Preparation and Evaluation of CuMnOx-Modified Activated Carbon Fibers for Indoor VOCs Removals

Abstract

This study aimed to develop a high-performance Modified Activated Carbon Fiber (ACF) filter for the effective removal of Volatile Organic Compounds (VOCs) generated in workplaces and for application in indoor VOCmitigation devices. ACF was modified with CuMnOx catalysts and evaluated for the removal of formaldehyde, acetaldehyde, and benzene. The modified ACF filter was prepared by introducing CuMnOx via an impregnation method using Cu(NO₃)₂⋅3H₂O and Mn(NO₃)₂⋅6H₂O precursors, followed by a crucial high-concentration oxygen plasma surface treatment (50 sccm gas flow) to effectively incorporate oxygen functional groups, thereby enhancing catalyst dispersion and activity. Characterization of the fabricated ACF/CuMnOx composite revealed that the optimized sample, now designated ACF-P-0.1 (representing both CuMnOx catalyst impregnation and O₂ plasma treatment), exhibited uniformly dispersed CuMnOx particles (<500 nm) on the ACF surface. This stability retained a high specific surface area (1342.7 m²/g) and micropore ratio (92.23%). H₂-TPR analysis demonstrated low-temperature reduction peaks at 140 °C and 205.8 °C, indicating excellent redox properties that enable high catalytic VOC oxidation near room temperature. The oxygen plasma treatment was found to increase the interfacial reactivity between the catalyst and ACF, contributing to further enhancement of activity. Performance tests confirmed that the ACF-P-0.1 sample provided superior adsorption–oxidation synergy. Benzene removal achieved a peak efficiency of 97.5%, demonstrating optimal interaction with the microporous ACF structure. For formaldehyde, a removal efficiency of 96.6% was achieved within 30 min, significantly faster than that of Raw ACF, highlighting the material’s ability to adsorb VOCs and subsequently oxidize them with high efficiency. These findings suggest that the developed ACF/CuMnOx composite filters can serve as promising materials for VOCs removal in indoor environments such as printing, coating, and conductive film manufacturing processes.

1. Introduction

Volatile organic compounds (VOCs) are emitted from various synthetic materials, construction products, and household goods, and are recognized as one of the major contributors to indoor air quality degradation. Due to their low boiling points and high vapor pressures, VOCs tend to accumulate indoors, and long-term exposure has been associated with headaches, respiratory diseases, central nervous system disorders, and increased carcinogenic risks [1]. In particular, benzene, toluene, and formaldehyde have been classified as Group 1 human carcinogens by the International Agency for Research on Cancer (IARC), underscoring the urgency of developing indoor VOC mitigation technologies from a public health perspective [2].

Traditional approaches for VOC removal include high-temperature thermal decomposition, catalytic oxidation, and wet scrubbing. However, these technologies are typically large-scale, externally installed systems that suffer from high equipment and energy costs as well as wastewater generation, rendering them unsuitable for direct indoor VOC treatment [3,4]. As an alternative, activated carbon fiber (ACF) has emerged as a promising adsorbent material. ACF possesses a high specific surface area (often exceeding 1000 m²/g), a uniform microporous structure, fast adsorption kinetics, and large adsorption capacity [5,6,7]. Moreover, its lightweight and modular form allows for efficient adsorption even under low-temperature conditions [8,9,10]. By tailoring surface functional groups, selective adsorption of specific VOCs can also be achieved [11]. In heating, ventilation, and air-conditioning (HVAC) or portable purification systems, ACF enables low pressure drops and high purification efficiency, thereby offering energy-saving benefits [8].

The efficiency of catalytic VOC oxidation over ACF is highly dependent on the catalyst incorporation method, which significantly influences dispersion and catalyst-support interaction [11]. The impregnation method is widely adopted for ACF due to its simplicity and scalability. Nitrate salts, such as Cu(NO₃)₂ and Mn(NO₃)₂, are preferred precursors as they readily convert into highly active CuMnOx oxides upon calcination [11]. Therefore, optimizing the catalyst loading is essential to achieve the optimal balance between the number of active sites and the preservation of the ACF’s porous structure.

Nevertheless, under high relative humidity (above 80%), competitive adsorption between water vapor and VOC molecules significantly deteriorates ACF performance [12]. To overcome this limitation, surface modification strategies such as plasma treatment and metal oxide doping have been investigated, and structural optimization has also been studied to improve adsorption–desorption recyclability [13]. However, performance degradation under humid conditions and insufficient selectivity toward specific VOCs remain unresolved challenges. To address these issues, this study focused on ACF modified with CuMnOx transition-metal catalysts. CuMnOx is known to efficiently oxidize VOCs into CO₂ at low temperatures [10], and since synthesis methods and conditions strongly influence its catalytic performance, careful optimization is essential [10]. Previous studies on CuO/ACF filters have demonstrated significant efficiency in formaldehyde removal [14], while the effectiveness of CuMnOx catalysts in CO oxidation further supports the scientific rationale of this work [15]. The impregnation method is a simple yet effective synthesis strategy that promotes low-crystallinity structures with large surface areas. While CuO alone exhibits limited reactivity at ambient conditions, its combination with Mn species in CuMnOx maintains excellent catalytic activity even at low temperatures [9].

Despite its advantages, the native ACF surface exhibits inherent hydrophobicity, which can hinder the uniform dispersion of polar metal salt precursors and limit catalyst-support interactivity [16]. To overcome this, various surface modification techniques are employed. Oxygen plasma treatment is recognized as a highly effective and non-destructive low-temperature method for introducing oxygenated functional groups (e.g., carbony, carboxyl) onto the carbon surface [16]. These newly formed groups significantly enhance the surface hydrophilicity and wetting properties of ACF, thereby promoting the homogeneous nucleation and nano-dispersion of catalyst particles onto the substrate.

Building on these insights, we synthesized CuMnOx-modified ACF using Cu(NO₃)₂⋅3H₂O and Mn(NO₃)₂⋅6H₂O precursors via the impregnation method. To maximize the effectiveness of the CuMnOx catalyst, we subsequently applied a high-concentration O₂ plasma surface treatment to the impregnated ACF, aiming to enhance catalyst dispersion and interfacial reactivity. The resulting materials were thoroughly characterized through surface analyses (EDX, XPS, etc.). Furthermore, adsorption and degradation performances were evaluated against three representative VOCs—formaldehyde, acetaldehyde, and benzene—commonly emitted from printing, coating, and film manufacturing processes.

A comprehensive understanding of the catalyst-support interaction and intrinsic activity requires detailed physicochemical characterization [17]. The textural properties are assessed via BET analysis, while the chemical state and surface composition of the catalyst species are analyzed by XPS [17]. Furthermore, the intrinsic catalytic properties such as reducibility and oxygen mobility are critical and are evaluated using H₂-TPR. These analyses collectively provide the mechanistic foundation for understanding the enhanced catalytic oxidation performance towards VOCs.

The objective of this study is to establish a technological basis for real-time removal of high-concentration VOCs in industrial environments through the development of ACF-based materials that exploit the synergistic effect of CuMnOx catalysis and O₂ plasma surface functionalization.

2. Materials and Methods

The overall experimental procedure used in this study is summarized in Figure 1. It outlines the sequential steps of CuMnOx precursor preparation, impregnation of activated carbon fiber (ACF), calcination, oxygen plasma surface treatment, and subsequent characterization and performance evaluation.

2.1. Synthesis of CuMnOx Catalysts and Incorporation into ACF via Impregnation

Cu(NO₃)₂·3H₂O and Mn(NO₃)₂·6H₂O (Daejung Chemical & Metals Co., Ltd., Siheung, Republic of Korea) were used as precursors, mixed at a 1:1 ratio (w/w), and impregnated into activated carbon fibers (ACF) to synthesize CuMnOx-loaded ACF. The ACF (Liaoning Carbon Fiber Technology Co., Ltd., Panjin, China) used for impregnation exhibited a specific surface area of 1000 m²/g, with a width of 1 m and a thickness of 1 mm.

Four types of CuMnOx-loaded ACF were prepared depending on the amount of precursor, and were designated as ACF-0.1, ACF-0.3, ACF-0.45, and ACF-0.6, based on the weight of Cu(NO₃)₂·3H₂O. Raw ACF without catalyst loading was used as a control. The detailed preparation process is summarized in Figure S1.

The CuMnOx catalyst precursors were first prepared by dissolving Cu(NO₃)₂⋅3H₂O and Mn(NO₃)₂⋅6H₂O in 100 mL of distilled water at the prescribed molar ratio (Table S1) to form a mixed solution. To promote uniform catalyst formation, Al(NO₃)₃⋅9H₂O was added as a secondary precursor, and the solution was then thoroughly stirred.

Next, the ACF was placed in a Petri dish (SPL Life Sciences Co., Ltd., Pocheon, Republic of Korea) and immersed in this precursor solution for 30 min to ensure complete absorption of the liquid into the fiber structure. Following the immersion, the impregnated ACF was subjected to a drying process, which involved placing the samples in a vacuum oven (OV-11, Jeio Tech Co., Ltd., Daejeon, Republic of Korea) at 80 °C for 24 h.

Finally, the dried samples were transferred to a tube furnace (TF1-600, Carbolite Gero Ltd., Hope Valley, UK) and calcined at 450 °C for 5 h under a nitrogen atmosphere. A heating rate of 10 °C/min was used to reach the calcination temperature. The physical appearance of the activated carbon fibers before and after CuMnOx impregnation is shown in Figure 2.

2.2. O₂ Plasma Surface Treatment and Final Composite Preparation

The ACF surface was functionalized by introducing oxygen functional groups using a high-concentration oxygen (99.999%) plasma surface treatment method, which is one of the radiation techniques. Figure S2 illustrates the schematic diagram of the plasma surface treatment equipment used in this study. The plasma system (Femto Science, Co. Hwaseong-si, Republic of Korea) consists of a vacuum pump (Ulvac Kiko Inc., Yokohama, Japan), an aluminum alloy reactor (200 × 240 × 160 mm), a gas flow controller (MKS Instruments Inc., Andover, MA, USA), and a power supply (Femto Science Co., Hwaseong-si, Republic of Korea). The Plasma surface treatment was performed for 1 to 5 min under the conditions of a gas flow rate of 50 sccm, a power of 200 W, and a frequency of 50 kHz. Samples ranging from ACF-01 ti ACF-05 were prepared by varying the treatment time in minute intervals.

2.3. Surface Characterization of ACF

The surface morphology and composition of the ACF were analyzed using field-emission scanning electron microscopy (FE-SEM, VEGA3, TESCAN, Brno, Czech Republic), energy-dispersive X-ray spectroscopy (EDX, Oxford Instruments, Abingdon, UK), inductively coupled plasma (ICP, Avio 220 Max, PerkinElmer, Waltham, MA, USA), and X-ray photoelectron spectroscopy (XPS, K-Alpha, Thermo Fisher Scientific, Waltham, MA, USA).

The surface morphology of the CuMnOx/ACF samples, which were prepared with varying precursor concentrations and plasma treatments while keeping a fixed Cu:Mn ratio of 1:1, was examined using Field Emission Scanning Electron Microscopy (FE-SEM, VEGA3, TESCAN, Brno, Czech Republic). Imaging was performed using a secondary electron detector (TESCAN, Brno, Czech Republic) at a 15 kV accelerating voltage and a 9–15 mm working distance, achieving magnifications of 2000×, 5000×, and 10,000× without requiring an additional conductive coating.

Energy Dispersive X-ray Spectroscopy (EDX, Oxford Instruments, Abingdon, UK) was employed to determine the elemental composition (C, O, Mn, Cu, Al) of the samples. This analysis was conducted at a 15 kV accelerating voltage, and the resulting elemental spectra were used to confirm the presence and ascertain the relative proportions of each element.

The specific surface area, total pore volume, and micropore volume were determined using the Brunauer–Emmett–Teller (BET, TriStar II 3020, Micromeritics, Norcross, GA, USA) method, which involved N2 adsorption measurements at 77 K. The specific surface area was calculated using the BET multipoint method, pore distribution was analyzed by the BJH method, and the micropore volume was found using T-Plot analysis. Additionally, the chemical composition of Cu and Mn in the most suitable sample, as identified from the BET results, was further quantified using Inductively Coupled Plasma (ICP, Avio 220 Max, PerkinElmer, Waltham, MA, USA).

X-ray Photoelectron Spectroscopy (XPS, K-Alpha, Thermo Fisher Scientific, Waltham, MA, USA) was utilized to analyze the surface chemical composition and elemental states of both the raw ACF and the catalyst-loaded ACF-P-0.1 sample.

The reducibility of the catalyst was investigated via H₂-Temperature Programmed Reduction (H₂-TPR, AutoChem II 2920, Micromeritics, Norcross, GA, USA), where the sample was heated from 50 °C to 900 °C under a flow of 5% H₂/Ar. The adsorption and desorption behavior of carbon monoxide was studied using CO-Temperature Programmed Desorption (TPD), which involved first adsorbing CO gas onto the sample, followed by ramping the temperature from 40 °C to 900 °C to analyze both the CO desorption profile and its conversion to CO₂.

2.4. Evaluation of VOC Removal Performance

The VOC removal performance of the CuMnOx/ACF composites was evaluated following the Korean Eco-Label Certification Standard (EL.608:2022) [18]. A gas mixture containing formaldehyde (15 ppm), acetaldehyde (14 ppm), and benzene (20 ppm) was prepared and introduced into a 5 L Tedlar bag containing each ACF sample (10 × 10 cm). A total of 3 L of the VOC mixture was injected to create a homogeneous atmosphere inside the bag, and the VOC concentration before and after the test was measured using a Gastec GV-100S gas-detection-tube system. The detailed experimental conditions—including gas composition, sample size, and injected gas volume—are summarized in Table S3 of the Supporting Information. All experiments were conducted at room temperature, and the reported values represent the average of replicate measurements.

The VOC removal performance of CuMnOx/ACF-P-0.1 was evaluated against a mixed gas stream containing formaldehyde (HCHO, 15 ppm), acetaldehyde (CH₃CHO, 14 ppm), and benzene (C₆H₆, 20 ppm). These concentrations were selected based on the regulated indoor air quality standards of Korea as well as reported exposure levels in occupational and indoor environments: average benzene concentrations of 2.7 ppm (maximum 11 ppm) at gas stations [19], 10–20 ppm of formaldehyde in anatomy laboratories [20], and 6.0–16.6 ppm of acetaldehyde from cooking processes [17]; additional indoor studies have reported 5–15 ppm levels [21].

For comparison, VOC removal was evaluated using three distinct samples over time intervals of 30, 60, 90, and 120 min (Table S3): raw ACF (#1), CuMnOx-impregnated ACF-Cat-0.1 (#2), and the dual-modified ACF-P-0.1 (#3). Tests were conducted following the Korean Eco-label Certification Standard (EL.608:2022) [18] using a gas detection tube method. Specifically, each sample (10 × 10 cm) was placed in a 5 L Tedlar bag(SKC Inc., Eighty Four, PA, USA), into which 3 L of the VOC mixture was injected. VOC concentrations were then measured using a detection tube system (GV-100S, Gastec Corp., Kanagawa, Japan) to determine removal efficiency.

All experimental techniques and analytical methods employed in this study were performed in accordance with internationally recognized standards. The specific surface area and pore parameters were determined using the Brunauer–Emmett–Teller (BET, TriStar II 3020, Micromeritics, Norcross, GA, USA) method following ISO 9277:2010 [16]. Elemental compositions were analyzed by energy-dispersive X-ray spectroscopy (EDX, Oxford Instruments, Abingdon, UK) in accordance with ISO 22309:2011 [22], and bulk metal contents were quantified by inductively coupled plasma optical emission spectrometry (ICP-OES) following ISO 11885:2007 [23]. X-ray photoelectron spectroscopy (XPS, K-Alpha, Thermo Fisher Scientific, Waltham, MA, USA) energy calibration was carried out according to ISO 15472:2010 [24], and the VOC removal tests were conducted in compliance with the Korean Eco-label Certification Standard (EL.608:2022) [18]. These reference standards ensure the reproducibility and methodological reliability of the results presented in this work.

3. Results and Discussion

3.1. Physicochemical Characteristics of ACF After CuMnOx Incorporation

Field-emission scanning electron microscopy (FE-SEM) analysis revealed that the diameter of the ACFs was generally uniform, with an average thickness of approximately 10 μm (Figure 3). CuMnOx catalyst particles were observed to be evenly dispersed across the ACF surface, and their size increased with higher precursor loading. The smallest particles, observed in the ACF-P-0.1 sample, exhibited sizes below 500 nm.

When the Cu precursor loading exceeded 0.45 g (ACF-0.45), excessive deposition of CuMnOx particles on the ACF surface was observed, leading to localized agglomeration. In the ACF-0.6 sample, the surface displayed relatively coarse and aggregated CuMnOx particles. This indicates that beyond a certain threshold, precursor overloading promotes catalyst agglomeration, which can cover portions of the ACF surface and significantly reduce the available surface area.

Therefore, optimizing the precursor amount is crucial for maintaining the structural advantages of ACF. Among the tested samples, ACF-P-0.1 and ACF-0.3 were identified as optimal conditions, exhibiting small, uniformly dispersed particles without noticeable agglomeration. This observation was further validated by BET surface area analysis, confirming the scientific rationale of the structural optimization.

Energy-dispersive X-ray spectroscopy (EDX) confirmed the presence of Cu, Mn, and Al originating from the precursors within the ACF structure (Figure S3). As the impregnation amount increased, the relative contents of these elements also increased consistently (

Table 1. Surface elemental composition (EDX) of CuMnOx-impregnated ACF under different preparation conditions.

	C (Atom%)	O (Atom%)	Cu (Atom%)	Mn (Atom%)	Al (Atom%)
Raw	91.31	8.69	-	-	-
ACF-P-0.1	81.40	13.18	2.77	1.37	1.28
ACF-0.3	75.29	14.39	4.15	3.76	2.41
ACF-0.45	68.65	15.29	7.12	6.10	2.83
ACF-0.6	36.82	20.92	26.11	11.47	4.69

). Overall, Cu exhibited the highest proportion among the detected elements. However, considering the limited quantitative accuracy of EDX, complementary analyses were deemed necessary to validate these findings.

A comparison of the oxygen content across samples showed that the atomic ratio of oxygen was more than twice as high as that of the metallic elements, suggesting the successful formation of CuMnO₂ catalysts on the ACF surface. In contrast, the ACF-0.6 sample exhibited a noticeably lower oxygen ratio than expected from its impregnation level. Such structural characteristics may restrict oxygen-mediated reactions compared with catalysts prepared under optimized conditions, potentially leading to reduced VOC removal efficiency.

This interpretation is consistent with the findings of Wen et al. [25], who reported that a decrease in oxygen vacancy concentration within MnO₂ catalysts resulted in over a 25% reduction in toluene removal efficiency.

BET analysis was performed on raw ACF and CuMnOx-loaded ACF samples with varying precursor concentrations to determine the specific surface area, total pore volume, and micropore volume (

Table 2. Textural properties (surface area, pore volume, microporosity) of CuMnOx-impregnated ACF.

Sample ID	1	2	3	4	5
Sample name	Raw	ACF-P-0.1	ACF-0.3	ACF-0.45	ACF-0.6
BET surface area (m2/g)	1740.1	1342.7	559.6	645.4	484.3
Total pore volume (cm3/g)	0.749	0.579	0.245	0.272	0.229
Micropore volume (cm3/g)	0.693	0.534	0.225	0.255	0.197
Micropore (%)	92.52	92.23	91.84	93.75	86.03

). The specific surface area of raw ACF was 1740.1 m²/g, which decreased upon CuMnOx loading. In the ACF-P-0.1 sample, the specific surface area was reduced to 1342.7 m²/g. A more pronounced decrease was observed for higher loading levels, with ACF-0.3 and ACF-0.6 exhibiting 559.6 and 484.3 m²/g, respectively, indicating a significant reduction compared to raw ACF. This reduction can be attributed to excessive CuMnOx deposition, where multilayer coverage on the ACF surface reduces the accessible surface area.

Notably, an anomalous increase in specific surface area was observed for ACF-0.45 (645.4 m²/g) compared to ACF-0.3 (559.6 m²/g), which contradicts the general trend of specific surface area reduction with increasing loading (Table 2). This transient increase is hypothesized to be due to a localized effect at the 0.45M concentration level, where the precursor solution or the subsequent calcination process may have induced a slight pore widening or activation effect within the ACF micropores, thereby temporarily increasing the accessibility of nitrogen molecules during BET measurement (Total Pore Volume for ACF-0.45 was 0.579 cm³/g vs. 0.245 cm³/g for ACF-0.3). Alternatively, the CuMnOx nanoparticles themselves may have formed a more porous structure at this specific concentration, contributing to the overall specific surface area.

However, the general conclusion is supported by the subsequent sharp drop at ACF-0.6 (484.3 m²/g), confirming that excessive loading ultimately leads to pore blockage. Ultimately, the ACF-P-0.1 Sample was chosen as it represented the optimal balance between retaining structural integrity and providing sufficient catalyst loading for performance evaluation.

Both the total pore volume and micropore volume decreased with increasing CuMnOx loading. In particular, the micropore volume (<2 nm), known to play a critical role in adsorption, decreased sharply from 0.693 cm³/g in raw ACF to less than one-third of this value in ACF-0.6. In contrast, ACF-P-0.1 maintained a micropore volume of 0.534 cm³/g and preserved a high microporosity (>90%), similar to raw ACF. This suggests that at lower loading levels, the reduction in micropores caused by catalyst incorporation was minimal.

These findings indicate that the CuMnOx loading concentration strongly influences the microporous structure of ACF, thereby affecting its adsorption and removal characteristics. Among the tested samples, ACF-P-0.1 demonstrated the most favorable balance between catalyst incorporation and preservation of surface area and micropore volume, suggesting that it represents the optimal loading condition.

The metal content of CuMnOx-loaded ACF, determined by ICP analysis (Table S4), showed that raw ACF contained Cu (17.99 ppm) and Mn (17.20 ppm), which are attributed to trace impurities inherent to the carbon fibers. In contrast, ACF-0.1 exhibited dramatically increased levels, with 232,426 ppm Cu and 171,311 ppm Mn, quantitatively confirming the successful impregnation of CuMnOx catalysts into the ACF. Although the theoretical molar ratio of Cu to Mn was 1:1, the actual analysis revealed higher Cu content compared to Mn. This discrepancy is likely due to the higher affinity of Cu for the ACF surface during the impregnation process.

XPS elemental analysis further supported these findings. In raw ACF, the major components were carbon (91.33 atom%) and oxygen (6.72 atom%), consistent with the EDX results (Table S5). In ACF-P-0.1, however, oxygen content markedly increased to 43.07 atom% due to the introduction of oxygen and metal ions, while carbon content decreased to 34.55 atom%. Additionally, Cu and Mn were detected at 5.98 atom% (Cu 2p) and 4.19 atom% (Mn 2p), respectively, indicating that CuMnOx was effectively incorporated into the ACF in proportions consistent with the initial impregnation ratio. These results suggest that CuMnOx incorporation plays a critical role in modifying the surface oxidation state and enhancing surface reactivity through the enrichment of oxygen species.

Detailed XPS peak analysis provided further insight into the chemical states of Cu and Mn in ACF-0.1 (Figure 4). The C1s and O1s spectra revealed that while raw ACF was dominated by C–C and C–H bonds, ACF-P-0.1 exhibited relatively stronger C=O and C–O peaks, indicating an increase in oxygen-containing surface functionalities. A major O1s peak observed around 531 eV corresponded to adsorbed oxygen (O_ads) associated with metal–oxygen bonds, as well as lattice oxygen (O_lat) coordinated with organic C–O structures. This confirms that CuMnOx catalysts were present on the ACF surface in oxygen-rich compound forms.

In the Cu 2p spectrum of ACF-P-0.1, two characteristic peaks were observed at 930–955 eV, corresponding to Cu 2p₁/₂ and Cu 2p₃/₂, respectively. The presence of Cu²⁺ (933–935 eV) along with a fraction of Cu⁺ indicated that Cu²⁺ ions existed predominantly as Cu₂O species on the surface. Similarly, the Mn 2p spectrum (639–654 eV) revealed two peaks, with Mn 2p₃/₂ showing contributions from both Mn³⁺ and Mn⁴⁺ states. These peaks suggest the coexistence of Mn₂O₃ and MnO₂ species, formed through bonding with surface oxygen. Taken together, these findings confirm the effective generation of catalytically active copper and manganese oxides on the ACF surface, which are expected to contribute to enhanced VOC oxidation performance.

The H₂-TPR profiles of CuMnOx-loaded ACF are shown in Figure 5. Reduction peaks were observed in both raw ACF and ACF-0.1. For raw ACF, a broad reduction peak appeared above 650 °C, which can be attributed to the intrinsic reduction behavior of the carbon fibers. In contrast, ACF-P-0.1 exhibited multiple reduction peaks at much lower temperatures, specifically at 140, 205.8, and 554.1 °C. This indicates that the incorporation of CuMnOx catalysts markedly improved the reducibility of the ACF. In particular, the peak observed around 554.1 °C suggests that hydrogen interaction with the ACF surface occurred more readily at lower temperatures due to the catalytic effect. These findings confirm that the oxide species of the CuMnOx catalysts were successfully impregnated into the ACF, imparting enhanced reducibility. The reduction peaks at 140 °C and 205.8 °C are attributed to the stepwise reduction in CuO → Cu₂O → Cu, while the broader peak at 554.1 °C is assigned to the reduction of Mn⁴⁺/Mn³⁺ lattice oxygen within the Cu–Mn mixed oxide. These assignments are consistent with literature-reported behaviors of Cu–Mn composite catalysts and with the XPS results, which confirmed the coexistence of Cu²⁺/Cu⁺ and Mn⁴⁺/Mn³⁺ species.

Therefore, the low-temperature reducibility is mainly governed by Cu species, whereas Mn contributes to lattice-oxygen mobility at higher temperatures. It has been well established in previous studies that catalysts exhibiting lower reduction temperatures generally correspond to higher catalytic activity [22].

The CO-TPD results are presented in Figure 3. Both raw ACF and ACF-P-0.1 showed CO desorption and subsequent CO₂ formation. Notably, ACF-P-0.1 exhibited strong and distinct peaks at 87, 248.6, and 402.7 °C. These peaks indicate that CO adsorbed on the surface of ACF-P-0.1 was effectively converted into CO₂ at the active catalytic sites. The presence of such pronounced peaks further demonstrates the strong interaction between CO molecules and the CuMnOx catalysts. The CO-TPD data were interpreted qualitatively to compare the desorption behavior and CO₂ formation profiles, reflecting the relative adsorption strength and redox activity of surface sites. Activation or binding energies were not calculated in this study, as such analysis would require additional assumptions and experiments under multiple heating rates. Future work will address the quantitative estimation of CO adsorption energetics.

SEM analysis of ACF before and after oxygen plasma treatment revealed that, unlike CuMnOx-incorporated samples, the overall surface morphology exhibited almost no visible changes (Figure 6). This indicates that oxygen plasma treatment effectively introduced oxygen functional groups while minimizing reductions in specific surface area and microporosity. These findings confirm that oxygen plasma treatment is an effective gas-phase surface modification method that enhances surface energy by generating chemically active sites without significantly altering the structural properties of ACF [22].

XPS elemental analysis further supported these results. In raw ACF, the primary elements were carbon (87.24 atom%) and oxygen (10.39 atom%). In contrast, oxygen plasma-treated ACF exhibited a markedly increased oxygen content (

Table 3. XPS elemental analysis of ACF after high-purity oxygen plasma treatment.

Samples	C1s (Atom%)	O1s (Atom%)
Raw	87.24	10.39
ACF-P-O1	62.51	31.26
ACF-O2	58.54	35.19
ACF-O3	55.70	36.75
ACF-O4	52.93	38.63
ACF-O5	54.03	39.23

Note: All values are expressed as atomic percentages (atom%).

). Moreover, the oxygen concentration increased progressively with longer treatment durations. This demonstrates that high-purity oxygen plasma treatment effectively introduced oxygen functional groups onto the ACF surface, and that the extent of functionalization increased with prolonged treatment time.

Since SEM analysis alone cannot adequately identify the bonding characteristics between ACF and oxygen functionalities, XPS peak deconvolution (C1s and O1s) was performed to investigate the detailed chemical structure of the plasma-treated ACF samples (Figure 7). Similarly to the results observed for CuMnOx incorporation, the high-purity oxygen plasma–treated ACF samples showed effective introduction of oxygen functional groups, such as C=O and C–O bonds, on the carbon surface (C1s peaks). Moreover, the relative intensities of C=O and C–O bonds increased with longer plasma treatment durations, confirming the progressive incorporation of oxygen functionalities.

These changes indicate that oxygen plasma treatment effectively induces chemical reconstruction of the ACF surface, resulting in oxygen-bonded structures embedded within the carbon matrix. Analysis of the O1s peaks further revealed that the oxygen species in raw ACF primarily existed as surface oxides and likely formed during the initial manufacturing process.

3.2. VOC Removal Performance

In this study, formaldehyde, acetaldehyde, and benzene were selected as model VOCs to evaluate the removal performance of the CuMnOx/ACF composites. This selection was based on actual emission characteristics identified in a printing factory, which has been designated as the target site for subsequent field application of this research. Field monitoring data from the printing process showed that these three compounds are the predominant VOCs emitted during ink drying and coating operations, making them representative of real industrial indoor environments. Furthermore, the selected VOCs encompass a wide range of chemical polarities and reactivities—formaldehyde and acetaldehyde are polar oxygenated compounds, while benzene is a non-polar aromatic hydrocarbon—allowing evaluation of the catalyst’s adsorption and oxidation behavior across chemically diverse species. This selection provides a realistic and application-oriented framework for assessing the CuMnOx/ACF catalyst’s performance and practical feasibility in VOC abatement from printing facilities.

Among the tested samples, ACF-P-0.1 exhibited more than 10% higher efficiency compared with raw ACF, demonstrating its superior combined performance (

Table 4. VOCs removal efficiencies of raw ACF and ACF-P-0.1.

Elapsed Time (min)	Removal (%)
	Formaldehyde		Acetaldehyde		Benzene
	Raw	ACF-P- 0.1	Raw	ACF-P- 0.1	Raw	ACF-P- 0.1
30	81.6	85.0	85.7	85.7	96.2	97.5
60	96.6	96.6	85.7	85.7	97.5	97.5
90	96.6	96.6	85.7	89.2	97.5	97.5
120	96.6	96.6	87.5	89.2	97.5	97.5

). On a component basis, benzene removal reached 97.5%, higher than that of raw ACF (96.2%), representing the most pronounced improvement. The enhanced benzene removal efficiency can be attributed to its nonpolar and stable molecular structure, which allows it to fit optimally into the microporous structure of ACF (<2 nm), resulting in stable adsorption. In addition, with a boiling point of 80.1 °C, benzene is less likely to desorb from pore surfaces compared with other VOCs, thereby increasing its retention.

In contrast, acetaldehyde removal efficiency was the lowest among the three VOCs, at 85.7%. This relatively poor performance can be explained by its high water solubility, low boiling point, high chemical reactivity, and difficulty in oxidative decomposition. These properties limit its adsorption onto ACF, facilitate its volatilization and diffusion at room temperature, and hinder effective capture. Furthermore, its molecular structure contains a reactive aldehyde group (–CHO) directly bonded to oxygen, which is more prone to oxidation or reaction with other species. This result is consistent with previous reports [26], where the hydrophobic nature of ACF surfaces favors adsorption of nonpolar VOCs such as benzene over polar VOCs such as acetaldehyde, clearly demonstrating that the physicochemical characteristics of ACF significantly influence removal efficiencies depending on VOC type.

The mechanistic interpretation of ACF-P-0.1 indicates that benzene removal is dominated by strong physical adsorption, whereas formaldehyde removal is primarily governed by catalytic oxidation. A clear difference in the initial removal rate was observed: after 30 min, raw ACF achieved 81.6% removal of formaldehyde, while ACF-0.1 reached 85.0%. This improvement is attributed to the high adsorption capacity of ACF-0.1 (1342.7 m²/g, confirmed by BET analysis) coupled with its high catalytic activity (low-temperature reducibility at 140 °C, demonstrated by H₂-TPR and XPS). This synergy enabled adsorbed formaldehyde to be rapidly oxidized into CO₂ and H₂O, regenerating active sites and accelerating removal.

For all three VOCs—formaldehyde, acetaldehyde, and benzene—the removal efficiencies stabilized after 60 min, with no further increase observed. This plateau indicates that ACF micropores reached saturation between 30 and 60 min, after which a dynamic equilibrium was established between adsorption and catalytic decomposition at CuMnOx active sites. Notably, ACF-P-0.1 reached this equilibrium state much faster than raw ACF, confirming its optimal synergistic performance.

The superior performance of ACF-P-0.1, compared with previous ACF-based VOC removal studies [19,20], highlights its strong competitiveness and high potential value as a composite material for VOC abatement.

4. Conclusions

In this study, CuMnOx catalysts were successfully introduced onto activated carbon fibers (ACF) via an impregnation method using Cu(NO₃)₂·3H₂O and Mn(NO₃)₂·6H₂O as precursors. The structural, chemical, and catalytic properties of the synthesized composites were comprehensively evaluated, and their VOC removal performances were assessed. Among the prepared samples, ACF-0.1 demonstrated the greatest significance, establishing the feasibility of developing a composite material that optimally combines the high adsorption capacity of ACF with the low-temperature oxidation activity of CuMnOx.

SEM analysis confirmed that CuMnOx particles in ACF-0.1 were uniformly dispersed on the fiber surface with particle sizes below 500 nm. This homogeneous distribution effectively preserved the inherent porous structure of ACF, maintaining a high specific surface area of 1342.7 m²/g and a micropore ratio of 92.23%, thereby minimizing pore blockage typically associated with catalyst loading. XPS results further demonstrated that Cu and Mn species were introduced in multiple oxidation states (Cu²⁺, Mn³⁺, Mn⁴⁺), while H₂-TPR analysis revealed low-temperature reduction peaks at 140 °C and 205.8 °C, directly confirming the excellent low-temperature catalytic activity of CuMnOx.

In VOC removal tests, ACF-0.1 exhibited markedly superior performance compared to raw ACF, attributed to a synergistic adsorption–oxidation mechanism. For benzene, ACF-0.1 achieved the highest efficiency of 97.5%, reflecting the favorable confinement of nonpolar VOCs within its microporous framework. In the case of formaldehyde, ACF-0.1 achieved a 96.6% removal efficiency within 30 min, significantly outperforming raw ACF. This was attributed to the rapid oxidation of adsorbed formaldehyde into CO₂ and H₂O at CuMnOx active sites, which enabled continuous regeneration of adsorption sites and accelerated removal rates. Thus, ACF-0.1 achieved an optimal balance between adsorption and catalytic oxidation, maximizing overall VOC removal efficiency.

Overall, the CuMnOx/ACF composites synthesized via impregnation demonstrated structural stability, well-dispersed catalysts, and outstanding removal performance across multiple VOCs, indicating strong potential as advanced filter materials. These composites are expected to serve as promising candidates for air purification systems targeting hazardous VOC emissions from industrial processes such as printing, coating, and conductive film manufacturing. Nonetheless, limitations were noted, including relatively lower removal efficiency for highly polar VOCs such as acetaldehyde due to the hydrophobic nature of ACF, and insufficient evaluation of long-term regeneration stability and durability. Future research should therefore focus on verifying removal efficiencies under diverse operating conditions (VOC concentration, flow rate, humidity) and conducting pilot-scale studies to assess long-term performance and practical applicability. In addition to its high removal efficiency, the CuMnOx/ACF composite offers a favorable cost-to-performance ratio, as it is synthesized from low-cost and readily available precursors through a simple impregnation–calcination process, indicating good potential for practical and economical industrial application.

In terms of scalability, the simple impregnation–calcination process and inexpensive precursors provide a strong basis for industrial-scale production and field implementation of CuMnOx/ACF composites. In large-scale production, maintaining uniform Cu–Mn dispersion on ACF and preventing catalyst detachment during long-term operation remain key engineering challenges. Additionally, optimization of coating thickness and air-flow resistance will be necessary to minimize pressure drop while preserving catalytic activity. However, prior to real-world application, potential environmental and safety issues—including metal leaching under humid or acidic conditions, particle release, and occupational exposure risks—should be systematically evaluated. Although oxygen plasma treatment was effective for enhancing surface oxygen functionalities and promoting metal dispersion at the laboratory scale, this batch process is not readily compatible with continuous industrial-scale production owing to the high equipment cost and limited reactor capacity. Alternative oxidative activation methods—such as thermal oxidation, ozone or H₂O₂ exposure, and chemical acid functionalization (e.g., HNO₃ or KMnO₄ treatment)—may offer more scalable and cost-effective routes for surface modification in future applications.

Future work will therefore focus on assessing these aspects to ensure the safe and sustainable use of the proposed VOC treatment system in industrial environments.