Study of the influence of Carbonyl iron particulate size as an electromagnetic radiation absorbing material in 12 . 4 to 18 GHz ( K u ) Band

Brazilian Microwave and Optoelectronics Society-SBMO received 30 Sept 2018; for review 14 Nov 2018; accepted 20 Nov 2018 Brazilian Society of Electromagnetism-SBMag © 2018 SBMO/SBMag ISSN 2179-1074 Abstract— This paper presents the influence of different sizes of carbonyl iron particles on the reflectivity measurements of Radar Absorbing Material (RAM). The electromagnetic characterization was performed with a vector network analyzer and a rectangular waveguide in the frequency range of 12.4 to 18GHz (Ku Band). The influence of different parameters such as thicknesses, particle sizes and concentration of carbonyl iron were evaluated. Reflectivity results showed the influence of these parameters on the performance of the RAM. The best reflectivity values (~ -18 dB) were obtained for samples with 60 wt% concentration and 5 mm thickness. We provide information about significantly reflection loss improvement by simply controlling carbonyl iron particulate size.


I. INTRODUCTION
Radar Absorbing Material (RAM) is a type of material designed to attenuate electromagnetic radiation on specific frequencies.Many researches have been made about RAM due its countless applications in electromagnetic compatibility and interference reduction.These materials can be applied on communication systems, modern electronic devices, anechoic chamber, military stealth technology, and so on [1].RAMs can be produced in different forms, such as paints or thin films [2].
Usually, RAMs are composite materials made with polymer (matrix) and absorptive material (mean).
In the literature, materials with dielectric and/or magnetic losses are commonly used as means.In this matter, materials like ferrite, carbonyl iron (CI), carbonaceous materials and conductive polymers has advantages over lossless materials [2], [3].Despite their high specific mass, composites made with ferrites or CI have advantages like thin thickness and broadband frequency absorption because of iron on their compositions [4].Carbonyl iron has a relatively low electrical conductivity, a high Curie temperature, and a high saturation magnetization.These properties make CI a good candidate to be used as absorption mean, especially in the frequency range between 2 -18 GHz [5].
The effectiveness of the absorptive material contributes to the reflectivity losses that are determined through the obtained values of the complex permittivity and permeability.However, the control of the complex permittivity and permeability are obtained by the addition of magnetic and dielectric additives.Understanding the effects caused by RAM processing require the investigation of several parameters, but the main ones are absorbent mean and matrix properties [6].Thus, this paper presents the influence of CI particle size on RAM absorption parameter over the frequency range from 2 to 18 GHz.Here, we propose the reflectivity control using different CI particle sizes, where we can significantly improve the absorption properties.

A. Material and sample preparation
Composites were prepared with commercial CI powder from BASF GmbH as additive.Commercial bi-component silicone was used as matrix.
Sample thickness ranged from 1 mm to 5 mm, but since the best results were obtained for 2, 3 and 5 mm, our discussions were based on these thicknesses.Composites were manually homogenized.Each sample of silicone-CI composite was mixed with ~0.6 ml catalyst until the beginning of reaction, which could be noticed by bubble formation and viscosity increasing.Before sample hardening, mixture was moved to K u -band mold with 15.7 mm width and 7.9 mm height.Full catalyst reaction of composite mixtures was about 30 minutes.

B. Morphological and electromagnetic characterization
Crystalline phases of samples were investigated through X-Ray Diffraction (XRD).A Panalytical X'Pert Powder system equipped with a CuKα (λ = 0.154 nm) was used.Composites were scanned from 20° to 90° with sampling intervals of 0.02°.Carbonyl iron particulate sizes were analyzed with a Field Emission Gun Scanning Electron Microscope (FEG-SEM).The equipment used was a TESCAN -Vega 3 operating with secondary electron detection.
Electromagnetic characterization was performed with a K u -band rectangular waveguide (Agilent WR-62 P11644A) coupled on a 50 GHz PNA-L vector network analyzer (Keysight N5232A).
Electromagnetic properties were measured in the K u band, i.e., from 12.4 to 18 GHz.Through scattering parameters (S parameters) it was possible to comprehend the interaction of electromagnetic wave on samples and calculate the permittivity and permeability of materials.The method used to obtain the electromagnetic properties of samples was Nicolson Ross Weir (NRW), which is also called Transmission and Reflection (TR) method [7].Fig. 1 illustrates the transmitted (S21 and S12) and reflected (S11 and S22) waves that can be measured with PNA-L.It is through these parameters that permittivity and permeability are extracted, allowing the material characterization over a frequency range.The electromagnetic properties of a material, i.e., the complex relative permittivity (εr = ε '-jε'') and permeability (μr = μ'jμ'') are determined by S-parameters measured over a frequency range [9].The real part of permittivity and permeability (ε' and μ') represents the capacity of energy storage in the sample, while the imaginary parts (ε" and μ") represent the electric and magnetic energy losses [10].

A. Field Emission Gun Scanning Electron Microscopy
Particulate sizes of pure carbonyl iron analyzed with FEG-SEM are presented in Fig. 2. All particulate sizes obtained after sifting presents spheres of different sizes.Although spheres apparently have similar sizes, there are different agglomerations of them over sifted powder.It is interesting to note that particulate sizes between 25 µm and 53 µm, Fig. 2a, presents grains more dispersed than particulates sizes bigger than 63 µm.This agglomeration caused the particle separation through the sieves, where the average size of spheres clusters was estimated based on the sieve weft spacing.

D. Reflectivity Measures
According to equations ( 1) and ( 2), the Reflection Loss (RL) of a RAM have influence of sample thickness and may have a frequency dependence inherited from complex permittivity and permeability [13].
(1) (2) Here, is the relative input impedance of material, t is the sample thickness, is the material relative magnetic permeability, is the material relative dielectric permittivity and is the wavelength of the incident wave in the free space.In Fig. 5a, which have RL plots of 40% concentration, it is possible to verify that the highest reflectivity value was obtained for sample with 5 mm thickness and particulate size bigger than 63 μm.The attenuation in this sample is -9.5 dB at 18 GHz, i.e., 90% of the incident wave was attenuated [14].It is also observed that RAM with the same particle size, but with thickness of 3 mm, tends to act as a microwave absorber in the X-band frequency range (8.2 -12.4 GHz).For these samples, the maximum reflectivity loss was -8.5 dB at 12.4 GHz.Thus, RAM with 40% CI concentration and 3 mm and 5 mm thicknesses tends to attenuate the incident wave in the X-band and K-band (18 -26.4 GHz), respectively.This behavior is not observed for samples with no sieving.In other words, granulometric separation made possible the choice of the frequency band to work.This effect can be attributed exclusively to particle size, since no structural changes that could justify such behavior was noticed, as it could be observed in XRD graphs in Fig. 3.
It is possible to observe that 5 mm thickness samples trends to attenuate K-band frequencies when samples present concentrations of 40% and 50%, Fig. 5b.However, a minimum RL value is observed at ~17 GHz for CI concentration of 60%, Fig. 5c.
Samples with thickness of 2 mm have greater attenuations, which occurs mainly with CI concentration of 60%.The best RL attenuation results are close to 13 GHz, Fig. 5d.
In summary, these results highlight different behaviors as a function of sample thickness and concentration.For the samples with 3 mm thickness, reflectivity presents a trend to have a better performance in the frequency band that precedes the K u -band, i.e., X-band.For samples with 5 mm thickness there is a trend to great reflectivity attenuation in the K-band.These results show that the thickness and the concentration have a strong influence on frequency range division.These results are very important since controlling particulate size and thickness enable controlling of frequency ranges.
For the same CI concentration (60%) and two different thickness (2 and 5 mm) it is possible to shift the RL from 13 GHz to 17 GHz.The reflectivity curves show the predominant influence of the particle sizes resulting in more efficient absorbers with attenuation values > 90%.

IV. CONCLUSIONS
We demonstrated the possibility to attenuate electromagnetic wave over a frequency range by controlling sample thickness, particle size and CI concentration of RAM.Values obtained for reflectivity (RL) showed that it is possible to have an efficient RAM (RL < -10 dB) in the frequency range of 12-18 GHz using a thin sample (2 mm) by simply controlling particulate size and concentration.Based on these results it is possible to design and manufacture different electromagnetic absorbers with the same material by simply controlling parameters of concentration particle size.Carbonyl iron separated in different particulates have great potential to be used as absorber materials in the K u -band.

Fig. 1
Fig.1 Schematic representation of a device with two ports [8].The symbols a and b represent the amplitude of the incident wave and the response.

Fig. 5a -
Fig. 5a -5d present the experimental reflectivity of evaluated samples.From Fig. 5a to Fig. 5c, reflectivity is highlighted for samples with 3 and 5 mm thickness, where all three particulate variations and CI concentrations are plotted.Since samples with 2 mm thickness presented the best RL results, all samples with this thickness is plotted in Fig. 5d for comparison.

Fig. 5 .
Fig. 5. (a), (b) and (c) reflectivity for samples of thicknesses of 3 mm and 5 mm with variation of the particulate, and (d) reflectivity for samples with variation of the particulate at 2 mm of thicknesses.

TABLE I .
SIZE OF THE PARTICULATES