Study on the catalytic mechanism and water resistance of Cu-Mn-Sn_x catalyst for CO elimination

Abstract

CO in coal mine underground spaces can easily cause casualties among miners. The humidity in coal mines is relatively high, and traditional Cu-Mn catalysts are prone to deactivation. Compared to traditional Cu-Mn catalysts, doping with Sn enhances the activity and water resistance of Cu-Mn catalysts. The Cu-Mn-Sn catalyst was tested using XPS, FTIR, and a self-made Activity Test Experimental System, with partial data processed using the Avantage software. The results show that Sn increases the number of active sites (Cu⁺) on the Cu-Mn catalyst and the proportion of surface adsorbed oxygen (O_ads) and surface lattice oxygen (O_lat) in O 1s, which is the reason for the improved activity of the Cu-Mn-Sn catalyst. Sn weakens the coordination bond between Cu²⁺ and H₂O, reducing the formation of CO₃²⁻, which is the reason for the improved water resistance of the catalyst.

Introduction

In the event of a mine fire or explosion within the subterranean confines of a coal mine, the concentration of CO gas undergoes a rapid escalation^1,2,3. Carbon monoxide (CO), a colorless and odorless toxic gaseous pollutant⁴, possesses a binding affinity with hemoglobin surpassing that of oxygen, thus significantly contributing to the poisoning and fatality risks for personnel underground^5,6,7.

Among the array of strategies for CO elimination, catalytic oxidation has garnered substantial interest from researchers owing to its elevated conversion rates and environmental compatibility^8,9,10,11. Notably, Cu-based catalysts have emerged as prominent contenders due to their commendable performance in facilitating the complete oxidation of CO (CO-COOX)^12,13,14. Mehta et al.¹⁵ explored diverse parameters such as magnetic stirrer speed, stirring duration, temperature variations, and pH levels during the synthesis process of low-temperature CO oxidation employing hopcalite catalysts. The study elucidated the influence of these parameters on the resultant catalyst’s activity. Furthermore, Dey et al.¹⁶ enhanced the dispersal capability of CuMnO_x catalysts by incorporating γ-Al₂O₃ as a supporting material. Another investigation conducted by Dey et al.¹⁷ involved the preparation of different hopcalite catalysts utilizing various methodologies (CP, SG, RG, I, and P), revealing that the co-precipitation method (CP) yielded the optimal catalytic performance towards CO at 80 °C. Additionally, Fan et al.¹⁸ adjusted the lattice structure of hopcalite (CuMnO_x) catalysts through modulation of the cerium content, thereby enhancing the catalyst’s performance.

One notable limitation of Cu-based catalysts in CO oxidation lies in their susceptibility to water vapor exposure¹⁹. The presence of H₂O can detrimentally impact the utilization of porous carrier materials by obstructing active sites, resulting in the formation of less reactive surface complexes on the catalyst’s surface⁸. Guan et al.²⁰ posited that during hydrothermal aging, Bronsted acid sites within the framework of Cu-SSZ-13 molecular sieve catalysts are vulnerable to degradation, leading to structural collapse. Consequently, a considerable proportion of active copper sites undergo conversion to copper oxide, often accumulating within the molecular sieve pores, thereby causing pore blockage and a consequent decline in catalytic performance. Li et al.²¹ and Zhou et al.²² further proposed that water molecules themselves and/or dissociated hydroxyl groups may occupy active sites on the catalyst surface, hindering reactant adsorption and impeding the initiation of the catalytic reaction cycle. However, in many subterranean passages of coal mines, humidity levels can be markedly elevated, often exceeding 95%^23,24,25, with damp air inevitably infiltrating, thereby compromising catalytic performance^8,26.

To enhance the water resistance of CO catalysts, Biemelt et al.²⁷ employed flame spray pyrolysis (FSP) to disperse a mixture comprising oleate, 2-ethylhexanoate, and nitrates onto surfactants, establishing a scalable production method for water-resistant hopcalite catalysts. Hazlett et al.²⁸ reported that water presence increased CO adsorption on Pd and Pt-Pd catalysts while decreasing it on Pt catalysts. Park et al.²⁹ designed a Co₃O₄@CuO-CeO₂ catalyst featuring an optimized core-shell structure, thereby enhancing hydrothermal durability and low-temperature CO oxidation performance. Chen et al.³⁰ modulated lattice distortion in Mn₃O₄ spinel through amorphous samarium oxide interface modulation, thereby facilitating lattice oxygen activation. Electron features and characterization outcomes indicated that complete filling of samarium shell orbitals shielded interface active sites from water interaction, resulting in a catalyst exhibiting outstanding water resistance.

Studies have revealed that doping Sn into conventional Cu-Mn catalysts can bolster CO catalyst water resistance^11,26,31,32. Nevertheless, prolonged exposure to moisture-rich conditions inevitably leads to water-induced carbonate formation, which can result in poisoning of the metal-oxide interface¹⁹. In light of these considerations, the present study incorporated Sn into the conventional Cu-Mn catalyst to enhance both its activity and water resistance. X-ray photoelectron spectroscopy (XPS) was employed in conjunction with Fourier transform infrared spectroscopy (FTIR) to investigate changes in the structural properties of the catalyst during CO catalytic oxidation, facilitating analysis of the characteristics of the Sn-doped Cu-Mn catalyst. Understanding the mechanism underlying the catalytic oxidation of CO and the effects of Sn on the water resistance of Cu-Mn catalysts lays a theoretical foundation for the utilization of Cu-Mn-Sn CO catalysts in coal mining applications.

Experimental

Preparation of catalysts

The preparation of the Cu-Mn-Sn_x catalyst primarily utilizes the co-precipitation method. Copper nitrate (Cu(NO₃)₂·3H₂O, aqueous solution, 0.58 mol L⁻¹), manganese nitrate (Mn(NO₃)₂, aqueous solution, 50 wt%), and tin(IV) chloride pentahydrate (SnCl₄·5H₂O, aqueous solution, 0.750 mol L⁻¹) are homogeneously mixed in the required experimental proportions. Then, they are added dropwise to a sodium carbonate (Na₂CO₃, aqueous solution, 2.5 mol·L⁻¹) solution, maintaining the temperature at 70 °C with vigorous stirring. Dilute nitric acid (aqueous solution, 2 mol L⁻¹) is dripped into the above mixture until the pH reaches 8.3. The mixture is then continuously stirred for 4 h under constant temperature at 70 °C. After this, the catalyst is washed, dried, and filtered, completing the catalyst preparation process. Preparation method followed the methodology outlined in our prior research³³.

Various samples were prepared with Cu: Mn: Sn mass fraction ratios of 1:2:0.158, 1:2:0.333, 1:2:0.529, 1:2:0.750, 1:2:1, and 1:2:1.304, denoted as Cu-Mn, Cu-Mn-Sn_0.158, Cu-Mn-Sn_0.333, Cu-Mn-Sn_0.529, Cu-Mn-Sn_0.750, Cu-Mn-Sn_1.000, and Cu-Mn-Sn_1.304, respectively. These samples were scrutinized to explore the influence of Sn on the catalytic mechanism and water resistance of the Cu-Mn catalyst in CO elimination.

In order to determine the actual atomic content of the catalyst, each sample was analyzed using an Agilent 5110 for ICP-OES (Inductively Coupled Plasma Optical Emission Spectrometry). The resulting elemental concentration of the solution was processed to obtain the proportion of each metal element in the total metal elements, as shown in Table 1.

Table 1 ICP-OES quantification results of the catalysts.

The catalyst’s water resistance of samples with varying water contents were assessed.

First, dry catalyst samples were precisely weighed and arranged into stainless steel trays, each containing 10 g. These trays, along with the samples, were then positioned within a sealed chamber furnished with a humidifier, thereby regulating the humidity level to 95%, simulating subterranean conditions. The samples were weighed at regular intervals until stabilization, signifying moisture saturation. Subsequently, the samples were transferred to centrifuge tubes. The experimental setup is depicted in Fig. 1, wherein trays 1–3 accommodate the samples, 4 denotes the humidifier, 5 represents the humidity sensor, and 6 signifies the humidity controller.

Fig. 1

Experimental device schematic diagram.

Using an electronic scale, weigh out three portions of the original sample (10 g each). First, all samples were dried at 110 °C until a constant weight was achieved, eliminating excess moisture. These dried samples were then stored in centrifuge tubes for subsequent use.

The samples within the centrifuge tubes were divided into three trays, distributed among trays 1, 2, and 3. These trays were then integrated into the experimental system depicted in Fig. 1 to establish samples with differing water contents. Based on experimental findings, the saturated water absorption capacity per unit mass of the sample (${m_2}$) was 0.10 g. The samples were withdrawn from the system upon reaching sample masses of 10 g, 17.51 g, and 21.94 g, respectively, thereby preparing samples with distinct water contents. Water content was calculated using the following formula ($\:1$):

$$\begin{array}{*{20}{c}} {w=\frac{{{m_1}}}{{{m_2}}} \times 100\% } \end{array}$$

(1)

where $\:w$ denotes water content (%),${m_1}$ denotes the mass of sample water absorbed before poisoning (g), and ${m_2}$ represents the mass of the water absorbed after complete poisoning (g).

This methodology facilitated the preparation of catalyst samples with varied moisture content for subsequent experiments.

Experimental methods and catalyst characterizations

The scanning electron microscope (SEM) images were taken using the ZEISS MERLIN Compact scanning electron microscope. Using the Micromeritics ASAP2020 fully automatic specific surface area and pore size analyzer, nitrogen adsorption isotherms (77 K, pressure range 0.05–0.25 MPa, P/P₀ = 0.99) were obtained, and the specific surface area was determined using the Brunauer-Emmett-Teller (BET) method.

The samples were analyzed through FTIR (Tensor37, Bruker, Germany) within resolution and measurement ranges of 0.5–1.0 cm⁻¹ and 400–4000 cm⁻¹, respectively. Before the experiment, the instrument was preheated for 30 min to stabilize electronic components and the light source. Measurements were conducted through the KBr pellet method, with 1.5 mg of sample and 150 mg of dry KBr powder weighed using a precision balance (with a precision of 0.1 mg), ground, and pressed into pellets. These pellets were subjected to transmission scans, yielding infrared spectra.

Surface oxygen activation performance of the catalyst was assessed through XPS under Al Kα radiation (1486.7 eV) with a 400 μm beam spot size. The obtained spectra were corrected against the standard C1s peak at 284.8 eV using Advantage software. The experimental setup comprised the following conditions: Al Kα radiation (0.6 eV) served as the X-ray excitation source, with a beam spot size of 400 μm. The analysis chamber maintained a vacuum of approximately 3 × 10⁻⁷ mbar, while the working voltage was set at 12 kV, with a filament current of 6 mA. A full spectrum scan was conducted with a pass energy of 150 eV and a step size of 1 eV, along with a narrow spectrum scan using a pass energy of 50 eV and a step size of 0.1 eV. Corrections were applied against the standard C1s peak at 284.8 eV to mitigate surface contamination.

Fig. 2

Activity test experimental system.

The catalyst used in the experiment is the same as that reported in our previous work³³. Details regarding the pore characteristics (Figs. S1, S2), surface microstructure (Fig. S3), and catalyst composition (Figs. S4, S5) are presented in the Supporting Information.

This article provides a brief description: Figs. S1 and S2 demonstrate that the addition of Sn increases the specific surface area of the catalyst, with the Cu-Mn-Sn_0.750 sample exhibiting the largest specific surface area. Figure S3 showcases the morphology of the Cu-Mn-Sn_0.750 sample, which exhibits a regular and uniform aesthetic. Figures S4 and S5 help us understand the surface composition and elemental valence states of the catalyst.

Catalytic reaction of CO oxidation

Within the reaction chamber of the Activity Test Experimental System, a 10 g sample was introduced, as depicted in Fig. 2. The prepared samples underwent assessment for their catalytic efficacy in CO oxidation. A continuous gas mixture, comprising 19.8 vol% O₂, 1 vol% CO, and N₂ to balance, flowed through a quartz tube at a constant rate of 80 mL/min. Gas composition analysis was conducted using an online gas analysis system provided by Shenzhen sentronictek Technology Co., Ltd., to determine the gas composition exiting the reactor. Before initiating the catalytic activity assessment, the reaction chamber containing the sample underwent evacuation and degassing. The CO elimination rate (R_CO, %) was computed using formula (2) to gauge catalyst performance.

$$\begin{array}{*{20}{c}} {{R_{CO}}=\frac{{{c_{in,CO}} - {c_{out,CO}}}}{{{c_{in,CO}}}} \times 100\% } \end{array}$$

(2)

where ${c_{in,CO}}$ denotes CO concentration at the inlet, and ${c_{out,CO}}$denotes the lowest concentration of CO at the outlet.

Catalytic activity and water resistance evaluation

Catalytic activity evaluation

The CO elimination rate of the samples was assessed using the Activity Test Experimental System, with a CO concentration of 1%. The catalytic oxidation performance of the Cu-Mn-Sn catalyst for CO is depicted in Fig. 3. The results indicated that the elimination rate followed the order: Cu-Mn-Sn_0.158 < Cu-Mn = Cu-Mn-Sn_0.333 = Cu-Mn-Sn_1.304 < Cu-Mn-Sn_1.000 < Cu-Mn-Sn_0.529 < Cu-Mn_0.750, suggesting that an optimal Sn content (0.750) can enhance Cu-Mn’s catalytic activity. This enhancement can be attributed to variations in the internal interactions of the catalyst, leading to increased oxygen vacancies and improved CO adsorption. The specific mechanism through which Sn promotes Cu-Mn in catalyzing CO will be explored in subsequent studies.

The samples were fully saturated with water (100% water content) and subsequently subjected to CO catalytic oxidation experiments to assess the catalyst’s water resistance. During the experimental phase, the saturated water samples containing Cu-Mn, Cu-Mn-Sn_0.333, Cu-Mn-Sn_1.000, and Cu-Mn-Sn_1.304 exhibited substantially reduced elimination rates, rendering them ineffective in CO elimination. In contrast, Cu-Mn-Sn_0.0475, Cu-Mn-Sn_0.529, and Cu-Mn-Sn_0.750, which exhibited higher activity in the dry state, retained some level of activity in the saturated water samples. Comparative analysis was conducted with 10 g of each hopcalite saturated water sample placed in a reaction chamber.

As depicted in Fig. 4, the hopcalite agent with saturated water content exhibited the lowest elimination rate, indicating severe deactivation upon water contact, with an elimination rate of only 15%. The increasing order of the elimination rate was hopcalite < Cu-Mn-Sn_0.0475 < Cu-Mn-Sn_0.529 < Cu-Mn-Sn_0.750, suggesting that Sn doping has a beneficial effect on enhancing the water resistance of Cu-Mn-type catalysts such as hopcalite. The specific mechanism underlying water resistance will be further investigated in subsequent studies.

Fig. 3

Elimination rates of catalysts with different Sn contents in a dry state.

Fig. 4

Elimination rates of catalysts with different Sn contents under saturated water conditions.

Effects of Sn on the catalytic process

During the CO elimination process, oxygen vacancies are consumed and replenished by O₂ molecules, following the Mars-Van-Krevlen mechanism²⁶. First, the catalyst adsorbs CO molecules, which then interact with lattice oxygen provided by the catalyst to produce CO₂ molecules, subsequently desorbed from the surface. Following CO₂ desorption, oxygen vacancies form on the catalyst surface, allowing O₂ molecules to enter and form adsorbed oxygen. This adsorbed oxygen binds to adsorbed CO molecules, leading to further CO₂ generation and subsequent desorption. This cycle involves the continuous filling of oxygen vacancies with O₂, completing the catalytic oxidation of CO. The influence of Sn doping on the catalyst’s catalytic oxidation performance was examined from four perspectives: CO adsorption sites, CO chemical adsorption properties, O₂ activation properties, and CO₂ desorption properties.

CO active sites

To explore the CO chemical adsorption sites of the Cu-Mn-Sn catalyst during CO catalytic oxidation, FTIR was employed to scan the spectra of the Cu-Mn-Sn_0.750 sample, exhibiting the highest CO elimination rate. The acquired spectra were compared with those of the sample before CO adsorption to pinpoint the adsorption sites.

The experimental procedure is outlined as follows: Firstly, 1 g of the Cu-Mn-Sn_0.750 sample was extracted for spectral scanning, resulting in Fig. 5a. Subsequently, the sample was placed on the CO oxidation experimental platform to carry out catalytic reactions (Fig. 2). CO gas with a concentration of 99.9% was continuously introduced for 20 min at a flow rate of 80 mL/min. Thereafter, the sample were purged with pure N₂ for 30 min to eliminate physically adsorbed gases before being retrieved for spectral scanning, resulting in Fig. 5b.

Fig. 5

FTIR spectra of Cu-Mn-Sn before (a) and after (b) CO adsorption.

The absorption peak of Cu ions is situated near 2100 cm^–1, indicative of its linear CO adsorption structure Fig. 5 displays the FTIR spectra of the Cu-Mn-Sn catalyst within the range of 2000–2200 cm^–1 before and after CO adsorption33. The Cu-Mn-Sn (CO) sample, post-CO absorption, showcases an absorption peak at 2109 cm^–1, falling within the 2100–2150 cm^–1 range and attributed to the CO absorption peak of Cu⁺, suggesting predominant Cu⁺-CO interaction during catalytic oxidation, i.e., Cu⁺-CO. Additionally, the Cu-Mn-Sn (CO) sample exhibits shoulder peaks at 2171 cm^–1 and 2057 cm^–1, respectively. These peaks, situated within the 2150–2200 cm^–1 and 2050–2200 cm^–1 ranges, correspond to the CO absorption peaks of Cu²⁺ and Cu⁰, albeit with low peak intensities, indicating minimal involvement of Cu²⁺ and Cu⁰ in CO adsorption.

As revealed by the XPS test results of the Cu 2p peak³³, indicating a Cu 2p binding energy of 933.62 eV, the primary form of Cu in the catalyst is Cu²⁺. However, the FTIR tests revealed that the CO absorption peak can be attributed to Cu⁺-CO interaction, suggesting substantial reduction of Cu²⁺ ions to Cu⁺ during the catalytic process. The presence of a minor amount of Cu⁰ indicated that a fraction of Cu²⁺ underwent further reduction to Cu⁰ ions. This can be attributed to the relatively strong interaction between Cu⁺ and Mn ions, hindering the reduction of Cu⁺ to Cu⁰.

CO adsorption property

The catalyst surface harbored numerous active sites for CO. Assessing the chemical adsorption performance of the catalyst toward CO gas was pivotal for understanding the influence of Sn on its CO oxidation performance. The catalyst’s CO chemical adsorption performance during catalytic oxidation was examined from a functional group perspective through FTIR spectroscopy to scan samples with varying Sn contents: Cu-Mn (CO), Cu-Mn-Sn_0.333 (CO), Cu-Mn-Sn_0.750 (CO), and Cu-Mn-Sn_1.304 (CO). The CO absorption peaks were compared, and the CO chemical adsorption performance was analyzed. The FTIR spectra of Cu-Mn (CO), Cu-Mn-Sn_0.333 (CO), Cu-Mn-Sn_0.750 (CO), and Cu-Mn-Sn_1.304 (CO) before and after CO adsorption are depicted in Fig. 6.

Fig. 6

FTIR spectra of samples with different Sn contents after CO adsorption.

Figure 6 illustrates the FTIR spectra of various catalysts after CO adsorption within the range of 2000–2200 cm^–1. All catalysts exhibited a prominent CO absorption peak around 2110 cm^–1, indicating predominant CO adsorption as Cu⁺. Additionally, a minor CO absorption peak at 2057 cm^–1 was observed, corresponding to CO adsorption by Cu²⁺. Furthermore, Cu-Mn and Cu-Mn-Sn_0.333 exhibited a slight CO absorption peak at 2171 cm^–1, attributable to CO adsorption by Cu⁰. Conversely, Cu-Mn-Sn_0.750 and Cu-Mn-Sn_1.304 did not exhibit a CO absorption peak attributed to Cu⁰, suggesting that increased Sn content cannot facilitate further reduction of Cu²⁺ to Cu⁰ but rather favored Cu²⁺ reduction to Cu⁺, enhancing their activity.

The CO absorption peak of the Cu-Mn catalyst is situated at 2113 cm^–1. Upon Sn doping, the absorption peaks of Cu-Mn-Sn_0.333, Cu-Mn-Sn_0.750, and Cu-Mn-Sn_1.304 shift to lower wavenumbers to varying degrees, approximately around 2109 cm^–1. This indicates that Sn addition strengthens the interaction between Cu-Mn oxides, reinforcing the Cu-C bond while weakening the C ≡ O bond energy, causing the CO absorption peak to shift towards lower wavenumbers. The wavenumber of the CO absorption peak remains consistent across samples with varying Sn contents, implying that Sn content does not influence the interaction between Cu-Mn oxides.

After Sn doping, the intensity of the absorption peaks for Cu-Mn-Sn_0.333, Cu-Mn-Sn_0.750, and Cu-Mn-Sn_1.304 varied to differing extents, with Cu-Mn-Sn_0.750 exhibiting the highest intensity, followed by Cu-Mn-Sn_0.333, and Cu-Mn-Sn_1.304 indicating the lowest intensity. Thus, an optimal Sn proportion increased active sites for CO adsorption on the catalyst’s surface, while excessive Sn can inhibit CO adsorption, resulting in reduced activity.

To quantitatively describe the effect of Sn addition on CO adsorption in the sample, CO temperature-programmed desorption (CO-TPD) experiments were conducted on the Cu-Mn and Cu-Mn-Sn_0.750 samples using the Xianquan TP-5080-B multipurpose adsorption instrument. In the experiment, 100 mg of the sample was placed in the reaction tube, heated at a rate of 10 °C/min from room temperature to 300 °C for drying and pre-treatment, purged with a 40 mL/min He gas stream for 1 h, cooled to 50 °C, then saturated with a 10% CO/He mixture gas (40 mL/min) for 1 h. After saturation, a He gas stream (40 mL/min) was used to purge for 1 h to remove weakly adsorbed CO on the surface. Finally, the temperature was raised to 700 °C at a heating rate of 10 °C/min in a He atmosphere for desorption, with the evolved gas being detected by a thermal conductivity detector (TCD).

As shown in Fig. 7, the Cu-Mn-Sn_0.750 and Cu-Mn catalysts exhibited distinct CO desorption peaks at 413 °C and 441 °C, respectively. Notably, the addition of Sn significantly shifted the CO desorption curve of the catalyst towards a lower temperature direction, indicating that the doping of Sn can significantly enhance the mobility of oxygen within the catalyst, a characteristic generally considered to improve catalyst performance. When the experimental temperature was raised to 700 °C, the CO desorption amount in the sample was close to zero, suggesting that the amount of CO desorbed during the CO-TPD experiment is equivalent to the amount adsorbed by the sample. After normalization and quantitative analysis, the adsorption capacity of Cu-Mn-Sn_0.750 reached 11.9728 mmol/g, while the adsorption capacity of Cu-Mn was 10.1127 mmol/g. A comparative analysis of these data shows that the CO desorption amount of Cu-Mn-Sn_0.750 is significantly higher than that of the unmodified Cu-Mn catalyst. This result also proves that the addition of Sn can increase the number of CO adsorption sites on the Cu-Mn catalyst, thus confirming that Sn doping is an important factor in enhancing catalyst activity.

Fig. 7

CO-TPD profiles of Cu-Mn-Sn_0.750 and Cu-Mn catalysts.

O₂ activation property

The oxygen activation performance on the catalyst surface and the influence of Sn doping on this aspect were investigated through XPS analysis on the four samples: Cu-Mn, Cu-Mn-Sn_0.333, Cu-Mn-Sn_0.750, and Cu-Mn-Sn_1.304. The XPS spectra of surface oxygen after carbon (C 1s) correction are depicted in Fig. 8.

Fig. 8

XPS spectra of surface oxygen for samples with different Sn contents. Cu-Mn (a), Cu-Mn-Sn_0.333 (b), Cu-Mn-Sn_0.750 (c), Cu-Mn-Sn_1.304 (d).

As illustrated in Fig. 8, both the unmodified Cu-Mn catalyst sample and the Sn-doped Cu-Mn-Sn_0.333, Cu-Mn-Sn_0.750, and Cu-Mn-Sn_1.304 catalyst samples exhibited a predominant O 1s peak at 529 eV. Additionally, a shoulder peak was observed at the same position across all catalyst samples, indicating the presence of various oxygen species with distinct chemical environments on the catalyst surfaces. These peaks represent the higher binding energy O1s peak attributed to surface adsorbed oxygen (O_ads) and the lower binding energy O1s peak attributed to surface lattice oxygen (O_lat).

For a deeper insight into the surface oxygen characteristics, the O1s peaks of these four samples were examined using Avantage software, and the O1s peaks into surface adsorbed oxygen (O_ads) and surface lattice oxygen (O_lat) were decomposed. The proportions of O_ads and O_lat in the O 1s peaks for the Cu-Mn, Cu-Mn-Sn_0.333, Cu-Mn-Sn_0.750, and Cu-Mn-Sn_1.304 samples with varying Sn contents are presented in Fig. 9.

Fig. 9

The ratio of adsorbed oxygen (O_ads) to lattice oxygen (O_lat) in samples with different Sn contents.

As indicated by Fig. 9, Sn doping enhances the presence of surface adsorbed oxygen (O_ads) on the catalyst, attributable to the atmosphere. This enhancement suggests that Sn doping augments the oxygen activation performance on the catalyst surface, thereby enhancing the catalyst’s activity. Moreover, for samples with different Sn contents (Cu-Mn-Sn_0.333, Cu-Mn-Sn_0.750, and Cu-Mn-Sn_1.304), a higher Sn content correlates with a greater presence of surface adsorbed oxygen (O_ads), implying a favorable effect of increased Sn content on the catalyst’s surface oxygen activation performance. However, as depicted in Fig. 8, surface adsorbed oxygen (O_ads) did not exhibit any direct correlation with catalyst activity. Therefore, an increase in surface adsorbed oxygen (O_ads) can enhance the catalyst’s activity to some extent but does not directly determine it.

CO₂ desorption performance

CO reacts with either lattice oxygen or adsorbed oxygen to form CO₂ through the Mars Van-Krevlen mechanism. After CO₂ desorption, the CO sites are reduced and continue to adsorb CO. The CO₂ desorption performance of the catalyst is crucial, as even if the catalyst surface has a large number of CO active sites, if the CO₂ desorption ability is weak, the catalytic performance may not reach its optimal state. Testing the catalyst’s adsorption and desorption performance for CO₂ gas is of great significance for exploring the factors influencing the activity differences of catalysts.

Using the experimental system depicted in Fig. 2, the reaction chamber temperature was maintained at 25 °C using a water bath. CO₂ and N₂ gas flows were controlled at 10 mL/min and 90 mL/min, respectively, through a flowmeter to achieve a 10% CO₂ concentration in the mixed gas entering the experimental system. Samples of Cu-Mn, Cu-Mn-Sn_0.0475, Cu-Mn-Sn_0.333, Cu-Mn-Sn_0.529, Cu-Mn-Sn_0.750, Cu-Mn-Sn_1.000, and Cu-Mn-Sn_1.304, each weighing 10 g, were placed in the reaction chamber, and CO₂ dynamic adsorption was conducted until the gas analyzer displayed a 10% CO₂ concentration, indicating saturation adsorption. Subsequently, the CO₂ gas valve was closed, and the N₂ flow was adjusted to 100 ml/min to study the catalysts’ desorption capacity at room temperature.

The catalyst’s CO₂ desorption rate evaluates its CO₂ desorption capacity, calculated as the ratio of the CO₂ desorption amount to the CO₂ adsorption amount. The CO₂ adsorption amount represents the proportion of CO₂ adsorbed by a unit mass of catalyst during the CO₂ adsorption process relative to the CO₂ in the reactant. The calculation formula is as follows (3):

$$\begin{array}{*{20}{c}} {{q_a}=\frac{{\mathop \smallint \nolimits_{0}^{{{t_{t1}}}} \left( {10 - {C_a}} \right) \cdot {l_{C{O_2}}} \cdot dt}}{m}} \end{array}$$

(3)

where ${q_a}$ denotes the CO₂ adsorption capacity per unit mass of the catalyst (mL/g), ${t_{t1}}$ denotes the total reaction time during the CO₂ adsorption process, ${C_a}$ denotes the CO₂ concentration varying function with time during the CO₂ adsorption process, $\:m$ denotes the catalyst mass, and ${l_{C{O_2}}}$ denotes the CO₂ gas flow rate.

The CO₂ desorption amount represents the amount of CO₂ desorbed from a unit mass of catalyst after N₂ purging, and the calculation formula is as follows (4):

$$\begin{array}{*{20}{c}} {{q_b}=\frac{{\mathop \smallint \nolimits_{0}^{{{t_{t2}}}} \left( {10 - {C_d}} \right) \cdot {l_{C{O_2}}} \cdot dt}}{m}} \end{array}$$

(4)

where ${q_b}$denotes the CO₂ desorption capacity per unit mass of catalyst (mL/g), ${t_{t2}}$denotes the total reaction time during the CO₂ desorption process (s), and ${C_d}$ represents the CO₂ concentration varying function with time during CO₂ desorption.

The CO₂ desorption rate was calculated according to formula (5) to evaluate CO₂ desorption capacity:

$$\begin{array}{*{20}{c}} {{y_{C{O_2}}}=\frac{{{q_a}}}{{{q_b}}}} \end{array}$$

(5)

where ${y_{C{O_2}}}$ denotes the CO₂ desorption rate.

Using Eqs. (3), (4), and (5), the CO₂ adsorption amount, CO₂ desorption amount, and CO₂ desorption rate for the Cu-Mn, Cu-Mn-Sn_0.333, Cu-Mn-Sn_0.750, and Cu-Mn-Sn_1.304 samples with varying Sn contents were calculated, allowing for comprehensive analysis of their CO₂ adsorption and desorption performance.

The CO₂ desorption rates of the different samples indicated that those of Cu-Mn, Cu-Mn-Sn_0.333, Cu-Mn-Sn_0.750, and Cu-Mn-Sn_1.304 were not completely desorbed. Consistent with the results reported by Shuai, the un-desorbed CO₂ combined with lattice oxygen on the catalyst surface to form CO₃²⁻, which reacted with metal elements present on the catalyst surface, generating stable carbonates that adhered to the surface, thereby affecting activity to some extent.

Figure 10a presents a quantitative analysis of the CO₂ adsorption amount for each sample, revealing a reduction in CO₂ adsorption amount with increasing Sn content. This decline can be attributed to the dominance of interactions with surface lattice oxygen during CO₂ adsorption, with the amount of lattice oxygen governing the adsorption capacity. As illustrated in Fig. 10a, the number of lattice oxygen sites decreased with increasing Sn content, resulting in diminished CO₂ adsorption capacity.

Fig. 10

Quantitative analysis of CO₂ adsorption amount (a) and CO₂ desorption amount (b) for each sample.

Conversely, the CO₂ desorption ability is closely related to the types of carbonates formed on the catalyst surface. Figure 10b demonstrates that Sn doping increased the CO₂ desorption rate, thereby enhancing the activity of the doped catalyst. This suggests that Sn doping reduced the bond energy of carbonates formed from CO₂, facilitating their desorption. This phenomenon likely contributed to the enhancement of catalyst activity. Nonetheless, the CO₂ desorption rate did not exhibit a linear correlation with activity, indicating that CO₂ desorption enhanced catalyst activity but did not play a deciding role.

Mechanism of CO catalytic oxidation and water resistance over Cu-Mn-Sn catalyst

Mechanism of CO catalytic oxidation

The catalytic mechanism of the Cu-Mn-Sn_0.750 catalyst in CO oxidation was investigated through gas analysis to monitor changes in CO, O₂, and CO₂ concentrations after introducing a mixed gas (1% CO, 19.8% O₂, 79.2% N₂) at 25 °C.

Figure 11 illustrates the temporal changes in O₂, CO, and CO₂ concentrations during the catalytic process. Based on these changes, the catalytic process can be categorized into stages I, II, III, and IV, elucidating the catalyst’s mechanism at a molecular level³⁵.

Fig. 11

Concentration changes of O₂, CO, and CO₂ gases during the catalytic process.

During stage I, CO concentration decreased notably, while O₂ and CO₂ concentrations increased. Throughout this stage, O₂ concentration gradually increased to a peak from its initial level, whereas CO concentration diminished rapidly without reaching its peak. The elevation in CO₂ concentration attributable to the initial catalytic reaction, where surface lattice oxygen [O²⁻] reacted with CO molecules, yielding CO₂. The O₂ concentration increased because a fraction of released free oxygen [O] molecules recombined to form O₂ molecules. This reaction can be depicted as in formulas (6–8).

$$\begin{array}{*{20}{c}} {CO+* \to C{O_{ad}}} \end{array}$$

(6)

$$\begin{array}{*{20}{c}} {\left[ {{O^{2 - }}} \right]+CO \to C{O_2}+\left[ O \right]} \end{array}$$

(7)

$$\begin{array}{*{20}{c}} {\left[ O \right]+\left[ O \right] \to {O_2}} \end{array}$$

(8)

In stage II, CO concentration continued to decline until it reached its minimum, while O₂ concentration rapidly decreased after peaking, and CO₂ concentration continued to increase after stage I. First, CO molecules adsorbed on the catalyst’s active sites directly engaged with surface lattice oxygen [O²⁻] to produce CO₂, followed by CO₂ desorption, generating vacancies on the catalyst’s active sites. Subsequently, in stage II, O₂ molecules infiltrated the catalyst’s surface, generating adsorbed oxygen (O_ads), which further catalyzed the oxidation reaction with adsorbed CO molecules to form CO₂. This process can be described using formulas (9) and (10).

$$\begin{array}{*{20}{c}} {{O_2}+{2^*} \to 2{O_{ad}}} \end{array}$$

(9)

$$\begin{array}{*{20}{c}} {{O_{ad}}+CO \to C{O_2}+{2^*}} \end{array}$$

(10)

In stage III, the CO concentration began to increase, indicating a weakening of the catalyst’s CO adsorption capacity as the reaction progressed. Simultaneously, the O₂ concentration continued to decrease, while the CO₂ concentration remained relatively stable with minimal fluctuations. As depicted in Fig. 11, during stage III, the CO concentration decreased from the initial 1–0.08%, resulting in a total CO reaction amount of 1.01 mmol. Conversely, the CO₂ concentration increased from 0 to 0.11%, yielding a total CO₂ generation amount of only 0.02 mmol. The ratio of the total change in CO₂ to CO substance was lower than the stoichiometric ratio, that is $\triangle n_{CO_{2}}/\triangle n_{CO}<1/1$, indicating incomplete CO₂ desorption. With the progression of the reaction, the incomplete desorption of CO₂ molecules intensified, occupying active sites on the catalyst’s surface and thereby weakening its CO adsorption capacity.

In stage IV, O₂ molecules persistently occupied the active sites, generating adsorbed oxygen (O_ads), and exhibited a declining trend. However, the CO₂ concentration increased, indicating enhanced CO₂ desorption capability. This enhancement stemmed from the weakened adsorption capacity of the catalyst’s active sites following the catalytic oxidation cycle, facilitating easier desorption of CO₂.

The catalytic CO process of the Cu-Mn-Sn catalyst adheres to the Mars-Van-Krevlen mechanism. Once CO is adsorbed on the active sites, it reacts with lattice oxygen [O²⁻], with O₂ replenishing the consumed lattice oxygen [O²⁻], thereby completing the cyclic catalytic oxidation process.

As suggested by Tang et al.³⁵, CuO serves as an oxygen acceptor, while MnO serves as an oxygen donor in the catalyst. When O₂ concentration is low, the high catalytic activity of CuMnOx in CO oxidation can be attributed to the resonance system: Cu²⁺—O²⁻—Mn³⁺$\:\rightleftharpoons\:$Cu⁺—□—Mn⁴⁺. In this study, the CO molecule adsorbed on Cu²⁺—O²⁻—Mn⁴⁺ to yield CO₂ molecules, while the O₂ molecule adsorbed on Cu⁺—□—Mn^{3+ 36}, with □ being the active site on the catalyst surface. Liu et al.²⁶ conducted redox property tests on CuMn-type catalysts containing Sn using H₂-TPR and O₂-TPD experiments. The results indicated changes in the valences of Cu and Mn ions but not Sn during the test, indicating that Sn did not participate in the redox reactions of the catalytic process but enhanced activity by strengthening the interaction between Cu and Mn oxides.

Given that the active sites for CO oxidation are predominantly supplied by Cu²⁺, it is observed that the majority of Cu²⁺ ions are reduced to Cu⁺ upon CO adsorption. The cyclic process of catalysis for CO oxidation by the catalyst is as illustrated in the Fig. 12. Initially, CO interacts with lattice oxygen (O_lat) on the catalyst surface, where the CO molecule adsorbs onto the Cu²⁺—O²⁻—Mn⁴⁺ site, leading to the formation of CO₂ molecules. Upon the desorption of CO₂, the catalyst surface reconstitutes oxygen vacancies (□). The oxygen vacancy (□) on the catalyst surface then adsorbs O₂, effectively restocking the oxygen deficit and yielding adsorbed oxygen (O_ads). This adsorbed oxygen subsequently engages in a reaction with CO, resulting in the production of additional CO₂. Following the desorption of CO₂, the catalytic oxidation cycle of the catalyst is completed, allowing for the ongoing and continuous catalysis of CO.

Fig. 12

CO catalytic oxidation reaction path on Cu-Mn-Sn_x catalyst.

Mechanism of water resistance

The effects of H₂O on the catalytic performance of the Cu-Mn-Sn catalyst were investigated using samples with the highest elimination rate, Cu-Mn-Sn_0.750, prepared with water concentrations of 100% and 60%, respectively. FTIR was employed to characterize the samples with different water contents after CO absorption, and their spectra were compared with those of the dry samples to assess the influence of H₂O on the catalyst’s CO adsorption performance.

To prepare the samples, 3 g of the most active sample Cu-Mn-Sn_0.750 was placed in an enclosed space, and the humidity was controlled using a humidity controller to maintain an average humidity level of 95% until the sample had absorbed water to its total water absorption capacity of 100% and 60%. Samples with varying water contents were then prepared. These samples were introduced into the reaction chamber of the experimental system depicted in Fig. 2, exposed to CO gas for 20 min, and subsequently purged with pure N₂ for 30 min to eliminate physically adsorbed gases. Finally, the samples were extracted and transferred to centrifuge tubes for subsequent experimentation. The FTIR spectra of the Cu-Mn-Sn_0.750 samples with different water contents after CO absorption are displayed in Fig. 13.

Figure 13 illustrates the FTIR spectra of Cu-Mn-Sn_0.750 samples with water contents of 0%, 60%, and 100% after CO absorption. These spectra reveal absorption peaks corresponding to lattice water, adsorbed oxygen [O²⁻], coordinated water, and CO₃²⁻ ions at 1638 cm⁻¹. With increasing water content, the absorption peaks of lattice water, coordinated water, and CO₃²⁻ ions intensify. This indicates that in a humid environment, water molecules combine with active sites to form surface hydroxyl groups, which then react with CO₂ produced during the catalytic process to form CO₃²⁻ ions^38,39. These ions may create stable carbonates with surface metals, impeding CO₂ desorption and occupying active sites for CO on the catalyst surface, thereby leading to catalyst deactivation. Additionally, higher water content correlates with a gradual decline in catalytic performance. The process of CO elimination by the hydrous catalyst is depicted as shown in Fig. 14.

Fig. 13

FTIR spectra of samples with different water contents after CO absorption.

Figure 15 presents the FTIR spectra of the Cu-Mn-Sn_0.750 samples with water concentrations of 0%, 60%, and 100% in the range of 2000–2200 cm⁻¹ after CO absorption. Across all samples with varying water contents, peaks indicating Cu⁺ adsorption of CO are observed. However, as water content increases, the CO absorption peak of Cu⁺ diminishes, indicating reduced CO adsorption capacity—a primary cause of catalyst deactivation in humid environments^39,40. Moreover, with increased water content, the disappearance of peaks indicating Cu²⁺ adsorption of CO suggests that high water content hinders CO adsorption by Cu²⁺, likely because of the formation of coordination compounds between Cu²⁺ on the catalyst surface and H₂O.

Fig. 14

The process of CO elimination by the hydrous catalyst.

Fig. 15

Effects of water content on the CO absorption peak.

To substantiate our findings, we conducted an H₂O-TPD (Temperature-Programmed Desorption) analysis to evaluate the hydrothermal stability of the catalyst using the Micrometrics AutoChem 2920-hiden HPR20 multipurpose adsorption instrument. A precise 0.05 g sample was loaded into a reaction tube and pre-treated in a 300 °C helium stream for 2 h, followed by cooling to 50 °C under a protective helium atmosphere. The pre-treated catalyst was then saturated in a 30 ml/min helium flow and passed through a 30 °C water bath for 1 h. Subsequently, the sample was purged with a 30 ml/min helium flow for 60 min to remove any physically adsorbed water at 50 °C. Thereafter, the H₂O-TPD was performed from 50 to 300 °C at a heating rate of 10 °C/min in a 30 ml/min helium flow. The results are depicted in Fig. 16.

Both the Cu-Mn and Cu-Mn-Sn catalysts exhibited a single H₂O desorption peak, indicative of a single strong adsorption site for water. Importantly, the introduction of tin (Sn) into the catalyst resulted in a significant reduction in the H₂O desorption peak for the Cu-Mn-Sn catalyst.

In conclusion, the FTIR and H₂O-TPD results collectively demonstrate that the incorporation of tin (Sn) markedly suppresses the adsorption of water on the catalyst surface. This finding elucidates the enhanced hydrothermal stability of the Sn-modified Cu-Mn catalyst.

Fig. 16

H₂O-TPD profiles of Cu-Mn-Sn and Cu-Mn catalysts.

Mechanism of catalytic oxidation of CO

The mechanism of catalytic oxidation of CO with Cu-Mn-Sn-type catalysts is depicted in Fig. 17.

Compared with traditional Cu-Mn-type catalysts, Cu-Mn-Sn-type catalysts, as per the catalytic oxidation mechanism, reveal the role of doped Sn. Sn doping shifts the main absorption peak of CO, Cu⁺-CO, to lower wavenumbers, indicating enhanced interaction between Cu-MnO_x, strengthened Cu-C bonds, increased bond energy, and weakened C ≡ O bonds. Consequently, the CO absorption peak shifts, granting the Cu-Mn-Sn-type catalyst greater chemisorption ability for CO. Sn doping also boosts O₂ activation, increasing adsorbed oxygen on the catalyst surface, thus facilitating catalytic oxidation reactions. Moreover, Sn doping enhances CO₂ desorption performance, ensuring cleaner CO active sites on the catalyst surface, conducive to CO and O₂ molecule adsorption.

Fig. 17

Mechanism of catalytic oxidation of CO with Cu-Mn-Sn-type catalysts.

The primary cause of catalyst deactivation in humid environments is the formation of coordination water with Cu²⁺. During catalytic oxidation, generated CO₂ reacts with adsorbed H₂O to form CO₃²⁻, leading to stable carbonate formation. This hinders CO adsorption and CO₂ desorption. Sn doping mitigates this by weakening the coordination bond between Cu²⁺ and H₂O, preventing Cu²⁺ and coordination water formation. Consequently, Sn doping inhibits CO₂ and coordination water combination, reducing CO₃²⁻ and carbonate formation, thereby enhancing catalyst activity in humid environments.

Conclusions

Traditional Cu-Mn catalysts doped with Sn exhibit enhanced catalytic oxidation performance and a degree of water resistance. This study investigated the dissolution mechanism of Cu-Mn-Sn-type catalysts, revealing that Sn doping enhances the interaction between Cu-MnO_x, strengthens Cu-C bonds, and improves CO chemisorption. Additionally, Sn doping promotes O₂ activation and CO₂ desorption, increasing adsorbed oxygen on the catalyst surface, thereby enhancing catalytic oxidation. In the context of water resistance, catalyst deactivation in humid environments can be attributed to two factors: first, the formation of coordination water by adsorbed H₂O with Cu²⁺, reducing active sites; second, the formation CO₃²⁻ through the reaction between CO₂ and adsorbed H₂O during dissolution, resulting in stable carbonates hindering CO adsorption and CO₂ desorption. Sn doping weakens the coordination bond between Cu²⁺ and H₂O, diminishing CO₃²⁻ formation and improving catalyst water resistance.