Influence of Sr Substitution on Catalytic Performance of LaMnO 3 / Ni Metal foam Composite for CO Oxidation

A series of Sr-substituted lanthanum manganite perovskites, La1–xSrxMnO3 (LSMO, x = 0, 0.1, 0.2, and 0.3), with mesoporous structures were prepared and coated onto a three-dimensional Ni metal foam (MF) as composite catalysts. The catalytic performances of La0.8Sr0.2MnO3/MF and La0.7Sr0.3MnO3/MF were found to be superior to those of La0.9Sr0.1MnO3/MF, LaMnO3/MF, and LaMnO3 powder in terms of catalytic oxidation of carbon monoxide with air. Under the reaction conditions (1.5 vol.% CO and air balance at a weight hourly space velocity of 90,000 hr), La0.8Sr0.2MnO3/MF reached 100% catalytic oxidation of CO, which is 27% higher than that of LaMnO3 powder. Sr substitution induced an increase of Mn and adsorbed surface oxygen species (O, O2, or O2), which increased the number of active centers for oxidation and thus enhanced the oxidizing ability of the catalyst. The high activity and excellent stability of La0.8Sr0.2MnO3/MF catalyst can be ascribed to a synergistic effect between the mesoporous structure and the high number of adsorbed oxygen species of the catalyst as well as the interconnected three-dimensional reticular configuration of the nickel metal support, which increases the number of active sites and enhances mass transfer for CO and O2. La0.8Sr0.2MnO3/MF composite can potentially be used in catalytic converters for CO removal of automotive exhaust gases.


INTRODUCTION
Automotive exhaust gases mainly contain nitrogen oxides (NO x ), hydrocarbons, and carbon monoxide (CO).Harmful CO is converted to carbon dioxide (CO 2 ) through catalytic oxidation by catalytic converters, which are commonly fabricated by coating precious-metal-based catalysts such as Pt and Pd compounds onto a monolithic honeycomb skeleton made of a ceramic or metallic material.Precious-metal-based catalysts are highly active; however, their high cost and low resistance to Cl poisoning are major shortcomings.Due to the rising cost of precious metals, perovskite-type oxides (with the general formula ABO 3 ) have attracted a lot of attention due to their low cost, good redox properties, and high poisoning resistance compared to those of preciousmetal-based catalysts (Seyfi et al., 2009;Zawadzki and Trawczynski, 2011).The redox ability and number of oxygen defects can be varied via a partial substitution with cations with similar oxidation state and ionic radius at the A-site or B-site to form A 1-x A' x B 1-y B' y O 3 , resulting in improved catalytic and other functional properties (Seyfi et al., 2009;Prasad et al., 2012;Wang et al., 2012).Compounds with lanthanum in the A position and Co or Mn in the position B, such as LaCoO 3 or LaMnO 3 , have been studied for application in catalytic oxidation (Prasad et al., 2012;Wang et al., 2012).Lanthanum manganite perovskites were found to be more active than the corresponding Co samples with the same composition, since they exhibit high oxygen transport.For example, partial substitution of La with divalent ions, e.g., Sr 2+ , can increase the average oxidation state of the cation in position B, and thus enhance the oxidation catalytic activity by facilitating oxygen mobility or by enhancing the redox activity of the B cation (Wei et al., 2008;Frozandeh-Mehr et al., 2012).Moreover, catalysts with partial substitutions at A and B, such as the series of La (1-x) Sr x Co (1-y) M y O 3 (M = Cu, V, Sn, Ti, Zr) catalysts, have been synthesized and reported to possess improved activities for CO oxidation in a temperature range of 300-400°C (He et al., 2001).During the catalytic combustion of CO, the catalyst provides adsorption and active sites to gaseous CO and O 2 during reaction.The catalyst is regenerated through an adsorption-dissociation-incorporation process of gaseous O 2 from the gas phase.Therefore, adsorption and diffusion often limit the overall performance of the catalyst.Usually, a high surface area and porous structure are beneficial for the adsorption and transportation of reactant and product molecules for catalytic combustion.However, calcination at high temperatures (> 600°C) is necessary for the formation of LaMnO 3 , which often results in a sharp decrease of the surface area of the catalyst and makes LaMnO 3 become nonporous.In particular, the nonporous structure of the catalyst limits the contact among reactants and active sites of the catalyst and strongly affects the catalyst's activity.Therefore, synthesizing LaMnO 3 with a mesoporous structure is expected to greatly increase the number of accessible active sites, enhancing catalytic efficiency.There have been several reports on the preparation of mesoporous and high-surface-area ABO 3 via the surfactant templating method.The introduction of P-123 is beneficial for the formation of mesoporous materials, which permit the rapid transport of reactants or redox couples to the high surface area provided by the mesopores (Hseu et al., 2013;Wang et al., 2013).In addition to the catalyst, the support material of the catalyst plays an important role in hightemperature catalytic combustion.A metal substrate made of nickel (Ni) metal foam (MF) has advantages such as low cost, a three-dimensional (3D) reticular configuration, high porosity, and high specific surface area (Yang et al., 2008).However, there have been no reports on the preparation of strontium-substituted lanthanum manganite, i.e., La (1-x) Sr x MnO 3 (LSMO), supported on Ni MF and its application in catalyzing the oxidation of CO.
In this work, to take advantage of the 3D reticular configuration of Ni MF and the superior oxidation activity of porous LSMO, LSMO/MF composite catalysts are prepared.Four formulations with nominal compositions of La (1-x) Sr x MnO 3 /MF (x = 0-0.3)were prepared by spin coating and the doctor blading method without the use of a binder.

Preparation of La 1-x Sr x MO 3 /Ni MF Composites
Ni MF (compressed, 110 pores per inch) was purchased from Changsha Lyrun New Material Co. Ltd.Before loading, the Ni MF was ultrasonically degreased in acetone for 30 min, treated with 2 M HCl for 20 min, washed with distilled water, and dried in an oven at 80°C overnight.The typical coating gel of strontium was synthesized as follows.La(NO 3 ) 3 •3H 2 O, Sr(NO 3 ) 2 •6H 2 O, and Mn(NO 3 ) 2 •6H 2 O were mixed with 50 mL of deionized water and stirred to form a homogeneous solution.An appropriate amount of citric acid (CA) was subsequently added to the above solution; the molar ratio of CA/(La + Sr + Mn) was 0.5.To produce P123 solution, a designed amount of P123 was successively added to 20 mL of deionized water, and the resulting solution was vigorously stirred at room temperature for 2 h.Then, solutions of lanthanum, strontium, and manganese nitrates were added to the P123 solution.The temperature of the mixture was then increased to 80°C and the solution was stirred for 3 h.A homogeneous gel was thus obtained.This coating gel was dropped onto the Ni MF, which was rotated at 200 rpm for 20 s using a spin coater.After the spin coating, the film was dried at 100°C for 10 min in a vacuum oven to evaporate the solvent and to remove organic residuals.The coating gel was then coated onto the previously obtained lanthanum, strontium, and manganese nitrates/MF substrate using the doctor blading method, and finally calcined at 600°C for 2 h as a composite sample.The molar composition of the mixture La(NO 3 ) 3 /Sr(NO 3 ) 2 /Mn(NO 3 ) 2 /CA/P-123/H 2 O was 1.0x/x/1.0/1.0/0.2/400(x = 0-0.3).Unsubstituted LaMnO 3 /MF is denoted as LMO/MF and samples with La substituted by Sr with La/Sr molar ratios of 0.9:0.1,0.8:0.2, and 0.7:0.3 are denoted as L9S1MO/MF, L8S2MO/MF, and L7S3MO/MF, respectively.For comparison, LaMnO 3 powder was also prepared.

Sample Characterization
The X-ray diffraction (XRD) patterns of the samples were measured using an X-ray diffractometer (PANalytical X'Pert PRO) with Cu radiation (λ = 0.15418 nm) in the 2θ range of 20° to 80°.The surface morphology and compositional characterization was performed using scanning electron microscopy (SEM, JEOL JSM-6700F) with energy-dispersive X-ray spectroscopy (EDS).Surface analysis of the chemical states of the lanthanum, strontium, manganese, and oxygen elements of the catalyst on the surface of Ni MF was carried out using X-ray photoelectron spectroscopy (XPS, Kratos Axis, Ultra DLD) with an Al Kα source.For calibration, the C 1s electron binding energy was set to 285.0 eV and the distributions of lanthanum, strontium, manganese, and oxygen functionalities were quantified after fitting the XPS La 3d, Mn 2p, and O 1s peaks, respectively, to Gaussian-Lorentzian component profiles.The surface area and pore volume of the as-prepared samples were determined using a volumetric sorption analyzer (Micromeritics ASAP 2020).The samples were degassed at 200°C under vacuum condition for at least 4 h prior to measurements.The nitrogen adsorption/desorption isotherms were measured over a relative pressure (P/P 0 ) range of approximately 10 -3 to 0.995.The surface areas were calculated using the Brunauer-Emmett-Teller (BET) equation using adsorption data in the P/P 0 range of 0.06 to 0.2.The pore size distributions were determined from the analysis of the adsorption isotherm using the Barrett -Joyner-Halenda (BJH) algorithm.The total pore volumes were estimated from the adsorbed N 2 amount at P/P 0 = 0.973.

Performance Test
The activity measurement was carried out in a stainless tubular reactor (i.d.: 20 mm; length: 120 mm) equipped with a porous plate.A 500 stainless steel mesh was placed horizontally 30 mm from the base of the tubular reactor.The temperature was measured using a K-type thermocouple, which was mounted on the wall of the tubular reactor, surrounded by the thermal jacket, and controlled by a proportional-integral-derivative temperature controller.In a typical experiment, the catalyst was regenerated prior to each CO oxidation test by heating the stainless tubular column to 150°C for 1 h while purging with a 5 vol.%H 2 and 95 vol.%N 2 stream at 200°C for 1 h, followed by cooling to room temperature.The reaction mixture was then introduced.
The feed mixture consisted of 1.5 vol.%CO and air balance at a weight hourly space velocity of 90,000 hr -1 .A catalyst loading amount of 10 mg was used for each measurement.The analysis of the feed and product mixture was performed using an online gas chromatograph (Thermo Trace GC Ultra) with a packed column (Carboxen-1010) and thermal conductivity detector.CO conversion was calculated using the following equation:

Catalyst Characterization X-ray Diffraction Analysis and Textural Property
Fig. 1 shows XRD patterns of LaMnO 3 powder and various La 1-x Sr x MnO 3 /MF (x = 0, 0.1, 0.2, and 0.3) composite catalysts.The XRD pattern of the LaMnO 3 powder shows several resolved peaks, namely a strong peak at 32.6° and three weak peaks near 40.4°,46.8°, and 58.2°, respectively, which can be indexed to the ( 110), ( 111), ( 200), and (211) reflections, respectively, of the cubic phase of LaMnO 3 according to ICDD card no.01-075-0440.Distinct NiO (ICDD 00-047-1049) and Ni (ICDD 00-004-0850) peaks can be observed for all samples due to the Ni MF substrate.All the characteristic diffraction peaks of La 1-x Sr x MnO 3 /MF composites belonging to the cubic LaMnO 3 structure were observed, indicating that La 1-x Sr x MnO 3 can form a perovskite structure that is not influenced by Sr substitution amount in the range of 0.1 to 0.3.Moreover, the main diffraction peak of (110) became sharper with increasing Sr substitution amount, demonstrating that the average crystallite size increased with increasing Sr substitution amount.The N 2 adsorption/desorption isotherms and pore size distributions of the obtained samples are shown in Fig. 2. As shown in Fig. 2(a), the LaMnO 3 powder exhibits a type IV curve with an H1-shaped hysteresis loop, which is an indication of cylindrical mesopores.The isotherms of La 1-x Sr x MnO 3 /MF (x = 0, 0.1, 0.2, and 0.3) composite samples are type II isotherms with combination forms of a well-developed H3type hysteresis loop in the relative pressure (P/P 0 ) range of 0.9-1.0.These results are confirmed by the pore size distribution of Fig. 2(b).As shown, all samples have bimodal distributions, centered at 5-8 nm and 12-17 nm, respectively.The LaMnO 3 powder has a widely distributed pore diameter (2-25 nm, maximum at 8 nm).All composite samples had smaller mesopores compared to that of the LaMnO 3 powder, with L8S2MO/MF having the highest distribution of mesopores among composite samples.The generation of the hierarchical mesoporous features of obtained samples might be attributable to the molecular structure of P-123, which was used as the template in this study.P-123, an amphiphilic triblock copolymer, is a nonionic surfactant, and thus its hydrophobic poly(propylene oxide) groups segregate into a hydrophobic microphase while its poly(ethylene oxide) (PEO) groups have affinity toward the polar phase of metal hydrates.The PEO groups of P-123 can adsorb onto the La-Mn particle surface during particle growth, and then organize themselves into hierarchical structures through a self-assembly process.All the obtained mesoporous samples were expected to have a greatly increased number of accessible active sites and thus enhanced catalyst activity.The surface areas of the obtained samples were in the range of 2-9 m 2 /g.LaMnO 3 powder had the largest BET surface area (9 m 2 /g) and LMO/MF had the lowest BET surface area (2 m 2 /g).

Morphology
To investigate the local morphology of samples, SEM images of bare Ni MF, LMO/MF, and La 1-x Sr x MnO 3 /MF

Catalytic Activity for Catalytic Combustion of CO
The catalytic activity results of CO conversion over the obtained samples are shown in Fig. 4. The CO conversion is compared at temperatures T 50 and T 100 , at which 50% and 100% of CO was converted to CO 2 , respectively.From Fig. 4, it can be clearly seen that the T 50 value of L8S2MO/MF is around 290°C, whereas those of the other samples are higher than 300°C.The T 100 value is about 400°C for L8S2MO/MF and those of the other samples are higher than 400°C.At 400°C, the percentages of CO conversion of LaMnO 3 powder, LMO/MF, L9S1MO/MF, and L7S3MO/MF are 73.2,81.5, 81.7, and 88.2, respectively.For Sr-substituted lanthanum manganite perovskites loaded on MF as composite catalysts, L8S2MO/MF and L7S3MO/MF have the highest CO conversion.From the lowest to highest T 50 and T 90 values of the catalyst, catalytic activity decreases in the order L8S2MO/MF > L7S3MO/MF > LMO/MF~ L9S1MO/MF > LaMnO 3 powder, revealing that the appropriate substitution of Sr into the lanthanum manganite perovskite structure improved the catalytic performance for the complete oxidation of CO.The LaMnO 3 powder sample had the highest specific surface area but showed poorer catalytic activity than that of the LMO/MF sample, which had the lowest specific surface area.Thus, the specific surface area did not influence the catalytic activity for CO oxidation in this study.Therefore, the observed phenomena indicate that the excellent catalytic performance of La 1-x Sr x MnO 3 /MF (x = 0-0.3)may be attributed to an enhanced mass transfer of CO and CO 2 with porous lanthanum manganite perovskites loaded onto the 3D macroporous structure of Ni MF, which reduces the mass transport resistance during the catalytic combustion of CO.

Surface Analysis by XPS
Since the catalytic combustion of CO takes place on the surface of the catalysts, the oxidation states and composition of surface elements were obtained via XPS.The La 3d, Mn 2p, and O 1s XPS spectra of La 1-x Sr x MnO 3 /MF (x = 0, 0.1, 0.2, and 0.3) samples were recorded.The La 3d XPS spectra have core peaks of La 3d 3/2 and satellite peaks of La 3d 5/2 located at 850.8, 855.0, 834.2, and 838.4 eV, respectively.The spin-orbit splitting between the core peaks of La 3d 3/2 and satellite peaks of La 3d 5/2 were identical (16.6 eV) for all composite samples, indicating that lanthanum ions were present in the trivalent state.
The Mn 2p 3/2 spectra of composite samples are shown in Fig. 5(a).The spectra show an asymmetrical Mn 2p 3/2 peak located at around 642.0 eV.This peak can be deconvoluted into two components at binding energies of 641.5 and 642.8 eV, respectively; the former component can be assigned to Mn 3+ ions and the latter one can be attributed to Mn 4+ ions, indicating that Mn 3+ and Mn 4+ ions coexisted in all samples (Alonso et al., 2000;Deng et al., 2008;Kucharczyk et al., 2008).By using the curve-fitting approach, the areas of the 641.5 and 642.8 eV peaks of Fig. 5(a) were compiled.The deconvolution results are listed in Table 1.For L9S1MO/MF, L8S2MO/MF, and L7S3MO/MF catalysts, the Mn 4+ /Mn 3+ ratios are 0.95, 1.19, and 1.11, respectively.With the substitution of Sr, the surface Mn 4+ /Mn 3+ ratios of the LSMO/MF samples increased, with the L7S3MO/MF and L8S2MO/MF samples showing higher Mn 4+ /Mn 3+ ratios (1.11 and 1.19, respectively) compared to that of the LMO/MF sample (0.99).Thus, the substitution of Sr into LaMnO 3 influenced the Mn 4+ /Mn 3+ molar ratio, implying that partial oxidation of Mn 3+ ions to Mn 4+ ions on the surfaces of catalysts increased with increasing Sr 2+ content in LSMO catalysts.
The O 1s XPS spectra of La 1-x Sr x MnO 3 /MF (x = 0, 0.1, 0.2, and 0.3) catalysts are shown in Fig. 5(b).A main peak at around 529.2 eV and a shoulder peak in the range of 530.0 to 533.0 eV can be observed.Gaussian and Lorentzian deconvolutions were performed on three peaks (the main band at ~529.2 eV, the shoulder band at ~530.8 eV, and the weak band at 532.1 eV).It has been reported that the peak at ~529.2 eV corresponds to lattice oxygen (O 2-) bonding to the metal cations (Sun et al., 2011).The peak at ~530.8 eV corresponds to peroxide ions (O - and O 2   2-) and the peak at 532.1 eV is ascribed to hydroxyl groups (Sun et al., 2011).These ions are ascribed to the adsorbed oxygen or oxygen in the hydroxyl groups (O -, O 2 -, or O 2 2-), whose content levels reflect the concentrations of the oxygen vacancies of a compound (Sun et al., 2011).Oxidation of CO on metal oxide catalysts is a superficial reaction where the CO reacts with the surface oxygen species, which transform on the surface of the metal oxide catalyst according to the general schemes: O 2 (adsorption) (Wang et al., 2012).From the literature, the adsorbed oxygen desorbs in the temperature range of 300-600°C and the lattice oxygen desorbs at a higher temperature (Yamazoe et al., 1990).By using the curve-fitting approach, the areas of the main and shoulder peaks of Fig. 5(b) were calculated.The content levels of the corresponding adsorption oxygen and lattice oxygen are listed in Table 2. Previous studies have reported that the relative ratio of the adsorbed oxygen to lattice oxygen, O adsorbed /O lattice , can be used as an index to evaluate catalytic activity (Hwang et al., 2010;Sun et al., 2011;Liang et al., 2013).As shown in Table 2, the obtained O adsorbed /O lattice ratio increases with increasing La/Sr molar ratio until the La/Sr molar ratio reaches 0.8:0.2.After the La/Sr molar ratio reaches 0.7:0.3(L7S3MO/MF), the O adsorbed /O lattice ratio begins to decrease.The relative ratio of the adsorbed oxygen to the lattice oxygen of the La 1-x Sr x MnO 3 /MF (x = 0, 0.1, 0.2, and 0.3) samples follows the sequence L8S2MO/MF (2.53) > L7S3MO/MF (2.40) > LMO/MF (0.75) > L9S1MO/MF (0.70), which is in agreement with the order of the catalyst activities shown in Fig. 4.Moreover, the O adsorbed /O lattice ratio follows the same sequence as that of the Mn 4+ /Mn 3+ ratio.Based on the results of the Mn 4+ /Mn 3+ ratio and O adsorbed /O lattice ratio, it can be seen that the substitution with Sr at the A-site had an effect on the surface Mn 4+ concentration and on the number of surface-adsorbed oxygen species.The higher catalytic activities of L8S2MO/MF and L7S3MO/MF compared to those of the other composite samples can be ascribed to appropriate Sr substitution.The improved catalytic activities may have resulted from the higher substitution of La 3+ by Sr 2+ being balanced with the higher conversion of the Mn valence state from Mn 3+ to Mn 4+ and the formation of oxygen vacancies, leading to an enhancement of the number of adsorbed oxygen species (Niu et al., 2006;Wei et al., 2010).Since adsorbed oxygen is loosely bound to the catalyst surface, the adsorbed oxygen species easily move to the reaction sites.More adsorbed oxygen traps more electrons, which is favorable for the formation of more O 2 -, which in turn leads to more active centers for oxidation and thus enhances the oxidizing ability of the catalyst.

Stability Tests
In order to investigate the stability of L8S2MO/MF and   stream had a detrimental effect on CO removal for LaMnO 3 powder.When 3 vol.%H 2 O was added, CO conversion over L8S2MO/MF remained at 100% for 60 h, whereas that over LaMnO 3 powder dropped to 70%.It has been reported that the presence of H 2 O in the feed has a negative effect on CO conversion, which results from the blockage of catalytic active sites by water adsorption, which limits the access of the reactant molecules to the sample surface (Park et al., 2004;Gamarra et al., 2009).For LaMnO 3 powder catalyst, H 2 O has an only slight negative influence on the catalyst activity due to the mesoporous structure of LaMnO 3 powder, which results from the addition of P123.Generally, specific surface area, B-site cation reducibility, and surface oxygen species are the key factors that influence the oxidation activity of perovskite catalysts (Prasad et al., 2012;Wang et al., 2012).From the results of particle size distribution, CO conversion, and O adsorbed /O lattice ratio, the high activity and good resistance of L8S2MO/MF catalyst to H 2 O are attributed to the synergistic effect between the MF support and suitable Sr substitution.The interconnected 3D network of Ni MF increases the mass transfer of CO and CO 2 and appropriate Sr substitution leads to an enhancement of the formation of surface oxygen, thus facilitating the oxidation of CO.In this study, LaMnO 3 powder had the largest BET surface area (9 m 2 /g) but poor catalyst activity, indicating that BET surface area does not play an important role in the catalytic oxidation of CO.

CONCLUSION
Mesoporous La (1-x) Sr x MnO 3 /MF composites for the catalytic oxidation of CO in air were prepared using a facile method of spin coating combined with the doctor blading method without the addition of a binder.The XRD results show that La 1-x Sr x MnO 3 /MF composites had a cubic LaMnO 3 structure that was not influenced by Sr substitution amount in the range of 0.1 to 0.3.The N 2 adsorption/desorption isotherms and pore size distribution results reveal that the generation of the hierarchical mesoporous features of obtained samples might be attributable to the addition of P-123.The existence of La 3+ , Mn 3+ , and Mn 4+ as well as adsorbed oxygen or oxygen in the hydroxyl groups (O -, O 2 -, or O 2 2-) were observed over the surfaces of La (1-x) Sr x MnO 3 /MF samples by XPS.The catalytic performances of L8S2MO/MF and L7S3MO/MF composites were superior to those of L9S1MO/MF and LMO/MF, indicating that the precise control of the Sr/La molar ratio is necessary to achieve good catalytic performance in CO catalytic oxidation.The interconnected 3D reticular configuration of the Ni MF support, mesoporous structure, and high Mn 4+ /Mn 3+ ratio are mainly responsible for the enhanced oxygen adsorption capacity, outstanding CO catalytic oxidation activity, and long-term stability of the L8S2MO/MF and L7S3MO/MF composites.The low-cost production techniques and low-cost materials (LaMnO 3 and Ni MF) make the proposed catalysts candidates for compact catalytic converters for CO removal of automotive exhaust gases.

Table 1 .
Binding energy values of Mn 2p spectra and Mn 4+ /Mn 3+ derived from XPS deconvolution analysis of samples.

Table 2 .
Binding energy values of O 1s spectra and oxygen content levels derived from XPS deconvolution analysis of samples.