Development and Kinetic Study of Novel Denitrification Catalysts Based on C₃H₆ Reductant

Abstract

With the acceleration of industrialization, the demand for NO_x abatement is becoming increasingly urgent. Finding safer and more stable reducing agent replacements and efficient catalysts is crucial for selective catalytic reduction (SCR) industrial NO_x abatement. Low-temperature hydrocarbon-assisted NO_x reduction (HC-SCR) remains attractive for industrial abatement. A series of industrial-grade TiO₂ support catalysts modified with a bimetallic MnCe active component, represented as TiO₂-ig, was prepared by the impregnation method to test the NO conversion performance under a 200–400 °C window with C₃H₆ as a reducing agent, and the physical properties were characterized using the BET and XRF methods. Under the feed of 150 ppm NO, 150 ppm C₃H₆, and 3%O₂—the optimal composition—Mn₁₅Ce₁₀/TiO₂-ig catalyst exhibited the highest NO_x conversion of 77.3% among industrial-grade TiO₂ support catalysts, with the corresponding temperature reduced to 275 °C. Furthermore, a slight improvement in catalytic activity was observed upon changing the TiO₂ support type. The industrial-grade and nano-sized TiO₂ supports predominantly exhibited mesoporous structures, while the anatase TiO₂ support contained a greater proportion of macropores. A steady-state kinetic model constructed for Mn₁₅Ce₁₀/TiO₂-ig catalyst indicates that the NO reaction rate is independent of C₃H₆ and O₂ concentrations at 200 and 250 °C. At 300 °C, C₃H₆ inhibits the reaction, while both O₂ and NO promote it. Changes in activation energy and the pre-exponential factor suggest a mechanistic shift from adsorption-limited at lower temperatures to reaction-limited at higher temperatures. Overall, using industrial-grade TiO₂ with MnCe promoters delivers meaningful NO_x reduction in a low-temperature regime and provides kinetic insights relevant to process design for industrial C₃H₆-SCR.

1. Introduction

As a major air pollutant, nitrogen oxides (NO_x) pose significant threats to both the atmospheric environment and human health [1]. With the rapid pace of industrialization, NO_x emissions from industrial sources have increased, making the need for NO_x reduction more urgent [2,3]. Selective Catalytic Reduction (SCR) is one of the most widely applied technologies for industrial denitrification [4,5,6]. Although SCR using ammonia as a reductant is one of the most effective methods for removing NO_x, and its catalyst development and mechanism are relatively well-established, it still faces limitations in industrial applications. Ammonia is toxic and corrosive, and it easily volatilizes and diffuses, which can lead to ammonia slip, posing risks to both the environment and workers. Moreover, the special storage and transportation requirements for liquid and gaseous ammonia increase operational costs [7]. Additionally, ammonia can react with sulfur oxides in flue gases, forming sulfates that may poison and deactivate the catalyst, block pipeline structures, and shorten catalyst lifespan, thereby hindering the long-term, efficient operation of industrial deNO_x processes [8]. To overcome these limitations, hydrocarbons (HCs), including olefins, have gained attention as viable alternatives due to their cost-effectiveness and availability as by-products from petroleum refining [9,10,11,12,13]. Among them, propylene stands out as a promising reducing agent for SCR systems due to its low cost and the avoidance of harmful by-products during the reaction. Utilizing propylene for NO_x reduction not only promotes the recycling of refining by-products but also holds substantial potential for industrial applications, especially at lower temperatures.

In the development of SCR catalysts, metal oxide supports offer advantages over other materials, such as molecular sieves and carbon-based supports, due to their low cost and excellent hydrothermal stability [14,15,16,17]. Liu et al. [18] prepared Mn-modified α-Fe₂O₃ catalysts and conducted experimental and density functional theory (DFT) studies. Their findings showed that Mn doping enhanced the adsorption of NO, facilitating electron transfer from Fe to the π* orbital of Mn-N-O. The Mn doping redistributes the electron density, improving the flexibility of electrons on the Fe atom and promoting N-O bond cleavage. This study demonstrates that incorporating unoccupied d orbitals with appropriate energy and symmetry facilitates a d-π interaction between the dopant and the reactant, thereby significantly enhancing catalytic efficiency. Zhang et al. [19] prepared a series of V₂O₅/TiO₂ catalysts with varying ratios of TiO₂ rutile phase and found that a small amount of TiO₂ rutile phase significantly improved NO conversion at temperatures below 270 °C. The rutile phase of TiO₂ lowered the band gap, especially the conduction band level, promoting the formation of reduced V species and superoxide ions, which are crucial for low-temperature SCR reactions. Liu et al. [11] explored the mechanistic basis for the enhanced C₃H₆-SCR activity of Ag/Al₂O₃ upon trace Pd doping and investigated the associated structure-property relationship. Under dry conditions, the 2Ag/Al₂O₃ catalyst achieved 80% NO_x conversion across a temperature range of 400–525 °C, demonstrating excellent SCR performance. Zhang et al. [3] also prepared Fe³⁺-doped MnCeO_x catalysts using the citric acid method (FeMnCeO_x-CA) and the CTAB-assisted template method (FeMnCeO_x-ST). Their results showed that the NH₃-SCR performance of FeMnCeO_x-CA exceeded 90% NO_x conversion between 75 and 150 °C, reaching 94% NO_x conversion at 125 °C, which was nearly 10% higher than that of the MnCeO_x-CA catalyst. Furthermore, the formation of N₂O was significantly suppressed in FeMnCeO_x-CA between 125 and 200 °C, showing a higher N₂ selectivity compared to MnCeO_x-CA. It can be seen that metal oxide supports have great application prospects in SCR NO_x reduction systems.

Mn and Ce elements have attracted widespread attention from researchers due to their exceptional properties [10,15,20,21]. Feng et al. [22] investigated the catalytic activity of Ce₃₀In₁₅/HBEA in CH₄-SCR technology within the temperature range of 300–600 °C. The NO_x conversion increased with temperature, maintaining over 97% at 500–600 °C. Compared to other catalysts, Ce₃₀In₁₅/HBEA achieved a high NO_x conversion of 99% at 500 °C, under a remarkable GHSV of 100,000 h⁻¹. Kashif et al. [23] synthesized a Mn/Ga-PCH catalyst using montmorillonite as the support via the impregnation method. The 8Mn/3Ga-PCH catalyst (8 wt% Mn and 24 wt% Ga) achieved 95% NO conversion at 300 °C and maintained over 70% conversion at temperatures between 150 and 250 °C. The enhanced low-temperature performance was attributed to the presence of active Mn, which increased surface acidity, promoted the formation of amorphous Ga and Mn phases, and influenced the Mn⁴⁺/Mn³⁺ ratio. Moreover, the catalytic activity followed the Mn³⁺/Mn⁴⁺ ratio, suggesting that high-valence manganese species are the active sites for NO conversion at low temperatures. Chen et al. [24] synthesized micron-sized spherical Ce_aMn_bO_x catalysts via the solvothermal method, achieving ultra-low-temperature NH₃-SCR deNO_x efficiency under challenging conditions, including SO₂ and high humidity. The optimized Ce₁Mn₇O_x-350 catalyst demonstrated outstanding performance, with over 91% NO_x conversion in the temperature range of 59–255 °C and over 78% N₂ selectivity below 140 °C. The Ce₁Mn₇O_x-350 microspheres, featuring the largest specific surface area and Mn₃O₄ as the dominant crystal phase, exhibited the lowest apparent activation energy (29.7 kJ/mol) for NH₃-SCR of NO, highlighting their optimal ultra-low-temperature deNO_x activity.

Catalytic kinetics research plays a pivotal role in advancing the understanding and optimization of SCR processes for NO_x abatement [25,26,27]. For industrial implementation of SCR technologies, especially those utilizing alternative reductants such as propylene (C₃H₆), it is crucial to establish reliable kinetic models that accurately describe the performance of catalysts over a wide range of conditions. Tan et al. [28] studied the kinetics of Mn-Ce oxide catalysts supported on nitrogen-doped graphene using the power law model. Their study found that at low temperatures, the reaction order with respect to NO was 1, the reaction order for NH₃ was zero, and the reaction order for O₂ was approximately 0.5. Under conditions with sufficient O₂, the apparent activation energy of the catalyst was 37.6 kJ/mol. Qin et al. [27] investigated the performance of iron-doped graphene catalysts with different nitrogen coordination structures in CO catalytic oxidation. Using density functional theory (DFT), they systematically analyzed reaction mechanisms, including Eley–Rideal (ER), Langmuir–Hinshelwood, and Termolecular Eley–Rideal, and calculated the reaction rate constants for each mechanism. Additionally, they derived the pre-exponential factor, reaction order, and activation energy for each elementary reaction step using the Arrhenius equation, providing data support for the development of a kinetic model. Tang et al. [25] used the transient and steady-state kinetics method to analyze the mechanisms of Cu/SAPO-34 for NH₃-SCR and found that the Cu/SAPO-34 catalyst with 4% Cu loading in the NH₃-SCR reaction process was mainly an E-R mechanism, and there was also an L-H mechanism.

Compared with high-purity TiO₂, industrial-grade TiO₂ (TiO₂-ig) offers cost advantages and a stable, large-volume supply. Despite extensive studies on MnCe/TiO₂, the industrial viability of C₃H₆-SCR using industrial-grade supports remains insufficiently established. A comprehensive study and kinetic modeling of propene-assisted SCR systems have yet to be fully explored, especially in the context of industrial-grade catalysts. This study developed an industrial-grade MnCe/TiO₂ catalyst for selective catalytic reduction (SCR) of NO at low temperatures in industrial applications. A series of MnCe-modified industrial-grade TiO₂ catalysts was synthesized, and the low-temperature deNO_x activity was investigated. The MnCe-modified catalyst demonstrated improved low-temperature deNO_x efficiency. X-ray fluorescence (XRF) and Brunauer–Emmett–Teller (BET) characterization techniques were used to analyze the physical properties of the catalyst. Additionally, a steady-state reaction kinetics model that resolves temperature-dependent rate orders was established for the catalyst formulation and indicates a mechanistic shift from adsorption-controlled to reaction-controlled regimes, supporting the engineering translation of catalytic systems into industrial-scale SCR systems.

2. Results and Discussion

2.1. Catalytic Performance

A series of catalysts was prepared using industrial-grade TiO₂ as the support with varying Mn/Ce ratios, and the NO conversion rates were evaluated to determine the optimal Mn/Ce loading concentrations. The catalytic activity test results are presented in Figure 1 and Figure 2. Based on the 5%Mn10%Ce catalyst, the NO conversion–temperature curves for catalysts with 2% higher and lower Mn and Ce loadings were tested. The results exhibited similar trends in conversion in that the NO conversion increased with temperature up to a certain point, after which it decreased as the temperature continued to rise. Under a constant 10% Ce loading, increasing the Mn loading by 2% to 7%Mn10%Ce enhanced the maximum conversion rate at 300 °C from 68% to 71.3%, with higher conversion than the 5%Mn10%Ce catalyst up to 300 °C. However, the conversion became similar above 300 °C. In contrast, when the Mn loading was reduced by 2% to 3%Mn10%Ce, the conversion rate decreased significantly across the entire temperature range, with the highest conversion temperature shifting to 350 °C. The maximum conversion was only 42.3%, indicating that Mn loading plays a critical role in the catalyst’s deNO_x activity. Under constant 5%Mn loading, increasing or decreasing Ce loading by 2% to prepare 5%Mn12%Ce and 5%Mn8%Ce catalysts, respectively, shifted the highest conversion temperature to 325 °C. The maximum conversion rates for these catalysts were lower than that of the 5%Mn10%Ce catalyst, at 65.5% and 64.9%, respectively. Therefore, the optimal Ce loading concentration is inferred to be 10%.

Based on previous findings, further increasing the Mn loading above 13% shifted the highest conversion temperature from 300 °C to 275 °C. As the Mn loading increased, the maximum conversion also increased, with the 15%Mn10%Ce catalyst reaching a maximum conversion of approximately 77.3%. After 300 °C, the conversion became similar when the catalysts had a loading of more than 5% Mn content. A comparison between the deNO_x activities of the 13%Mn10%Ce and 15%Mn10%Ce catalysts showed only a modest improvement in deNO_x activity across the entire temperature range. The catalytic activity of the 15%Mn10%Ce catalyst was further tested by changing the TiO₂ support to anatase and nano-sized TiO₂. The results are shown in Figure 3. The more expensive anatase TiO₂ support exhibited the highest NO conversion of 83.3%, which, compared to the nano-sized TiO₂ (78.7% deNO_x efficiency) and the cost-effective industrial-grade TiO₂ (77.3% deNO_x efficiency), showed only a modest improvement in overall performance.

2.2. Physical Property Analysis

2.2.1. Porosity Structure Analysis

BET characterization tests were conducted on five different catalysts: Mn₅/TiO₂-ig, Mn₅Ce₁₀/TiO₂-ig, Mn₁₅Ce₁₀/TiO₂-ig, Mn₁₅Ce₁₀/TiO₂-nano, and Mn₁₅Ce₁₀/TiO₂-anatase. The specific surface area and pore structure parameters, including the pore size distribution and adsorption–desorption curves, are shown in

Table 1. Specific surface area and pore structure parameters of different catalysts.

Catalysts	Specific Surface Area (m2/g)	Total Pore Volume (cm3/g)	Average Pore Diameter (nm)
Mn5/TiO2-ig	68.57	0.36	15.93
Mn5Ce10/TiO2-ig	63.03	0.29	14.06
Mn15Ce10/TiO2-ig	56.9	0.28	14.2
Mn15Ce10/TiO2-nano	48.66	0.17	10.76
Mn15Ce10/TiO2-anatase	16.84	0.18	42.61

and Figure 4. As observed in Table 1, modifications to the catalyst’s active components and their ratios, as well as the type of TiO₂ support, resulted in significant changes in the catalyst’s specific surface area, pore volume and pore diameter. The strong synergistic effect between Mn³⁺/Mn⁴⁺ and Ce³⁺/Ce⁴⁺ significantly changes the pore structure of the catalyst. As the loading of the metal active components, Mn and Ce, increases, the catalyst’s specific surface area gradually decreases, from 68.57 m²/g for Mn₅/TiO₂-ig to 56.9 m²/g for Mn₁₅Ce₁₀/TiO₂-ig. The pore volume also decreases from 0.36 m³/g to 0.29 m³/g and 0.28 m³/g. Compared to the industrial-grade TiO₂ support, both the nano-sized TiO₂ and anatase TiO₂ supports show a decrease in specific surface area and pore volume, with the reduction being most pronounced for the anatase TiO₂ support. Further analysis based on Figure 4 shows that the catalysts Mn₅/TiO₂-ig, Mn₅Ce₁₀/TiO₂-ig, Mn₁₅Ce₁₀/TiO₂-ig and Mn₁₅Ce₁₀/TiO₂-nano exhibit broad mesoporous distributions, primarily concentrated in the smaller pore size range below 25 nm. Among them, Mn₅/TiO₂-ig has the highest proportion of mesopores, followed by Mn₅Ce₁₀/TiO₂-ig, Mn₁₅Ce₁₀/TiO₂-ig and Mn₁₅Ce₁₀/TiO₂-nano. In contrast, the crystal structure of Mn₁₅Ce₁₀/TiO₂-anatase undergoes significant changes, exhibiting a structure dominated by macropores, with the coexistence of mesopores. The average pore width increases to 42.61 nm, as seen in Table 1. Such pore structure characteristics correlate with the BET surface areas reported in Table 1.

The adsorption–desorption curves of all five catalysts exhibit a type IV isotherm with an H3-type hysteresis loop [21]. From the perspective of catalytic activity, increases in specific surface area and pore volume generally contribute to providing more active sites and facilitating the diffusion of reactants. However, excessively small pore sizes may lead to increased mass transfer resistance. Based on the BET characterization results and experimental data, it can be inferred that the type and loading amount of active metals, as well as the type of TiO₂ support, can significantly influence the pore structure of the catalysts. Although the specific surface area gradually decreases across Mn₅/TiO₂-ig, Mn₅Ce₁₀/TiO₂-ig and Mn₁₅Ce₁₀/TiO₂-ig, the increase in the number of active metal sites compensates for the slight loss in surface area, thereby enhancing the overall catalytic performance.

2.2.2. Support Composition Analysis

To further understand the specific composition and constituents of the industrial-grade TiO₂ support, the XRF characterization results are shown in

Table 2. Relative content of each component in industrial-grade TiO₂.

Components	Content (%)
TiO2	98.237
SO3	1.312
P2O5	0.164
Cl	0.080
CaO	0.070
Nb2O5	0.060
SiO2	0.058
ZrO2	0.010
SrO	0.010

. The TiO₂ content in the support is 98.237%, with the remaining impurities primarily consisting of SO₃, P₂O₅, Cl, CaO, Nb₂O₅, SiO₂, ZrO₂, and SrO. Among these, SO₃ is the most prevalent impurity, which can easily form sulfates at elevated temperatures. This may lead to the coverage of active catalytic sites or the aggregation of active components, ultimately decreasing catalytic activity.

2.3. Kinetic Study in C₃H₆-SCR

2.3.1. Reaction Order Determination

The reaction orders of the individual reactants with Mn₁₅Ce₁₀/TiO₂-ig catalyst were determined by measuring the reaction rate under varying concentrations of NO, C₃H₆, and O₂ under 200 °C, 250 °C and 300 °C while keeping other conditions constant. To determine the reaction orders of each reactant gas, the natural logarithms of the reactant gas concentrations were plotted on the x-axis, while the natural logarithms of the NO conversion were plotted on the y-axis. A linear fit was applied to the data, and the slope of the resulting line represents the reaction order for each reactant gas. As shown in Figure 5, at 200 °C, the NO conversion remains constant despite variations in the concentrations of C₃H₆ and O₂, indicating a zero-order reaction with respect to both C₃H₆ and O₂ at this temperature, which is affirmed in Figure 5d,f. The reaction rate is independent of their concentrations. In contrast, the NO conversion rate increases linearly with the concentration of NO. As shown in Figure 5b, the linear fit of the NO data yields the equation of the line $y=-9.36+1.87x$, with a coefficient of determination of 0.96. Therefore, the reaction order with respect to NO at 200 °C is approximately 1.87.

At 250 °C, the relationship between the NO conversion rate and the concentrations of the reactant gases, along with the linear fit results, is shown in Figure 6. As seen in Figure 6a–f, the NO conversion still remains unchanged with variations in the concentrations of C₃H₆ and O₂, indicating that the reaction order with respect to C₃H₆ and O₂ is 0, which is consistent with the results observed at 200 °C. The NO conversion increases linearly with the concentration of NO. The linear equation for the NO conversion is $y=-7.89+1.61x$, with a coefficient of determination of 0.98. Therefore, the reaction order with respect to NO at 250 °C is approximately 1.61.

However, at 300 °C, the NO conversion under different concentrations of C₃H₆ and O₂ changed significantly compared to those at 200 °C and 250 °C, as shown in Figure 7. As the concentration of C₃H₆ increases, the NO conversion rate decreases linearly, suggesting that an excess of C₃H₆ competes with other reactants, thereby covering the active sites and leading to a negative reaction order. The natural logarithms of the NO conversion and C₃H₆ concentration were linearly fitted, yielding the equation $y=4.22-0.55x$, with a coefficient of determination of 0.99. Therefore, the reaction order with respect to C₃H₆ at 300 °C is −0.55. The shift in reaction order from 0 to −0.55 can be attributed to changes in the adsorption–reaction–desorption equilibrium of C₃H₆. At lower temperatures, the catalyst’s reaction is likely controlled by the adsorption process on the catalyst surface, with C₃H₆ adsorption reaching saturation, resulting in a constant reaction rate independent of C₃H₆ concentration, hence a zero-order reaction. At higher temperatures, the desorption rate of C₃H₆ may exceed the adsorption rate, reducing the concentration of reactants on the active sites of the catalyst surface. Furthermore, at 300 °C, the thermal decomposition of C₃H₆ may lead to coke formation, which can block some active sites, reducing the effective reaction sites for NO and C₃H₆ and consequently decreasing the reaction rate, thus lowering the reaction order. The linear fit of the reaction order with respect to O₂ at 300 °C is shown in Figure 7f, with the equation $y=-3.42+0.46x$ and a correlation coefficient of 0.99. The reaction order for O₂ increased from 0 to 0.46, which can be attributed to the enhanced catalytic activity at high temperature (300 °C), leading to an accelerated consumption of O₂ and, consequently, an increase in its conversion. And the linear equation for the NO conversion is $y=-3.84+1.03x$, with a coefficient of determination of 0.99.

2.3.2. Determination of the Apparent Activation Energy and the Pre-Exponential Factor

Since the reaction orders of C₃H₆ and NO are 0 at both 200 °C and 250 °C, while a change in the reaction order occurs at 300 °C, it is inferred that the reaction mechanism may have changed within the 250–300 °C temperature range. Therefore, the temperature range was divided into two segments: 200–250 °C and 250–300 °C. The fitted straight lines lnK and 1/T are shown in Figure 8 and Figure 9.

Within the temperature range of 200–250 °C, the reaction orders were determined to be zero for both C₃H₆ and O₂, whereas NO exhibited a reaction order of 1.74. Linear regression analysis against 1/T yielded an Arrhenius plot from which the pre-exponential factor was derived through intercept analysis. The optimized rate constant expression was obtained as $K=2.97\times e^{-{\displaystyle \frac{10376.8}{RT}}}$. Based on the established methodology, the activation energy for the 250–300 °C reaction regime was calculated as 45.2 kJ/mol, with a corresponding pre-exponential factor of 8266.8. The complete Arrhenius equation is expressed as $K=8266.8\times e^{-{\displaystyle \frac{45226.8}{RT}}}$.

The variation in activation energy and pre-exponential factor across two temperature ranges suggests a dynamic evolution in the surface reaction mechanism of the catalyst. In the low-temperature range (200–250 °C), the reaction is primarily governed by physical adsorption and weak chemisorption of the reactants, rather than the surface oxidation reactions activated at higher temperatures. Although the energy barrier to be overcome is lower in this range, the reactivity of the adsorbed species is limited due to the lower temperature, resulting in a reduced frequency of effective molecular collisions and a correspondingly smaller pre-exponential factor. The reaction order with respect to C₃H₆ and O₂ is zero, while that of NO is 1.74, indicating that O₂ and C₃H₆ have reached adsorption saturation on the catalyst surface. The dissociation or activation of C₃H₆ appears to be limited, with the adsorption rate exceeding the reaction decomposition rate. Both lattice oxygen and chemisorbed oxygen on the catalyst surface participate in the reaction, and the overall reaction rate is primarily controlled by the adsorption rate of NO.

In the temperature range of 250–300 °C, the adsorption and desorption processes gradually reach a dynamic equilibrium. The activation energy increases to 45.2 kJ/mol, indicating that the rate-determining step shifts from adsorption to surface chemical reactions, which require overcoming a higher energy barrier. The reaction order with respect to C₃H₆ is approximately −0.28, while those for NO and O₂ are about 1.32 and 0.23, respectively. The negative reaction order of C₃H₆ suggests that excessive concentrations may lead to the accumulation of hydrocarbon species on the catalyst surface, thereby inhibiting NO conversion. The decrease in NO reaction order implies that the rate of the deNO_x reaction is no longer solely controlled by NO adsorption but may instead be influenced by surface oxide species or reaction intermediates. At higher temperatures, lattice oxygen is consumed more rapidly, necessitating the replenishment of oxygen vacancies by O₂, which accounts for the observed change in O₂ reaction order. The kinetic behavior of the catalyst across different temperature regions reveals a mechanistic transition from NO adsorption control to surface redox process control, closely relating to competitive adsorption among reactants and the occurrence of side reactions.

3. Methods

3.1. Materials and Preparation

A series of Ti-based catalysts supported by varying Mn and Ce content was synthesized by the impregnation method. Taking MnxCey/TiO₂-ig catalysts as an example, x and y represent the mass fraction of active components, and ig indicates that the catalyst carrier is an industrial-grade powder, which was provided by Lanzhou Petrochemical Research Center of PetroChina Company Limited. All the corresponding metal nitrates were of analytical grade and from Macklin (Shanghai, China). The detailed preparation method has been published in a former research paper [29]. Firstly, the required mass of Mn(NO₃)₂·4H₂O and Ce(NO₃)₃·6H₂O, which needs to be calculated based on the mass proportion of the active component in the carrier and the proportion in nitrate, was added to deionized water. The Ce(NO₃)₃·6H₂O could not be added unless the former solution was stirred uniformly. The TiO₂-ig powder was then impregnated with the mixed solution and stirred at 80 °C until a uniform paste formed and no free liquid remained, followed by bath sonication for 30 min to break agglomerates. The paste was dried at 105 °C for 12 h, gently ground, and sieved to 80–120 mesh prior to calcination. Finally, the powder was calcined in flowing air at 500 °C for 3 h.

3.2. Catalytic Activity Test

The tested catalyst was placed onto the distribution plate within a fixed quartz tube reactor with an inner diameter of 20 mm. And the total gas flow rate of the inlet was 1200 mL/min. The catalytic activity between 200 °C and 400 °C at each 25 °C was measured by simulating the flue gas under the concentrations of [NO] = [C₃H₆] = 150 ppm, [O₂] = 3% and N₂ balance. The gas hourly space velocity (GHSV) was 8000 h⁻¹. The NO conversion was calculated by the following equation:

$$\mathrm{NO} \mathrm{conversion}={\displaystyle \frac{{[NO]}_{\mathrm{in}}-{[NO]}_{\mathrm{out}}}{{[NO]}_{\mathrm{in}}}}\times 100\%$$

(1)

In which [NO]_in represents the inlet concentration of NO, and [NO]_out represents the outlet concentration of NO.

3.3. Catalysts Characterization

The Brunauer–Emmett–Teller (BET) characterization was carried out by the Micromeritics ASAP 2460 multi-station fully automated specific surface and pore size analyzer (Norcross, GA, USA). X-ray fluorescence (XRF) utilizes high-energy X-rays to excite the inner electrons of atoms in a sample, causing them to be ejected from their atomic orbitals and creating inner-shell vacancies. To fill these vacancies, electrons from outer shells transition to the inner shells, releasing energy in the form of X-rays with specific wavelengths, known as fluorescent X-rays. By analyzing the wavelength of these fluorescent X-rays, the elements present in the sample can be identified. Additionally, the intensity of the fluorescence can be used for quantitative analysis of the element’s concentration. To determine the specific composition of the catalyst support provided by Lanzhou Petrochemical Research Center of PetroChina Company Limited, a quantitative analysis was performed using the BrukerS8 TIGER X-ray fluorescence spectrometer (Berlin, Germany).

3.4. Steady-State Kinetics

The apparent activation energy and reaction order for NO reduction were measured in the fixed-bed reactor. Under conditions where the reaction rate constant and effective diffusion coefficient remain unchanged, the internal diffusion efficiency factor approaches 1, which indicates that the reaction rate is predominantly controlled by intrinsic kinetics when the particle size is reduced to a certain range, with minimal influence from internal diffusion [30]. Consequently, controlling the particle size range can effectively mitigate the impact caused by internal diffusion. Previous studies have shown that when the particle size is smaller than 0.38 mm, internal diffusion in catalysts with a pore size volume of 0.24 cm³/g and an average pore diameter of 12.22 nm can be considered negligible [25]. In this study, the catalyst was selected with a particle size range of 80–120 mesh, corresponding to an average particle size between 0.12 mm and 0.18 mm. The total pore volume was 0.28 cm³/g, and the average pore diameter was 14.2 nm. As a result, the effect of internal diffusion can be considered negligible.

External diffusion refers to the process by which reactants diffuse from the gas phase to the external surface of catalyst particles, and products diffuse back into the bulk fluid phase from the catalyst surface [31]. Therefore, it is essential to eliminate the influence of external diffusion in catalytic reaction kinetic studies to ensure that the measured reaction rates reflect the intrinsic behavior of the catalyst rather than apparent rates limited by mass transfer. If not properly accounted for, external diffusion can lead to a reduction in the observed reaction order and apparent activation energy. Moreover, it may cause the reaction rate to be affected by factors such as flow rate or particle size, thereby distorting the derived kinetic parameters. Li et al. [32] eliminated the influence of external diffusion by using 0.3 g of catalyst under a total flow rate of 300 mL/min. According to previous studies [28,33], the effect of external diffusion can be considered negligible when the reaction has a large space velocity. Based on these reports, the present study employed 1 mL of catalyst with a total flow rate of 2000 mL/min, corresponding to a space velocity of 120,000 h⁻¹, to eliminate the influence of external diffusion.

The reaction rate equation is widely used in catalytic kinetics to describe the relationship between the reaction rate and reactant concentrations. In this study, the reaction rate equations describing the dependence of the rate on the concentrations of the reactants were provided in Equations (2) and (3). By taking the natural logarithm of both sides of Equation (2), Equation (4) is obtained.

$$r_{\mathrm{NO}}=k_{\mathrm{NO}}\cdot {[C_3{H}_6]}^a\cdot {[NO]}^b\cdot {[O_2]}^c$$

(2)

$$r_{\mathrm{NO}}={\displaystyle \frac{\eta C_{\mathrm{NO}}V}{22.4\mathrm{W}}}$$

(3)

$$ln r_{\mathrm{NO}}=ln k_{\mathrm{NO}}+aln[C_3{H}_6]+bln[NO]+cln[O_2]$$

(4)

In the equation, $r_{NO}$ is the reaction rate (μmol/(g∙min)); [C₃H₆], [NO] and [O₂] represent the initial concentrations of C₃H₆, NO, and O₂, respectively; a, b and c are the reaction orders of C₃H₆, NO, and O₂ with respect to the catalyst, respectively; η is the deNO_x efficiency; C_NO is the initial concentration of NO; V is the flue gas flow rate (L/min); W is the catalyst mass (g).

Based on the Arrhenius equation (Equation (5)) and the established relationship between the apparent reaction rate and NO concentration (Equation (6)), the pre-exponential factor (A) and activation energy (E_a) can be obtained through logarithmic transformation and linear fitting.

$$K=A{e}^{-{\displaystyle \frac{E_{\mathrm{a}}}{RT}}}$$

(5)

$$K=-{\displaystyle \frac{V}{W}}ln(1-\eta )$$

(6)

In these Equations: E_a is the apparent activation energy (kJ/mol); A is the pre-exponential factor; R is the ideal gas constant, taken as 8.314 J/(mol∙K); T is the reaction temperature (K).

In this study, the catalyst Mn₁₅Ce₁₀/TiO₂-ig, which had the best performance of all tested industrial catalysts, was selected to test the reaction order for steady-state kinetics by changing the concentration of different reaction gases at 200 °C, 250 °C, and 300 °C. The specific composition of the flue gas feed was 100–400 ppm C₃H₆ at each 50 ppm, 100–400 ppm NO at each 50 ppm, and 1–7 vol% O₂ at each 1 vol%.

4. Conclusions

Compared to traditional ammonia-based NO_x SCR, C₃H₆ SCR technologies over industrial-grade TiO₂-supported MnCe catalysts do not rely on ammonia as a reducing agent, offering a safer and operationally simpler alternative for industrial NO_x abatement. Refineries naturally contain hydrocarbons such as olefins, which offer a significant advantage for the implementation of olefin-based deNO_x technologies.

(1) Mn₁₅Ce₁₀/TiO₂-ig catalyst, the optimal Mn/Ce ratio, exhibits the best deNO_x efficiency among industrial-grade TiO₂ support catalysts, achieving a NO conversion of 77.3%. The temperature corresponding to the highest NO conversion rate is reduced to 275 °C, effectively enhancing the catalyst’s low-temperature deNO_x performance. Replacing TiO₂-ig with anatase yields only a modest improvement in peak conversion, indicating that industrial-grade TiO₂ is a technically viable support for HC-SCR with favorable cost and supply considerations.

(2) Catalysts supported on industrial-grade and nano-sized TiO₂ exhibit predominantly mesoporous structures, while the introduction of anatase TiO₂ support results in a shift toward macroporous structures with the coexistence of mesopores.

(3) Under low-temperature conditions (200, 250 °C), the NO reaction rate is independent of propylene and oxygen concentrations. At 300 °C, an increase in propylene concentration exhibits an inhibitory effect on the C₃H₆-SCR reaction, while both O₂ and NO show a promoting effect. The variations in activation energy and pre-exponential factor across the 200–250 °C and 250–300 °C temperature ranges indicate a shift in the reaction mechanism, from surface adsorption-limited at lower temperatures 200–250 °C to chemically reaction-limited at 250–300 °C.

Leveraging readily available refinery olefins (C₃H₆) avoids NH₃ storage/handling and potential slip, while industrial-grade TiO₂ provides a scalable, cost-effective support, promoting the practical viability of HC-SCR for industrial NO_x reduction.