Introduction

Hazardous gases in post-fire smoke pose great threat to human safety and task performances in typical environmental control and life support systems (ECLSS) such as spacecrafts, submarines, and naval ship compartments. By far, poisoning induced by the high-concentration CO is considered as the prime reason for life casualties in fire [1–3]. Binary copper manganese oxides composites have been extensively utilized for catalytic removal of CO from ECLSS [4–8].

Recently, copper manganese oxides catalysts have been reported to exhibit good activity for CO oxidation at ambient temperature [9–24]. The catalysts might show promise for removing CO from post-fire smoke, and this has not been extensively reported. One significant concern of using copper manganese oxides catalysts for CO depollution treatment from post-fire smoke is that their CO oxidation activities might be adversely affected by the high concentration CO2. There are some previous contributions highlighting the deactivation process of CO catalysts in the presence of CO2. Hoflund et al. reported that Au exhibited decreased CO oxidation activities when exposed in the atmosphere containing CO2 [25]. Liang et al. also found that CO2 had an adverse effect on the stability of Pd/CeO2-TiO2 [26]. Parinyaswan et al. reported that CO2 molecules could occupy the adsorption active sites over Pt–Pd/CeO2. The catalyst deactivation process was deduced as the fact that the migration of mobile oxygen in the redox process had been inhibited by CO2 molecules [27]. Furthermore, in the work of Wang et al., the inhibiting effect of CO2 on the activity of Pd–Cu catalyst was attributed to several reasons as the competitive adsorption between CO and CO2, the occlusion of adsorption active sites and the impediment effect on the transformation of intermediates [28]. Although catalyst deactivation for CO oxidation in the presence of CO2 has received much attention, more efforts on improving the CO oxidation performances and mitigating the catalyst deactivation in the presence of CO2 are necessary.

For supported catalysts, supporting materials with porous structures and basic/acid surface would affect their catalytic performances [29, 30]. Generally, surfaces, grooves, pores, and channels could be present in supporting materials, as shown in Fig. S1. The textural properties of the supports play significant roles in their CO oxidation performances. A large surface area and benign surface chemistry provide more active sites for adsorption. The dispersion state of active components over the surface of the supports also affect their CO oxidation behaviors. Grooves, pores, and channels play significant roles in that they provide space to accommodate more active components and offer contact zone for gas diffusion. Nevertheless, excessive loading of active components is expected to cause pore blockage and the covering of active sites, which are not conducive to enhanced CO oxidation performances. Moreover, the catalyst synthesis method influences the loading and dispersion of active components and therefore affect their CO oxidation performance. Furthermore, operating parameters such as CO concentration, CO2 concentration, gas hourly space velocity (GHSV) and particle size could also be important factors affecting their CO oxidation activities. Hence, the effects of supporting material, synthesis method and operating parameters on CO oxidation performances of the catalysts deserve more understanding.

In this work, copper manganese oxides precursors with different activated carbon supports are synthesized using the co-precipitation and high-pressure impregnation methods. The precursors are further modified by KOH to obtain the bi-functional catalysts with improved CO2-resistance. The first objective is to evaluate the CO oxidation performances of the catalysts in the presence of CO2. The effects of supporting material and synthesis method on CO oxidation performances are expounded, and the optimal candidate with high activity and excellent long-term working stability is screened. Furthermore, the effects of CO concentration, CO2 concentration, GHSV, and particle size on CO oxidation performances are demonstrated.

Experimental

Materials

Coconut-shell activated carbon, coal activated carbon and wood activated carbon with an average particle diameter of 300 μm were selected as the supporting materials. They were purchased from Shandong Yimei Machinery Equipment Manufacturing Co., Ltd., Weihai Mingxiang Medical Devices Co., Ltd., and Jining Luyi Material Co., Ltd. Copper acetate (C4H6CuO4·H2O, AR 99 %), manganese acetate (C4H6MnO4·4H2O, AR 99 %), and potassium hydroxide (KOH, AR 99 %) were determined as the active components. They were provided by Tianjin Damao Chemical Reagent Factory, Tianjin Guangfu Fine Chemical Research Institute, and Shanghai Jiuyi Chemical Reagent Co., Ltd.

Catalyst synthesis

Synthesis of copper manganese oxides precursors

Copper manganese oxides precursors with different supports were synthesized by the co-precipitation and high-pressure impregnation methods. To facilitate description, the as-synthesized precursors were labeled as CM-P-CSAC, CM-P-CAC, CM-P-WAC, and CM-I-WAC. CM denoted copper manganese oxide compounds, P and I denoted the co-precipitation and high-pressure impregnation methods. CSAC, CAC, and WAC denoted the coconut-shell activated carbon, coal activated carbon and wood activated carbon.

CM-P-CSAC, CM-P-CAC, and CM-P-WAC were synthesized with the co-precipitation method. 10 g of KOH was first dissolved in 100 mL of deionized water under stirring to form KOH solution. 50 g of the supports (CSAC, CAC, and WAC) were mixed with certain grams of C4H6CuO4·H2O and C4H6MnO4·4H2O dissolved in 200 mL of deionized water under stirring at ambient temperature for 2 h. Then, KOH solution was added drop wise into the mixture under stirring for 4 h. The precipitates were then filtered and washed several times by deionized water followed by drying in a rotary vacuum evaporator at 85 °C overnight and calcination in a muffle furnace at 350 °C for 6 h to obtain the desired precursors.

CM-I-WAC was synthesized with the high-pressure impregnation method. 50 g of the support (WAC) was mixed with certain grams of C4H6CuO4·H2O and C4H6MnO4·4H2O dissolved in 200 mL of deionized water under stirring overnight in a high-pressure vessel. The mixtures were then dried in a rotary vacuum evaporator at 85 °C overnight followed by calcination in a muffle furnace at 350 °C for 6 h to obtain the desired precursor.

Synthesis of the bi-functional catalysts

The bi-functional catalysts of KCM-P-CSAC, KCM-P-CAC, KCM-P-WAC, and KCM-I-WAC were prepared with the impregnation of KOH on the as-synthesized precursors. A certain gram of the precursors were mixed with 25 g of KOH dissolved in 250 mL of de-ionized water under stirring overnight at ambient temperature. The mixtures were subsequently dried in a rotary vacuum evaporator at 85 °C overnight and calcined in a muffle furnace at 350 °C for 6 h to obtain the desired bi-functional catalysts.

Catalyst characterization

X-ray diffraction (XRD) patterns of the catalysts were recorded on an X’Pert PRO diffractometer (Philips, Netherlands) using nickel-filtered Cu Kα radiation at 30 kV and 150 mA (2θ angle ranging from 10° to 70°, wavelength λ = 0.15406 nm and 0.02° sampling width). Texture parameters including the BET surface areas, pore volumes, and average pore diameters were determined from N2 adsorption–desorption tests performed on an automatic surface area and porosity analyzer Tristar II 3020 M (Micromeritics, USA). Particle morphologies of the catalysts were observed by field emission scanning electron microscopy (FESEM) SIRION200 (Philips, Netherlands). Chemical compositions of the catalysts were determined by inductively coupled plasma-mass spectrometry (ICP-MS) X Series 2 (Thermo Fisher Scientific, USA).

Catalytic testing

CO oxidation tests were performed in a fixed-bed reactor with an inner diameter of 0.02 m and a height of 0.2 m. 5 g of the catalysts were packed in the reactor. Prior to each test, the reactor was heated to 200 °C in a pure N2 stream to eliminate the adsorbed water vapor. Then, the reactor was cooled to ambient temperature. Gaseous mixtures of 0.4 %CO + 1 %CO2 and balanced air with a total flow rate of 500 mL/min (GHSV = 4000 h−1) then flowed through the reactor. The reactor was subsequently heated to the desired temperature with a ramping rate of 10 °C/min for CO oxidation test. By measuring the change of CO concentration at the outlet of the reactor with a gas analyzer, CO oxidation performances of these catalysts could be evaluated. By changing the operating parameters, the effects of CO concentration, CO2 concentration, GHSV, and particle size on CO oxidation performances could also be evaluated.

Catalytic activity evaluation

The as-synthesized catalysts were screened on the basis of their CO oxidation activities and long-term working stabilities. The CO conversion could be expressed as:

$$X = \frac{{Q_{{in}} C_{{CO}}^{0} - Q_{{out}} C_{{CO}}^{f} }}{{Q_{{in}} C_{{CO}}^{0} }} \times 100\;\%$$
(1)

Here \(X\) is the CO conversion, \(C_{CO}^{0}\) and \(C_{CO}^{f}\) are the CO concentrations at the inlet and outlet of the reactor. \(Q{}_{in}\) and \(Q{}_{out}\) are the total flow rates at the inlet and outlet of the reactor. Besides, the temperatures for half and total CO conversions (T 50 and T 100 ) were recorded to evaluate their CO oxidation activities. Long-term working stabilities were characterized by the durations for keeping their stable CO conversions.

Results and discussion

Catalyst characterization

Chemical compositions

The chemical compositions of KCM-P-CSAC, KCM-P-CAC, KCM-P-WAC, and KCM-I-WAC were measured by ICP-MS, and the results are shown in Table 1. The contents of Cu and Mn species in KCM-I-WAC synthesized by the high-pressure impregnation method are high as 13.20 and 30.80 wt%. By contrast, the contents of Cu and Mn in KCM-P-CSAC, KCM-P-CAC, and KCM-P-WAC synthesized by the co-precipitation method are rather low. This indicates that the high-pressure impregnation method benefits the catalysts higher loading amounts of the copper manganese oxides. The contents of K in the four catalysts are almost at the same level.

Table 1 Textural properties of the supports and catalysts
Full size table

N2 adsorption–desorption tests

N2 adsorption–desorption tests are performed to acquire the texture parameters. N2 adsorption–desorption isotherms and pore size distributions of the catalysts are plotted in Figs. S2 and S3. According to the IUPAC (International Union of Pure and Applied Chemistry) classification, the isotherm of KCM-P-CSAC presents typical type-IV behavior with hysteresis loop of type H4, while those of KCM-P-CAC, KCM-P-WAC, and KCM-I-WAC reveal type-IV behavior with hysteresis loop of type H3. These indicate that the catalysts present the characteristics of mesoporous materials. The hysteresis loops located in the relative pressure range of 0.45–0.99 are associated with the capillary condensation of nitrogen molecules in mesopores. It should be noticed that KCM-P-CSAC shows higher N2 adsorption quantity, indicating that the catalyst owns better texture property.

Pore size distributions of the catalysts are displayed in Fig. S3. Most pores of KCM-P-CAC, KCM-P-WAC, and KCM-I-WAC are located in the pore diameter range of 2–10 nm. By contrast, KCM-P-CSAC shows a broader pore size distribution with pore diameter ranging from 2 to 100 nm. This should indicate a greater total volume for KCM-P-CSAC, compared to the other three catalysts.

Texture parameters including the BET surface areas, pore volumes and average pore diameters of the catalysts are listed in Table 1. For the sake of comparison, texture parameters of the supports are also provided. The BET surface areas of CSAC, CAC, and WAC are 887.95, 734.95 and 253.32 m2/g. Their total pore volumes are calculated as 0.87, 0.62 and 0.17 cm3/g. Both the surface areas and pore volumes show significant decrease after the loading of the active components. The surface areas of the four supported catalysts are 160.96, 79.57, 67.67 and 67.08 m2/g. Their total pore volumes are determined as 0.48, 0.09, 0.08 and 0.08 cm3/g. The loading of the active components over the supports has reduced the number of the active sites available for adsorption and caused pore blockage. The average pore diameters of the supports slightly increase, due to the reduction in their micropore volumes by the filling of the active components.

XRD analysis

The crystallinity and phase of the as-synthesized catalysts are determined by the XRD patterns, as shown in Fig. 1. The XRD profiles of the four catalysts show resembling diffraction peaks which could be assigned as CuO, CuMn2O4, Mn2O3, and KOH. The XRD spectra of KCM-P-CSAC, KCM-P-CAC, and KCM-P-WAC show another three peaks of high intensity, which should be attributed to SiO2, CaCO3, and Al2O3 in the supports. In contrast, these phases are not detectable in the profile of KCM-I-WAC. This should be ascribed to the fact that the high-pressure impregnation method has resulted in more uniform dispersion of copper manganese oxides over the support. As indicated in Table 1, the contents of Cu and Mn species in KCM-P-CSAC, KCM-P-CAC, and KCM-P-WAC are rather low. However, the XRD spectra of these catalysts show obvious diffraction peaks of high intensity for bulk CuO and Mn2O3, implying a weak interaction between the supports and the active components. The diffraction peaks for bulk CuO and Mn2O3 in the profile of KCM-I-WAC show lower intensity, despite the fact that the contents of Cu and Mn are high. This indicates that the active components of CuO and Mn2O3 are highly dispersed over the support for KCM-I-WAC.

Fig. 1
figure 1

XRD patterns of the catalysts. Square-CuO Circle-CuMn2O4 Diamond-Mn2O3 Filled triangle-SiO2 Filled square-CaCO3 Filled diamond-Al2O3 Triangle-KOH

Full size image

FESEM analysis

Active components in the catalysts are generally utilized in the dispersed state, and the distribution characteristics of the active components over the supports affect their catalytic performances. To observe the dispersion characteristics of the active components, particle morphologies of the catalysts are examined, and the FESEM images are presented in Fig. 2.

Fig. 2
figure 2

FESEM images of the catalysts. a KCM-P-CSAC; b KCM-P-CAC; c KCM-P-WAC; d KCM-I-WAC

Full size image

Fig. 2 shows that many small grey blocks and white aggregates are dispersed on the surface of the catalysts. They are deduced as copper manganese oxides and KOH. For the morphology of KCM-P-CSAC, most of the active components are coated on the surface, channel wall, and grooves of the support. Although the catalyst remains considerable pore structures that may contribute to the CO oxidation process, the active components are distributed unevenly with rather larger size. Particle morphology of KCM-P-CAC shows even surface, and the active components are uniformly distributed thereon. Although the micropores have been filled with the active components, the mesoporous structures are partially well remained, which are conducive to the gas diffusion and the contact with active components. It can be observed from the FESEM image of KCM-P-WAC that the loaded KOH aggregate as large white blocks and copper manganese oxides with large size scatter non-uniformly on the surface. These have induced the blockage of pore structures. Particle morphology of KCM-I-WAC shows that large quantities of copper manganese oxides granules combine with each other to form a plate-shaped precursor, and then the small white KOH aggregates are scattered over the precursor. Although the active components show more uniform distribution over the supports, their loadings have caused significant pore blockage.

FT-IR analysis

The as-synthesized catalysts are further characterized by FT-IR, and the spectra are presented in Fig. 3. The spectra of the catalysts exhibit IR bands at 510, 570, 700, 1400, and 3550 cm−1. The characteristic absorption peak of CuO is reported to be located around 500 cm−1 [31]. The IR band at 510 cm−1 is thus deduced as the vibrational feature of CuO. The weak IR band observed at 570 cm−1 could be attributed to Mn2O3. It is reported that compounds with spinel structure (AB2O4) would show broad absorption band in wavenumber range of 400–700 cm−1 due to the vibration of B3+–O bond [32]. The absorption band observed at 700 cm−1 would correspond to the particular bond of CuMn2O4 phase. The significant absorption peak located around 1415 cm−1 is assigned as the symmetric stretching vibration of carbonate ions. The presence of carbonate ions could be deduced as the formation of K2CO3 via the carbonation reaction of KOH with CO2 in open air atmosphere. The stretch vibrational frequency of hydroxyl ions in alkaline hydroxide is reported in the wavenumber range of 3550–3720 cm−1. The IR band located at 3550 cm−1 can thus be assigned as KOH.

Fig. 3
figure 3

FTIR spectra of the catalysts

Full size image

Catalytic activity

CO oxidation activities of the precursors and catalysts are tested in 0.4 %CO + 1.0 %CO2 with balanced air, and the results are shown in Fig. 4. As indicated in Fig. 4a, CO conversions of CM-P-CSAC, CM-P-CAC, and CM-P-WAC increase slowly with the increasing temperature, whereas that of CM-I-WAC shows rapid increase. CM-I-WAC achieves its total CO conversion before the temperature reaches 200 °C. By contrast, CO conversions of CM-P-CSAC, CM-P-CAC, and CM-P-WAC are less than 50 % at 200 °C. These indicate that CM-I-WAC exhibits relatively higher CO oxidation activity.

Fig. 4
figure 4

CO oxidation activities of the precursors (a) and catalysts (b) tested in 0.4 %CO + 1.0 %CO2 and balanced air

Full size image

Fig. 4b depicts that CO conversions of the catalysts versus temperature show similar variation with those of the precursors. Compared to the precursors, CO conversions of the catalysts are relatively higher. This indicated that CO oxidation activities of the precursors have been improved with the modification of KOH. It has been reported that KOH modification possesses a double functionality, namely, to mitigate the adverse effect of CO2 by reducing its concentration driving force, and to provide more oxygen adsorption sites to replenish the depleted oxygen in the CO oxidation process [33]. The temperature required for half CO conversion (T 50 ) must be focused on, since it is an important parameter for evaluating the catalytic activity. T 50 values for KCM-P-CSAC, KCM-P-CAC, KCM-P-WAC, and KCM-I-WAC are 178, 190, 188 and 110 °C. The CO oxidation process should be more economically promising when operated at lower temperatures. Amongst the four catalysts, KCM-I-WAC exhibits the best catalytic activity toward CO oxidation with the lowermost T 50 value.

Catalytic behaviors of the as-synthesized catalysts under different reaction conditions

Effect of support and synthesis method

CO oxidation activities of the catalysts prepared with different supports and methods have been elucidated in Fig. 4. To investigate the effect of support and synthesis method on their long-term working stabilities, CO conversions as a function of the time-on-stream are presented in Fig. 5.

Fig. 5
figure 5

Effect of support and synthesis method on long-term working stability (CO concentration: 0.4 %, CO2 concentration: 1.0 %, air: balanced, GHSV: 4000 h−1, particle size: 100 μm)

Full size image

As depicted in Fig. 5, when the final temperature is maintained at 200 °C, CO conversions of the four catalysts keep stable first and then decrease slightly. Within 90 min, CO conversions of KCM-P-CSAC, KCM-P-CAC, and KCM-P-WAC slightly decrease from 72.0, 65.7 and 68.8 % to 66.8, 63.7 and 61.3 %. KCM-I-WAC maintains its total CO conversion for 118 min, and its conversion slightly decreases to 95.9 % within 150 min. This indicates that KCM-I-WAC exhibits better long-term working stability for CO oxidation.

The difference in the CO oxidation activities amongst these catalysts could be ascribed to the texture properties of the supports and the coupling behaviors of the active components with the supports, due to the nature of the different activated carbons and the catalyst synthesis methods. Although KCM-P-CSAC exhibits better texture properties with higher surface area and greater pore volume, the active components are not uniformly dispersed and the utilization efficiency for the active sites is rather low. For KCM-P-CAC and KCM-P-WAC, the loading of the active components has resulted in pore structure blockage and non-uniform dispersion of the active components on the supports. These could be possible reasons for the poor CO oxidation activities of KCM-P-CSAC, KCM-P-CAC, and KCM-P-WAC. Another major reason could be that the loading amounts of copper manganese oxides over the catalysts synthesized by the co-precipitation method are rather low. For KCM-I-WAC, although most of the pore structures have been destroyed by the loading of the active components, the high-pressure impregnation method has benefited the catalyst more uniform dispersion of the active components. This means that more active sites are readily available for the CO adsorption and catalytic oxidation process. Moreover, the high-pressure impregnation method could be favorable for the effective loading of copper manganese oxides. Therefore, these could be responsible for the preferable CO catalytic activity of KCM-I-WAC. Compared to the nature of the different activated carbon supports, catalyst synthesis method has exerted more significant impact on the CO oxidation activity.

Effect of CO concentration

The aforementioned analysis indicates that KCM-I-WAC exhibits the best CO oxidation activity in the presence of CO2. Therefore, KCM-I-WAC is determined as a candidate to further investigate the effects of CO concentration, CO2 concentration, GHSV, and particle size on CO oxidation activity. To investigate the effect of CO concentration, KCM-I-WAC is tested in CO + 1.0 %CO2 and balanced air, whereby the CO concentration changes in the range of 0.1–0.5 %.

As shown in Fig. 6, the CO conversion of the catalyst decreases with the increase of CO concentration from 0.1 to 0.5 %. As shown in Table S1, T 50 and T 100 increase from 93.1 and 140 °C to 115.8 and 180 °C, when the CO concentration increases from 0.1 to 0.5 %. CO conversions of the catalyst in different CO concentrations are correlated with the reaction rate, as shown in Fig. S4. CO oxidation reaction rate increases first and then keeps stable with the increasing reaction temperature. With the increase of CO concentration from 0.1 to 0.5 %, the CO oxidation reaction rate increases from 0.89 × 10−7 to 2.03 × 10−7 mol g−1 s−1, when the catalyst is tested at 120 °C. Luo et al. reported that, with the increasing CO partial pressure from 0.3040 to 3.0398 kPa, CO conversion of Pt/TiO2 decreased from 17.4 to 3.6 %, and its reaction rate increased from 1.55 × 10−7 to 3.21 × 10−7 mol g−1 s−1, when the catalyst was tested at 40 °C [34]. On the one hand, the increase in CO concentration enhances the CO concentration driving force, and benefit the catalyst increased reaction rate. On the other hand, more CO2 is generated under higher CO partial pressures, and they compete with CO for the adsorption active sites to form surface carbonate species, and therefore to inhibit the CO oxidation process by reducing the CO conversion [35].

Fig. 6
figure 6

Effect of CO concentration on CO oxidation activity. (Catalyst: KCM-I-WAC, CO concentration: 0.1–0.5 %, CO2 concentration: 1.0 %, air: balanced, GHSV: 4000 h−1, particle size: 100 μm)

Full size image

Effect of CO2 concentration

It has been confirmed that CO2 would exert adverse effect the oxidation activities of CO catalysts. To investigate the effect of CO2, KCM-I-WAC is tested in 0.4 %CO + CO2 and balanced air, whereby the CO2 concentration varies from 0.5 to 15.0 %.

As shown in Fig. 7, CO conversion of the catalyst decreases with the increasing CO2 concentration from 0.5 to 15.0 %. T 50 and T 100 increase from 105.9 and 160 °C to 142.1 and 200 °C when the CO2 concentration increases from 0.5 to 15.0 %. CO oxidation activity could hardly be affected when the catalyst is tested under lower CO2 concentrations of 0.5 and 1.0 %. Nevertheless, CO conversion shows significant decrease when the CO2 concentration further increases in the range of 5.0–15.0 %. The presence of CO2 competes with CO and O2 to take up the adsorption active sites and therefore inhibit the CO adsorption process [36, 37]. KOH modification can effectively mitigate the adverse effect of CO2 under lower CO partial pressures. When the CO2 partial pressure increases to a high value, more bicarbonate species are formed. The generated bicarbonate species cover the surface active sites and cause pore blockage, and therefore impair the CO oxidation process.

Fig. 7
figure 7

Effect of CO2 concentration on CO oxidation activity. (Catalyst: KCM-I-WAC, CO concentration: 0.4 %, CO2 concentration: 0.5–15.0 %, air: balanced, GHSV: 4000 h−1, particle size: 100 μm)

Full size image

Effect of GHSV

To study the effect of GHSV on CO oxidation activity, KCM-I-WAC is tested in 0.4 %CO + 1.0 %CO2 and balanced air, whereby the GHSV changes from 2400 to 5600 h−1.

Fig. 8 shows that CO conversion of the catalyst decreases with the increase of GHSV. With the increase of GHSV from 2400 to 5600 h−1, T 50 and T 100 increase from 103.4 and 170 °C to 125.4 and 180 °C. The effect of GHSV on CO oxidation reaction rate is also presented in Fig. S5. With the increase of GHSV from 2400 to 5600 h−1, CO oxidation reaction rate increases from 1.24 \(\times\) 10−7 to 2.10 \(\times\) 10−7 mol g−1 s−1, when the catalyst is tested at 120 °C. The increase in GHSV means that more reactants are replenished to the CO catalytic reaction. Besides, the increasing GHSV would be conducive to decreasing the thickness of the boundary layer and benefits the catalyst facilitative mass-transfer process. These improve the CO oxidation reaction rate. In contrast, the increasing GHSV also indicates reduced reaction residence time and therefore the overall CO conversion is decreased. These are in line with the results reported in previous publications [35, 37–39].

Fig. 8
figure 8

Effect of GHSV on CO oxidation activity. (Catalyst: KCM-I-WAC, CO concentration: 0.4 %, CO2 concentration: 1.0 %, air: balanced, GHSV: 2400–5600 h−1, particle size: 100 μm)

Full size image

Effect of particle size

Particle size is an important factor affecting the activities of heterogeneous catalysts [40]. To study the particle size effect, the catalysts with particle diameter ranging from 60 to 300 μm are tested in 0.4 %CO + 1.0 %CO2 and balanced air.

Fig. 9 depicts that the CO conversion of KCM-I-WAC decreases with the increase of particle diameter. With the increasing particle diameter from 60 to 300 μm, T 50 and T 100 increase from 108.6 and 170 °C to 157.8 and 200 °C. Internal diffusion is one significant parameter that affect the CO oxidation activity. The internal diffusion effect is characterized by the intra-particle diffusion efficiency factor η, which is defined as the ratio of the real reaction rate affected by internal diffusion to that of the limit reaction rate free from the influence of internal diffusion [41]. A greater value for η should indicate non-significant impact of internal diffusion on CO oxidation activity. Meanwhile, the Thiele modulus φs is the only influential factor for η:

Fig. 9
figure 9

Effect of particle size on CO oxidation activity. (Catalyst: KCM-I-WAC, CO concentration: 0.4 %, CO2 concentration: 1.0 %, air: balanced, GHSV: 4000 h−1, particle size: 60–300 μm)

Full size image
$$\eta = \frac{1}{{\varphi_{s} }}\left[ {\frac{1}{{th\left( {3\varphi_{s} } \right)}} - \frac{1}{{3\varphi_{s} }}} \right]$$

For the first order CO catalytic reaction, φs could be correlated with the particle diameter:

$$\varphi_{s} = \frac{R}{3}\sqrt {\frac{k}{{D_{e} }}}$$

Here R is the particle diameter, k is the kinetic constant per volume, and D e is the effective diffusion coefficient.

With the increase of particle diameter R, the Thiele modulus φs is increased, and the intra-particle diffusion efficiency factor η is decreased. This indicates that the intra-particle diffusion process exerts significant impact on the CO oxidation activity. Therefore, reducing the particle size of the catalyst could be the most effective means for reducing the intra-particle diffusional resistance. Therefore, to increase the CO oxidation activity and to decrease the internal diffusion resistance, the particle diameter of the catalyst should be controlled at a lower value in practical operations.

Conclusions

In this work, copper manganese oxides precursors with different activated carbon supports are synthesized using the co-precipitation and high-pressure impregnation methods. The precursors are further modified by KOH to synthesize the bi-functional catalysts with improved CO2-resistance. The CO oxidation performances of the catalysts are evaluated in the presence of CO2 using a fixed-bed reactor. The effects of the support, catalyst synthesis method and operating parameters on CO oxidation performances are demonstrated. Compared to the nature of the different activated carbons, the catalyst synthesis method exerts more significant effect on CO oxidation performances. Amongst all these catalysts, KCM-I-WAC synthesized by the high-pressure impregnation method with wood activated carbon as support exhibits excellent CO oxidation activity and long-term working stability. The high-pressure impregnation method benefits the catalyst effective loading and uniform dispersion of the active components. The CO conversion decreases with the increasing CO and CO2 partial pressures, due to the covering of the adsorption active sites and the formation of surface carbonate/bicarbonate species. Increasing GHSV shortens the reaction residence time and therefore decrease the overall CO conversion. The increase in particle diameter is detrimental to high CO conversion, due to the adverse effect of internal diffusion. The results provide insights into optimizing the synthesis method and operation parameters for enhanced CO oxidation performances. The bi-functional catalyst with high CO2-resistance shows promise as a scavenger agent for post-fire atmosphere recovery application. Future work will be focused on evaluating CO oxidation performances of the catalyst in the atmosphere more close to a real post-fire smoke scenario, considering the impurities such as H2O, NO x , SO x , HCl and HCN.