Anisotropic X-Band Microwave Properties of Amine-Functionalized Carbon Fibers Derived from Polyacrylonitrile

Abstract

Carbon fibers derived from carbonized and activated polyacrylonitrile (CFPAN) were sequentially brominated and subsequently functionalized with selected primary and secondary amines to engineer a directional electromagnetic (EM) response. Besides bromine incorporation, bromination introduced oxygen-containing surface groups (e.g., carboxyl, lactone), enabling nucleophilic substitution by amines. Surface characterization (SEM-EDS, FTIR ATR) confirmed successful amine grafting, while thermal analysis (TGA, TPD MS) revealed increased weight loss in the 150–450 °C range due to the decomposition of covalently bonded nitrogen- and oxygen-containing moieties, evidencing strong surface functionalization. Microwave characterization in the X-band (8.2–12.4 GHz) demonstrated that functionalization strongly influences the EM response of CFPAN fibers. The measured reflection coefficient varied from −1.0 to −2.5 dB for sulfonylethylenediamine (SuEn)-functionalized fibers and from −2.0 to −4.0 dB for ethylenediamine (En)-treated ones, depending on frequency and fiber orientation. The frequency-averaged absorption coefficient of pure CFPAN amounted to 32–41%, with absorption maxima and minima corresponding to orientations differing by 90°. SuEn modification decreased absorption to 21–35%, while En functionalization enhanced it to 32–51%. Pure CFPAN exhibited the lowest absorption anisotropy (factor 1.28), whereas piperazine- and En-modified samples showed the highest anisotropy (1.57 and 1.59, respectively). Across all compositions, the attenuation constant remained within 1.5–4.5 mm⁻¹. The observed anisotropic behavior is governed primarily by orientation-dependent variations in characteristic impedance and, to a lesser extent, by anisotropic attenuation constants. Such tunable anisotropy is particularly advantageous for EM shielding textiles, where fiber alignment can be tailored to enhance interaction with polarized fields. Among the tested amines, En-functionalized CFPAN exhibited the highest nitrogen content (up to 10.1 at%) and the most significant enhancement in microwave absorption, positioning it as a promising candidate for advanced orientation-sensitive shielding applications.

1. Introduction

The rapid advancement of communication technologies has not only enhanced various aspects of human life but has also contributed to significant electromagnetic (EM) pollution due to the extensive use of electronic devices [1,2,3]. EM absorbing materials, which attenuate or dissipate EM waves, are of considerable interest for mitigating these effects. The study of chemical modifications and their impact on EM wave interactions is highly relevant, as high-frequency EM radiation and electromagnetic interference (EMI) pose potential risks to human health and safety [4,5,6] and can cause severe disruptions to both military and civilian infrastructure [7,8]. Consequently, the development of highly efficient shielding materials with tailored EM properties has become a critical research focus. Recent studies have increasingly explored the EM response of materials designed to reflect or absorb microwave radiation [9,10,11,12]. Research efforts have been directed toward the mitigation of EMI, the enhancement of stealth technology [13,14], and the development of protective materials for EM shielding applications [15,16]. Carbonaceous materials, including carbon fibers, possess physicochemical properties that make them highly suitable for EM absorption. Their high electrical conductivity, large specific surface area, and low density, combined with their resistance to corrosion, chemical degradation, and mechanical stress, render them advantageous for use in EM shielding [17,18,19]. Due to these favorable characteristics, carbon fibers are often incorporated into composite materials, typically consisting of a polymeric matrix and a carbonaceous filler, thereby combining the structural and functional benefits of both components. A fundamental aspect of carbon materials, including carbon fibers (CFs), is their ability to undergo chemical modification of carbon surface layer. By introducing atoms different from carbon into the surface carbon layer, one can significantly alter their physical and chemical properties [20,21,22,23] and, consequently, their interaction with EM radiation [23,24,25].

Our rationale for selecting CFs is that they combine high electrical conductivity, tunable surface chemistry, and excellent mechanical stability with relatively low density. In particular, CFs can be directly functionalized (e.g., by amination), which not only modifies their electronic structure but also introduces dipolar sites that promote absorption losses. Such features make them uniquely suitable for developing lightweight and absorption-dominant shielding materials, in contrast to carbon nanotubes (CNTs) or graphene nanoplatelets (GNPs) that are often studied in polymer or cement composites. Therefore, while Jang et al. [26] provide a valuable perspective on filler comparison, our focus on aminated CFs allows us to highlight the role of direct chemical modification of fibers themselves—a direction less explored in the literature and complementary to composite-based studies. CFs composed of thin filaments primarily made of carbon atoms, which are arranged into microscopic crystals aligned parallel to each other. This alignment imparts high tensile strength to the fibers. CFs also exhibit low density, low coefficient of thermal expansion, and chemical inertness. The polyacrylonitrile-derived carbon fibers (CFPAN) used in this work combine high thermal and chemical stability with the ability to undergo chemical modification, unlike fully graphitized CFs. Their large specific surface area, developed porous structure, and resistance to pH variations make them particularly suitable for functionalization studies. Modified CFPAN fibers are expected to retain higher thermal stability of grafted functional groups compared to other carbon materials, such as activated carbon (AC) or carbon nanomaterials, making them an excellent model system for investigating chemically modified carbon materials [27,28,29,30,31,32].

Despite recent advances in the development of carbon-based EM absorbers, clear correlations between surface chemical modifications and anisotropic microwave properties have not yet been systematically established. Prior studies have largely focused on bulk composites or isotropic responses, leaving the influence of controlled heteroatom doping and surface functionalization of carbon fibers underexplored. This gap arises largely from the difficulty of disentangling surface chemistry contributions from bulk or matrix effects. In particular, orientation-dependent absorption and reflection behaviors of carbon fibers in the X-band have received little direct attention, despite their importance for anisotropic EM performance.

The present work addresses this gap by introducing a selective amination strategy for CFPAN, enabling the incorporation of nitrogen functionalities without compromising structural integrity. This approach allows us to directly probe how specific amino groups influence dielectric losses, absorption anisotropy, and impedance matching. To our knowledge, this is the first systematic study that links surface amination of carbon fibers to their anisotropic EM performance, providing new opportunities for the design of next-generation radio-protective textiles and high-efficiency EM shielding materials.

2. Materials and Methods

2.1. Materials

Pristine CFPAN belong to the family of CFs with well-developed nanoporosity. Commercial CFPAN fabric (SkyCarbon, Kharkov, Ukraine) was used as received. The material consists of a narrow tape approximately 4 cm wide and several meters long (Figure 1). Individual CFs have diameters of 6–10 µm, and a single fabric layer has a thickness of approx. 340 µm. The fabric features a longitudinal carbon backbone with filler fibers woven perpendicular to it, providing a stable structure suitable for chemical modification. Carbonization in an argon atmosphere followed by steam activation was performed at 800 °C, producing fibers with a relatively low degree of graphitization, which ensures sufficient chemical reactivity. The narrow width of the fabric allows small pieces to be cut without fraying or disassembly, an advantage over conventional meter-wide carbon fabrics.

This CFPAN can be considered as an ideal candidate for studying the effects of chemical modification on physicochemical and microwave properties. Their inherent anisotropy in microwave response must be considered when designing composites or functionalized materials. Furthermore, understanding how chemical modification and particle size affect microwave anisotropy is critical, as the literature on these effects remains limited.

2.2. Reagents

The reagents from Sigma-Aldrich (Burlington, MA, USA) included bromine (Br₂, 99.99%) as the brominating agent, potassium oxalate (K₂C₂O₄, ≥99%) to fix physisorbed bromine, and sodium carbonate (Na₂CO₃, ≥99.5%) as an agent for neutralization and regeneration of amino groups. Isopropyl alcohol (i-C₃H₇OH, 99.5%) and deionized (DI) water (ASTM Type II, HARLECO^® Purified Water, Merck KGaA, Darmstadt, Germany) were used as solvents. Hydrochloric acid (HCl, 32% v/v) was used to desorb physisorbed amines. The reagents from Merck KGaA (Darmstadt, Germany) included sodium hydroxide (NaOH, ≥99%) and sodium nitrate (NaNO₃, ≥95%) and were used to decompose samples prior to potentiometric and titrimetric analyses. Concentrated reagents were diluted with DI water, and all samples were washed with double-distilled water (DDW), which was prepared using a laboratory bidistillator.

The selected amines (for chemical formulas see Figure S1 in Supplementary Materials) for chemical modification included ethylenediamine (En), sulfonylethylenediamine (SuEn), diethylamine (Et₂N), monoethanolamine (MEA), and piperazine (PIP) (all reagent purity, Enamine, Kyiv, Ukraine and Reagent grade or Laboratory reagent of Sigma-Aldrich, Burlington, MA, USA), with their respective abbreviations used in sample labeling.

Argon gas (99.95%) was supplied in a cylindrical vessel by Linde Gas (Dublin, Ireland).

2.3. Preparation

Bromination with liquid bromine, followed by amination, was performed according to the procedure described in [33], with modifications. Prior to liquid-phase amination, CFPAN sample was subjected to bromination. For this process, 5 g of CFPAN sample was immersed in 100 mL of an aqueous bromine solution adjusted to provide 1 mmol of Br per g of fabric and kept at room temperature (22 °C) for 60 min under constant stirring. To remove physisorbed bromine, the brominated PCFPAN sample was treated with 200 mL of a 10% w/v aqueous K₂C₂O₄ solution until CO₂ evolution ceased. The resulting material, hereafter referred to as PCFPAN/Br₂, was washed with DI water for couple days until no bromide ions were detected in the rinse water by analytical AgNO₃ test. The washed PCFPAN/Br₂ sample was finally dried at 120 °C for 2 h using electrical resistance heating.

For the amination step, 1 g of PCFPAN/Br₂ sample containing ~1 mmol Br per g was placed in a beaker and treated with an excess of a 30% (v/v) i-C₃H₇OH solution of the selected amine. The mixture was transferred into a Teflon-lined autoclave, sealed and heated at 120 °C for 15 h to promote chemisorption and desorption of physisorbed amine. After cooling to room temperature, the aminated PCFPAN/Br₂ samples were washed sequentially with (i) DDW, (ii) 0.1 N HCl solution (to remove unbound and physisorbed amine), and (iii) 1% (w/v) Na₂CO₃ solution (to regenerate amino groups), followed by rinsing with DI water until its neutral pH was reached. The neutralized samples were dried in hot air at 120 °C for 10 h.

Subsequent studies were conducted using both the unmodified CFPAN and a series of aminated PCFPAN/Br₂ samples, denoted as CFPAN/Br₂/X, where X = MEA, En, SuEn, or PIP, corresponding to the respective amine residues.

2.4. Methods

A high-resolution analytical scanning electron microscope (SEM), Tescan Mira3 (TESCAN group, Brno, Czechia), equipped with a high-brightness field emission electron source, was employed for imaging. Beam deceleration technology was utilized to enhance resolution at low accelerating voltages. The SEM column incorporates an additional condenser lens that optimizes the electron beam spot size for higher beam currents while enabling Depth and Wide-Field™ observation modes. High-resolution imaging at short working distances was conducted using the in-column secondary electron detector of the Tescan Mira system.

Additionally, a Jeol JSM-6490-LV SEM (JEOL Ltd., Tokyo, Japan) was employed for image capture, while elemental analysis was performed using an AZtec (Advanced Inca Energy 350) energy-dispersive X-ray spectroscopy (EDS) system (Oxford Instruments, High Wycombe, UK) operated with the AZtec 3.3 software package for X-ray mapping and elemental analysis. This EDS system is equipped with a large area X-Max 80 mm² elemental silicon drift detector (ESDD), which enhances the sensitivity and accuracy of elemental composition analysis.

For further analysis, the Inspect S SEM from (FEI Company/ThermoFisher Scientific from FEI Manufacture, Hillsboro, OR, USA) was utilized. This model featured a low-vacuum secondary electron detector for high-resolution imaging and an EDS system with a Back Scattered Electron Detector (BSED) for compositional contrast. The SDD in this SEM had a detection area of 30 mm², providing a resolution of 129 eV on Mn Ka at a counting rate of 10,000 counts per second. It achieved a throughput exceeding 300,000 counts per second and a peak/background ratio greater than 10,000:1, allowing for the detection of elements from Beryllium (Be) to Americium (Am). Sample preparation involved the use of aluminum stages with double-sided carbon tape to secure the powder samples. The samples were poured onto the adhesive surface and gently pressed with a spatula to ensure adequate contact, followed by the removal of excess powder using compressed air. The SEM analyses were conducted in low-vacuum mode at 80 Pa, with a voltage of 20 kV applied during mapping mode to accurately determine the elemental composition. This comprehensive methodology ensured precise imaging and elemental analysis of the powder samples under controlled conditions.

The total bromine concentration (C_Br) was determined by potentiometric and titrimetric methods of chemical analysis following standard procedures [34,35].

The CFPAN was analyzed using transmission electron microscopy (TEM) on a Jeol JEM 1230 TEM microscope (JEOL Ltd., Tokyo, Japan). Prior to visualization, an aliquot of the sample was deposited onto a copper grid coated with a formvar film. The sample was analyzed 24 h after the preparation procedure.

Fourier Transform Infrared (FTIR) spectra were taken to identify the characteristic vibrations in the samples produced. The spectra, recorded with a resolution of 4 cm⁻¹ in the spectral range between 4000 and 500 cm⁻¹, were collected using an FTIR Prestige-21 spectrometer (Shimadzu Co., Kyoto, Japan) in the attenuated total reflection (ATR) mode, using 30 scans. At least three measurements were taken for each sample using the ATR MIRacle accessory (PIKE Technologies, Madison, WI, USA) at 20 °C. Using a metal rod and continuous mechanical pressure, the samples were pressed against the diamond surface to ensure perfect contact.

Temperature-programmed desorption mass spectrometry (TPD MS) analysis was conducted under ultra-high vacuum (8 × 10⁻⁶ Pa in the operating mode) to minimize interference from re-adsorbed gases. Samples were heated from 30 to 800 °C at a linear heating rate of 10 °C min⁻¹, enabling both detection and identification of volatile species released during thermal decomposition. Thermal desorption profiles were acquired on a MX 7304A mass spectrometer (Selmi, Sumy, Ukraine) using electron ionization [36]. Volatile species in the m/z 12–200 range were continuously monitored as a function of temperature, providing real-time insight into desorption kinetics and decomposition pathways.

In parallel, thermogravimetric analysis (TGA) was carried out to monitor weight loss as a function of temperature using a computer-controlled thermal analyzer [37]. Samples (~0.1 g) were heated in a dry argon atmosphere from 30 to 1000 °C at a linear heating rate of 10 °C min⁻¹. Thermograms: thermogravimetric (TG) and differential thermogravimetric (DTG) profiles were recorded with a sensitivity of 50 mg, providing precise characterization of weight loss and decomposition stages.

The complementary use of TPD MS and TGA provides a comprehensive understanding: TPD MS identifies the gaseous species evolved and their desorption temperatures, whereas TGA quantifies the weight changes. The complementary application of TPD MS and TGA allowed correlation of evolved gas species with weight-loss events, providing a comprehensive understanding of thermal stability and decomposition pathways of functional groups in CFPAN and its aminated derivatives.

Nitrogen (N₂) adsorption–desorption isotherms were measured at liquid nitrogen temperature using a Quantachrome Autosorb-6 nitrogen sorption analyzer (Quantachrome, Boynton Beach, FL, USA). Tested sorbents were degassed overnight at 120 °C under vacuum prior to measurements.

2.5. Microwave Property Measurements

Microwave studies of the high-frequency EM properties of synthesized composite materials were performed using the X-band (8–12 GHz) scalar network analyzer. Measurements were made using a test cell consisting of two sections of a standard metal rectangular waveguide with a cross-section of 23 × 10 mm². Figure 2 illustrates some experimental setup for X-band microwave spectroscopy measurements and including (a) the principal measurements schema, (b) schema of axis positioning relative to the sample orientation, (c, d, e) installation views and (f) sample’s rotation setup used. The sample was a single layer of chemically modified carbon fiber fabric with an average thickness d, positioned between the waveguide flanges (Figure 2c) to fully cover the cross-section (see in Figure 2a). Due to this positioning, the plane of the sample was perpendicular to the propagation vector $\overrightarrow{k}$ of the H10 EM mode of the waveguide. In addition, the orientation of the linearly polarized E-field vector of the incident EM wave was exactly in the plane of the sample. The measurement setup allowed continuous and controllable rotation of the sample of CFs around the longitudinal axis of the waveguide. Thus, the angle between the axis of CFs and the direction of the E-field polarization of the EM wave varied smoothly between 0° and 90°. This makes it possible to study the anisotropic absorption properties of given materials.

To do measurements at different angles, according to positioning shown in Figure 2b, the sample rotation setup consists of two interlocking pieces that allow free rotation of one relative to the other was used. Both pieces are marked on their surfaces to track relative orientation in 15° steps. In a typical experiment, one piece securely holds the sample under investigation, while the other is fixed to the waveguide flange (Figure 2c–e). During the experiments, after each measurement, the sample holder was removed from the waveguide section, rotated to the required angle, and then reattached to the anchored piece. In this way, a controlled rotation of the sample axis relative to the waveguide axis was achieved.

In a typical experiment, the network analyzer allows measuring the absolute values (modules) of microwave scattering parameters (S-parameters) of various microwave devices, components and samples in logarithmic scale. They are defined as follows [38]:

$$S_{21}=10lg\left({\displaystyle \frac{P_T}{P_I}}\right)=10lg\left({\displaystyle \frac{E_T^2}{E_I^2}}\right),$$

(1)

$$S_{11}=10lg\left({\displaystyle \frac{P_R}{P_I}}\right)=10lg\left({\displaystyle \frac{E_R^2}{E_I^2}}\right),$$

(2)

where P_I is the initial (incident) power of the EM wave, P_T is the power transmitted through the sample, and P_R is the reflected power (see Figure 2), and correspondingly, E_I, E_T, and E_R are the electric field strengths of the incident, transmitted, and reflected waves. These quantities have a simple physical interpretation: S₁₁ quantifies the fraction of EM energy reflected by the sample, the CFPAN-based samples in our case; S₂₁ shows the fraction of energy that is transmitted through it.

In addition to the scattering parameters, we will also use the dimensionless EM power transmission T and reflection R coefficients introduced as

$$T={\displaystyle \frac{P_T}{P_I}}={10}^{{\displaystyle \raisebox{1ex}{$S_{21}$}\!\left/ \!\raisebox{-1ex}{$10$}\right.}},$$

(3)

$$R={\displaystyle \frac{P_R}{P_I}}={10}^{{\displaystyle \raisebox{1ex}{$S_{11}$}\!\left/ \!\raisebox{-1ex}{$10$}\right.}}.$$

(4)

Using the measured data of the scattering parameters, the EM wave absorption coefficient A can be found from the power balance equation:

$$A=1-R-T.$$

(5)

Prior to measurements, the scalar network analyzer was carefully calibrated, and reference planes were placed on the flanges of the test cell (see Figure 2). In this way, the transmission losses of the wave-guide adapters and connecting sections were properly subtracted, and the measured S-parameters represented only the specific reflection/absorption properties of the tested material. After calibration, the sample was inserted as described above and the 8–12 GHz frequency spectra of the EM energy reflected and transmitted by the studied modified CFPAN samples (i.e., S₁₁ and S₂₁ parameters) were recorded. The measured data were further subjected to mathematical post-processing using low-pass Fast Fourier Transform filtering to reduce spurious noise and obtain more reliable curves. During the investigation, the samples of modified CFPANs were manually rotated around the direction of EM wave propagation so that the angle φ between the axis of CFs and the direction of EM wave E-field polarization varied from 0° to 90° in 15° steps.

3. Results

3.1. HRTEM

HRTEM images (Figure 3a,b) show the nanostructures of CFPAN; the latter have a highly detailed, layered structure with fine, wavy and closely packed lines. Dark regions in the images correspond to areas of higher electron density or thicker/more ordered carbon stacking, as more atoms scatter the electrons and reduce transmitted intensity. The images present a textured pattern with no distinct large objects, and the nanostructures appear to be graphitic or carbonaceous layers with some degree of disorder [39,40].

The pattern is like the turbostratic carbon or disordered graphitic layers common in PAN-Based CFs and AC materials [41,42,43,44]. The fine wavy lines indicate stacked graphene-like layers, but with irregularities suggesting a presence of partially amorphous or disordered structures. In addition, the fine wavy lines in the micrographs represent stacked graphitic layers, but their curvature [45] is irregular rather than perfectly straight. The waviness and curvature of the carbon layers suggest a turbostratic carbon structure, meaning that the carbon layers are misaligned rather than in a highly ordered graphitic form [46]. Curvature is typically affected by defects, oxidation, and thermal treatment-in this case, presumably by oxygen functional groups that rather intensively increase disorder [47,48,49]. The radius of curvature appears to vary at different points, with some regions showing more tightly bent layers and others appearing more relaxed. The visible graphitic layers have an interlayer spacing that appears larger than that of pure crystalline graphite, which is ~0.335 nm for ideal graphite [50].

Based on the 5 nm scale bar (Figure 3a,b), the individual layers appear to be a few nanometers thick, with the interlayer spacing likely in the range of 0.34–0.38 nm and suggesting defect-induced strain [51]. The wavelength of the waviness varies, but on average the waviness is in the range of 1–2 nm, suggesting that these are nano-structured features. The lack of long-range order distinguishes it from highly crystalline graphite. The wavy, layered structure in the image suggests the presence of slit-like pores formed between misaligned or curved graphitic layers [52].

These nanoscale voids are caused by structural defects, partial oxidation, or incomplete stacking of carbon layers, and the dark and light contrasts in the image indicate regions where the layers separate, forming narrow, elongated pores rather than round or irregular pores. Based on the 5 nm scale bar, the slit-like pores appear to be in the sub-nanometer (ultramicropore) range (below ~0.5–0.7 nm wide). In the images (Figure 3a,b), the carbon walls (stacked graphene layers) of pores appear as darker fringes, while the voids or slit-like pores are observed as light (bright) elongated regions between the fringes. Although HRTEM images do not resolve pore structures, the porosity parameters (see Section 3.2) indicate that some pores extend several nanometers, consistent with the presence of both mesopores (2–50 nm) and micropores (<2 nm). The pore width is likely governed by interlayer spacing, which can increase during pyrolysis due to the incorporation of oxygen-containing functional groups.

According to Ref. [53], pyrolysis at moderate high temperatures typically removes oxygen-containing functional groups, leading to densification and increased graphitic ordering, thereby reducing the interlayer d-spacing. At higher temperatures or under defect-rich oxidation, however, structural disorder and layer bending cause the d-spacing to rebound to values in the range of ~0.36–0.39 nm. Residual oxygen-rich functional groups and structural defects, originating from incomplete pyrolysis or intentional oxidation, further contribute to interlayer expansion and the formation of micropores and nanoscale distortions in the carbon matrix [54,55]. This structural evolution is consistent with the features observed in the prepared CFPAN.

The PAN-derived CFPAN sample inherently contains nitrogen, which has significant influence on the structural evolution of graphite crystallites. According to the literature [56,57], nitrogen atoms can be present in carbon materials in several forms: substitutionally incorporated into aromatic rings, as interplanar bridges within the graphite lattice, and at the terminal positions of graphite planar edges. The latter readily thermodesorbs to form highly reactive dangling bonds that facilitate graphite crystallite growth via sp²/sp³ carbon interactions [58].

It should be noted that, typically, transforming PAN precursor into CFPAN requires stabilization and carbonization, while graphitization is optional. The precursor is stabilized at 200–400 °C, during which cyclization, dehydrogenation, and oxidation convert the linear PAN chains into a thermally stable conjugated ladder structure [59]. This dense aromatic framework limits inward diffusion of oxygen, leading to retention of oxygen in the fiber sheath and promoting structural densification. As a result, the density of specimens depends on the initial oxygen content and distribution, ranging from 1.167 to 1.380 g cm⁻³ prior to carbonization [59]. Progressive heat treatment further densifies the material through impurity removal and contraction of graphite planes, increasing the density up to 1.76–1.88 g cm⁻³ after carbonization [59,60].

HRTEM analysis (Figure 3a,b) reveals nanometer-scale graphitic domains embedded within an amorphous carbon matrix. These domains exhibit short-range ordering of graphene layers, but their curved and misaligned fringes indicate a turbostratic structure without long-range crystalline order. In addition, no distinct crystalline–amorphous phase boundary can be discerned, which is characteristic of turbostratic carbons. In such structures, the gradual loss of registry between adjacent graphene layers and the presence of curved, misaligned fringes result in a continuous transition between ordered and disordered regions rather than sharp interfaces. The graphite-like planes exhibit irregular curvilinear alignment, with amorphous regions containing sp²-bonded carbon clusters. Such curvilinear fringes are typically associated with partial orientation relative to the fiber axis, although this cannot be unambiguously resolved from the present images. Structurally, carbon fibers consist of sp² carbon layers (crystalline graphite) and sp² clusters (disordered carbon-like components), with amorphous carbon distributed parallel to the stacking planes and at interplanar cross-linking sites [60]. Crystallization proceeds through an amorphous-to-graphitic transition, manifested as the gradual ordering of turbostratic graphene layers into a more regularly stacked graphite-like structure with emerging interlayer (three-dimensional) correlations. Compared to the outer layer, the core region exhibits disordered graphite stacking with presumable axial wrinkling. As reported in Ref. [61], graphite crystallites with aligned orientations can form percolating branched networks, whereas misaligned domains give rise to grain boundaries, dislocations, and stacking faults. In the present micrographs (Figure 3a,b), however, only short-range turbostratic order is observed, without clear evidence of such extended crystalline defects. Notably, crystallite thickening and elongation correlate with the spatial distribution and structural evolution of amorphous carbon, highlighting its key role in microstructural refinement. As can be seen in Figure 3c, the presence of multiple concentric SAED rings suggests that the material is nanocrystalline or turbostratic, i.e., it consists of many small, randomly oriented crystallites rather than a highly ordered single crystal. The broad and diffuse nature of the rings indicates that the material has some degree of disorder, consistent with partially graphitized carbon structures such as those found in biochar, AC, or amorphous carbon. The intensity (brightness) of the rings can be associated with specific interplanar d-spacings that help to identify the crystallographic structure of the material. Based on the reciprocal scale, the positions of the rings suggest characteristic d-spacings typically found in carbon-based materials. For the CFPAN sample, the common d-spacings are the first ring (~0.34 nm) corresponds to the (002) plane of graphite, indicating a layered carbon structure. The second ring (~0.21 nm) probably corresponds to the (100) plane, associated with in-plane hexagonal carbon structures [62]. The third ring (~0.12 nm) corresponds to the (110) plane, associated with more disordered carbon structures. The (002) ring shift or broadening suggests partial oxidation, which can increase reactivity but decrease electrical conductivity. This effect corresponds to layered graphitic structures. The d-spacing of pure graphite d(002) is 0.335 nm, so a slightly larger value (~0.34 nm) indicates increased interlayer spacing due to defects or oxygen functional groups, and this observation may indicate partial amorphization in the carbon matrix. An increase in d(002) indicates increased disorder and possible oxidation, which is common in PAN CFs prepared at high temperatures. The second ring (~0.21 nm, (100) plane) is consistent with graphene-like domains present in CFPAN. The third ring (~0.12 nm, (110) plane) represents disordered or turbostratic carbon where dislocations and defects dominate. The presence of the (100) and (110) planes confirms a mixture of graphitic and disordered carbon, indicating that pyrolysis alters the stacking order. Although pyrolysis largely eliminates oxygen-containing groups, residual oxygen functionalities can remain at edges and defects, where they contribute to disorder in the carbon lattice. The observed disorder is also consistent with incomplete carbonization and nitrogen incorporation, since substitutional, interplanar, and edge-type N atoms introduce lattice distortions that hinder the development of long-range crystalline order.

It should be emphasized that from the elemental composition of CFPAN (see Table 1, Section 3.2), it is evident that during pyrolysis of PAN-based fibers most oxygen-containing groups (–OH, –COOH, –C=O, etc.) are eliminated as volatile species (CO, CO₂, H₂O), particularly above ~600 °C where the oxygen content decreases sharply. Nevertheless, incomplete pyrolysis at lower temperatures or shorter dwell times can leave residual oxygen functionalities at the edges and defect sites of carbon sheets. Furthermore, re-oxidation during post-pyrolysis handling may also reintroduce oxygen at edge sites. These residual groups, which tend to localize in amorphous regions, edges, and grain boundaries, indicate that while pyrolysis is primarily an oxygen-removal process, structural disorder and incomplete carbonization allow some oxygen to persist and influence stacking order.

3.2. SEM, SEM-EDS, Specific Surface Area and Porosity Studies

Figure 4 presents SEM images of CFPAN/Br₂/En and CFPAN/Br₂/SuEn samples obtained through bromination followed by amination. The original CFPAN consists of CFs approximately 9.3 μm in diameter. A relatively smooth carbon surface is best seen in Figure 4b,d. This surface is the result of mechanical forming of CFPAN textile using spinnerets. As a result of carbonization and release of gaseous pyrolysis products, some depressions and irregularities appear on the surface. However, in general, the surface of the carbon fiber does not contain protruding sharp elements. Amination introduces only minor changes to the surface morphology after bromine treatment [33], and in some cases, appears to smooth or even heal preexisting surface imperfections, giving, as a result, a relatively smooth carbon surface (Figure 4c,d).

Cross-sectional SEM images (Figure 4a,c) reveal a dense outer layer and a more porous inner core, a characteristic feature of PAN-originated CFs after pyrolysis (see also Figure S1, Supplementary Materials). This demonstrates some structural heterogeneity between the outer shell and inner core. In Figure S1, the enlarged central area (left panel) shows numerous macropores, although their exact dimensions cannot be determined from SEM data; they are estimated to be in the range of 50–100 nm. Macropores are predominantly concentrated in the central cross-sectional region.

By contrast, the near-surface region (lower right corner, right panel, Figure S1, Supplementary Materials) contains almost no visible macropores, confirming the presence of a compact outer shell enclosing a porous interior (see also Figure S2, Supplementary Materials). The smooth exterior suggests a high degree of surface carbonization, whereas the porous core likely originates from gas evolution and thermal decomposition during pyrolysis. This combination of voids and shell–core heterogeneity reflects non-uniform heat distribution within the fiber during thermal treatment.

SEM imaging revealed that amination, following bromination, did not alter the smooth CFPAN surface, possibly because the modification process is gentle and does not damage the CF structure or create observable defects or surface erosion. Figures S3 and S4 (see Supplementary Materials) show the distribution of carbon (C), nitrogen (N), and oxygen (O) in the original CFPAN and the typical aminated CFPAN/Br₂/MEA samples. At the macro level, both samples exhibit a uniform distribution of C, N, and O on the outer surface of the CFs, regardless of modification. The inner surface of the CFs is also significantly colored, indicating their involvement in the amination process. Notably, the aminated CFPAN sample shows stronger coloration for N and O than the original CFPAN sample. This finding is consistent with the SEM-EDS elemental composition results presented in Table 1. Closer examination of the images suggests that surface modification by N and O atoms occurs in an island-like manner. Disrupting the CFs’ ideal “aromatic” structure likely increases their reactivity by involving substantial surface regions (≥100 nm) in the modification process. Overall, the distribution of N and O is heterogeneous at the micro level but homogeneous at the macro level.

The SEM-EDS data presented in Table 1 demonstrate the successful surface functionalization of CFPAN through bromination and subsequent amination, as evidenced by clear changes in elemental composition and textural parameters. Upon modification, the carbon content decreases from 91.1 at% to 84.0–87.1 at%, while nitrogen content increases significantly, reaching up to 10.1 at% in CFPAN/Br₂/En, confirming effective amine grafting. Oxygen content also rises in CFPAN/Br₂/En, indicating the introduction of oxygen-containing groups such as amide and hydroxyl groups. This increase originates from the bromination step, which was carried out in open air. During this stage, surface oxidation occurs in parallel with bromination, leading to the incorporation of additional oxygen functionalities. Such oxidative side-reactions are intrinsic to liquid-phase bromination under atmospheric conditions and are difficult to avoid completely. It should also be noted that although pyrolysis typically reduces the oxygen content of PAN-derived CFs by driving off oxygen as CO and CO₂, the subsequent bromination step reintroduces oxygen-containing species at the surface. Trace amounts of bromine (0.1–0.4 at%) across all modified samples confirm its role as an intermediate, with most Br atoms replaced by amines, while sulfur is detected only in CFPAN/Br₂/SuEn due to the nature of the modifying agent. Concurrently, as compared to [63], the specific surface area determined by the Brunauer–Emmett–Teller (BET) equation (S_BET) and the total pore volume (V_tot) decrease moderately after modification, reflecting partial pore blockage or surface coverage by the newly introduced functional groups [64]. Despite these reductions, the aminated samples retain substantial porosity, with CFPAN/Br₂/En exhibiting the highest nitrogen content alongside a well-preserved texture. Overall, these findings confirm the chemical modification of the carbon fiber surface and highlight the tunability of surface chemistry and porosity, which are essential for tailoring material performance in adsorption, catalysis, or EM applications. Preliminary bromination of CFPAN is presumed to allow a more selective modification of the carbon surface with amino groups compared to the procedure described in [65].

Table 1. SEM-EDS elemental composition and N₂ adsorption-derived specific surface area (S_BET) and total pore volume (V_tot) parameters of studied samples.

Sample	Elemental Composition, at%				Specific Surface Area and Total Pore Volume Parameters
	C	N	O	Br	SBET, m2 g−1	Vtot, cm3 g−1
CFPAN	91.1	5.3	3.6	–	521	0.249
CFPAN/Br2/Et2N	87.1	6.1	6.4	0.4	479	0.232
CFPAN/Br2/SuEn a	84.6	8.5	6.9	0.3	424	0.205
CFPAN/Br2/MEA	84.3	7.8	7.5	0.4	451	0.219
CFPAN/Br2/PIP	85.2	9.5	5.2	0.1	438	0.213
CfPAN/Br2/En	84.0	10.1	5.8	0.1	447	0.218

Note: ^a The CFPAN/Br₂/SuEn sample also contains 0.6 at% S.

Table 1. SEM-EDS elemental composition and N₂ adsorption-derived specific surface area (S_BET) and total pore volume (V_tot) parameters of studied samples.

Sample	Elemental Composition, at%				Specific Surface Area and Total Pore Volume Parameters
	C	N	O	Br	SBET, m2 g−1	Vtot, cm3 g−1
CFPAN	91.1	5.3	3.6	–	521	0.249
CFPAN/Br2/Et2N	87.1	6.1	6.4	0.4	479	0.232
CFPAN/Br2/SuEn a	84.6	8.5	6.9	0.3	424	0.205
CFPAN/Br2/MEA	84.3	7.8	7.5	0.4	451	0.219
CFPAN/Br2/PIP	85.2	9.5	5.2	0.1	438	0.213
CfPAN/Br2/En	84.0	10.1	5.8	0.1	447	0.218

Note: ^a The CFPAN/Br₂/SuEn sample also contains 0.6 at% S.

Further, the unmodified CFPAN and the aminated brominated CFPAN samples were analyzed by FTIR ATR in order to gain insight into their surface chemistry and structural modifications.

3.3. FTIR ATR

Figure 5 presents the FTIR ATR spectra of the original CFPAN and aminated CFPAN samples. The most intense absorption, observed between 1590 and 1490 cm⁻¹ (with a maximum at 1515 cm⁻¹ and a shoulder at 1549 cm⁻¹), corresponds to the skeletal vibrations of aromatic and/or conjugated C=C bonds [66,67]. These bands remain largely unchanged after amination, indicating that the chemical modifications are confined to surface functionalities, while the underlying sp²-hybridized carbon framework of the fibers remains intact.

Absorption in the 1640–1800 cm⁻¹ spectral region, with prominent bands at 1749 (1788) and 1696 cm⁻¹, is attributed to C=O stretching vibrations of functional groups such as anhydride, lactone, carboxyl, and quinone groups [67,68]. Amination leads to a redistribution of intensity within this range, though the overall spectral shape and total absorption remain largely consistent across different aminated CFPAN samples. These observations imply that the amination process contributes minimally to the overall concentration of C=O groups, and that the spectral differences between aminated and original CFPAN are primarily due to the prior bromination step. This bromination induces surface oxidation and the formation of carboxyl, anhydride, and lactone groups.

After amination, several low-intensity absorption bands emerge in the 1450–1290 cm⁻¹ range. These bands are likely associated with vibrations of C–N, C–O, C–C [67,69], and C–H (possibly including S=O) bonds in grafted amine and amide residues, resulting from the reaction between amines and surface carboxyl groups [70]. Broad absorption bands at 1189 and 1145 cm⁻¹ correspond to C–OH stretching in phenolic groups and C–O–C linkages. As previously reported [24], phenolic groups form extensively during bromination of CFPAN. It is also well established that C–N bonds in aliphatic and aromatic amines exhibit medium-intensity absorption in the 1360–1180 cm⁻¹ and 1240–1020 cm⁻¹ regions. Therefore, for the aminated CFPAN, the spectra changes in the 1240–1110 cm⁻¹ region are attributed to grafted amine groups, with C–N bonds contributing to absorption—both directly and through frequency shifts caused by local structural constraints and intermolecular interactions at the fiber surface. The most prominent spectral changes after amination are observed in the 1100–900 cm⁻¹ region. Broad bands at 1021 and 945 cm⁻¹ are associated with C–O bond absorption and O–H/C–O–C stretching vibrations in surface anhydrides and lactones. These changes reflect chemical interactions between carboxyl groups and amines, leading to the formation of amides and salts, and potentially involving lactone ring opening in alkaline amine environments. Additional absorption bands between 1100 and 850 cm⁻¹, observed in samples such as CFPAN/Br₂/En, CFPAN/Br₂/PIP, CFPAN/Br₂/SuEn, and CFPAN/Br₂/MEA, indicate the presence of C–N bonds from primary and secondary amines. However, due to their inherently low intensity, the specific nature of nitrogen-containing surface groups remains difficult to resolve. Notably, the absence of characteristic C–Br stretching bands in the 880–660 cm⁻¹ region confirms the successful substitution of bromine with amino groups. Overall, the spectral evidence suggests that the amination process involves chemisorbed bromine as well as carboxyl, anhydride, and lactone functionalities, which possess reactive carbonyl centers readily attacked by nucleophilic amines. In contrast, phenolic –OH groups are far less reactive toward amines under the applied conditions, and their absorption bands remain essentially unchanged, indicating minimal participation in the chemisorption process.

In fact, phenolic groups are less involved in amine chemisorption compared to carboxyl, anhydride, or lactone groups due to acidity and electrophilicity reasons. Carboxyl, anhydride, and lactone groups are more acidic and electrophilic. They readily undergo nucleophilic substitution or addition reactions with amines, forming amide bonds or ammonium carboxylates. Lactones can even undergo ring-opening reactions in the presence of nucleophilic amines. By contrast, phenolic –OH groups are weak acids, stabilized by resonance in the aromatic ring. Their oxygen is less electrophilic, and the –OH bond is relatively strong. This means that they do not easily react with amines under mild conditions. If considering reaction pathways, amines attack carbonyl carbons in –COOH, –COOR, –C(=O)–O–C– (anhydrides/lactones), where a reactive carbonyl center is available. But phenolic groups lack such a carbonyl carbon; the only possible route would be condensation (–OH substitution), which requires stronger activation or catalysts not present in our system. Finally, from spectroscopic evidence, after amination, the FTIR ATR spectra show clear changes in the C=O stretching region and the C–N/C–O–C regions, consistent with reactions at carboxyl, anhydride, and lactone sites. However, bands assigned to phenolic –OH remain relatively unchanged in position and intensity. This indicates they are preserved rather than consumed during amination.

Further, the unmodified CFPAN and aminated brominated CFPAN samples were analyzed to understand their thermal decomposition and surface properties using TGA and TPD MS methods.

3.4. TGA and TPD MS

According to the TG-DTG data (Figure 6a,b), the total weight loss in the temperature range of 30–850 °C (Δm) for the original CFPAN/Br₂ sample is about 4.84%. Amination of CFPAN by the proposed route leads to an increase in Δm by 1.7–4.3 times, and this increase occurs due to weight loss processes at high (T > 200 °C) temperatures (

Table 2. TGA data of pristine and aminated CFPAN: total weight loss (Δm), weight loss effect of water (Δm (H₂O)), weight loss effect (Δm₂), temperature range of second effect (ΔT₂), the maximum temperature at the second effect peak (T_2,max), and concentration of N-containing groups (C_N) and TPD MS temperature at the peak maximum of m/z 30 signal (T_max).

Sample	TGA				Temperature, °C
	Weight Loss, g/g			CN, mmol g−1	TGA		TPD MS
	Δm	Δm (H2O)	Δm2		ΔT2	T2,max	Tmax
CFPAN	0.0484	0.012	0.0058	–	162–412	–	–
CFPAN/Br2/Et2N	0.1136	0.033	0.0462	0.55	203–426	286	314
CFPAN/Br2/SuEn	0.2084	0.039	0.1069	0.57	154–394	232	343
CFPAN/Br2/MEA	0.0884	0.016	0.0411	0.58	171–420	295	328
CFPAN/Br2/PIP	0.0980	0.009	0.0603	0.63	146–428	262	307
CFPAN/Br2/En	0.1052	0.019	0.0463	0.67	156–411	237	347

). Three temperature intervals can be distinguished from the analysis of DTG curves (Figure 6b).

The first effect of weight loss corresponds to the desorption of physisorbed water; this process is usually completed at 150–200 °C. The magnitude of this effect (Δm(H₂O)) for the aminated samples is larger compared to the original CFPAN sample. Since water is better adsorbed on the polar surface, it can be concluded that the surface of the aminated CFPAN samples is somewhat hydrophilized compared to the unmodified CFPAN.

The second weight loss event, Δm₂, is observed at moderate temperatures (ΔT₂ = 146–428 °C, Table 2) and varies depending on the type of amine used. This effect is minimal for the original CFPAN (0.0058 g/g) but increases significantly in aminated CFPAN samples. The temperature range ΔT₂ differs slightly among the studied samples due to variations in the amine residues nature, concentration, and ratio of amino and other surface functional groups that undergo transformation within this temperature range. In the 150–430 °C temperature range, carboxyl, amine, and amide groups decompose, while amide groups may also form through reactions between carboxyl groups (such as anhydrides or lactones) and amines. Despite differences in molecular weight and boiling points for used amines, all aminated CFPAN samples exhibit thermal decomposition within this same temperature range of 150–430 °C, indicating that the amines are chemisorbed onto the CFPAN surface. The broad, asymmetric peaks observed in the DTG curves suggest that the amines are grafted onto various carbon surface sites or that the modified surface layer decomposes in a stepwise manner with temperature rise. Elemental analysis of aminated samples preheated in argon at 450 °C reveals nitrogen contents comparable to those of the original CFPAN, supporting the suitability of this temperature range for estimating amino group content. Since bromination generates few carboxyl groups, the amino group content C_N was calculated based on the difference in Δm₂ between aminated and unmodified CFPAN samples.

Table 2 shows that the aminated CFPAN samples can be ranked based on the estimation of strength of the bonding between amine residues and the CF surface in the following order:

CFPAN/Br₂/En > CFPAN/Br₂/PIP > CFPAN/Br₂/MEA > CFPAN/Br₂/SuEn > CFPAN/Br₂/Et₂N.

This ranking is an indirect estimate based on thermal stability: the stronger the amine residue is bound to the carbon surface, the higher the temperature (and activation energy) required for decomposition of the grafted amine residues or desorption of their decomposition products. Thus, higher decomposition/desorption temperatures or larger activation energies correspond to stronger binding.

This order indicates that the proposed amination method is highly effective, enabling the introduction of up to 0.7 mmol g⁻¹ of amino groups into the surface layer of the carbon fibers. Notably, the efficiency of amination appears to be independent of the molecular weight or other physical properties of the amine, supporting the conclusion that the process involves chemisorption of the amines onto the CFPAN surface. The thermal stability of amino groups as chemisorbed N-containing species is higher as compared to that reported in [71].

To gain a deeper understanding of the surface layer structure of the aminated CFPAN/Br₂ samples, the thermal decomposition process was investigated using the TPD MS method. As an example, Figure 7 shows selected regions of the TPD MS spectra for the CFPAN/Br₂/MEA sample.

The positive ions with mass-to-charge ratios (m/z) of 18 (H₂O⁺), 28 (CO⁺, C₂H₄^•+), and 44 (CO₂⁺, CH₂CH₂N⁺H₂) show the highest intensities in the MS spectra of aminated CFPAN/Br₂ samples (Figure 7a). These gaseous products are formed by the decomposition of surface oxygen-containing groups during vacuum heating and are typical of carbonaceous materials [72]. The release of water was observed over a wide temperature range, indicating an intense dehydration process occurring over the carbon surface, especially between 100 and 400 °C. The release of CO₂ gas in the range of 150–400 °C corresponds to the thermal decomposition of carboxyl groups [73]. In addition, a significant release of CO gas is observed at elevated temperatures, which is attributed to the thermal decomposition of phenolic groups [74].

However, the peak temperatures (T_max) observed in the TPD MS profiles at m/z 18, 28, and 44 and corresponding to H₂O, CO, and CO₂ gassing under vacuum differ significantly within the temperature range of 100–400 °C (Figure 7a). A remarkable difference is also observed between the T_max values of H₂O⁺ (m/z 18) and OH⁺ (m/z 17) ions. These variations suggest the occurrence of several parallel thermal decomposition processes involving both the thermal decomposition of oxygen-containing functional groups and amino groups. In particular, the thermal decomposition of amino groups (originating from MEA () residues) leads to the formation of C₂H₄^•+ (m/z 28) and CH₂CH₂N⁺H₂ (m/z 44) ions, whose TPD MS profiles overlap with those of CO (m/z 28) and CO₂ (m/z 44), respectively (Figure 7a). In the 100–400 °C temperature range, the MS spectra also revealed the fragmentation, and several lower-intensity fragments were detected, including ions with m/z 27 (C₂H₃⁺), 30 (CH₂=N⁺H₂), 31 (CH₂=O⁺H), 43 (CHCH₂N⁺H₂), and 60 (HOCH₂CH₂N⁺H), with the latter being an analog of a molecular ion (Figure 7b). These fragments are products of vacuum pyrolysis of amino groups [68], and their temperature-resolved evolution provides insights into the decomposition stages of the surface layer. The molecular ion analog (m/z 60) showed two distinct temperature profile peaks with temperature maxima at 157 °C and 286 °C in the TPD profile, confirming the chemisorption of this amine. These temperatures, exceeding the normal boiling point of MEA (170 °C), indicate the participation of multiple active sites on the CFPAN surface in the amination process. In parallel with the presence of these larger MEA-derived fragments in the MS spectra, a partial thermal decomposition of amino groups attached to the carbon surface was also observed. This is evidenced by the appearance of ions such as OH⁺ (m/z 17) with a peak maximum at 299 °C, CH₂=N⁺H₂ (m/z 30) with a peak maximum at 328 °C, CH₂=O⁺H (m/z 31) with a peak maximum at 275 °C, and the fragment ions of C₂H₃⁺ and CHCH₂N⁺H₂ with a peak maximum at 309 °C in the MS spectra. Among these positive ions, the nitrogen-containing fragment ions observed by MS at high temperatures showed the highest decomposition temperatures, underscoring the good thermal stability of the grafted amino groups and crediting their anchoring to the surface via the nitrogen atom. The temperature corresponding to the profile maximum of the CH₂=N⁺H₂ ion (m/z 30), which are characteristic of all aminated CFPAN samples, is given in Table 2. These decomposition patterns were consistent for all aminated CFPAN samples, regardless of the amine used, confirming the substantial and stable chemisorption of amines onto the CFPAN surface. Combined thermogravimetric and TPD MS data indicate that, after the amination reaction, nitrogen-containing species are covalently bound to the carbon surface rather than merely physisorbed, as the decomposition and vacuum thermodesorption temperatures of amine-derived fragments exceed the boiling points of the free amines, providing strong evidence of chemisorption.

From TGA data in an inert atmosphere, the weight loss in the range of 140–430 °C (precisely) is mainly because of the thermal decomposition of chemisorbed amines, nitrogen-containing surface functionalities residues of the chemical reaction of selected amines with surface groups of CFs (grafted –NH₂/–NH– and their reaction products, chiefly amides), along with concurrent breakdown of accompanying carbon-oxygen containing groups.

This decomposition stimulates further releases of small molecules (e.g., they can be presumably NH₃, small organics, H₂O, CO, and CO₂ in the thermal decomposition process in TGA experiments, and amine fragments observed in TPD MS spectra), causing the increased Δm₂ events observed for the aminated brominated CFPAN versus the pristine CFPAN fabric. The common temperature window—despite different amine boiling points—supports a chemisorption/grafting rather than physisorption process.

From comparative analysis of TGA and TPD MS results, as in the first temperature range of the TG weight loss event, water is desorbed from all samples; in the second one, thermal decomposition of amino groups occurs. (This range can tail, and at maximum, it can be 140–465 °C). The amino groups decomposition caused future release of adsorbed decomposition products, the weight of the sample decreases accordingly, and on the DTG curves, we see maxima corresponding to a certain weight loss. Knowing which oxygen-containing groups are desorbed at this time and their quantity, one can determine the concentration of grafted amines by the difference.

The identification of the decomposed species as amino groups, as well as the detection of their decomposition products within this temperature range under vacuum conditions, was confirmed by the MS profiles and the fragment ions observed in the TPD MS experiments.

3.5. Microwave Properties

Figure 8a–f shows the frequency dependencies of the microwave properties (scattering parameters and absorption coefficient) for both the initial (non-modified) CFPAN and En-aminated sample of CFPAN in the whole X-band. Note that each S-parameter magnitude is given in decibels, whereas the dimensionless absorption coefficient is given in arbitrary units.

The presented data demonstrate strongly anisotropic behavior for both compositions; namely, with an increase in angle φ between carbon fibers and E-field polarization, the reflection coefficient R decreases and the transmission coefficient T increases. This means that less EM energy is reflected, and more energy is transmitted. However, the transmission coefficient significantly depends on the impedance matching between the free space and considered material, and therefore it does not fully represent the absorption properties of the carbon fiber. Indeed, the transmission coefficient T represents the total loss of EM wave energy upon interaction with a material, including both reflection and absorption contributions. Reflection losses are mainly governed by impedance matching, whereas absorption losses depend on the imaginary components of the material’s complex permittivity and permeability. Because the transmission coefficient T alone does not allow one to distinguish between these two contributions, it is not possible to compare the intrinsic absorption of different materials based solely on the results of transmission measurements.

A more appropriate characteristic is the absorption coefficient A defined as in Equation (3). It describes the actual losses inside the sample and thus allows the different materials to be compared without concerns for their different impedance matching.

The analysis of absorption dependencies (see Figure 8c,f) confirms the pronounced anisotropic character of EM energy absorption. Absorption coefficient A varies with φ by almost 30% for CFPAN and by almost 60% for CFPAN/Br₂/En (

Table 3. Average X-band microwave absorption coefficient A, angle at A measurements φ, and anisotropy coefficient K in the aminated CFPAN at different angles of sample orientation. Values in parentheses indicate the sample average thickness (d).

φ, °	Average X-Band Microwave Absorption Coefficient, A
	CFPAN (340 μm)	X in Aminated CFPAN/Br2/X
		SuEn (370 μm)	PIP (350 μm)	Et2N (350 μm)	MEA (400 μm)	En (350 μm)
0	0.32 ± 0.02	0.21 ± 0.04	0.23 ± 0.05	0.29 ± 0.02	0.27 ± 0.02	0.32 ± 0.02
15	0.33 ± 0.02	0.24 ± 0.04	0.24 ± 0.05	0.31 ± 0.02	0.28 ± 0.02	0.35 ± 0.02
30	0.35 ± 0.01	0.27 ± 0.03	0.27 ± 0.04	0.33 ± 0.02	0.30 ± 0.02	0.42 ± 0.02
45	0.36 ± 0.01	0.30 ± 0.03	0.30 ± 0.04	0.33 ± 0.02	0.34 ± 0.02	0.49 ± 0.01
60	0.38 ± 0.01	0.35 ± 0.04	0.32 ± 0.03	0.34 ± 0.02	0.35 ± 0.02	0.49 ± 0.01
75	0.41 ± 0.01	0.32 ± 0.03	0.34 ± 0.03	0.37 ± 0.02	0.39 ± 0.02	0.49 ± 0.01
90	0.41 ± 0.01	0.32 ± 0.03	0.36 ± 0.03	0.38 ± 0.02	0.40 ± 0.02	0.51 ± 0.01
90/0	Anisotropy Coefficient, K
	1.28	1.52	1.57	1.31	1.48	1.59

). Moreover, it was found that whereas for φ = 0° the absorption is almost equal (A ≈ 0.3) for both compositions, for large values of φ = 90° it is noticeably different (A ≈ 0.4 for CFPAN and A ≈ 0.5 for En). Thus, in this case, the amination of the initial CFs increases the energy absorption only for the specific microwave E-field polarization while leaving it approximately the same for the other polarizations. In general, the comparison between aminated and non-aminated CFs shows that aminated CFs demonstrate similar characteristics for small φ values, but for large φ it has lower reflection coefficient and larger absorption losses for each orientation angle φ from this range.

The transmission and reflection frequency dependence data for the CFPAN/Br₂/SuEn, CFPAN/Br₂/PIP, CFPAN/Br₂/Et₂N and CFPAN/Br₂/MEA are given in Figure 9. The pronounced anisotropic behavior (just as on Figure 8) was found for all of the compositions. Similarly, the increase in angle φ between CFs and E-field polarization resulted in a decreased reflection coefficient and increased transmission coefficient. However, the magnitude of the angle-dependent variation in microwave parameters in all cases was less than for the En-aminated CFPAN and comparable to the pure CFPAN sample.

As shown in Figure 9, the transmission is relatively low (1–10%), indicating that the incident EM energy is primarily distributed between reflection and absorption. A closer inspection reveals that the absorption contribution does not exceed 50%, while reflection consistently remains above this value (corresponding to >−3 dB in log scale). Thus, the studied materials demonstrate a nearly balanced EM response with a slight predominance of reflection. Although this differs from absorption-dominant systems recently reported for other material classes [75,76], such analytical approaches provide useful perspectives that could be adapted for a more detailed evaluation of absorption mechanisms in future works.

The results of absorption coefficient calculations for other amine-modified CFPAN are given in Figure 10 and Table 3. Since in almost all cases the frequency spectrum of A only weakly varies within the 8–12 GHz frequency range, a frequency-averaged value of A was introduced and is presented in the table. Note that the error margins in table cells represent not experimental errors but the standard deviation of data calculated during the statistical averaging.

The same qualitative tendency for anisotropic absorption was observed for all other modified fibers. However, the variation in absorption level with respect to the reference material is different, dependent on the actual chemical modification agent. Thus, CFPAN/Br₂/SuEn and CFPAN/Br₂/PIP samples showed lower absorption than the initial CFPAN sample did, whereas CFPAN/Br₂/Et₂N and CFPAN/Br₂/MEA samples were almost at the same level (within the experimental error and frequency variation), and only the CFPAN/Br₂/En sample showed a noticeable increase in absorption coefficient. To characterize the magnitude of anisotropy, we introduced the anisotropy coefficient:

$$K={\displaystyle \raisebox{1ex}{$A_{90}$}\!\left/ \!\raisebox{-1ex}{$A_0$}\right.},$$

(6)

where A₉₀ is the absorption at φ = 90° and A₀ is the absorption at φ = 0°. The anisotropy coefficient K shows the ratio of the lowest and highest frequency-averaged absorption coefficients (see Table 3). From this data it is seen that the K values for all compositions do not differ much and fall within 1.3–1.6 range with the largest anisotropy taking place in the CFPAN/Br₂/En sample and the smallest anisotropy being measured for the starting unmodified CFPAN.

Thus, the analysis presented above shows that the microwave dielectric losses in investigated carbon-fiber-based materials are indeed anisotropic and absorption strongly depends on the polarization of the incident EM wave. All samples with amination have demonstrated the improved anisotropy properties in comparison with the initial CFPAN (K values have increased); however, the absolute values of absorption coefficients changed ambiguously: for some samples, A values decreased and, for some, increased (with respect to those of CFPAN), leading to the conclusion that only specific amino groups are suitable for the radio-shielding materials with the enhanced microwave absorption.

3.6. Attenuation Characteristics and Anisotropic Behavior in Microwave Absorbing Aminated CFPAN

The absorption coefficient, as a practical metric for evaluating radio-absorbing materials, is inherently dependent on external experimental parameters such as sample thickness and impedance matching with the incident EM wave. Variations in measurement configurations—whether in coaxial, co-planar waveguides, or free space—introduce differences in absorption due to altered matching conditions between the sample and its surrounding medium. Consequently, to facilitate a reliable comparison across samples with differing thicknesses and measurement setups, an intrinsic material property, one should consider the attenuation constant α, which is the real part of the complex propagation constant. This parameter quantifies EM wave attenuation in a uniform medium, following the Beer–Bouguer–Lambert extinction law:

$$P\left(d\right)=P(0)e^{-\alpha d}.$$

(7)

In our case, P(0) is the power of the EM wave at the front air–sample interface (see Figure 2) and P(d) is the power of EM wave at the back interface. Considering the reflection on sample’s interfaces R, we get

$$P\left(0\right)=P_I\left(1-R\right)$$

(8)

and

$$P(d)={\displaystyle \frac{P_T}{(1-R)}},$$

(9)

therefore,

$$\alpha =-{\displaystyle \frac{1}{d}}ln\left({\displaystyle \frac{T}{{(1-R)}^2}}\right)$$

(10)

where d is the average sample thickness, with numerical values given in Table 3. Equation (10) is approximate since it accounts only for a single pass of the EM wave through the material, and the contribution from multiple reflections is ignored. However, for radio-absorbing material with large values of absorption A (or attenuation α), this model has to be quite adequate since the effect of multiple reflections may be reasonably neglected.

The calculated frequency dependencies of the attenuation constant in the X-band for all investigated samples are presented in Figure 11.

The lowest attenuation was found for the unmodified CFPAN and Et₂N-aminated CFPAN. Other compositions demonstrate a larger attenuation with SuEn and PIP, showing the most dissipating proprieties among them. As for the anisotropy, CFPAN and MEA-aminated CFPAN showed negligible variations with E-field orientation angle φ; for PIP, SuEn, and Et₂N, the anisotropy is more expressed; however, the angular behavior of α is irregular, and only in the case of En-aminated CFPAN was a pronounced anisotropic behavior registered with the attenuation constant growing from φ = 0° to φ = 45° and then decreasing for φ from 45° to 90°. This data led to the conclusion that the anisotropic radio-absorptive behavior of the investigated composites (as in Figure 8) is mostly determined by the angle-dependent variation in the sample’s characteristic impedance (which in turn affects the matching conditions and the reflection R of EM at interface) and not by anisotropic internal losses.

However, in the specific case of En-aminated carbon fiber, the effect may be due to the combined action of both anisotropic impedance and anisotropic attenuation. The physical background for the suggested anisotropic impedance phenomenon is yet to be clarified. Regardless of that, the data in Figure 11 represents the intrinsic parameter of the respective materials and therefore can be used for the meaningful comparison with published results for materials with similar compositions, and also in theoretical calculations for the properties of various radio-absorption and radio-shielding coatings that can be designed on the basis of materials investigated in this work.

Additional key findings on the attenuation constant α and its anisotropy are summarized below. The attenuation constant α plays a critical role in the evaluation of microwave absorption in materials, particularly carbon-based composites. Unlike the absorption coefficient A, which can vary with sample thickness and impedance matching conditions, α offers a more intrinsic measure of a material’s microwave absorption capabilities. Attenuation constant α is a complex function of both the material’s permittivity (ε) and permeability (μ), and thus it is strongly related to the material’s chemistry via the dependence of ε and μ on the sample’s chemical composition. Understanding α and its anisotropic behavior is essential for optimizing the design and application of the CFPAN-based materials in various technological fields. The behavior of the attenuation constant α is governed by the Beer-Bouguer-Lambert law, which describes the exponential decrease in the power of EM waves as they propagate through a medium. For highly absorbing materials, the assumption of neglecting multiple reflections is justified, allowing for the use of approximations as indicated in Equation (10).

Experimental findings on the attenuation constant α are as follows: The α values vary among the samples. The lowest values of attenuation α were observed in the case of unmodified CFPAN and CFPAN/Br₂/Et₂N, indicating a reduced capacity for microwave absorption. In contrast, the highest attenuation constant α was recorded for the CFPAN/Br₂/SuEn and CFPAN/Br₂/PIP samples, suggesting that the surface amination with SuEn and PIP enhances energy dissipation mechanisms. The anisotropy of α exhibits variability across different carbon-based samples. The different amination results in different complex values of ε and μ, which in turn affect all microwave properties of the sample, including attenuation, reflection, and absorption. This influences both the average values of these characteristics and their variation with sample orientation (anisotropy). Therefore, amination does affect anisotropy indirectly, through modification of the material’s physical parameters, which themselves are anisotropic. The CFPAN and CFPAN/Br₂/MEA samples show negligible dependence on the electric field orientation angle φ. Aminated CFPAN samples, including CFPAN/Br₂/PIP, CFPAN/Br₂/SuEn, and CFPAN/Br₂/Et₂N, show an irregular φ angular dependence for the attenuation constant α. Notably, the CFPAN/Br₂/En sample demonstrates a clear anisotropic trend, where α increases from φ = 0° to 45° and subsequently decreases from 45° to 90°.

The observed anisotropic behavior of absorption A can be interpreted through several mechanisms: (i) Angle-dependent impedance variations. For most studied samples, the anisotropy in radio wave absorption primarily arises from angle-dependent changes in the characteristic impedance, which significantly affects the reflection coefficient R at the material interface; (ii) Intrinsic attenuation anisotropy. In the case of CFPAN/Br₂/En, both anisotropic impedance characteristics and intrinsic attenuation anisotropy may contribute to the observed behavior. Despite these findings, the underlying physical mechanisms that govern the anisotropic impedance variations remain inadequately understood. Further comprehensive studies are essential, including accurate broadband EM characterization—through precise frequency-dependent measurements of complex permittivity (ε′, ε″) and permeability (μ′, μ″) over a wide microwave range using vector network analyzers with suitable fixtures (coaxial line, waveguide, free-space, etc.)—together with computational EM simulations, such as finite-difference time-domain, finite element method, and commercial solvers like CST Microwave Studio or Ansys High-Frequency Structure Simulator, to elucidate these complex interactions, correlate microstructural features with macroscopic EM behavior, and predict shielding and absorption performance in advanced carbon-based materials.

4. Discussion

The most representative is the chemical reduction of GO to form reduced graphene oxide (RGO), often via strong or mild reducing agents with vacuum or heat treatment to restore conductivity, reduce defects, and increase graphitic domains. The latter causes greatly improved conductivity, which increases shielding effectiveness, and provides good shielding efficiency in the X-band with lower mass loadings [77]. For example, treating carbon fibers with H₂ plasma to convert C–O–H to C–O–C groups enhances surface carbon content, perhaps reducing defects or adjusting work function and conductivity. Here, plasma/chemical treatment led to changes in the types of surface functional groups that improved electrical conductivity and total shielding (SE > 45 dB over a certain frequency range) due to better surface electron availability and possibly enhanced interfacial losses [78]. Numerous studies exist on EM shielding, most of which focus on composites where carbon materials are combined with other fillers. In contrast, there are relatively few reports that specifically address the chemical modification of carbon materials themselves. Chemical modification of carbon materials has been explored as an effective route to enhance their EM shielding performance. A representative example is the chemical reduction of graphene oxide (GO) to produce reduced graphene oxide (RGO), typically achieved through chemical reducing agents, thermal annealing, or vacuum treatment. Such reduction processes restore graphitic domains, remove oxygenated defects, and significantly improve conductivity, resulting in enhanced shielding efficiency even at low filler loadings [77].

Another strategy involves plasma or chemical surface treatments of carbon fibers. For instance, H₂ plasma treatment can convert surface C–O–H functionalities into C–O–C groups, increase the surface carbon content, and adjust the electronic structure. This modification improves electrical conductivity and facilitates interfacial polarization losses, leading to higher total shielding effectiveness (SE > 45 dB over a broad frequency range [78]). Heteroatom doping represents another powerful approach. Nitrogen doping of graphene or carbon nanotubes introduces electron-rich sites, modifies carrier density, and creates defect dipoles, which not only increase conductivity but also promote dielectric losses. For example, N-doped graphene frameworks have been shown to reach high absorption-dominant shielding efficiencies due to synergistic conduction and polarization contributions [79,80]. Taken together, these—although relatively rare—examples demonstrate that targeted chemical modifications such as reduction, plasma/chemical treatment, and heteroatom doping can restore or enhance conductivity, tailor interfacial interactions, and introduce additional loss mechanisms, thereby substantially improving the shielding performance of carbon-based systems.

The results presented above contribute to comprehensive understanding of the structural, thermal, chemical, and EM properties of heteroatom-doped (aminated) CFPAN samples and provide a valuable extension to the synthesis, absorption behavior, and findings previously reported in [81,82]. Our present findings highlight the effectiveness of a two-step surface functionalization approach—bromination followed by amination—which not only tailors the structural, chemical, and thermal characteristics of CFPAN but also has a pronounced impact on its microwave absorption performance.

HRTEM analysis revealed the characteristic turbostratic carbon nanostructure of CFPAN, comprising misaligned, curved graphitic layers with interlayer spacings larger than that of ideal graphite. This structural disorder originates from the pyrolysis process: although oxygen is largely removed during carbonization, residual nitrogen- and oxygen-containing groups remain and, together with pyrolytic strain, disrupt long-range graphitic ordering. The presence of curved, layered carbon domains result in slit-like nanopores (0.5–2 nm wide), which are beneficial for increasing the specific surface area and potentially enhancing adsorption [63] and interfacial polarization—key factors in EM wave attenuation. The SAED patterns confirmed the nanocrystalline nature of CFPAN, with broad diffraction rings corresponding to (002), (100), and (110) planes. A shift and broadening of the (002) ring indicate defect-induced interlayer expansion and partial oxidation. These structural features, including the turbostratic carbon domains, are common in disordered carbons such as AC and biochar [83,84,85], so some correlation between structural features and enhanced microwave absorption have been reported, highlighting that misaligned graphitic layers, enlarged interlayer spacings, and structural defects can strongly facilitate interfacial polarization, multiple scattering, and dipolar relaxation, which are crucial for enhancing dielectric losses during EM absorption.

Surface amination, confirmed through SEM-EDS and FTIR ATR analyses, introduced nitrogen-containing groups onto the CF surface. SEM observations suggest that some surface irregularities were reduced, leading to a somewhat smoother fiber exterior, although the overall morphology of the CFs remained largely unchanged as in Ref. [69]. Elemental analysis showed increased nitrogen and oxygen content post-amination, with nitrogen levels reaching up to 10.1 at% for the CFPAN/Br₂/En sample. This confirms the successful grafting of amine functionalities, likely through reactions with surface carboxyl, lactone, and anhydride groups generated during bromination. TGA and TPD MS supported the chemical evidence of functionalization by revealing increased mass loss at elevated temperatures (150–450 °C) in aminated samples—attributable to decomposition of amino and oxygen-containing groups [86]. The TPD-MS spectra revealed consistent evolution of CO, CO₂, and amine-derived fragments (e.g., CH₂=N⁺H₂, CH₂CH₂NH₂⁺), providing clear evidence for the covalent attachment of amine groups to the carbon surface. Notably, the decomposition temperatures of the amine-related fragments were significantly higher than the boiling points of the corresponding free amines, indicating strong chemisorption rather than mere physisorption.

Among the samples, CFPAN/Br₂/En exhibited the highest nitrogen loading (0.67 mmol g⁻¹) and the most thermally stable amine functionalities, indicating an efficient and robust surface modification. Microwave absorption behavior and anisotropy studies revealed a strong correlation between surface functionalization and microwave absorption behavior.

All aminated brominated CFPAN samples exhibited increased dielectric losses (as indicated by higher attenuation constants α) relative to the pristine CFPAN, confirming enhanced internal energy dissipation. However, their absorption coefficient A—the more application-relevant metric—varied depending on the amine used. These differences can be rationalized by the chemical modifications introduced via surface amination: the grafted nitrogen-containing groups alter the complex permittivity (ε′, ε″) and permeability (μ′, μ″) of the fibers, thereby enhancing dielectric losses and modifying impedance matching with free space. As a result, energy dissipation is improved, particularly for CFPAN/Br₂/En, which exhibits the most favorable balance of ε and μ, leading to superior microwave absorption (especially at φ = 90°) and reduced reflection. This implies superior impedance matching and effective attenuation of incident EM radiation. The anisotropic nature of microwave absorption was also evident, with A varying by up to 60% depending on fiber orientation φ. This anisotropy stems primarily from orientation-dependent changes in impedance matching between the sample and the incident wave, while the contribution from attenuation anisotropy is minor. Notably, only in the case of CFPAN/Br₂/En was this angular dependence in attenuation well defined, suggesting a synergistic effect of chemical modification and orientation-dependent EM response.

The findings align with previous reports on nitrogen-doped carbon materials, which often exhibit enhanced dielectric polarization, thermal stability, and surface reactivity. However, the anisotropic behavior and tunability achieved via selective amination of brominated CFPAN present a novel approach. Unlike metal-filled or hybrid composites, the system here remains entirely carbon-based and leverages controlled surface chemistry to tailor EM response—providing a lighter, corrosion-resistant, and flexible solution for EM shielding. Moreover, the method demonstrates versatility in tuning both porosity and functional group content while maintaining structural integrity. This offers an advantage over many traditional conductive polymer composites or metal oxides, which can suffer from mechanical brittleness or agglomeration issues.

Table 4. Summarized microwave absorption performance, in the minimum reflection loss (RL_min = −20 log|S₁₁|) in the selected frequency range, of carbon-based materials. Abbreviations: rGO—reduced graphene oxide, Gr—graphene, NS—nanosheets, MWCNTs—multi-walled carbon nanotubes, MWCNTs–OH—oxidized multi-walled carbon nanotubes, CNT oxide—oxidized carbon nanotubes, PEG—polyethylene glycol, carbon nanoparticles—CNPs, PCNS—porous carbon NS, CMS—carbon microspheres, GO—graphene oxide, hex-BN—hexagonal boron nitride; Matrix: PN—paraffin, SR—silicone rubber, EP—epoxy, and NBR—nitrile butadiene rubber.

Filler	Matrix	Filler Content, wt%(at%)	Frequency Band, GHz	RLmin, dB	d, mm	Ref.
3D-rGO	PN	1	2–18	−24.9	5.0	[87]
Gr microflowers	PN	10	2–18	−30.8	2.0	[88]
Hollow Gr aerogel spheres	PN	5	2–18	−52.7	2.3	[89]
rGO NS	SR	2	2–18	−37.8	2.5	[90]
Holey Gr NS	SR	1	2–18	−32.1	2.0	[91]
Layered and porous rGO	NBR	10	2–18	−57.0	3.0	[92]
rGO	EP	1	2–18	−47.2	3.0	[93]
MWCNTs	EP	8	1–26.5	−20.0	3.0	[94]
MWCNTs–OH	EP	33.3	8–18	−20.0	3.0	[95]
PAN CFs	PN	33	2–18	−23.0	3.0	[96]
PAN CFs	EP	5	2–18	−14.0	2.5	[97]
Mesoporous hollow CMS	PN	50	2–18	−50.9	1.7	[98]
Yolk-shell C@CMS	PN	50	2–18	−39.4	1.5	[99]
Hollow porous CNPs	PN	2	2–18	−18.1	1.6	[100]
Intrinsic defects-rich PCNS	PN	5	2–18	−56.8	3.4	[101]
2D graphene-like PCNS	PN	5	2–18	−56.5	3.4	[102]
Biomass-derived rGO	PN	15	2–18	−27.5	1.5	[103]
Mesoporous hollow CMS	PN	25	2–18	−34.6	3.2	[104]
Wrinkled rGO-wrapped polymer-derived CMS	PN	25	2–18	−55.2	1.9	[105]
Reduced Gr/CNT oxide	EP	1	2–18	−61.8	1.9	[106]
Polyaniline/Gr sheet	PN	5	2–18	−51.8	2.1	[107]
Polypyrrole/GO	PN	30	2–18	−58.1	2.96	[108]
Gr/PEG composite	PN	50	2–18	−43.2	2.35	[109]
N-doped Gr nanofakes	PN	2	2–18	−21.0	2.0	[110]
rGO/hex-BN	PN	25	2–18	−40.5	1.6	[111]
N/B co-doped rGO	PN	0.05	2–18	−52.0	2.8	[112]
Amino groups	CFPAN	6.1–10.1 (at%)	8–12	From −1.02 to −2.59	0.34–0.40	Our study

summarizes reported data on microwave absorption performance (expressed as minimum reflection loss, RL_min) of a range of carbon-based materials with different matrices and fillers, enabling direct comparison with the present CFPAN samples. In our case, only amino functionalities were introduced, and the fiber thickness was extremely small (0.34 mm). Under these conditions, the obtained RL_min values are consistent and reasonable, reflecting the intrinsic contribution of surface chemical modification without additional fillers.

Most carbon-based microwave absorbers reported in the literature require relatively high filler loadings (typically 5–50 wt%) and millimeter- to centimeter-scale thicknesses to achieve strong reflection loss values (RL_min < −30 dB). By contrast, the present study demonstrates that amino-functionalized CFPAN fibers, even at minimal thickness and without secondary fillers, exhibit measurable absorption in the X-band (RL_min from −1.02 to −2.59 dB). Although the absolute RL values are lower than those of bulk composites, these results highlight the unique advantage of anisotropic PAN-based CFs, where surface functionalization enables a tailored microwave response at very low thickness. This comparison underscores the potential of CFPAN derivatives as lightweight, tunable building blocks for advanced EM shielding applications.

Importantly, these results highlight the intrinsic contribution of surface chemical modification without the need for additional fillers. With an increase in fiber thickness and the introduction of other fillers or modifiers (e.g., magnetic or dielectric additives), a further decrease in reflection and enhancement of absorption can be expected. However, this was beyond the scope of the present work, which specifically aimed to isolate and demonstrate the effect of amino functionalization alone.

In this paradigm, future investigations should focus on clarifying the physical mechanisms underlying the observed impedance anisotropy, particularly in the CFPAN/Br₂/En system. Advanced simulations and impedance spectroscopy could shed light on the orientation-dependent dielectric behavior and its correlation with fiber morphology and surface chemistry. Moreover, extending the functionalization strategy to heteroatom-containing amines or combining CFPAN fibers with magnetic additives—such as ferromagnetic and superparamagnetic metallic nanoparticles (Fe, Co, and Ni), carbonyl iron, or magnetic dielectrics (e.g., ferrimagnetic spinels, garnet ferrites)—could create synergistic effects in achieving both dielectric and magnetic losses, opening opportunities for multi-band EM shielding. Finally, studies on long-term stability, environmental durability, and scalability will be essential to advance these materials toward practical applications.

CFPAN modified with specific amino groups exhibited pronounced anisotropic microwave properties in the X-band frequency range, particularly in terms of the anisotropic absorption coefficient. All aminated samples demonstrated higher internal losses compared to the unmodified material, with CFPAN/Br₂/En showing the most significant enhancement, characterized by both an increased absorption coefficient and reduced reflection—two key parameters for effective EM shielding. Notably, the angular dependence of absorption in En-functionalized CFPAN displayed the greatest variation with respect to the orientation of the EM wave E-field. These findings indicate that anisotropically functionalized PAN-based fibers have strong potential for microwave-absorbing radio-protective coatings, enabling optimized attenuation of incident EM waves through proper material orientation.

Taken together, the results demonstrate a robust strategy for tuning the structural, thermal, and EM properties of CFPAN through selective amination. The resulting materials exhibit enhanced microwave absorption and anisotropic behavior, with CFPAN/Br₂/En emerging as a promising candidate for next-generation EM shielding applications.

5. Conclusions

This study demonstrates that sequential bromination and amination of CFPAN fibers is an effective strategy to modify their surface chemistry and tailor their EM response without compromising structural integrity, while increasing nitrogen content. Bromination not only introduced reactive bromine atoms but also oxygen-containing groups, enabling efficient nucleophilic substitution with amines. TPD MS detects oxygen-containing groups, which decompose at 100–500 °C, releasing carbon oxides. Mass spectra showed amine fragments’ series observed at 230–310 °C, indicating successful functionalization of brominated CFPAN. TGA shows increased weight loss after amination, confirming the incorporation of amine groups, ~0.44–0.56 mmol g⁻¹.

Microwave investigations revealed that amine functionalization significantly altered reflection, absorption, and anisotropy in the X-band, strongly dependent on the type of amine and fiber orientation in the waveguide. Microwave measurements showed that the reflection coefficient varied from −1.0 to −2.5 dB for SuEn-functionalized and from −2.0 to −4.0 dB for En-functionalized samples, depending on frequency and fiber orientation in the waveguide. The frequency-averaged absorption of pure CFPAN was 32–41%, with maximum and minimum values corresponding to orientations differing by 90°. Absorption decreased to 21–35% upon SuEn functionalization but increased to 32–51% for En-functionalized samples. Pure CFPAN exhibited the lowest absorption anisotropy (factor 1.28), while PIP- and En-functionalized samples were the most anisotropic (1.57 and 1.59, respectively). The attenuation constant for all compositions ranged between 1.5 and 4.5 mm⁻¹.

In summary, the incorporation of amino groups imparted pronounced anisotropic microwave absorption in the X-band, governed by both increased dielectric losses and orientation-sensitive impedance. Ethylenediamine functionalization proved most effective, producing the highest nitrogen content and strongest enhancement of absorption while reducing reflection. These findings demonstrate that aminated CFPAN, particularly CFPAN/Br₂/En, is a promising material for advanced EM shielding and radio-protective textiles, where fiber orientation can be optimized to maximize energy dissipation and attenuation of polarized EM fields.