFSS and SFSS (S11 to S14) could be further increased, while the absorbing peak frequencies are almost still kept unchanged compared with the cases using one single FSS.

Figure 4
Fig. 4. (a) Reflectivity results of the samples proposing both SFSS and AFSS plates. (b) The corresponding impedances.

First, the samples with SFSS connecting to AFSS S11 and S12 were investigated. The SFSS and AFSS layers, combined together without any space between them, forms a two-metal-layer FSS and it is located in the middle based on the results mentioned earlier that MMAs with FSSs locating in the middle behavior the best absorbing performances. The peak values are increased to −8.42 and −9 dB, and at the same time, the required bandwidths (under −5 dB) are broadened to 5.28 and5.36 GHz, respectively. The sample S12 with the ranging order of SFSS over AFSS presents better absorption performances.

Table 3
Table 3 Microwave absorption properties of the samples.
Figure 5
Fig. 5. Microwave absorption properties of the samples:(a) Absorbing peak frequencies; (b) peak values; (c) bandwidths under −5dB.

Then, guided from the perfect FSS absorber researched by Landy [2008], the introduced SFSS and original AFSS plates were arranged separated from each other by spacing traditional microwave absorbers in MMAs, producing the samples S13 and S14. Similarly to the former results, the introducing SFSS layer leads to enhance the absorbing performances significantly. The absorbers S13 and S14 have the peak values of −7.98and −9.84 dB, and meanwhile the bandwidths under −5 dB of 4.08 and 4.56 GHz, respectively. Both of S12 with the widest bandwidth under−5 dB and S14 with the minimum peak value indicate that the introduced SFSS plate need to be placed over the original AFSS plate to exhibit the strongest absorbing performances, and this may be caused by the fact that SFSS behaviors better impedance matching performances than AFSS.

All the microwave absorbing properties of the samples S5S14 are demonstrated in Table 3. The variations of absorption characteristics after introducing SFSS into traditional MAMs are shown more directly and clearly in the form of column diagrams in Fig. 5. Obviously in Fig. 5(a), the SFSS plate nearly has no effect on the absorbing peak frequencies of MAMs because of the extrathin thicknesses. After substituting the SFSS plate instead of the AFSS plate in the composites, the absorption characteristics could be improved strongly. Moreover, the absorption properties could be further enhanced by proposing both the SFSS and AFSS plates. Remarkably, the sample S14 with AFSS and SFSS separated in Fig. 5(b) has the minimum peak value −9.84 dB and S12 with AFSS and SFSS connected together in Fig. 6(c) has the widest required bandwidth 5.36 GHz.



In summary, this letter offers a new way to develop the absorbing properties of traditional MMA using a novel kind of FSS fabricated by magnetic films. The controllably resistivity, high-complex permeability and magnetic loss of magnetic films make them ideal candidates for manufacturing FSSs to satisfy the impedance matching condition and minimize the reflectivity of MMAs. The experimental results show that the introducing SFSS plate will not change the absorbing peak frequencies, but extremely strengthen the peak values and broaden the required bandwidths of MMAs. The sample with SFSS placed in the middle can increase the peak values by 50% and the bandwidths under −5 dB by 100% compared to the corresponding case of the sample using AFSS. By proposing both the FSS plates, the absorbing properties could be further enhanced. The fact that SFSS behaviors better impedance matching performances than AFSS makes the best order for ranging the two FSS layers is SFSS over AFSS. Thus, at a thickness of 0.67 mm and surface density of 3.57 g/cm3, the samples with this kind of arrangement could get the minimum peak vale of −9.84 dB and the widest bandwidth under −5 dB of 5.36 GHz, respectively. In future, MMAs with FSS plates fabricated by other kinds of magnetic films, i.e., FeCoNiB and FeCoSm, will be researched to optimize the best impedance matching and microwave absorbing performances.


Corresponding author: Y. Nie (


1. Permeability enhancement of soft magnetic films through metamaterial structures

O Acher

J. Magn. Magn. Mater., Vol. 320, pp. 3276–3281, 2008, 10.1016/j.jmmm.2008.06.039

2. Dipole arrays printed on ferrite substrates

N E Buris, T B Funk, R S Silverstein

IEEE Trans. Antennas Propag., Vol. 41, pp. 165–176, 1993, 10.1109/8.214607

3. Soft-magnetic properties of Fe-Co-B thin films for ultra-high-frequency applications

L H Chen, T J Klemmer, K A Ellis, R B Dover, S Jin

J. Appl. Phys., Vol. 87, pp. 5858–5860, 2000, 10.1063/1.372546

4. Ultrathin frequency permeability of sputtered Fe-Co-B thin films

T J Klemmer, K A Ellis, L H Chen, B Dover, S Jin

J. Appl. Phys., Vol. 87, pp. 830–833, 2000, 10.1063/1.371949

5. Implementation of broadband microwave absorber using FSS screens coated with Ba(MnTi)Fe10O19 ferrite

J C Liu, C Y Liu, C Y Wu, H Y Hsu

Microw. Opt. Techn. Lett., Vol. 41, pp. 323–326, 2004, 10.1002/mop.20131

6. High-frequency permeability of sputtered Fe–Co–B-based soft magnetic thin films

J S Liao, Z K Feng, J Qiu, Y Q Tong

Phys. Stat. Sol (a)., Vol. 205, pp. 2943–2947, 2008, 10.1002/pssa.200824316

7. Perfect metamaterial absorber

N I Landy, S Sajuyigbe, J J Mock, D R Smith, W J Padilla

Phys, Rev. Lett., Vol. 100, 207402, 2008, 10.1103/PhysRevLett.100.207402

8. Frequency Selective Surface: Theory and Design

B A (2000) Munk

New York
Frequency Selective Surface: Theory and Design, Wiley

9. B-concentration dependence on anisotropy field of CoFeB thin film for gigahertz frequency use

M Munakata, S-I Aoqui, M Yagi

IEEE Trans. Magn., Vol. 41, pp. 3262–3264, 2005, 10.1109/TMAG.2005.854666

10. Ohmic loss in frequency-selective surfaces

J E Raynolds, B A Munk, J B Pryor, R J Marhefka

J. Appl. Phys., Vol. 93, pp. 5346–5358, 2003, 10.1063/1.1565189

11. Experimental investigations of microwave absorber with FSS embedded in carbon fiber composites

Y N Sha, K A Jose, C P Neo

Microw. Opt. Techn. Lett., Vol. 32, pp. 245–249, 2002, 10.1002/mop.10144

12. Radar Absorbing Materials

K J Vinoy, R M (1996) Jha

Norwell, MA
Radar Absorbing Materials, Kluwer

13. Effect of FSS on microwave absorbing properties of hollow-porous carbon fiber composites

W Xie, H F Cheng, Z Y Chu

Mater. Design., Vol. 30, pp. 1201–1204, 2009, 10.1016/j.matdes.2008.06.018

14. Enhancing and tuning absorption properties of microwave absorbing materials using metamaterials

Y H Zou, L Y Jiang, S C Wen

Appl. Phys. Lett., Vol. 93, 261115, 2008, 10.1063/1.3062854


No Photo Available

Zhangqi Liao

No Bio Available
No Photo Available

Yan Nie

No Bio Available
No Photo Available

Wenyi Ren

No Bio Available
No Photo Available

Xian Wang

No Bio Available
No Photo Available

Rongzhou Gong

No Bio Available

Cited By

No Citations Available


No Corrections


No Content Available

Indexed by Inspec

© Copyright 2011 IEEE – All Rights Reserved