Specifically designed 10-Fe₁Mn₁-USY hydrophobicity material for acetone adsorption under high humidity condition

Abstract

Acetone is a solvent used in many laboratories and factories. Serious problems will occur when it is exposed to the environment. Therefore, a new design hydrophobicity bimetallic metal material (10-Fe₁Mn₁-USY) was prepared for acetone adsorption under conditions of high humidity. Hydrophobic Y type zeolite was used to prepare bimetallic metal materials and the effect of different operating parameters (including different material, humidity, temperature, pollutant concentration, residence time, and regeneration) on the efficiency of acetone treatment was examined. Isothermal adsorption model, kinetics and thermodynamic model analysis were also used to establish the reaction mechanism. The 10-Fe₁Mn₁-USY material has good adsorption capacity (133 mg g⁻¹) for acetone under a relative humidity of 50%. The main factors affecting the adsorption capacity are the contact angle, hydrophobicity, specific surface area, and Si/Al of the material. The isothermal adsorption and the kinetic adsorption behavior of 10-Fe₁Mn₁-USY material for acetone are more suitable for the Temkin isotherm adsorption model and the pseudo-first-order kinetic model. The adsorption of acetone by 10-Fe₁Mn₁-USY material is dominated by intra-particle diffusion. According to the thermodynamic analysis results, the adsorption behavior of 10-Fe₁Mn₁-USY material for acetone is a type of physical adsorption, and a spontaneous and non-sequential reaction.

1 Introduction

As one of the representative ketones of Volatile Organic Compounds (VOCs), acetone is an important chemical raw material and solvent [1, 2]. It is extremely harmful to the environment and the human body, making the prevention and control of acetone pollution critical [3, 4]. At present, common treatment methods of VOCs include the condensation [5], absorption [6], plasma destruction [7], thermal incineration [8], catalyst incineration [9], photocatalytic [10, 11] and adsorption [12, 13] methods. The condensing method has limited treatment efficiency and cannot treat exhaust gas containing various pollutant components. Absorption methods are prone to wastewater treatment problems and high maintenance costs. The plasma destruction method is prone to interference, requires expensive equipment and has high maintenance costs. The thermal incineration and catalyst incineration methods require higher temperatures to destroy pollutants, wasting energy, and creating secondary pollutant problems that increase the cost of secondary treatment. Therefore, the adsorption method is the most promising technology in the treatment of VOCs. The adsorption method can simultaneously treat multiple components of pollutants, but is easily affected by the humidity of the exhaust gas, which absorbs water vapor easily, resulting in material adsorption saturation. Hence, a new design for a hydrophobicity bimetallic metal material was prepared for adsorption VOCs under high humidity conditions in this study.

Several studies on the adsorptive removal of various VOCs have been reported. Zhang et al. [14] conducted a comparative study on the adsorption of toluene over MCM-41 (Mobil Composition of Matter No. 41) and SBA-15 (Santa Barbara Amorphous-15). Their results showed that SBA-15 exhibited adsorption capacities superior to those of MCM-41, which is probably attributable to its two-dimensional biporous (micro-/mesoporous) system [14, 15]. Hung et al. [16] investigated the adsorption behaviors of acetone vapors over MCM-41 and ZSM-5 (Zeolite Socony Mobil–5). They demonstrated that MCM-41 revealed better adsorption capability and regeneration ability than the ZSM-5 zeolite, an adsorbent used in commercialized zeolite rotor concentrators [15, 16]. The adsorption of acetone and toluene on dealuminated Y-zeolite was studied by Lee et al. [17]. Their results showed that the adsorption amount of acetone was higher than that of toluene at low temperatures, but its adsorption affinity was weaker. Su et al. [4] conducted a comparative study on the adsorption of acetone over activated carbons (AC). Their results demonstrated that different functional groups under the same conditions provided different non-electrostatic contribution and electrostatic contributions to acetone adsorption capacity. The electrostatic contribution of oxygen-containing functional groups was better than that of nitrogen-containing functional groups, and functional groups containing oxygen and hydrogen were better than the functional groups containing only oxygen. Humidity resulted in the pronounced decrease in the non-electrostatic contribution to all functional groups. However, electrostatic contribution is less affected by humidity [4].

In recent years, hydrophobic zeolites have been safely used in catalytic oxidation processes at high temperatures due to their non-flammability, thermal stability, and moisture resistance. Hydrophobic zeolites have high adsorption affinity for acetone under humid conditions [18]. The affinity of hydrophobic and hydrophilic molecules and the ratio of Si/Al atoms affect the adsorption performance of zeolite, and the adhesive may also affect the adsorption performance of zeolite. When the Si/Al ratio in the zeolite is larger in a humid environment, the hydrophobicity and the adsorption capacity of the adsorbent will be enhanced. Bhatia et al. [19] prepared AgY (Si/Al = 40) and AgZSM-5 (Si/Al = 140) with different Si/Al ratios, and carried out experiments on the adsorption of butyl acetate under a relative humidity (RH) of 35%. Results showed that the adsorption capacity of AgZSM-5 was about 35% better than that of AgY material [3]. Huang et al. [13] prepared ZSM-5 (Si/Al = 50–300) zeolite adsorbents with different Si/Al ratios to adsorb toluene at RH = 50% condition. The adsorption capacity (75 mg g⁻¹) of ZSM-5–300 (Si/Al = 300) is higher than that of ZSM-5–50 (Si/Al = 50) (49 mg g⁻¹) by about 35% [13]. Therefore, it appears that increasing the Si/Al ratio of zeolite can improve its hydrophobicity, increase the adsorption capacity of pollutants, and enable it to be less affected by moisture.

Consequently, in this study, a new design hydrophobicity monometallic material and bimetallic metal material was prepared and applied for the adsorption acetone at lower temperature (298–328 K) and RH = 50%. The effect of different operating parameters (including different materials, humidity, temperature, pollutant concentration, residence time, and regeneration) on the efficiency of acetone treatment is discussed in this study.

2 Materials and methods

2.1 Materials preparation

In this study, Fe–Mn mixed bimetallic oxide materials were prepared by immersing 5 g of granular zeolite carrier (including NaY, HY, USY and ZSM-5 zeolite) in 250 mL of an aqueous solution containing 10% Fe(NO₃)₃·9H₂O or 10% Mn(NO₃)₂, and soaking for 1 h, and then placing the mixture in a sonic shock tank on ultra-high and shaking it for 30 min. The liquid was then filtered through a screen to obtain the solid content. After that, the solid content was placed in an oven, dried at a temperature of 105 °C for 2 h, and then placed in a crucible and calcined at 550 °C for 6 h in a high-temperature furnace. The prepared material was then cooled to room temperature to obtain monometallic materials loaded with Fe or Mn (abbreviated as10-Fe-NaY, 10-Fe-HY, 10-Fe-USY, 10-Mn-USY and 10-Fe-ZSM-5), and bimetallic material loaded with Fe and Mn (10-Fe_xMn_y-USY, x/y is the ratio of Fe content to Mn content, which can be 1:4, 1:2, 1:1, 2:1 and 4:1).

2.2 Materials characterization

The morphology of the material was measured using a Scanning Electron Microscopy (SEM) via Energy Dispersive X-ray Spectroscopy (EDS) analysis. The specific surface area, pore size distribution, and pore volume were determined by performing nitrogen adsorption–desorption measurements with an ASAP 2020 apparatus using Brunauer–Emmett–Teller (BET) calculation methods. The Fourier Transform Infrared (FTIR) of the material was recorded to study the functional groups at room temperature. A Thermo Gravimetric Analysis (TGA) instrument with a heating rate of 20 °C min⁻¹ was used to study the thermal properties of the material. The change in the metal valence state in the material, the interaction between the metal oxide and its support, and the activation energy of the oxide reduction reaction were determined by the Hydrogen-Temperature Programmed Reduction (H₂-TPR) technique.

2.3 Adsorption performance assessment

An acetone adsorption measurement was performed for NaY zeolite, HY zeolite, USY zeolite, ZSM-5, and AC in a fixed bed reactor (length: 70 cm, inside diameter: 3 cm, outside diameter: 3.4 cm) at 800 ppm under humid conditions. The hydrophobic adsorbent with the best adsorption capacity was then selected, and different experimental parameters such as different acetone concentrations (400, 600, 800 and 1000 ppm), the effect of different residence time (0.18, 0.36, 0.54, 0.72 s), different operating temperatures (298, 308, 318, 328 K) and regeneration were examined for their effects on the adsorption capacity of acetone. The adsorption run was performed using high purity acetone (99.9%) gas and the feed flow was controlled by a mass flow controller connected to the fixed bed reactor in the experimental setup. The breakthrough curve of acetone was obtained by a gas chromatograph. In this study, the accuracy and reliability of the standardized or internationally recognized method were verified by analysis of certified standards. Analytical accuracy, determined from replicate analysis of laboratory standards and samples was 97%. Method detection limits under the conditions described are 0.1 ppm for acetone.

2.4 Isothermal adsorption model

2.4.1 Langmuir isotherm model

Langmuir adsorption model use to describe adsorption of solution species onto solid adsorbents. Based on the theory derive an equation, which depicted a relationship between the number of active sites of the surface undergoing adsorption and pressure. Langmuir model is followed in Eq. (1) [20, 21]:

$$\begin{aligned} Q=\left(\frac{ab\times C_{e}}{1+aC_{e}}\right)\end{aligned}$$

(1)

Taking the inverse of both sides gives Eq. (2) as follows:

$$\begin{aligned} \frac{1}{Q}=\frac{1}{b}+\frac{1}{ab\times {C}_{e}}\end{aligned}$$

(2)

Again, C_e is the equilibrium acetone initial concentration in the solution and Q is the amount of acetone adsorbed by the additive. a and b are Langmuir constants related to adsorption capacity and adsorption energy, respectively. Plotting Q⁻¹ versus C_e⁻¹ and measuring the slope and intercept of the plot, a and b can be directly interpreted [20, 21].

2.4.2 Freundlich isotherm model

The Freundlich equation or Freundlich adsorption isotherm, an adsorption isotherm, is an empirical relation between the adsorbate concentrations on the surface of an adsorbent to the adsorbate concentration in the air with which it is in contact. The Freundlich model is followed in Eq. (3) [20, 21]:

$$\begin{aligned} Q=K{{C}_{e}}^{\frac{1}{n}}\end{aligned}$$

(3)

Taking the natural logarithm of both sides gives Eq. (4) as follows:

$$\begin{aligned} \text{ln} {Q}=\text{ln} K+(\frac{1}{n({\mathrm{lnC}}_{\mathrm e})})\end{aligned}$$

(4)

Plotting lnQ versus lnC_e, the values of the Freundlich constants of K and n can be directly interpreted from the plot as the intercept and slope if the data of adsorption experiment matches the above empirical model [20, 21].

2.4.3 Temkin isotherm model

This isotherm contains a factor that explicitly taking into the account of adsorbent–adsorbate interactions. By ignoring the extremely low and large value of concentrations, the model assumes that heat of adsorption of all molecules in the layer would decrease linearly rather than logarithmic with coverage. As implied in the equation, its derivation is characterized by a uniform distribution of binding energies was carried out by plotting the quantity sorbed q_e against lnC_e and the constants were determined from the slope and intercept. The model is given by the following Eqs. (5) and (6) [22]:

$$\begin{aligned} \text{q}_{\text{T}}= {\rm q}_{\text{e}}= { } \frac{{\rm RT}}{\text{b}} {\rm ln} \left({{\rm A}}_{{\rm T}} {\text{C}}_{\text{e}}\right)\end{aligned}$$

(5)

$$\begin{aligned} {\text{q}}_{\text{e}}= { } \left(\frac{{\rm RT}}{\text{b}}\right){\text{lnA}}_{\text{T}} \text{ + }\left(\frac{{\rm RT}}{\text{b}}\right){\text{lnC}}_{\text{e}}\end{aligned}$$

(6)

where, q_T is adsorption capacity simulated by Temkin isotherm model (mg g⁻¹), q_e is adsorption capacity at reaction equilibrium (mg g⁻¹), R is universal gas constant (8.314 J mol⁻¹ K⁻¹), T is temperature at 298 K, b is Temkin isotherm constant, A_T is Temkin isotherm equilibrium binding constant (L g⁻¹) and C_e is adsorption concentration at reaction equilibrium (ppm).

2.5 Adsorption kinetic model

2.5.1 Pseudo-first-order

A first order reaction depends on the concentration of only one reactant (a unimolecular reaction). The pseudo-first-order equation is given as following Eq. (7):

$$\begin{aligned} \frac{{dq}_{t}}{dt} ={k}_{1}\left({q}_{e}-{q}_{t}\right)\end{aligned}$$

(7)

where, q_t is the amount adsorbed at time t (mg g⁻¹). In addition, k₁ is the pseudo-first-order rate coefficient (min⁻¹) and t is the contact time (min). The integration of Eq. (7) with the initial condition, q_t = 0 at t = 0, leads to the following Eq. (8) [21, 23]:

$$\begin{aligned} log\left(q_{e}-q_{t}\right)=\log\;q_{e}-\frac{k_1}{2.303}t\end{aligned}$$

(8)

2.5.2 Pseudo-second-order

A second order reaction depends on the concentrations of one second order reactant, or two first order reactants. The pseudo-second-order model is represented as following Eq. (9):

$$\begin{aligned} \frac{{dq}_{t}}{dt} ={k}_{2}{\left({q}_{e}-{q}_{t}\right)}^{2}\end{aligned}$$

(9)

where, k₂ is the pseudo-second-order rate coefficient (mg g⁻¹ min⁻¹). Integrating Eq. (9) and q_t = 0 at t = 0, the following Eq. (10) is obtained [21, 23]:

$$\begin{aligned} \frac{t}{{q}_{t}}=\frac{1}{{k}_{2}{{q}_{e}}^{2}} + \frac{1}{{q}_{e}}t\end{aligned}$$

(10)

2.5.3 Intra-particle diffusion model

The intra-particle diffusion model was also tested with rate processes usually expressed in terms of square root of time. So q_t or fraction metal sorbed is plotted against t^1/2 as follow [24, 25]:

$$\begin{aligned} {q}_{t} =f\left({t}^{1/2}\right)\end{aligned}$$

(11)

The rate constant for intra-particle diffusion (k_i) at different initial concentrations is determined using the following Eq. (12):

$$\begin{aligned} q_{t}=k_{i}\mathrm{t}^{1/2}\end{aligned}$$

(12)

where, k_i is in mg g⁻¹ media min^−1/2.

3 Results and discussion

3.1 Characterization

In this study, a contact angle meter was used to analyze the contact angle between the surface of the material and the water droplet to understand whether the material was hydrophilic or hydrophobic. The results are shown the contact angles of 10-Fe-NaY, 10-Fe-HY, 10-Fe-USY, 10-Fe-ZSM-5 and 10-Fe₁Mn₁-USY are 89°, 113°, 126°, 63° and 110 ^o, respectively, as shown in Fig. 1. Therefore, 10-Fe-HY, 10-Fe-USY and 10-Fe₁Mn₁-USY are hydrophobic materials (contact angle > 90°), while 10-Fe-NaY and 10-Fe-ZSM-5 are hydrophilic materials (contact angle ˂ 90°). The main reason is that 10-Fe-USY has the largest Si/Al ratio (10.7), as shown in Table S1 of Supplementary Materials, resulting in 10-Fe-USY having the largest contact angle (126°), while 10-Fe-ZSM-5 has the smallest Si/Al ratio (4.8), resulting in 10-Fe-ZSM-5 having the smallest contact angle (63°). A larger Si/Al ratio results in a larger contact angle and thus a stronger hydrophobicity, since Si–OH groups can make water molecules stay on the pore wall surface. However, free water weakly interacts with material pore surface, because water molecules form amorphous groups through limited intermolecular hydrogen bonding [26]. Li et al. [26] show that the contact angle of ZSM-5/siliceous zeolite composites is 17°, which is much smaller than that of the 10-Fe-USY.

Fig. 1

Material contact angle analysis results

To understand the thermal stability of different iron materials and its feasibility at high temperature, this study conducted a thermo gravimetric analysis of 10-Fe-USY, 10-Fe-HY, 10-Fe-NaY, 10-Fe-ZSM-5 and 10-Fe₁Mn₁-USY, and analyzed the weight loss change of the material after heating. The results of this analysis are shown in Fig. 2. These results show that the weight loss changes in two stages at temperatures of 50–200 °C and 350–500 °C, respectively. In the first stage, the initial weight loss of 10-Fe-USY, 10-Fe-HY, 10-Fe-NaY, 10-Fe-ZSM-5 and 10-Fe₁Mn₁-USY is the weight loss of decomposed adsorbed water. This weight loss is about 2.56, 8.30, 12.13, 17.54 and 1.23 wt%, respectively. The weight loss of the second stage changed by about 1.20, 4.44, 8.22, 6.78 and 0.7 wt%, respectively, mainly due to the removal of the adhesive and the aggregation of silanol groups to form siloxane bond condensation. The total weight loss at 500 °C is 3.76, 12.74, 20.35, 24.32 and 1.12 wt%, respectively. The weight loss of each stage is summarized in Table S2. Therefore, the 10-Fe-USY and 10-Fe₁Mn₁-USY in this study has the best thermal stability and the operating temperature will not affect the structure of the 10-Fe-USY and 10-Fe₁Mn₁-USY material. These results are similar to those of Tsai [27]. The main reason is that the content of Si in USY is quite high (Si/Al: 10-Fe-USY > 10-Fe₁Mn₁-USY > 10-Fe-HY > 10-Fe-NaY > 10-Fe-ZSM-5), therefore, it does not easily lose weight due to high temperature.

Fig. 2

Thermal stability analysis results for different iron materials

To understand the microstructure and morphology characterization of the USY zeolite, 10-Fe-USY, 10-Mn-USY, and 10-Fe₁Mn₁-USY materials, SEM was used to observe their appearance. The results show that the USY zeolite appears to comprise irregular quadrilateral particles with a smooth surface, as shown in Fig. 3a. The surface of the 10-Fe-USY, 10-Mn-USY, and 10-Fe₁Mn₁-USY is rough with fine particles, as shown in Fig. 3b-d. According to EDS analysis results, these particles are iron or manganese (Fig. S1). This demonstrates that the iron-manganese-containing composite materials have been successfully prepared in this study. The results of the analysis show that 10-Fe₁Mn₁-USY contains more iron and manganese than 10-Fe-USY and 10-Mn-USY, as shown in Table 1. The main reason is the synergistic relationship of Fe and Mn, which makes it easier for Fe and Mn to form oxides on the surface of the material.

Fig. 3

SEM image of adsorption material: a USY Zeolite, b 10-Fe-USY, c 10-Mn-USY and d 10-Fe₁Mn₁-USY

Table 1 EDS analysis results of iron and manganese composites

To understand the redox properties of ferromanganese material, this study conducted an H₂-TPR analysis and thermal conductivity detector (TCD), with the results shown in Fig. 4. The 10-Fe-USY material exhibits two reduction peaks. The first peak at 384 °C represents the reduction of Fe₂O₃ to Fe₃O₄, and the second peak at 496 °C represents the reduction of Fe₃O₄ to FeO. The 10-Mn-USY material also exhibits two reduction peaks, the first 354 °C corresponding to the reduction of MnO₂ to Mn₂O₃, and the second at 459 °C corresponding to the reduction of Mn₂O₃ to Mn₃O₄. There are four reduction peaks with the bimetallic 10-Fe₁Mn₁-USY material. The first peak at 303 °C is the reduction of MnO₂ to Mn₂O₃, the second at 320 °C is the reduction of Fe₂O₃ to Fe₃O₄, the third at 442 °C is the reduction of Mn₂O₃ to Mn₃O₄, and the fourth at 484 °C corresponds to the reduction of Fe₃O₄ to FeO. These results are similar to those of Qin et al. [28] who also observed that 10Fe-15Mn-Al₂O₃ has four peaks at 310, 380, 530 and 580 °C, mainly due to the interaction between Fe and Mn elements, meaning that the iron-manganese-containing material improves the redox performance and reduces the reduction temperature. Therefore, the use of iron-manganese-containing material can treat acetone gas at low temperatures and improve the conversion efficiency.

Fig. 4

H₂-TPR analysis results for different materials

To investigate the contribution, chemical state and electronic structure of the prepared iron-manganese-containing material, XPS measurements were performed on 10-Fe-USY, 10-Mn-USY and 10-Fe₁Mn₁-USY. Figure S2(a) shows the peaks of the 10-Fe-USY fitting results. The peak at 710.5 and 723.9 eV are the binding energies of Fe²⁺, and the peak at 713.4 and 727.1 eV are the binding energies of Fe³⁺. Figure S2 (b) shows the peaks of the 10-Mn-USY fitting results. The peaks at 641.8 and 654.4 eV are the binding energies of Mn³⁺, while the peak at 644.7 eV is the binding energy of Mn⁴⁺. Figure S2 (c) shows the peaks of the 10-Fe₁Mn₁-USY fitting results. The peaks at 710.6 and 723.6 eV are the binding energies of Fe²⁺, while the peaks at 713.4 and 726.9 eV represent the binding energies of Fe³⁺. The peaks at 642.3 and 645.2 eV are the binding energies of Mn³⁺, and 654.6 eV is the binding energy of Mn⁴⁺, as shown in Fig. S2 (d).

We integrated the results of the peak area to calculate the proportion of the valence states of each element. The results, presented in Table S3, show that the Fe³⁺/Fe²⁺ ratios of 10-Fe-USY and 10-Fe₁Mn₁-USY are 0.79 and 0.88, respectively. The Mn⁴⁺/Mn³⁺ ratios of 10-Mn-USY and 10-Fe₁Mn₁-USY are 0.75 and 0.84, respectively. The O_latt/O_ads ratios of 10-Fe-USY, 10-Mn-USY and 10-Fe₁Mn₁-USY are 1.16, 1.23 and 1.51, respectively. The proportion of Fe³⁺, Mn⁴⁺ and O_latt valence states that contribute to the acetone reaction with bimetallic catalyst 10-Fe₁Mn₁-USY is higher than that with single metal catalysts (10-Fe-USY, 10-Mn-USY). This is largely because the interaction between Fe and Mn is beneficial in increasing electron transfer and promoting vacancies, further increasing the ratio of Fe³⁺, Mn⁴⁺ and O_latt containing bimetallic catalysts. That in turn improves the catalyst's ability to convert acetone. The results of this study are similar to those of Qin et al., who used iron and manganese metals modified onto γ-Al₂O₃ [28]. Through XPS analysis, it was found that the ratios of Fe³⁺/Fe²⁺, Mn⁴⁺/Mn³⁺ and O_latt/O_ads were 0.89, 0.86 and 1.30, respectively, which was similar to the 10-Fe₁Mn₁-USY prepared in this study [28]. The ratios of Fe³⁺/Fe²⁺, Mn⁴⁺/Mn³⁺ and O_latt/O_ads are 0.88, 0.84 and 1.51, respectively.

In this study, nitrogen isothermal adsorption and desorption instrument was used to understand the surface area, pore size, and pore volume of different materials. The surface area of 10-Fe-USY, 10-Mn-USY, and 10-Fe₁Mn₁-USY was 709, 398, 422, and 444 m² g⁻¹, respectively. The pore volume of the materials was 0.41, 0.30, 0.32, and 0.46 cm³ g⁻¹, respectively. The pore size of the materials was 2.41, 2.98, 2.85, and 2.95 nm, respectively. Results show that the modification of material by Mn/Fe metal will reduce its surface area and pore volume. This is because these metal ions fill in the pore walls and surfaces of the material, as shown in Table 2. The specific surface area of the 10-Fe₁Mn₁-USY materials prepared in this study is higher than that of 1Fe2MnC material (135 m² g⁻¹) prepared by Chen et al. [29].

Table 2 Specific surface area and pore size analysis results for different materials

3.2 Effect of different materials on acetone adsorption capacity

To understand the adsorption performance of different materials, the adsorption capacity of acetone was evaluated under the condition of containing water vapor (50%), and the adsorption materials of AC, ZSM-5, NaY zeolite, HY zeolite, and USY zeolite were compared at an adsorption temperature of 298 K, 0.18 s residence time, and 800 ppm inflow concentration. The breakthrough time is defined as the time when C/C_e is 0.1, and the saturation time is defined as the time when C/C_e is 0.9. The experimental results show that the breakthrough time of AC, ZSM-5, NaY zeolite, HY zeolite, and USY zeolite were 30, 35, 40, 45, and 50 min, respectively, and the saturation times were 75, 95, 110, 160, and 175 min, respectively. The capacities were 25.6, 44.3, 45.0, 59.0, and 65.3 mg g⁻¹, respectively, as shown in Fig. 5. Although AC has a high specific surface area (1551 m² g⁻¹), the affinity of AC with water is higher than that of acetone, which leads to the early adsorption of AC to reach saturation, lowering its adsorption capacity. However, USY zeolite has a higher Si/Al ratio (Si/Al = 10.7) and a larger contact angle (θ_c = 130°), resulting in good hydrophobicity of the material. Therefore, USY zeolite has a late adsorption penetration and reaches saturation later, giving it better adsorption capacity. Hsu [30] used Y-type zeolite and ZSM-5 molecular sieve to adsorb acetone, with adsorption capacities of 63.3 and 52.5 mg g⁻¹, respectively, showing that material containing ZSM-5 zeolite had poor adsorption capacity for acetone.

Fig. 5

Adsorption of acetone of different materials: a Removal efficiency, and b Adsorption capacity

3.3 Effect of humidity conditions on acetone adsorption capacity

In this study, to understand whether the hydrophobic 10-Fe₁Mn₁-USY will affect its ability to adsorb pollutants in the state of water gas, the adsorption capacity of 10-Fe₁Mn₁-USY under dry and wet conditions was tested. The acetone adsorption tests were compared between dry (RH = 5%) and wet (RH = 50%) influent gases at an adsorption temperature of 298 K, a residence time of 0.18 s, and an influent acetone concentration of 800 ppm. Figure 6 shows that when the inflow gas is dry, the breakthrough time is 55 min, the saturation time is 180 min, and the adsorption capacity is 146 mg g⁻¹. When it is wet, the breakthrough time is 50 min, the saturation time is 175 min, and the adsorption capacity is 133 mg g⁻¹. The results showed that RH had no obvious influence on the acetone adsorption performance, which suggests that this material has good hydrophobic ability. Su et al. [4] has shown the humidity resulted in the pronounced decrease in the adsorption contribution to AC materials. However, adsorption contribution of 10-Fe₁Mn₁-USY material is less affected by humidity in this study.

Fig. 6

Effects on acetone adsorption under dry and wet conditions: a Removal efficiency, and b Adsorption capacity

3.4 Effect of different influent concentrations on acetone adsorption capacity

To understand the effect of different influent concentrations (400, 600, 800, and 1000 ppm) on the adsorption capacity of acetone in this study, 10-Fe₁Mn₁-USY material was used to conduct adsorption tests under the conditions of an adsorption temperature of 298 K and a residence time of 0.18 s. Figure 7 shows that as the inflow concentration increases, the adsorption penetration time and adsorption saturation time both decline. The penetration times were 120, 80, 55, and 35 min, respectively, while the saturation times were 220, 195, 180, and 155 min, respectively. The adsorption capacities were 116, 134, 146, and 157 mg g⁻¹, respectively, when the concentrations were 400, 600, 800, and 1000 ppm. These results show that when the acetone inflow concentration is higher, the adsorption capacity of the material improves, but the penetration time and saturation time fall. The main reason is that with the increase of the acetone inflow concentration, the number of molecules that can be adsorbed by the material increases, raising the chance of contact between the material and acetone molecules, making it easy for the adsorption material to reach saturation earlier. The experimental results are similar to those of Hong et al. [31] and Ma and Ruan. [32]. As the inflow concentration of pollutants increases, the number of molecules that can be adsorbed by MCM-41 in the airflow also increases, resulting in a relative increase in the adsorption capacity of MCM -41, but the adsorption material will reach saturation earlier.

Fig. 7

Effects of different influent concentrations on acetone adsorption: a Removal efficiency, and b Adsorption capacity

3.5 Effect of different residence time on acetone adsorption capacity

In this study, to understand the effect of different residence times (0.18, 0.36, 0.54, and 0.72 s) on the adsorption capacity of acetone, 10-Fe₁Mn₁-USY material was used to conduct adsorption tests under the conditions of an adsorption temperature of 298 K and an acetone inflow concentration of 800 ppm. Figure 8 shows that as the residence time increases, the adsorption penetration time and adsorption saturation time also increase. The penetration times were 55, 90, 115, and 130 min, respectively, while the saturation times were 180, 245, 310, and 365 min, respectively, and the adsorption capacities were 146, 139, 124, and 107 mg g⁻¹, respectively, when the residence times were 0.18, 0.36, 0.54, and 0.72 s, respectively. These results show that when the acetone residence time is shorter, the adsorption capacity of the material rises, but the penetration time and saturation time fall. The main reason is that the surface pressure of the material increases, which increases the driving force for the acetone to enter the pores, which is conducive to the capture of acetone molecules, giving the 10-Fe₁Mn₁-USY material an increased adsorption capacity for acetone.

Fig. 8

Effect of different residence time on acetone adsorption: a Removal efficiency, and b Adsorption capacity

3.6 Effect of different temperatures on acetone adsorption capacity

This study examined the effect of different temperatures (298, 308, 318, and 328 K) on the adsorption capacity of 10-Fe₁Mn₁-USY material with an acetone inflow concentration of 800 ppm and a residence time of 0.18 s. The results are shown in Fig. 9. The adsorption capacity of 10-Fe₁Mn₁-USY material for acetone gradually fell as temperature rose. The penetration times were 45, 40, 35, and 30 min, respectively, and the adsorption capacities were 146, 136, 119, and 109 mg g⁻¹, respectively. The penetration time is earlier and the adsorption capacity is reduced when the temperature increases. The main reason is that under high temperature adsorption, the acetone gas is prone to violent collisions, meaning that the surface adsorption sites of the 10-Fe₁Mn₁-USY material cannot effectively capture the acetone gas, resulting in the rapid passage of the acetone gas through the adsorption material. These results are similar to those of Tsai [27], who found that a lower reaction temperature is beneficial to the adsorption reaction.

Fig. 9

Effect of different temperatures on acetone adsorption: a Removal efficiency, and b Adsorption capacity

3.7 Adsorbent material regeneration test

To determine whether the 10-Fe₁Mn₁-USY material can be reused, the adsorbed saturated material was desorbed and regenerated at high temperature (473 K, heating rate: 5 °C min⁻¹) under air condition. In this study, the adsorption capacity of 10-Fe₁Mn₁-USY before adsorption and after three cycles of desorption and regeneration were compared at an adsorption temperature of 298 K, a residence time of 0.18 s, and an influent concentration of 800 ppm. The results are shown in Fig. 10. The adsorption capacity of the fresh adsorbent was 146 mg g⁻¹, and the adsorption capacities of the materials after the first, second, and third desorption and regeneration were 123, 110, and 108 mg g⁻¹, respectively. The adsorption capacity of 10-Fe₁Mn₁-USY material decreased by about 27% after three cycles of desorption and regeneration. The adsorption behavior of 10-Fe₁Mn₁-USY material is a form of reversible physical adsorption behavior, making it a regenerable adsorbent. The reason for the decrease in adsorption capacity is that during the heating and desorption process, the 10-Fe₁Mn₁-USY material is lost due to the heating of the material surface, resulting in the loss of adsorption sites.

Fig. 10

Comparison of acetone adsorption capacity of materials after adsorption and desorption: a Removal efficiency, and b Adsorption capacity

3.8 Isothermal adsorption model simulation results

Langmuir, Freundlich, and Temkin models have been proven to be the most reliable and widely used in adsorption systems. Therefore, Langmuir, Freundlich, and Temkin models were used to perform the adsorption capacity of ferromanganese material in this study. The experimental results are fitted to the models and the most appropriate model is determined to estimate the maximum amount of adsorption associated with ferromanganese material. In addition, the simulation results can be judged whether the adsorption reaction is physical adsorption or chemical adsorption. The saturated adsorption capacity of 10-Fe₁Mn₁-USY at an adsorption temperature of 298 K and a residence time of 0.18 s and inflow concentrations of 400, 600, 800, and 1000 ppm was substituted into the Langmuir, Freundlich, and Temkin models to obtain the isothermal adsorption model of 10-Fe₁Mn₁-USY for acetone. The linear regression results of the Langmuir model, the Freundlich model, and the Temkin model are shown in Fig. S3. The regression correlation coefficients are 0.999, 0.998, and 0.999 respectively. The parameters of the correlation isotherm adsorption formula are summarized in Table 3 and Fig. S4. The actual adsorption capacity of various acetone concentrations was compared with the predicted adsorption capacity of the Langmuir, Freundlich, and Temkin models, when using 10-Fe₁Mn₁-USY, as shown in Table S4. The results show that the regression correlation coefficient of the Temkin model is relatively high (R² = 0.999), and the calculated value of the Temkin model is very close to the actual value. It can be concluded that the adsorption of acetone using 10-Fe₁Mn₁-USY is more consistent with the Temkin model. The adsorption behavior of the 10-Fe₁Mn₁-USY for acetone is both monolayer and multi-layer adsorption simultaneously.

Table 3 Parameters for isothermal adsorption analysis of acetone by Langmuir, Freundlich, and Temkin models

3.9 Adsorption kinetic model simulation results

The purpose of the kinetic adsorption model simulation is to understand the change of the adsorption capacity of acetone with the adsorption time, and to verify the possible adsorption mechanism by the model operation of the empirical formula. In this study, two adsorption kinetic models, the pseudo-first-order adsorption kinetic model and the pseudo-second-order adsorption kinetic model, were used to simulate and discuss the adsorption kinetics of the 10-Fe₁Mn₁-USY for acetone, and the most suitable adsorption kinetic model was obtained. The adsorption capacity per unit time of 10-Fe₁Mn₁-USY at an adsorption temperature of 298 K, a residence time of 0.18 s, and acetone concentrations of 400, 600, 800, and 1000 ppm was substituted into the adsorption kinetic model, and the appropriate kinetic model evaluated. Table S5 shows the linear regression parameters of the adsorption kinetic model. The estimation results show that the correlation coefficient R² value of the pseudo-first-order kinetic model ranges from 0.886 to 0.907 and the reaction rate (k₁) is about 0.040 – 0.09 min⁻¹. The correlation coefficient R² value of the pseudo-second-order adsorption kinetic model is in the range of 0.523 – 0.862 and the reaction rate (k₂) is about 1.2 × 10^–7 – 1.7 × 10^–5 g mg⁻¹ min⁻¹. The kinetic analysis results show that the k₁ value has limited differences under various concentrations, but the k₂ value has larger differences, which indicates that the acetone adsorption kinetic behavior is more suited to the pseudo-first-order adsorption kinetic model, as shown in Fig. 11 a and b

Fig. 11

Analysis results of (a) pseudo-first-order and (b) pseudo-second-order adsorption kinetic models and (c) internal diffusion models

To understand the internal diffusion during the adsorption of acetone using 10-Fe₁Mn₁-USY material, an analysis of its adsorption behavior was performed using the intra-particle diffusion model. Adsorption capacities of 10-Fe₁Mn₁-USY material at acetone concentrations of at 400, 600, 800, and 1000 ppm, an adsorption temperature of 298 K, and a residence time of 0.18 s were analyzed. Taking the adsorption capacity (q_t) as the ordinate, and taking the root of the adsorption time (t) as the abscissa, and performing the regression analysis, q_t and t^0.5 were obtained, as shown in Fig. 11c and Table S6. Figure 11c shows that the adsorption behavior curve has a three-stage curve. The first curve is the film diffusion rate (K_Fd), the second curve is the intra-particle diffusion rate (K_Id), and the third curve is the equilibrium phase diffusion rate (K_Ed). The reaction rate constant of diffusion in the process of the 10-Fe₁Mn₁-USY material adsorbing acetone may then be calculated. In this study, the K_Fd and the K_Id increased with an increase in the acetone concentration. These results indicated that the diffusion of acetone controlled the adsorption rate. These results are similar to those of Shiue et al. [33] and Shiue et al. [34] who found that in the adsorption of toluene using granular AC, the intra-particle diffusion constant also increased as the toluene concentration increased from 10 to 70 ppm.

3.10 Thermodynamic model simulation results

To investigate the activation energy of 10-Fe₁Mn₁-USY for acetone adsorption reaction at different temperatures, the pseudo-first-order kinetic model was used to calculate the adsorption reaction rate constants of 10-Fe₁Mn₁-USY at 298, 308, 318, and 328 K (k₁; 0.00067, 0.00087, 0.0013, and 0.0016 s⁻¹, respectively). The natural logarithm value of the adsorption reaction rate constant (lnk₁) was used as the ordinate and the reciprocal temperature (1/T) as the abscissa. Figure S5 was obtained after the regression analysis. Results show that the activation energy (Ea) is 24.9 kJ mol⁻¹ and the collision frequency factor (A) is 15.3 s⁻¹, as shown in Table 4.

Table 4 Arrhenius equation parameters of 10-Fe₁Mn₁-USY and comparison with other material

In addition, the thermodynamic properties of 10-Fe₁Mn₁-USY for acetone adsorption at different temperatures were also investigated in this study. The changes in free energy (ΔG), enthalpy (ΔH) and entropy (ΔS) at the four temperatures of 298, 308, 318, and 328 K were investigated when the influent concentration was 800 ppm and the residence time was 0.18 s. The reciprocal temperature (1/T) was used as the abscissa and the natural logarithm value of the distribution coefficient (k_v) was used as the ordinate. Figure S6 was obtained after the regression analysis. Table 5 shows that the ΔG is negative, meaning that the adsorption reaction process is a feasible spontaneous adsorption reaction, and the ΔG value is between 0 and -20 kJ mol⁻¹, meaning that the adsorption is a form of physical adsorption. The ΔH value is negative (-2.2 kJ mol⁻¹), indicating that the adsorption process is an exothermic reaction. The ΔS value (25.2 J mol⁻¹ K⁻¹) is greater than 0, which shows that the adsorption process is not self-sequential, and the entropy of the adiabatic system will not automatically decrease.

Table 5 Thermodynamic model parameters of 10-Fe₁Mn₁-USY material

4 Conclusions

The 10-Fe₁Mn₁-USY material exhibits good adsorption capacity for acetone at a relative humidity of 50%. The main factors affecting the adsorption capacity are the contact angle, hydrophobicity, specific surface area, and Si/Al of the material. The adsorption capacity of 10-Fe₁Mn₁-USY rises with higher influent concentration and shorter residence time. Conversely, the adsorption capacity of 10-Fe₁Mn₁-USY material falls as the temperature increases. The isothermal adsorption and the kinetic adsorption behavior of the 10-Fe₁Mn₁-USY material for acetone is more suited to the Temkin isotherm adsorption model and the pseudo-first-order kinetic model. The adsorption of acetone by the 10-Fe₁Mn₁-USY material is dominated by intra-particle diffusion. Based on the thermodynamic analysis results, the adsorption behavior of the 10-Fe₁Mn₁-USY material for acetone is a form of physical adsorption, and a spontaneous and non-sequential reaction. TGA results show that 10-Fe₁Mn₁-USY materials have high stability under high temperature conditions, so future work projects can be extended to understand the conversion efficiency of 10-Fe₁Mn₁-USY materials to acetone.