Synthesis and characterization of Co–Al–Fe nonstoichiometric spinel-type catalysts for catalytic CO oxidation

Abstract

A series of CoAlFe nonstoichiometric spinel-type oxides were synthesized from hydrotalcite precursors prepared through a co-precipitation method, and their catalytic activities for CO oxidation were investigated. The solids were characterized by XRD, BET, SEM, TG-DTG, H₂-TPR and in situ FTIR. The calcined hydrotalcite-like precursors were composed of spinel-like CoAlFe mixed oxide with crystallite sizes in the range 8–10.5 nm. The nanosized spinel oxide catalysts showed higher surface area as calcination led to dehydroxylation and carbonate decomposition of anions in interlayer spaces. FTIR results showed two vibrational frequency bands (ν1 and ν2) for tetrahedral and octahedral sites, confirming the formation of the Co₃O₄ spinel. Iron ions were introduced into the spinel system leading to improved redox properties, as confirmed by TPR. Furthermore, the CoAlFe ternary oxide nanoparticles exhibited superior catalytic performance in CO oxidation in contrast with CoAl and CoFe binary oxides, which can be ascribed to the improved reducibility. According to the in situ FT-IR analysis, CO adsorbed on the catalyst surface reacted with surface lattice oxygen to form CO₂. In addition, CO₂ could adsorb on the surface and form intermediate carbonate species.

---

1. Introduction

Carbon monoxide is a kind of strongly toxic gas. It directly takes part in the formation of ground-level ozone and increases the greenhouse effect due to transformation to CO₂ and stabilization of CH₄ in the atmosphere. There are numerous and different sources of carbon monoxide formation, e.g. transportation, energy production, agriculture, and the chemical and steel industries.¹ Catalytic oxidation of CO into CO₂ is considered to be a good route for the reduction of CO emission in motor engines and other industrial processes.² Currently, various kinds of catalysts such as supported noble metals,^3–8 and transition metal oxides,^8–12 are being widely used for catalytic oxidation of CO. Among them, the supported noble metals have been generally regarded as the most desirable catalysts, in terms of their catalytic activity, selectivity and stability for catalytic oxidation.¹³ However, transition metal based oxides can replace traditional noble-metal-based catalysts, because of their low cost and stability.¹⁴ Recently, some hydrotalcites^15–18 have been reported as precursors for catalytic elimination of CO and they exhibit promising results.

The hydrotalcites (HT) belong to a great group of natural or synthetic inorganic lamellar compounds, named layered double hydroxides or anionic clays. The general formula, which specifies this class of materials is [M_(1−x)²⁺M_x³⁺(OH)₂]^x+[A_x/nⁿ⁻]·mH₂O, where M²⁺ is bivalent (Mg²⁺, Ni²⁺, Co²⁺, Zn²⁺, etc.) and M³⁺ is trivalent (Al³⁺, Cr³⁺, Fe³⁺, etc.) metal cation, x represents the fraction of the M³⁺ cation, namely x = M³⁺/M³⁺ + M²⁺ (the most reliable composition range corresponds approximately to 0.2 ≤ x ≤ 0.4) and m is the number of water molecules.¹⁹ Hydrotalcite-like compounds (HTLcs or HLCs), consist of positively charge metal hydroxide layers separated from each other by anions and water molecules. The layers contain metal cations of at least two different oxidation states.^20,21 Indeed, after calcination treatment the formed mixed oxides possess unique properties like high surface area and porosity, good thermal stability, basic properties, and high metal dispersion.²² Hydrotalcite or hydrotalcite-like compounds, also called layered double hydroxides (LDH), are widely applied in catalysis and adsorption. In recent years, the interests in application of hydrotalcite derived mixed oxides as environmental catalysts have significantly increased.²³ Spinel-type oxides is one-type of mixed metal oxides like spinel ferrites which can be obtained from calcination of LDH at a given temperature.²⁴ Spinels of the type M²⁺M₂³⁺O₄ (which is also named as AB₂O₄) attract a great deal of interest because of their diverse practical applications.²⁵ In the lattice of a normal spinel AB₂O₄ A²⁺ ions occupy the sites with tetrahedral oxygen environment while B³⁺ ions have octahedral oxygen coordination.²⁶ Co₃O₄ with a spinel structure containing Co³⁺ in an octahedral coordination and Co²⁺ in a tetrahedral coordination^12,27 is a very active oxide catalyst for the oxidation of CO. In such catalysts, octahedrally coordinated Co³⁺ is considered to be the active site, while tetrahedrally coordinated Co²⁺ is assumed to be basically inactive.²⁸ Haruta and co-workers²⁹ reported that Co₃O₄ exhibited excellent low-temperature activity for CO oxidation. The interest has then persisted in the next years, searching for promising catalyst formulations for low temperature oxidation carbon monoxide and continues up to now for the unsupported Co₃O₄ oxide^30–34 as well as the cobalt mixed oxides^6,15,35,36 or supported catalysts.^6,7,37

In this study, we have prepared a series of CoAlFe mixed oxides with nonstoichiometric spinel structure derived from hydrotalcites for CO oxidation, and detailed microstructural characterization and evaluation of redox properties of the prepared oxides have been carried out. Additionally, the in situ FTIR tests were performed in the presence of CO to explore possible intermediates generated during oxidation reactions.

2. Experimental

2.1 Catalyst preparation

The CoAlFe hydrotalcites samples with Co²⁺/(Al³⁺ + Fe³⁺) molar ratio fixed at 3 followed the ref. 38 by co-precipitation method. An aqueous solution containing nitrates of the metallic salts Co(NO₃)₂·6H₂O, Al(NO₃)₃·9H₂O and Fe(NO₃)₃·9H₂O (M²⁺/M³⁺ molar ratio = 3), was added drop-wise into a vigorously stirred deionized water solution. During the synthesis, the pH was maintained constant at 10 by drop-wise addition of a solution of Na₂CO₃ 1 M and NaOH 2 M. Then, the resulting slurry was stirred at 80 °C for 18 h. It was then filtrated and washed with distilled water several time until the filtrate had a pH of 7. The resulting solid was dried in an oven at 120 °C for 12 h to obtain HT precursors with Co : Al : Fe molar ratios of 3 : 1 : 0, 3 : 0.75 : 0.25, 3 : 0.5 : 0.5, 3 : 0.25 : 0.75 and 3 : 0 : 1 (assigned as CAF0-HT, CAF1-HT, CAF2-HT, CAF3-HT and CAF4-HT, respectively). The corresponding mixed oxides were obtained by thermal decomposition of precursors at 500 °C in air atmosphere for 5 h with a ramp rate of 5 °C min⁻¹. These oxides samples were denoted with CAF0, CAF1, CAF2, CAF3 and CAF4, respectively.

2.2 Catalyst characterization

The crystalline phase structure was measured by BRUKER-AXS D8 Adance X-ray diffractometer (XRD) using Cu Kα radiation in the 2θ range from 5° to 80°. The step size was 0.02° and time step was 0.2 s. The JCPDS database (International Centre for Diffraction Data, 1996.) was used for the phase identification.

The textural properties of the mixed oxides samples were analyzed by N₂ adsorption/desorption isotherms which were performed using a Micromeritics ASAP 2020 surface area analyzer after outgassing at 300 °C for 5 h prior to analysis. The specific surface areas were calculated with the BET equation. The pore size distributions were obtained by the Barrett–Joyner–Halenda (BJH) methods using the desorption branch of the isotherms.

Scanning electron microscopy (SEM) analysis was performed on QUANTA FEG250 instrument to study the surface morphology of the catalysts. Samples were adhered on aluminum stubs covered with 12 mm carbon adhesive tabs. Typical working parameters were an accelerating voltage of 20 kV, and a beam current of 60 pA.

Thermal gravimetry analysis (DTG) was carried out using a METTLER TGA/DSC1/1600HT system. The rate of heating was maintained at 10 °C min⁻¹ and the mass of the sample was about 10 mg. The measurement was carried from RT to 800 °C under argon flowing at the rate of 100 ml min⁻¹.

Temperature-programmed reduction with H₂ (H₂-TPR) experiments were performed with a XIANQUAN tp-5080 instrument. A 50 mg sample was pretreated at 500 °C for 30 min in a flow of N₂ and cooled to room temperature. TPR was conducted at 10 °C min⁻¹ up to 900 °C in a 30 ml min⁻¹ flow of 10 vol% H₂ in Ar.

Infrared spectra were recorded on a Bruker Tensor 27 spectrometer. The powders were mixed with KBr at the mass ratio of 1 : 100, and then pressed into thin and transparent slices for analysis. Spectra were collected from 400 to 4000 cm⁻¹ at 4 cm⁻¹ resolution with 16 scans acquired for each sample.

2.3 Catalytic activity measurements

The catalytic activity for CO oxidation was measured using a continuous flow fixed-bed reactor system. A feed gas containing 0.2 vol% CO and 5.0 vol% O₂ balanced with He was allowed to pass through the catalyst sample of 100 mg (40–80 mesh) at a flow rate of 50 ml min⁻¹ (corresponding to a space velocity of 30 000 ml h⁻¹ g_cat⁻¹). Prior to the reaction, the catalyst was pretreated with N₂ at 500 °C for 1 h, and then cooled to the reaction temperature under N₂. The outlet gas compositions were on-line analyzed by a GC-2080 gas chromatograph equipped with a TDX-01 column and a FID.

2.4 In situ FTIR characterization of CO adsorption

In situ FTIR spectra were collected from 400 to 4000 cm⁻¹ at a spectral resolution of 4 cm⁻¹ (16 scans) on a Bruker Tensor 27 spectrometer, and equipped with a high-sensitive detector cooled by condensate water.

(1) The adsorption and desorption reaction of CO. The CAF2 sample was pressed into a thin self-supporting wafer with a thickness of 7.5 mg cm⁻². The samples were pretreated in situ at 500 °C under a flow of He for 1 h to eliminate any adsorbed species. After cooling to room temperature under inert gas, the sample was exposed to 0.4% CO/He flow at room temperature until gas equilibration. Desorption studies were performed under the inert gas.

(2) CO and O₂ temperature-programmed reaction (TPR). The CAF2 sample was pressed into a thin self-supporting wafer with a thickness of 7.5 mg cm⁻². The samples were pretreated in situ at 500 °C under a flow of He for 1 h to eliminate any adsorbed species. Then, the sample background of each target temperature was collected during the cooling process. Reactions were performed by heating the sample in CO + O₂, and the spectra were recorded at various target temperatures at a rate of 10 °C min⁻¹ from room temperature to 500 °C by subtraction of the corresponding background reference spectrum.

3. Results and discussion

3.1 Characterization of hydrotalcite-like precursors

3.1.1 The XRD analysis of precursor. The XRD patterns of the hydrotalcite precursors are presented in Fig. 1. The patterns show common features, with reflections located at the typical angles of a hydrotalcite phase,¹⁹ containing carbonate anions in the interlayer space: sharp and symmetric reflection at small diffraction angles (peaks close to 2θ = 11°, 23° and 35°; ascribed to diffraction by basal planes (003), (006) and (012), respectively) and broad, less intense peaks at higher angles (peaks close to 2θ = 38°, 46°, and 60°; ascribed to diffraction by (015), (018) and (110) planes). For CAF0-HT sample, all primary diffraction peaks can be indexed to a well-crystallized hydrotalcite phase of Co₆Al₂CO₃(OH)₁₆·4H₂O (JCPDS Card no. 51-0045). No other phases were detected after iron incorporation, which suggests that aluminum may be substituted progressively by iron in the brucite-like layers. However, there are differences in the peaks intensities from one sample to another, indicating different degrees of crystallinity or structural order when the cationic composition varies.

Fig. 1 XRD patterns of CoAlFe hydrotalcite precursors.

The lattice parameters, a and c, typical of hydrotalcite structures with rhombohedral 3R symmetry were calculated for all samples and listed in Table 1. The “a” parameter corresponds to the average distance cation–cation in the layers of brucite type, which is obtained by a = 2d(110). The “a” values are slightly increased after iron incorporation, which is due to Al³⁺ substitution by the larger radii cation Fe³⁺ (rAl³⁺ = 0.53 Å, rFe³⁺ = 0.63 Å). In the case of “c” parameter, it is related to the thickness of the layer brucite type and interlayer distance, which is usually calculated using the relationship c = 3d(003). It depends upon several factors such as the amount of interlayer water, the size of the interlayer anion and of the M²⁺–M³⁺ cations, and the strength of the electrostatic attractive forces between the layer and the interlayer. The results shown in Table 1 must probably reflect the influence of these various factors.

Table 1 Chemical composition and structural parameters of the hydrotalcite precursors

Hydrotalcites precursors	Compositions (molar ratio)	FWHM (2θ)	a (Å)	c (Å)	Xs (Å)
CAF0-HT	Co/Al/Fe = 3 : 1 : 0	0.772	3.072	22.879	108
CAF1-HT	Co/Al/Fe = 3 : 0.75 : 0.25	0.726	3.074	22.722	112
CAF2-HT	Co/Al/Fe = 3 : 0.5 : 0.5	0.439	3.087	22.686	189
CAF3-HT	Co/Al/Fe = 3 : 0.25 : 0.75	0.654	3.103	22.583	112
CAF4-HT	Co/Al/Fe = 3 : 0 : 1	0.591	3.119	22.687	131

---

The 2θ reflection in the 11–12.5° range is attributed to (003) reflection of rhombohedral unit cell and indicates the formation of hydrotalcite-like compounds having carbonate as interlayer anion.³⁹ The full width at half maximum (FWHM) of the (003) plane (2θ ≈ 11.5°) is included in Table 1 as a measure of crystallinity of the hydrotalcite phase in the c-axis direction. The average crystallite size calculated from d(003) and d(110) planes using Debye–Scherrer equation varied in the range 100–190 Å.

3.1.2 The infrared spectra of precursors. The FTIR spectra of the precursors are showed in Fig. 2. The hydrotalcite structure can be characterized using infrared analysis by three vibration bands corresponding to hydroxyl groups, octahedral layers and interlayer species. The broadest and most intense band (around 3450 cm⁻¹) which looks asymmetry in shape is due to the OH stretching mode (water molecules in interlayer, structural hydroxy group or OH group attached to the cation in the layers).¹⁷ The shoulder at lower wave number (3000 cm⁻¹) is due to the stretching vibration of hydrogen bonds between interlayer water and carbonate anion also in interlayer.²⁰ The band located at around 1630 cm⁻¹ is attributed to the H₂O scissoring mode. The sharp band at 1360 cm⁻¹ is due to the mode antisymmetric stretching of interlayer carbonate, which shifted to higher frequencies due to the ferric incorporation. Another band at around 860 cm⁻¹ is attributed to hindered rotations and translations of carbonates. These bands are present on the different samples confirming the presence of carbonate anions in the interlayer zone. The bands observed in the lower wave number region inferior of 800 cm⁻¹ are due to the interaction between metal and oxygen (M–O)⁴⁰ or between metal and hydroxyl (M–OH).¹⁵ The assignment of these bands was difficult but the band at around 430 cm⁻¹ is attributed to the Al–O vibration. The band at 650 cm⁻¹, 1000 cm⁻¹ might be caused the Fe–O vibration¹⁷ and FeO–OH,⁴¹ which increases markedly after Fe doped. These observations were in agreement with XRD data.

Fig. 2 FTIR spectra of CoAlFe-HT precursors.

3.1.3 Thermal analysis (DTG). The differential thermal analysis curves for all samples are included in Fig. 3. As is known, the destruction of the hydrotalcite structure is physisorbed water, (2) elimination of the interlayer water, (3) brucite layer dehydroxylation and (4) loss of interlayer anions.⁴² It can be seen from the Fig. 3 that the hydrotalcites samples display two similar mass loss stages, in agreement with previous study.⁴³ The first weight loss at lower temperatures (50–250 °C) corresponded to the removal of inter particle pore water, crystallization water, and weakly bonded hydroxyl groups without collapse of the hydrotalcite structure.⁴⁴ The second peak at higher temperatures (250–400 °C) is due to the dehydroxylation of the brucite-like sheets as well as the decomposition of the interlayer carbonate anions during which the layered structure was partially destroyed (partial overlap).⁴⁵ The weight loss stages are further categorized according to the actual temperature of the weight loss shown in Table 2. The replacement of trivalent cation in the hydrotalcite structure influences the thermal stability. By comparing the different DTG signals, the most stable sample is CAF0-HT, indeed, its decomposition in metallic oxide ends in a close temperature of 356 °C while the CAF4-HT decomposition ends at 290 °C. Moreover, after Fe introduced into hydrotalcite structure, the two weight loss stages were shifted to low temperature range. With the increase of Fe content, the weight loss in the second stages were decreased from 15.9% for the CAF0-HT to 10.1% CAF4-HT, which means that Fe introduced into hydrotalcite structure significantly accelerate decomposition of interlayer anions. The accelerated decomposition can be attributed to distortion of brucite-like sheets modified by substitution of Al cations with transition metals (Fe), resulting in weakening of anions bounding by hydrotalcite layers.⁴⁶ When the temperature exceeds 500 °C, these hydrotalcites samples were transformed into the corresponding mixed oxides.

Fig. 3 DTG of CoAlFe-HT precursors.

Table 2 Weight loss associated with DTG of the hydrotalcite precursors

Sample	Weight loss (%)		Peak temperature (°C)
	1^st stage	2^nd stage	1^st stage	2^nd stage
CAF0-HT	8.7	15.9	196	297
CAF1-HT	10.2	15.3	198	291
CAF2-HT	5.9	12.7	191	273
CAF3-HT	7.1	11.6	189	261
CAF4-HT	5.6	10.1	182	246

---

3.2 Characterization of calcined CoAlFe hydrotalcites

3.2.1 XRD analysis. After calcination at 500 °C, the CoAlFe mixed oxides were obtained as depicted by XRD patterns in Fig. 4. Obviously, calcination has destroyed the layered structure of hydrotalcites as no characteristic diffractions of hydrotalcites are present in the XRD patterns for the calcined materials. No other crystalline phases but spinles like oxides are detected in the XRD patterns, which indicates that all the samples are single phases and the positions of the characteristic (111), (220), (311), (222), (400), (422), (511) and (440) reflections along with relative intensity of peaks are similar. For CAF0, a mixture of three cobalt-spinel phases, Co₃O₄ (JCPDS 43-1003), CoAl₂O₄ (JCPDS 44-0160), and Co₂AlO₄ (JCPDS 38-0814), may be present because it is difficult to differentiate by XRD, which is in line with the previous reports.^15,17,19 After Fe introduced, the presence of Fe₃O₄ (JCPDS 26-1136) and CoFe₂O₄ (JCPDS 22-1086) can't be excluded due to the similar diffraction peaks. The presence of such spinel phases was also confirmed by H₂-TPR study (see below). Thus, a gradual incorporation of Fe into the spinel lattice at moderate calcinations temperature can be obtained in the Co–Al–Fe systems via hydrotalcites routes. Taking the larger contents of Co in the precursors into account, spinel-like Co₃O₄ should be predominate oxides phase. The crystallite sizes of the oxides samples were in the range 8–10.5 nm, as estimated from the peak width using Debye–Scherrer equation.

Fig. 4 XRD patterns of CoAlFe mixed oxides.

3.2.2 SEM analysis. The morphological properties of the catalysts were investigated by SEM as shown in Fig. S1.† From the SEM images, it can be seen that after calcined at 500 °C, the layered structure of the precursor material HT collapses, and the pretended needle-like morphology of the particles is corresponding to projections of platelets.⁴⁷ The sample features a shaggy surface, composed of small needle-like particles with a large degree of porosity. To reveal more information about the pore structure of the calcination samples, N₂ physisorption has been studied in the present work.

3.2.3 FTIR analysis. The FTIR spectra, which help to identify the formation of spinel structure is shown in Fig. 5. The two main bands at approximately 1000–400 cm⁻¹ are observed and attributed to the M–O stretching vibration in the spinel structures.⁴⁸ For the CAF0 sample, the two bands at 667 and 564 cm⁻¹ can be attributed to Co²⁺ in a tetrahedral environment and Co³⁺ in an octahedral environment, respectively.²⁰ Compared with ν(Co–O) modes typical of Co₃O₄ (670 and 571 cm⁻¹), the red shift of IR frequency in CAF0 can be explained by some dissolution of Al³⁺ into Co₃O₄ to form a Co(Al)O solid solution.⁴⁹ After Fe introduced into the cobalt-spinels, the M–O stretching vibration was shifted gradually towards lower frequency region, which may be due to substitution of the Al with Fe altering the nature of the interaction and thereby influencing the band position.⁵⁰ Besides, small bands at around 945 and 490 cm⁻¹ became intense with the increase of Fe content. The bands at around 945 cm⁻¹ in the samples are characteristic of the cobalt ferrite systems and could be due to FeO–OH.⁴¹ The bands around 490 cm⁻¹ are assigned to the ν(Fe–O) mode, which belongs to the characteristic peaks of Fe₂O₃.⁵¹

Fig. 5 FTIR spectra of CoAlFe-500 sample.

3.2.4 N₂ adsorption and desorption characterization. N₂ physisorption experiments were carried out to examine the texture characteristics of the calcined mixed oxides in Fig. S2.† All the samples presented adsorption isotherm type IV in the IUPAC characterization. Adsorption isotherms of this type are representative of mesoporous materials with no or few micropores and strong interaction between adsorbent and adsorbate molecules. The hysteresis loops at a high relative pressure indicate a capillary condensation of adsorbate in the mesoporous and macropores of the solids. The hysteresis loop for the mixed oxides samples is of the H3 type, which is usually given by adsorbents containing slit-shaped pores, being in agreement with the morphology of HT materials. As calculated from the pore size distribution curves, most of the pores fall in meso size range (2 nm < d < 50 nm). The BET specific surface area (S_BET), total pore volume and pore size of the mixed oxides determined from isotherms are listed in Table 3. Obviously, the composition slightly affects the three textural parameters. The surface areas of the prepared catalysts with different Al : Fe ratios ranged from 90 to 117 m² g⁻¹. The introduction of iron leads to the increase of the surface area of and the decrease of pore size.

Table 3 Textural properties and catalytic performances of the oxides catalysts

Samples	SBET (m^2 g^−1)	Pore volume (cm^3 g^−1)	Pore size (nm)	T50 (°C)
CAF0	90.1	0.46	20	76
CAF1	98.3	0.47	19	50
CAF2	116.9	0.43	14	52
CAF3	111.2	0.38	13	54
CAF4	103.7	0.35	13	86

---

3.2.5 H₂-TPR analysis. The H₂-TPR profiles of CoAlFe mixed oxides, presented in Fig. 6, show that the materials possess considerably different reducibilities. For all the samples, two reduction domains were observed in low-temperature (LT) and high-temperature (HT) ranges. The former part consists of two reduction peaks associated with reduction of surface oxygenated species (∼140 °C)⁵² followed by reduction of Co₃O₄ to CoO (∼365 °C). The latter broad peak (450–900 °C) is attributed to the reduction of CoO to Co as well as Co cations surrounded by Al or Fe ones in a spinel-like phase.²⁰ After incorporation of Fe into the system, the oxide samples displays a shoulder peak at about 390 °C in the LT domain assigned to the reduction of Fe₃O₄, which is visible on higher Fe content samples. In the case of ternary mixed oxides, the overlapping reduction peak indicates the strong interactions among Co, Al and Fe. The complexity of the profiles is due to the multiphase compositions of the investigated mixed oxides as discussed in XRD results. Additionally, it can be seen from Fig. 6(b) that the reduction of surface oxygenated species was shifted to lower temperature suggesting a better reducibility after Fe introduced. The improvement in low-temperature reducibility of CoAlFe mixed oxides would be beneficial for the enhancement in catalytic oxidation reactions.

Fig. 6 H₂-TPR profiles of CoAlFe mixed oxides samples.

3.2.6 Activity in CO oxidation. Fig. 7 displays the temperature dependence of CoAlFe mixed oxide catalysts in CO oxidation. As expected the increase in the reaction temperature increased the CO conversion. All the samples can completely oxidize CO into CO₂ at certain temperatures. The light-off curves rise slowly with increasing temperature, which is typical for a Mars-van Krevelen (MVK) mechanism involving the consumption and regeneration of surface and lattice oxygen species.⁵³ In the case of the binary oxides, CoAl oxide (CAF0) showed higher activity than CoFe oxide (CAF4), which may be related to the larger amount of oxygen species reducible at ∼140 °C. Compared with the binary oxides, all the ternary oxides exhibited higher activity of CO oxidation in the whole temperature range, complete CO conversion at about 80 °C. At the same time, the ternary oxides exhibited higher activity than Co₃O₄ either. This indicated the promotion effect of Fe incorporation into the Co-based nonstoichiometric spinels, which can be ascribed to the improved low temperature reducibility and enlarged surface areas. The temperature at 50% CO conversion (T₅₀) was chosen as a measure of catalytic of the spinel oxides. The catalytic performance order can be established as: CAF1 > CAF2 > CAF3> CAF0 > CAF4. It can be seen from Fig. 7 and Table 2 that the ternary oxides displayed close activities. Combined with XRD and TPR results, the active oxygen species responsible of CO oxidation in low temperature ranges may mainly come from surface Co³⁺ ion. The transition metal ion coordination has a great influence on the catalytic activity of CO oxidation over the Co-based nonstoichiometric spinels.

Fig. 7 Efficiency of the catalytic oxidation of CO over CoAlFe-500.

3.2.7 In situ FTIR study. In order to identify the surface species of CoAlFe mixed oxides responsible for the exceptional high activity, CO was used as probe molecule to conduct in situ FTIR measurement as shown in Fig. 8. In order to present the change of catalyst surface we exposed it into stream of reactant gas (0.4 vol% CO/He) for about 30 min which was named A-xmin, and then the CO concentration was switched from 0.4 vol to 0.0 vol% which was named D-xmin. As shown in IR spectra for the adsorption and desorption of carbon monoxide at room temperature, two large absorption bands centered at 2176 and 2118 cm⁻¹ can be observed immediately after the CO gas was introduced. The band of 2176 cm⁻¹ lies at a higher frequency than that of gaseous CO (2143 cm⁻¹), which suggests the CO molecule is absorbed in a relatively high oxidation state cation through donating its electrons in this σ-type coordination bond.^54,55 This band can be therefore assigned to the CO adsorption in the coordinatively unsaturated Co³⁺ sites.⁵⁶ The band centered at 2118 cm⁻¹, as reported previously in the literature can be ascribed to the CO adsorbed on Co²⁺ sites. These assignments are also in line with the fact that Co₃O₄ is a predominant surface phase in the CAF2 sample. After CO adsorption saturation at room temperature, the intensities of cationic CO vibration peaks (2176 and 2118 cm⁻¹) are reduced significantly upon the helium purging.

Fig. 8 Evolution of different IR bands in the interval 1000–2500 cm⁻¹ for a CAF2 catalyst exposed to 0.4 vol% CO in He (which named A-xmin) while switching the CO concentration from 0.4 to 0.0 vol% at t = 30 min (which named D-xmin).

CO adsorption and desorption on catalyst surfaces at room temperature show some remarkable effects on the CO₂ IR intensities at 2363 and 2338 cm⁻¹. The observed oscillation in IR peak intensities indicates a complex CO adsorption and its reaction with lattice oxygen. At the same time, another two bands are observed at 1638 and 1263 cm⁻¹, which are the characteristic vibrations of surface bidentate carbonate species. They are formed through the adsorption of CO₂ on the catalysts surface. Thus, the CO adsorption and desorption on catalyst surfaces at room temperature are accompanied by the obvious formation and decomposition of surface carbonate species. Through observation, it is not difficult to see the surface carbonate remain present after evacuation.

In situ FTIR spectra of the CAF2 samples with CO adsorbed at different temperatures in the presence of 2000 ppm CO are shown in Fig. 9. When the temperature was increased, the bands at 2176 and 2118 cm⁻¹ ascribed to cationic CO vibration were also increased gradually, reaching its strongest intensity at about 250 °C. On heating to 300 °C, this band became weaker due to desorption of adsorbed CO. The dynamic equilibrium also exists after the temperature heating to 300 °C. After the sample was heated, the intensities of the adsorbed CO decreased gradually after 250 °C due to the reaction with the weakly bound active oxygen on Co₃O₄, but the formation of CO₂ appeared gradually. The band at 2363 and 2338 cm⁻¹ have been attributed to the asymmetric stretch of physisorbed CO₂. It is noticed that the intensity of the bands increases in proportion to the temperature. This observation may therefore result from a partial reduction of Co₃O₄ by CO. The weakly bound active oxygen on the sample surface can easily be “swept out” simultaneously by the weakly adsorbed CO to produce CO₂. The thermal treatment clearly affects the mobility of the active oxygen. The higher the temperature the more mobile the active oxygen becomes and the easier the oxidation of CO. Also, some of the adsorbed CO migrates to form stable surface bidentate carbonate species which remain present after evacuation with He. After the CAF2 sample was heated, the intensities of the band at 1263 cm⁻¹ attributed to surface bidentate carbonate increased gradually. Apparently, the slight change of the band intensities indicates the influence of the reaction of CO with the active oxygen on sample. This behavior indicates that parts of the CO₂ was formed probably due to the interaction of bidentate carbonate with oxygen.

Fig. 9 In situ FTIR spectra of CAF2 sample with CO adsorbed at various temperatures.

4. Conclusions

In the present study, CoAlFe nonstoichiometric spinel type mixed oxides derived from hydrotalcites were shown to be highly active catalysts for the CO oxidation. Thermal decomposition hydrotalcites precursors at 500 °C leads to spinel-type oxides with specific surface area ranged from 90 to 117 m² g⁻¹.

Compared with the binary oxides and Co₃O₄, CoAlFe ternary oxides displayed much larger catalytic activity for CO oxidation, with complete CO conversion at about 80 °C. The promotion effect of Fe incorporation into the Co-based nonstoichiometric spinels can be ascribed to the improved low temperature reducibility and enlarged surface areas.

From in situ FTIR spectrum, when the catalyst was exposed to stream of reactant gas (0.4 vol% CO/He) at room temperature, CO can be adsorbed on the catalysts surface as cationic CO vibration peaks, accompanied by the obvious formation and decomposition of surface carbonate species. At the same time, the small physisorbed CO₂ peaks were observed which means the adsorbed CO reacted with surface lattice oxygen in oxide catalyst with formations of CO₂. Except for the slightly increased intensity of CO₂, CO/O₂ co-adsorption was studied by testing the sample. After increasing temperatures, reactions between the adsorbed CO and active oxygen have been observed in the oxide catalyst.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (no. 21007019 and 21277059), Shandong provincial key research project (no. 2014GSF117039 and 2015GSF117025) and Shandong provincial natural science foundation, China (2014ZRB01268).

References

1. A. Biabani-Ravandi, M. Rezaei and Z. Fattah, Chem. Eng. Sci., 2013, 94, 237–244.
2. S. M. El-Sheikh, F. A. Harraz and K. S. Abdel-Halim, J. Alloys Compd., 2009, 487, 716–723.
3. X. Liu, R. Wang, L. Song, H. He, G. Zhang, X. Zi and W. Qiu, Catal. Commun., 2014, 46, 213–218.
4. A. Sandoval, A. Aguilar, C. Louis, A. Traverse and R. Zanella, J. Catal., 2011, 281, 40–49.
5. V. P. Santos, S. A. C. Carabineiro, J. J. W. Bakker, O. S. G. P. Soares, X. Chen, M. F. R. Pereira, J. J. M. Órfão, J. L. Figueiredo, J. Gascon and F. Kapteijn, J. Catal., 2014, 309, 58–65.
6. J. Luo, M. Meng, X. Li, X. Li, Y. Zha, T. Hu, Y. Xie and J. Zhang, J. Catal., 2008, 254, 310–324.
7. Y. Liu, H. Dai, J. Deng, S. Xie, H. Yang, W. Tan, W. Han, Y. Jiang and G. Guo, J. Catal., 2014, 309, 408–418.
8. A. Holmgren, B. Andersson and D. Duprez, Appl. Catal., B, 1999, 22, 215–230.
9. L. Cai, Z. Hu, P. Branton and W. Li, Chin. J. Catal., 2014, 35, 159–167.
10. O. P. Tkachenko, A. A. Greish, A. V. Kucherov, K. C. Weston, A. M. Tsybulevski and L. M. Kustov, Appl. Catal., B, 2015, 179, 521–529.
11. I. V. Lukiyanchuk, V. S. Rudnev, I. V. Chernykh, I. V. Malyshev, L. M. Tyrina and M. V. Adigamova, Surf. Coat. Technol., 2013, 231, 433–438.
12. X. Xie, Y. Li, Z. Q. Liu, M. Haruta and W. Shen, Nature, 2009, 458, 746–749.
13. S. C. Kim and W. G. Shim, Appl. Catal., B, 2009, 92, 429–436.
14. K. Everaert and J. Baeyens, J. Hazard. Mater., 2004, 109, 113–139.
15. M. Mokhtar, S. N. Basahel and Y. O. Al-Angary, J. Alloys Compd., 2010, 493, 376–384.
16. C.-T. Chang, B.-J. Liaw, Y.-P. Chen and Y.-Z. Chen, J. Mol. Catal. A: Chem., 2009, 300, 80–88.
17. E. Genty, R. Cousin, S. Capelle, C. Gennequin and S. Siffert, Eur. J. Inorg. Chem., 2012, 2012, 2802–2811.
18. O. Saber and T. Zaki, J. Chem. Sci., 2014, 126, 981–988.
19. M. Gabrovska, R. Edreva-Kardjieva, K. Tenchev, P. Tzvetkov, A. Spojakina and L. Petrov, Appl. Catal., A, 2011, 399, 242–251.
20. C. Gennequin, S. Siffert, R. Cousin and A. Aboukaïs, Top. Catal., 2009, 52, 482–491.
21. K. Jirátová, P. Čuba, F. Kovanda, L. Hilaire and V. Pitchon, Catal. Today, 2002, 76, 43–53.
22. F. Kovanda, K. Jirátová, J. Rymeš and D. Koloušek, Appl. Clay Sci., 2001, 18, 71–80.
23. Z. Wang, X. Zhang, L. Wang, Z. Zhang, Z. Jiang, T. Xiao, A. Umar and Q. Wang, Sci. Adv. Mater., 2013, 5, 1449–1457.
24. J. Zhao, L. Yang, T. Chen and F. Li, J. Phys. Chem. Solids, 2012, 73, 1500–1504.
25. F. Li, J. Liu, D. G. Evans and X. Duan, Chem. Mater., 2004, 16, 1597–1602.
26. A. G. Kochur, A. T. Kozakov, K. A. Googlev, S. P. Kubrin, A. V. Nikolskii, V. I. Torgashev, A. A. Bush, V. Y. Shkuratov and S. I. Shevtsova, J. Alloys Compd., 2015, 636, 241–248.
27. Y.-Z. Wang, Y.-X. Zhao, C.-G. Gao and D.-S. Liu, Catal. Lett., 2007, 116, 136–142.
28. D. Gu, C.-J. Jia, C. Weidenthaler, H.-J. Bongard, B. Spliethoff, W. Schmidt and F. Schüth, J. Am. Chem. Soc., 2015, 137, 11407–11418.
29. D. A. H. Cunningham, T. Kobayashi, N. Kamijo and M. Haruta, Catal. Lett., 1994, 25, 257–264.
30. N. R. E. Radwan, M. S. El-Shall and H. M. A. Hassan, Appl. Catal., A, 2007, 331, 8–18.
31. L. Simonot, F. o. Garin and G. Maire, Appl. Catal., B, 1997, 11, 167–179.
32. F. Grillo, Appl. Catal., B, 2004, 48, 267–274.
33. C.-B. Wang, C.-W. Tang, S.-J. Gau and S.-H. Chien, Catal. Lett., 2005, 101, 59–63.
34. M. Pollard, B. Weinstock, T. Bitterwolf, P. Griffiths, A. Piersnewbery and J. Paineiii, J. Catal., 2008, 254, 218–225.
35. J.-y. Luo, M. Meng, Y. Qian, Z.-Q. Zou, Y.-n. Xie, T.-d. Hu, T. Liu and J. Zhang, Catal. Lett., 2007, 116, 50–56.
36. J. Jansson, J. Catal., 2000, 194, 55–60.
37. C.-B. Wang, C.-W. Tang, H.-C. Tsai, M.-C. Kuo and S.-H. Chien, Catal. Lett., 2006, 107, 31–37.
38. Z. Jiang, Z. Hao, J. Yu, H. Hou, C. Hu and J. Su, Catal. Lett., 2005, 99, 157–163.
39. E. Bolshak, S. Abelló and D. Montané, Int. J. Hydrogen Energy, 2013, 38, 5594–5604.
40. J. Pérez-Ramírez, G. Mul and J. A. Moulijn, Vib. Spectrosc., 2001, 27, 75–88.
41. M. H. Habibi and H. J. Parhizkar, Spectrochim. Acta, Part A, 2014, 127, 102–106.
42. J. Pérez-Ramírez, G. Mul, F. Kapteijn and J. A. Moulijn, J. Mater. Chem., 2001, 11, 821–830.
43. V. Vágvölgyi, S. J. Palmer, J. Kristóf, R. L. Frost and E. Horváth, J. Colloid Interface Sci., 2008, 318, 302–308.
44. W. Cheng, W. Wang, Y. Zhao, L. Liu, J. Yang and M. He, Appl. Clay Sci., 2008, 42, 111–115.
45. N. Voyer, A. Soisnard, S. J. Palmer, W. N. Martens and R. L. Frost, J. Therm. Anal. Calorim., 2009, 96, 481–485.
46. L. Chmielarz, P. Kuśtrowski, A. Rafalska-Łasocha, D. Majda and R. Dziembaj, Appl. Catal., B, 2002, 35, 195–210.
47. P. Gao, R. Xie, H. Wang, L. Zhong, L. Xia, Z. Zhang, W. Wei and Y. Sun, J. CO2 Util., 2015, 11, 41–48.
48. F. Saffari, P. Kameli, M. Rahimi, H. Ahmadvand and H. Salamati, Ceram. Int., 2015, 41, 7352–7358.
49. L. Obalová, K. Jirátová, F. Kovanda, K. Pacultová, Z. Lacný and Z. Mikulová, Appl. Catal., B, 2005, 60, 289–297.
50. S. Kannan and C. S. Swamy, J. Mater. Sci., 1997, 32, 1623–1630.
51. J.-L. Cao, G.-J. Li, Y. Wang, G. Sun, X.-D. Wang, B. Hari and Z.-Y. Zhang, J. Environ. Chem. Eng., 2014, 2, 477–483.
52. L. Xue, C. Zhang, H. He and Y. Teraoka, Catal. Today, 2007, 126, 449–455.
53. J. Tian, H. Peng, X. Xu, W. Liu, Y. Ma, X. Wang and X. Yang, Catal. Sci. Technol., 2015, 5, 2270–2281.
54. G. Busca, R. Guidetti and V. Lorenzelli, J. Chem. Soc., Faraday Trans., 1990, 86, 989–994.
55. J. E. Bailie, C. H. Rochester and G. J. Hutchings, J. Chem. Soc., Faraday Trans., 1997, 93, 2331–2336.
56. G. Busca, V. Lorenzelli and V. Bolis, Mater. Chem. Phys., 1992, 31, 221–228.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra02204a

---