Electromagnetic Properties of Carbon Nanotube/BaFe12−xGaxO19/Epoxy Composites with Random and Oriented Filler Distributions

The microwave properties of epoxy composites filled with 30 wt.% of BaFe12–xGaxO19 (0.1 ≤ x ≤ 1.2) and with 1 wt.% of multi-walled carbon nanotubes (CNTs) were investigated in the frequency range 36–55 GHz. A sufficient increase in the microwave shielding efficiency was found for ternary 1 wt.%CNT/30 wt.% BaFe12–xGaxO19/epoxy composites compared with binary 1% CNT/epoxy and 30 wt.% BaFe12–xGaxO19/epoxy due to the complementary contributions of dielectric and magnetic losses. Thus, the addition of only 1 wt.% of CNTs along with 30 wt.% of barium hexaferrite into epoxy resin increased the frequency range where electromagnetic radiation is intensely attenuated. A correlation between the cation Ga3+ concentration in the BaFe12–xGaxO19 filler and amplitude–frequency characteristics of the natural ferromagnetic resonance (NFMR) in 1 wt.%CNT/30 wt.% BaFe12–xGaxO19/epoxy composites was determined. Higher values of the resonance frequency fres (51.8–52.4 GHz) and weaker dependence of fres on the Ga3+ concentration were observed compared with pressed polycrystalline BaFe12–xGaxO19 (fres = 49.6–50.4 GHz). An increase in the NFMR amplitude on the applied magnetic field for both random and aligned 1 wt.% CNT/30 wt.% BaFe12–xGaxO19/epoxy composites was found. The frequency of NFMR was approximately constant in the range of the applied magnetic field, H = 0–5 kOe, for the random 1 wt.% CNT/30 wt.% BaFe12–xGaxO19/epoxy composite, and it slightly increased for the aligned 1 wt.% CNT/30 wt.% BaFe12–xGaxO19/epoxy composite.


Introduction
Composites are considered multifunctional materials having suitable structural, microstructural, magnetic, electromagnetic, and other properties for certain applications. In particular, composites work as a material for protective coatings and shields which could be applied as microwave absorbers. Investigation of microwave absorbing materials is important since such developments allow product appliances that reduce electromagnetic interference, protecting devices and biological tissues from undesirable radiation. Electromagnetic energy can be absorbed completely when magnetic and dielectric losses are combined in the material. Microwave absorbers are effective when electromagnetic impedance matching and attenuation of electromagnetic waves are achieved within the material. Improving the effectiveness of microwave absorbing materials is possible by changing their magnetic, conductive, or dielectric components. The current trend is the

Materials and Methods
M-type BaFe 12-x Ga x O 19 (x = 0.1-1.2) hexagonal ferrites were prepared by the method of solid-state reaction. High-purity Ga 2 O 3 and Fe 2 O 3 oxides and BaCO 3 carbonate were used in a stoichiometric ratio [25]. The synthesis was conducted at 1200 • C for 6 h. Epoxybased composite materials (CMs) with a magnetic nanofiller (BaFe 12-x Ga x O 19 ) and carbon nanotubes (CNTs) were prepared by the method of mixing in solution. Multi-walled carbon nanotubes (CNTs, length of 10-30 µm, outer diameter of 10-30 nm) were purchased from CheapTubes Ins, (Grafton, WV, USA) (Figure 1a). Low-viscosity epoxy resin Larit285 (abbreviated L285) (Lange&Ritter, Gerlingen, Germany) with hardening agent H285 was used as a polymer matrix. The main stages of the investigated CMs' preparation were as follows. A mixture of L285 epoxy resin and appropriate BaFe 12-x Ga x O 19 (x = 0.1−1.2) powder was subjected to initial ultrasound action (in a BAKU 9050 ultrasonic cleaner, Guangzhou Hanker Electronics Technology Co., Ltd., Guangzhou, China, 40 kHz, 50 W), for 1 h. In the case of CMs with a nanocarbon component, CNTs were then added, and the mixture was ultrasonicated for an extra hour. After addition of H285, the liquid composite mixture was carefully mixed and then poured into a mold made of a nonmagnetic silicon material. Further, for CMs with a uniform filler distribution, the samples were polymerized under normal conditions in air for one day, followed by drying of the cured CMs at a stepwise increasing temperature from 40 to 800 • C for 5 h. As for the CMs with an aligned BaFe 12−x Ga x O 19 (x = 0.1-1.2) filler distribution in the polymer matrix, alignment was performed by the placement of a mold containing a liquid CM mixture in a magnetic field of~0.64 T. Molds were left in the magnetic field until full epoxy polymerization was achieved, followed by drying (according to the above-described scheme).

R REVIEW 3 of 13
The aim of this work was to study the effect of the addition of carbon nanotubes on the electromagnetic properties of epoxy composites filled with substituted hexaferrites BaFe12-xGaxO19 (0.1 < x < 1.2) in the frequency range 36-55 GHz.

Materials and Methods
M-type BaFe12-xGaxO19 (x = 0.1-1.2) hexagonal ferrites were prepared by the method of solid-state reaction. High-purity Ga2O3 and Fe2O3 oxides and BaCO3 carbonate were used in a stoichiometric ratio [25]. The synthesis was conducted at 1200 °C for 6 h. Epoxybased composite materials (CMs) with a magnetic nanofiller (BaFe12-xGaxO19) and carbon nanotubes (CNTs) were prepared by the method of mixing in solution. Multi-walled carbon nanotubes (CNTs, length of 10-30 μm, outer diameter of 10-30 nm) were purchased from CheapTubes Ins, (Grafton, WV, USA) ( Figure 1a). Low-viscosity epoxy resin Larit285 (abbreviated L285) (Lange&Ritter, Gerlingen, Germany) with hardening agent H285 was used as a polymer matrix. The main stages of the investigated CMs' preparation were as follows. A mixture of L285 epoxy resin and appropriate BaFe12-xGaxO19 (x = 0.1−1.2) powder was subjected to initial ultrasound action (in a BAKU 9050 ultrasonic cleaner, Guangzhou Hanker Electronics Technology Co., Ltd., Guangzhou, China, 40 kHz, 50 W), for 1 h. In the case of CMs with a nanocarbon component, CNTs were then added, and the mixture was ultrasonicated for an extra hour. After addition of H285, the liquid composite mixture was carefully mixed and then poured into a mold made of a nonmagnetic silicon material. Further, for CMs with a uniform filler distribution, the samples were polymerized under normal conditions in air for one day, followed by drying of the cured CMs at a stepwise increasing temperature from 40 to 800 °С for 5 h. As for the CMs with an aligned BaFe12−xGaxO19 (x = 0.1-1.2) filler distribution in the polymer matrix, alignment was performed by the placement of a mold containing a liquid CM mixture in a magnetic field of ~0.64 T. Molds were left in the magnetic field until full epoxy polymerization was achieved, followed by drying (according to the above-described scheme).  As can be seen from Figure 1b,c, a certain dispersion of particle sizes was for BaFe12-xGaxO19 powders with Ga 3+ concentrations of x = 0.3 and 0.9. Th BaFe11.7Ga0.3O19 particles changes in the range 0.5-12 μm, the average particle siz and some agglomerates of barium hexaferrite particles are observed. In the BaFe11.1Ga0.9O19 powders, the particles' size is slightly higher (0.7-14 μm, average μm), and a larger number of agglomerated barium hexaferrite particles are form Epoxy composites with the combined filler CNT/BaFe12-xGaxO19 were prepar tailed description of the composite fabrication method with random and aligned tributions was presented in our previous paper [26]. The use of ultrasonic disp the composite mixture allows de-agglomeration of the barium hexaferrite filler a form distribution of fillers in the epoxy matrix. The contents of fillers in epoxy co were as follows: BaFe12-xGaxO19-30 wt.%, CNT-1 wt.%.
Microwave scalar network analyzers P2-67 within a 36-55.5 GHz frequen were used for measurements of the standing wave ratio (SWR) and transmissio cient T of the investigated CMs at room temperature. Measurements using s work analyzers were performed for specimens with dimensions of 5.2 × 2.6 × 2.6  As can be seen from Figure 1b,c, a certain dispersion of particle sizes was observed for BaFe 12-x Ga x O 19 powders with Ga 3+ concentrations of x = 0.3 and 0.9. The size of BaFe 11.7 Ga 0.3 O 19 particles changes in the range 0.5-12 µm, the average particle size is 6 µm, and some agglomerates of barium hexaferrite particles are observed. In the case of BaFe 11.1 Ga 0.9 O 19 powders, the particles' size is slightly higher (0.7-14 µm, average size is 7 µm), and a larger number of agglomerated barium hexaferrite particles are formed.
Epoxy composites with the combined filler CNT/BaFe 12-x Ga x O 19 were prepared. A detailed description of the composite fabrication method with random and aligned filler distributions was presented in our previous paper [26]. The use of ultrasonic dispersion of the composite mixture allows de-agglomeration of the barium hexaferrite filler and a uniform distribution of fillers in the epoxy matrix. The contents of fillers in epoxy composites were as follows: BaFe 12-x Ga x O 19 -30 wt.%, CNT-1 wt.%.
Microwave scalar network analyzers P2-67 within a 36-55.5 GHz frequency range were used for measurements of the standing wave ratio (SWR) and transmission coefficient T of the investigated CMs at room temperature. Measurements using scalar network analyzers were performed for specimens with dimensions of 5.2 × 2.6 × 2.6 mm 3 .
The measurement configuration was such that the direction of alignment of the filler in the sample was across the direction of the incident wave. The shielding effectiveness SE T (in dB) is related to the measured EMR transmission index T using the following equations: where T = |E T /E I | 2 , E I , E T are the electric field strengths of the incident and transmitted waves.

Amplitude-Frequency Characteristics of NFMR
The frequency dependencies of the electromagnetic response (shielding efficiency SE T ) for epoxy composites with BaFe 12-x Ga x O 19 and 1% BHT/BaFe 12-x Ga x O 19 for random and oriented distributions of fillers are shown in Figure 2, which also shows the curve of SE T ( f ) for the 1% CNT/epoxy composite [27,28] and pressed BaFe 12-x Ga x O 19 samples for comparison. As can be seen from the figure, at the frequency f ≈ 50 GHz, there is a minimum on the SE T ( f ) curve for all studied samples corresponding to the lower-order natural ferromagnetic resonance (NFMR) modes. The minimum of SE T is most clearly pronounced for pressed powders of nanocrystalline BaFe 12-x Ga x O 19 , although it is much wider than that observed for single crystal samples [7]. For 30 wt.% BaFe 12-x Ga x O 19 /epoxy composites, the value of |SE Tmin | is sufficiently lower and less pronounced in comparison with pressed samples of BaFe 12-x Ga x O 19 , which is explained by the small volume content of BaFe 12-x Ga x O 19 (~8.5 vol.%). clearly pronounced for pressed powders of nanocrystalline BaFe12-xGaxO19, although it is much wider than that observed for single crystal samples [7]. For 30 wt.% BaFe12-xGaxO19/epoxy composites, the value of min T SE is sufficiently lower and less pronounced in comparison with pressed samples of BaFe12-xGaxO19, which is explained by the small volume content of BaFe12-xGaxO19 (~8.5 vol.%). 35 40 45 50 55 -20   The addition of 1 wt.% of CNTs to 30 wt.% BaFe 12-x Ga x O 19 /epoxy composites leads to a significant increase in EMR shielding SE T ; however, the shape of the SE T ( f ) curves with a wide minimum changes only slightly. As it is known, the main parameters that are responsible for the excellent EMR shielding properties of the materials are their electrical conductivity σ and electrodynamic parameters, such as complex permittivity ε * r = ε r − iε r and magnetic permeability µ * r = µ r − iµ r . The EMR shielding efficiency SE T (in dB) is defined by the following expression [29,30]: where SE A is the shielding factor due to the EMR absorption; SE R and SE I are the shielding factors due to reflection and multiple reflection, respectively; n = k z /k 0 is the complex index of refraction; k 0 = 2π/λ 0 is the wave vector in free space; λ 0 = C 0 / f ; λ 0 and f are the wavelength and the frequency; is the propagation constant of the electromagnetic waves; β is the phase constant; α is the attenuation index; and d is the sample thickness.
The higher the electrical conductivity-and, accordingly, the imaginary part of the dielectric permittivity ε r = σ/(ω · ε 0 )-the higher the degree of EMR shielding, both due to the high reflection coefficient and effective absorption of EMR. It is obvious that the introduction of highly conductive carbon nanotubes into the polymer matrix leads to an increase in the electrical conductivity of the material and, accordingly, to a weakening of EMR. Table 1  As can be seen from the presented data, the addition of 1% CNTs to epoxy leads to an increase in electrical conductivity, but the percolation threshold has not yet been reached. The introduction of BaFe 12-x Ga x O 19 alone does not lead to significant changes in electrical conductivity, since BaFe 12-x Ga x O 19 is a dielectric. Moreover, as can be seen from Figure 2a and Table 1, the SE T values correlate with the data on the electrical conductivity of these CMs: SE T is minimal for epoxy and BaFe 12-x Ga x O 19 and increases for 1% CNT/epoxy, since the electrical conductivity and complex dielectric permittivity ε * r increase, especially the imaginary part of the dielectric permittivity ε r = σ/(ω · ε 0 ), which is responsible for the absorption of EMR.
The use of a combined filler, 1% CNT/30% BaFe 12-x Ga x O 19 , leads to a further slight increase in electrical conductivity (up to 5.0 × 10 −8 S/m) compared to 1% CNT/epoxy CM; however, a significant increase in shielding SE T is observed. Such an increase in SE T for ternary CMs is related not only to increased conduction loss but also to the occurrence of magnetic loss due to the presence of magnetic particles of BaFe 12-x Ga x O 19 . In addition, it may be assumed that the use of CNT and BaFe 12-x Ga x O 19 fillers in combination results in an increase in the real part of dielectric permittivity ε r due to the formation of a large number of dipoles and strong interfacial polarization [32,33]. This increase in ε r promotes an increase in shielding due to the reflection of EMR. Thus, the use of CNTs in combination with magnetic BaFe 12-x Ga x O 19 particles as fillers in epoxy matrices results in an increase in dielectric permittivity ε r , a slight increase in magnetic permeability µ r , and also an increase in dielectric ε r and magnetic µ r losses. Such changes in electrodynamic parameters of CMs lead to an increase in the EMR attenuation coefficient α, which is responsible for the attenuation of incident electromagnetic radiation [32]: where C is the velocity of light. The high dielectric ε r and magnetic loss µ r could result in a high value of α. Figure 3 displays the resonance frequency of NFMR for various types of composites with BaFe 12-x Ga x O 19 . The NFMR frequency f res was measured at half of the bandwidth W res /2. As can be seen in Figure 3, the frequency of the NFMR resonance for epoxy composites containing 1 wt.%CNT/30 wt.% BaFe 12-x Ga x O 19 is higher compared with the pressed BaFe 12-x Ga x O 19 sample, and such a change is similar to the 30 wt.%BaFe 12-x Ga x O 1 /epoxy composites investigated in our previous paper [34]. combination with magnetic BaFe12-xGaxO19 particles as fillers in epoxy matrices results in an increase in dielectric permittivity r ε ′ , a slight increase in magnetic permeability r μ ′ , and also an increase in dielectric r ε′ ′ and magnetic r μ ′′ losses. Such changes in electrodynamic parameters of CMs lead to an increase in the EMR attenuation coefficient α , which is responsible for the attenuation of incident electromagnetic radiation [32]: where C is the velocity of light.
The high dielectric r ε′ ′ and magnetic loss r μ′′ could result in a high value of α . Figure 3 displays the resonance frequency of NFMR for various types of composites with BaFe12-xGaxO19. The NFMR frequency res f was measured at half of the bandwidth 2 res W . As can be seen in Figure 3, the frequency of the NFMR resonance for epoxy composites containing 1 wt.%CNT/30 wt.% BaFe12-xGaxO19 is higher compared with the pressed BaFe12-xGaxO19 sample, and such a change is similar to the 30 wt.%BaFe12-xGaxO1/epoxy composites investigated in our previous paper [34].   [37]; it was shown that the magnetic parameters of 30 wt.% BaFe12-xGaxO19/epoxy composites are higher than the corresponding parameters of pure BaFe12-xGaxO19 (0 ≤ x ≤ 0.1) polycrystalline samples. It was concluded that f res is determined by the magneto-crystalline anisotropy field H a and magnetic saturation M s of BaFe 12-x Ga x O 19 [35,36]: where γ/2π = 2.8 MHz/Oe is the gyromagnetic ratio. Following from Equation (4), the increase in f res may be related to the increase in H a at M s = constant or to the decrease in M s at H a = const. It may be concluded that the increase in f res in the case of the 1% CNT/30 wt.% BaFe 12-x Ga x O 19 /epoxy composite is the result of the H a increase and M s decrease observed for 30 wt.% BaFe 12-x Ga x O 19 /epoxy CMs in our previous research [37]; it was shown that the magnetic parameters of 30 wt.% BaFe 12-x Ga x O 19 /epoxy composites are higher than the corresponding parameters of pure BaFe 12-x Ga x O 19 (0 ≤ x ≤ 0.1) polycrystalline samples. It was concluded that the polymer coating on magnetic particles obviously affects the contributions of the surface anisotropy, shape anisotropy, and interface anisotropy to the total anisotropy [38,39]. The slightly higher values of f res for the aligned 1% CNT/30 wt.% BaFe 12-x Ga x O 19 /epoxy composite may be related to the higher value of the magneto-crystalline anisotropy field H a due to a change in the shape anisotropy at the formation of the elongated barium hexaferrite chains under magnetic field alignment [36,40].
This f res also depends on the cation Ga 3+ concentration in BaFe 12-x Ga x O 19 . The concentration dependencies of the resonance frequency for BaFe 12-x Ga x O 19 and 1% CNT/30% BaFe 12-x Ga x O 19 /epoxy are nonmonotonic and have a minimum at x = 0.6. As shown for BaFe 12-x Ga x O 19 , this dependence can be satisfactorily approximated by the second-order polynomial f res = 50.04 + 3.37x 2 − 3.73x [41]. This concentration behavior is observed during a monotonic decrease in the magnetic parameters, such as the Curie temperature, the remnent magnetization, and the coercive force, when the cation Ga 3+ concentration increases. Thus, the increase in the resonance frequency at x ≥ 0.6 is thought to be caused by an increase in the magneto-crystalline anisotropy field H a and a decrease in the saturation magnetization M s with the Ga 3+ content increase. For 1% CNT/30% BaFe 12-x Ga x O 19 /epoxy CMs with an oriented distribution of fillers, the dependence of f res on the Ga 3+ concentration is weaker than for random 1% CNT/30% BaFe 12-x Ga x O 19 /epoxy CMs.
The amplitude of the resonance for 1% CNT/30% BaFe 12-x Ga x O 19 /L285 is lower compared with a pure pressed sample of BaFe 12-x Ga x O 19 and also changes with the Ga 3+ concentration: firstly, it decreases with the Ga 3+ concentration up to x = 0.6, and then it sharply increases for x = 0.9 and decreases again for x = 1.2. For the pressed samples of BaFe 12-x Ga x O 19 , the opposite behavior of A res on the Ga 3+ concentration is observed. It should be noted that the determination of A res for 1% CNT/30% BaFe 12-x Ga x O 19 /L285 is approximate, since the resonance peaks are less pronounced.

Amplitude-Frequency Characteristics of NFMR for 1% CNT/30% BaFe 12-x Ga x O 19 /Epoxy Composites at Applied Magnetic Field
Figures 4 and 5 present the results of the NFMR study in which a DC magnetic field was applied to 1% CNT/30% BaFe 12-x Ga x O 19 /L285. As can be seen from Figure 4a, an applied DC magnetic field leads to a decrease in the amplitude of the NFMR resonance for the 1% CNT/30% BaFe 12-x Ga x O 19 /L285 composite with a random filler distribution for all Ga 3+ concentrations (x = 0.1-1.2). Regarding the frequency of the NFMR resonance, this does not change with the application of a DC magnetic field.  In the case of 1% CNT/30% BaFe12-xGaxO19/L285 with an aligned filler distribution (Figure 5a), an increase in the amplitude of NFMR was also observed; however, the ( )   CNT/30% BaFe12-xGaxO19/L285 with 0.6 Ga 3+ , res f increases from 51.9 to 52.3 GHz in the ext H range 0-5 kOe.
where ext H is the applied DC magnetic field.
As shown in [41] for BaAlxFe12-xO19 samples, the behavior of the resonance frequency In the case of 1% CNT/30% BaFe 12-x Ga x O 19 /L285 with an aligned filler distribution (Figure 5a), an increase in the amplitude of NFMR was also observed; however, the A res (H ext ) dependencies are more complicated. Firstly, A res increases with H ext up to 2.5 kOe and then does not change with the magnetic field increase. Contrary to the random 1% CNT/30% BaFe 12-x Ga x O 19 /L285 composite, for the aligned 1% CNT/30% BaFe 12-x Ga x O 19 /L285 composites, a slight increase in f res is observed. For example, for 1% CNT/30% BaFe 12-x Ga x O 19 /L285 with 0.6 Ga 3+ , f res increases from 51.9 to 52.3 GHz in the H ext range 0-5 kOe.
It was noted that such an increase in f res was sufficiently lower compared with pressed polycrystalline BaFe 12-x Ga x O 19 , where f res increased from 49 to 54 GHz in the H ext range 0-3.5 kOe; these dependencies were almost linear for all samples [40]. As concluded in [40] for pressed polycrystalline BaFe 12-x Ga x O 19 , the resonance frequency increased with the magnetic field as the internal magnetic field related to the anisotropy increased.
Such behavior of the minimums of the SE T ( f ) dependencies and changes in the amplitudes of the SE T peaks with the variation in the magnetic field values confirms their ferromagnetic nature.
Within the theory of hexagonal ferrites, the NFMR frequency f res at the applied external magnetic field H ext may be described by the following expression [42]: where H ext is the applied DC magnetic field. As shown in [41] for BaAl x Fe 12-x O 19 samples, the behavior of the resonance frequency f res versus the applied magnetic field H ext is determined by the value of the saturation magnetic field H sat . For the range of the external magnetic field H ext < H sat , the resonance frequency f res is approximately constant at the applied DC magnetic field. If the value of the applied magnetic field H ext is higher than H sat , f res of NFMR linearly increases with H ext . Table 2 shows the data on magnetic parameters for the pressed polycrystalline BaFe 12-x Ga x O 19 samples and 30 wt.% BaFe 12-x Ga x O 19 /epoxy composites (x = 0.1-1.2), which were studied in our previous papers [37,40]. As can be seen from Table 2, for the pressed polycrystalline BaFe 12-x Ga x O 19 samples, H sat ≈ 20 kOe [40], which is why the approximately linear dependencies f res (H ext ) in the range of H ext = (0-4) kOe have a slope (γ/2π = 1.5-2) that is lower than the theoretical value of 2.8.
In the case of 30% BaFe 12-x Ga x O 19 /L285 (for x = 0-1.2), the values of the saturation magnetic field H sat are higher (23-33) kOe [39], which results in independence (for random composites) or only a slight increase (for aligned epoxy CMs with a lower saturation field H sat compared with random composites) in f res with the applied magnetic field H ext . This statement is also correct for both random and aligned 1% CNT/30% BaFe 12-x Ga x O 19 /L285 composites. Thus, we need to highlight that the quality of the dispersion/alignment of fillers in composites is very important for the electrodynamic properties of CMs [42,43].

Conclusions
Epoxy composites with random and aligned magnetic field distributions of 1 wt.% of CNTs and 30 wt.% of BaFe 12-x Ga x O 19 (0.1 < x < 1.2) were fabricated. It was found that adding 1 wt.% CNTs along with 30 wt.% of BaFe 12-x Ga x O 19 into epoxy resin resulted in an increase in electrical conductivity up to 5.0 ×·10 −8 S/m that is explained by the high electrical conductivity of CNTs. The observed sufficient increase in the microwave shielding efficiency of ternary random and aligned 1% CNT/30 wt.% BaFe 12-x Ga x O 19 /epoxy composites in the frequency range 36-55 GHz compared with binary 1% CNT/epoxy and 30 wt.% BaFe 12-x Ga x O 19 /epoxy was explained by the increased complementary contributions of dielectric and magnetic losses and the increase in the EMR attenuation constant. The higher values of the natural ferromagnetic resonance (NFMR) frequency f res (51.8-52.4 GHz) and weaker dependence of f res on the Ga 3+ concentration in 1 wt.% CNT/30 wt.% BaFe 12-x Ga x O 19 /epoxy composites compared with pressed polycrystalline BaFe 12-x Ga x O 19 ( f res = 49.6-50.4 GHz) may be related to the higher values of the magnetic parameters of 30 wt.%BaFe 12-x Ga x O 19 in the epoxy matrix compared with the corresponding parameters of pure pressed BaFe 12-x Ga x O 19 (0 ≤ x ≤ 0.1) polycrystalline samples. The slightly higher values of f res for the aligned 1% CNT/30 wt.% BaFe 12-x Ga x O 19 /epoxy composite compared with the random composites may be related to the higher value of the magneto-crystalline anisotropy field H a due to a change in the shape anisotropy at the formation of the elongated barium hexaferrite chains under magnetic field alignment. The approximately constant value of the NFMR frequency f res in the range of the applied magnetic field, H = 0-5 kOe, for the random 1 wt.% CNT/30 wt.% BaFe 12-x Ga x O 19 /epoxy composite and slightly increased f res for the aligned 1 wt.% CNT/30 wt.% BaFe 12-x Ga x O 19 /epoxy composite were explained by the lower saturation field H sat compared to the pressed