Formaldehyde (HCHO) was classified as a human carcinogen by the International Agency for Research on Cancer in June 2004 [1]. Long-term exposure to parts per million (ppm) HCHO concentrations thus represents a health hazard, and the short-term (30 min) exposure limit recommended by the World Health Organization is 0.1 mg/m3 [2]. HCHO is released from various products, such as chipboard, textiles, and decorative materials, and can also contribute to the formation of photochemical smog [3]. Many efforts have been made to devise methods of reducing indoor HCHO concentrations to satisfy the stringent environmental regulations [4-6]. Conventional absorbents (such as activated carbon) are a ready means of eliminating HCHO, but the associated challenges of waste disposal and frequent replacement cannot be ignored. Photocatalysis allows the ongoing degradation of low HCHO concentrations, although the limited capacity of this approach and the inevitable formation of undesirable by-products restricts its practical application [7]. As such, high efficiency, low-temperature catalytic oxidation is believed to represent the most promising technology for HCHO removal [8, 9].
Initially, supported noble metal catalysts were applied to HCHO oxidation at room temperature [10-13], but the high cost of these metals restricts their large-scale application. Fortunately, oxides of transition metals, including Mn, Co, and Cu, also exhibit outstanding catalytic performance for low- temperature HCHO oxidation. Among these, Co3O4 has been widely investigated as a component of heterogeneous catalysts because of its high performance [12-18]. Many researchers have found that the catalytic activity of Co3O4 varies depending on the method used to prepare the catalyst. Zhu et al. [19] synthe- sized a material consisting of Co3O4 supported on ZSM-5 (Co3O4/ZSM-5) for the oxidation of propane, using impregnation (IM), deposition precipitation (DP), and hydrothermal (HT) methods. The catalytic activities of the resulting materials were in the order of HT > DP > IM, and it was also determined that the catalyst prepared with ammonium bicarbonate as the precipitant was superior to one prepared with NaOH. Shi et al. [20] produced a MnxCo3-xO4 catalyst for HCHO oxidation as a solid solution via both co-precipitation and citric acid methods. They found that the sample prepared by co-precipitation exhibited the best catalytic performance and was able to completely oxidize HCHO at 75 ℃. The addition of specific alkali elements, such as K and Na, has also proven to be effective for promoting the catalytic oxidation of HCHO, by increasing the concentration of OH? on the catalyst surface [21-23].
The precipitation method is popular and also practical with regard to eventual scaled-up applications. Therefore, it is necessary to both research and develop the synthesis of transition metal-based catalysts utilizing different precipitation methods. It would also be beneficial to further research the effect of the precipitation method on catalyst performance for low- temperature HCHO oxidation. Therefore, in the present study, various precipitants were used to generate Co3O4 catalysts, and the effects of these precipitants on the subsequent HCHO oxidation performance were discussed.
All chemicals used in this work were analytical grade and were purchased from the Sinopharm Chemical Reagent Co. (SCRC). The NH3-Co, KOH-Co, NH4HCO3-Co, K2CO3-Co, and KHCO3-Co catalysts were prepared using a precipitation method, employing five precipitants: NH3·H2O, KOH, NH4HCO3, K2CO3, and KHCO3. In this method, a solution of the chosen precipitant (2 mol/L) was added dropwise (10 mL/min) to a solution of Co(NO3)2·6H2O (100 mL, 0.1 mol/L) with rapid stirring at room temperature until the reaction solution had a pH value of 9. After standing for 4 h, the resulting precipitate was removed by filtration and washed with deionized water until the wash water was neutral, then dried at 100 ℃ for 10 h and calcined at 400 ℃ for 2 h. For comparison purposes, a sample designated as PC/AHC-Co was prepared by dispersing HN4HCO3-Co in a K2CO3 solution (2 wt%) with stirring for 30 min, followed by drying of the resulting product overnight at 100 ℃.
All samples were pretreated at 200 ℃ for 1 h prior to catalytic trials. X-ray diffraction (XRD) patterns of the catalysts were acquired using a D8 ADVANCE (Bruker) X-ray diffractometer with Cu Kα radiation. The K content of each sample was determined by atomic absorption spectroscopy (AAS). N2 adsorption- desorption isotherms were obtained used the Barrett- Joyner-Halenda (BJH) method with a TriStar II 3020 apparatus (Micromeritics) and X-ray photoelectron spectroscopy (XPS) spectra were acquired on an AXIS Ultra DLD instrument (Kratos) at 300 W using Mg Kα excitation. H2 temperature- programmed reduction (TPR) data were obtained with a Chemisorb 2720 TPX apparatus (Micromeritics). In these tests, a 0.1-g sample (40?60 mesh) was pretreated under a N2 flow at 400 ℃ for 30 min in a quartz reactor. The cooled sample was subsequently reduced under a flow (40 mL/min) of 5 vol% H2 in Ar from 20 to 700 ℃ (10 ℃/min). In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) data were acquired on a Nicolet 6700 FTIR spectrometer equipped with an MCT detector and a DRIFTS cell (Harrick), scanning from 4000 to 800 cm-1 with a resolution of 4 cm-1. Samples (30 mg) were assessed under a flow of 100 ppm HCHO and 21 vol% O2 in N2 at 100 mL/min total flow. Prior to the FTIR characterization, catalysts were pretreated under N2 for 1 h at 300 ℃.
Catalytic reactions were carried out in a U-shape fixed-bed quartz tubular reactor with an inner diameter of 4 mm. The catalyst sample (100 mg, 40-60 mesh) was placed between two quartz wool layers in the tube and mass flow controllers were used to set the gas flow rates. HCHO vapor was obtained by passing a flow of N2 through a paraformaldehyde (99%, SCRC) solution in a container within a water bath. The concentration of HCHO was adjusted by varying the N2 flow rate and/or the temperature of the water bath. By mixing this N2 stream with a flow of O2 in N2, a typical feed gas composition (100 mL/min, GHSV of 69000 h-1) containing 100 ppm HCHO and 21 vol% O2 was obtained.
The concentrations of CO and CO2 in the outgoing gas stream were measured by gas chromatography (GC 9560, HUAAI, flame ionization detector with a CH4 conversion oven, 5Å molecule sieve and TDX-01 packed columns). Experimental data were recorded beginning at the point at which the reaction stabilized at each reaction temperature and data acquisition continued over the span of 1 h. The HCHO conversion values were calculated using the equation
HCHO conversion = [CO2]out/[HCHO]in × 100%
where [CO2]out is the CO2 concentration in the outlet stream (vol%), and [HCHO]in is the inlet HCHO concentration (vol%).
Figure 1 summarizes the catalytic activities during HCHO oxidation over different catalysts. It is evident that 100% of the HCHO was oxidized to CO2 over the KHCO3-Co and PC/AHC-Co at 90 ℃, and so these two catalysts had the best low- temperature activity in this study. In contrast, the KOH-Co and NH3-Co catalysts achieved 100% HCHO conversion at 120 and 130 ℃, respectively. The light off temperature (T10) obtained with the KHCO3-Co was only 60 ℃, much lower than the values for the NH4HCO3-Co (90 ℃) and NH3-Co (110 ℃). The PC/AHC-Co performance was significantly better than that of the NH4HCO3-Co because of the immersion process applied during the preparation of the catalyst.
The textures of the catalysts were characterized by XRD and N2 physisorption, and the XRD patterns of the as-prepared Co3O4 samples are shown in Fig. 2. The peaks at 2θ = 31.3°, 36.8°, 38.5°, 44.8°, 55.6°, 59.5°, 65.2° and 77.3° correspond to the (220), (311), (222), (400), (422), (511), (440) and (533) planes, respectively. All these diffraction peaks are in good agreement with those of spinel Co3O4 (JCPDS 65-3103) and do not indicate any impurity phases. The average crystal sizes of these catalysts, as calculated from the XRD data (FMWH of the (311) peak) using the Scherrer equation, ranged from 17 to 23 nm (Table 1).
The N2 adsorption-desorption isotherms of the catalysts are shown in Fig. 3. Each catalyst generated a typical IV isotherm and a type H1 hysteresis loop, indicating that the catalyst pores had a narrow distribution and that these materials each had a mesoporous (2 to 50 nm) structure [24]. The BET surface areas, total pore volumes and pore sizes were found to vary depending on the precipitant. Comparing the NH3-Co and KOH-Co with catalysts precipitated using carbonate and bicarbonate reagents, the latter had the higher physical parameters. The significant increase in the adsorption capability indicates that the specific surface area was enhanced due to the decomposition of cobalt carbonate during the calcination process. The PC/AHC-Co catalyst shows reduced structural parameters compared with the NH4HCO3-Co, which is attributed to either clogging or covering of the NH4HCO3-Co pores after K2CO3 loading [16]. The residual K concentration in the KHCO3-Co was 15 times that of the K2CO3-Co, although the peak position demonstrates that the K+ cations did not enter the bulk phase Co3O4 or replace the Co2+ ions because of the space (rK+ = 0.152 nm, rCo2+ (hs) = 0.0885 nm and rCo3+ (hs) = 0.075 nm) [25] and charge factors. That is, if the Co2+ ions in the Co3O4 lattice had been replaced by K+, the lattice parameters of the Co3O4 would have been altered accordingly. These results are similar to those reported previously by Wu et al. [26] and Park et al. [27]. As a result of the formation of carbonic acid and hydroxide ions, a hydrogen carbonate solution will slowly generate CO2 at room temperature [28]. In the presence of Co2+, the hydrolyza- tion rate will increase substantially and so more CO2 will be released. The CO2 gas bubbles thus produced could provide aggregation centers for the precipitation reaction, creating loose hydroxide carbonate precursors that might include K+ [29]. During the calcination process, these hydroxide carbonate precursors will decompose to CO2 and H2O, refining the particle size and increasing the porosity of the powder [30].
In Fig. 4, two obvious reduction peaks can be observed in the low (Ⅰ) and high (Ⅱ) temperature regions. There are different opinions concerning the reduction mechanism of Co3O4. Arnoldy et al. [31] stated that the reduction of Co3O4 consisted of only a single step, while many other researchers [32-34] regarded the reduction of Co3O4 as a two-step process involving the intermediate reduction of CoO. In the low temperature region, the Co3+ in Co3O4 was presumably reduced to Co2+ with the sample transitioning to CoO as an intermediate. In the high temperature section, this intermediate was further reduced to metallic cobalt. Peak Ⅰ (at 290 ℃) is therefore attributed to the reduction of Co3+ to Co2+, while peak Ⅱ (at 400 ℃) resulted from the reduction of CoO to Co. The KHCO3-Co showed the lowest reduction temperature (266 ℃) among all the samples examined, while the NH3-Co had the highest peak. The reduction temperatures of the others were in the order of KOH-Co > NH4HCO3-Co > K2CO3-Co > PC/AHC-Co, which is in good agreement with the observed differences in their catalytic HCHO oxidation activities. There have been several literature reports that the amount of Co3+ on the catalyst surface is related to the catalytic activity [35-37]. In the present work, it can be seen that the reduction temperature shifts from 300 to 283 ℃ upon applying a coating of K2CO3 (2 wt%) on the NH4HCO3-Co catalyst. The reason for this shift is explained below, based on XPS analysis.
Taking into account the structures and residual K contents of these catalysts, the KHCO3-Co, NH4HCO3-Co and PC/AHC-Co were chosen to investigate the effect of the precipitant on the surface chemical state. The Co 2p, K 2p and O 1s XPS data for these materials are presented in Fig. 5, and associated values are summarized in Table 2. As can be seen from Fig. 5(a) and Table 2, K was found in the KHCO3-Co and PC/AHC-Co, but not in the NH4HCO3-Co. The amounts of K on the surfaces of the KHCO3-Co and PC/AHC-Co were both approximately 3.3 at%, demonstrating the presence of residual K following the use of KHCO3 as the precipitant. Fig. 5(b) indicates two major oxygen species with O 1s binding energy (BE) values of 529.8 and 530.8 eV. The former results from surface lattice oxygen (Olatt) contained in the Co3O4 [38], while the latter is attributed to adsorbed surface oxygen (Oads) [39]. The Oads/Olatt ratio in the KHCO3-Co was 0.96 (Table 2), a value that is slightly lower than that of the NH4HCO3-Co (1.08). The Oads/Olatt ratio of the PC/AHC-Co (1.33) was relatively high, which is attributed to the presence of surface hydroxyl (OH) groups resulting from the hydrolysis of K2CO3 [22]. The peaks in the Co 2p XPS spectra (Fig. 5(c)) at 795.0 and 780.1 eV could represent Co 2p1/2 and Co 2p3/2 spin-orbital peaks [19, 40]. In this work, the Co 2p spin-orbit splitting value of the Co3O4 catalysts was 14.9 eV, which is close to that of Co3+ (15.0 eV) [41], therefore, the catalysts likely consisted of Co3O4. The Co3+/Co2+ ratio for the NH4HCO3-Co was 0.24, a value that is less than that of the PC/AHC-Co (0.42), indicating that a greater amount of Co3+ was present on the PC/AHC-Co surface exposed after the K2CO3 immersion. This could explain the shift in the reduction temperature, which decreased from 300 to 283 ℃ when K2CO3 was coated onto the NH4HCO3-Co catalyst. The Co3+/Co2+ ratio of the KHCO3-Co was 0.35, which is between the values for the NH4HCO3-Co and PC/AHC-Co. It appears that less of the KHCO3-Co surface was covered with K2CO3 compared with the PC/AHC-Co. Despite the much higher Co3+/Co2+ ratio of the PC/AHC-Co, the relatively low specific surface area of this material (Table 1) might explain why its catalytic activity was similar to that of the KHCO3-Co.
Figure 6 shows the in situ DRIFT spectra acquired during HCHO oxidation over the NH4HCO3-Co and KHCO3-Co catalysts. In Fig. 6(a), no peaks associated with adsorbed HCHO are seen in the NH4HCO3-Co spectrum, thus the HCHO was oxidized as soon as it was adsorbed on the surface of the catalyst. The peaks at 1593 and 1575 cm-1 are attributed to the asymmetric (COO) stretching vibration of formate species, while those at 1360, 1315 and 1276 cm-1 are assigned to the corresponding symmetric (COO) stretch [42]. The absorbance bands at 1480, 1434, 1132, 1097 and 1055 cm-1 are ascribed to dioxymethylene (DOM) species [43, 44] and are seen to increase significantly over the first few minutes, while those due to OH species at 3400 and 3342 cm-1 increased slightly. In addition, the absorptions at 3633, 3568, 3542, and 3373 cm-1 ascribed to other hydroxyl groups decreased to give negative peaks, presumably as these groups were consumed during the oxidation. When O2 was introduced and the temperature increased, similar spectra were obtained, as shown in Fig. 6(b). The bands resulting from DOM species were significantly weakened as the temperature increased. As well, the monodentate formate species (1593 and 1276 cm-1) appear to have been converted to bidentate formate (1575 and 1360 cm-1) [45] as the temperature was raised to 200 ℃, with the simultaneous appearance of a new monodentate species (1506 and 1248 cm-1). Surface carbonate and bicarbonate species (1543, 1335 and 1210 cm-1) [45, 46] were formed at 80 ℃ and the absorbance of the two types of hydroxyl groups returned to their initial values.
The behavior of the KHCO3-Co was simple. As shown in Fig. 6(c), the peak due to one type of hydroxyl group (3623 cm-1) decreased slightly with increasing temperature, while the other type (3373 cm-1) increased greatly during HCHO adsorption. The intensities of the DOM peaks (1434, 1412, 1132, 1099 and 1060 cm-1) increased rapidly in the initial stage, and the peaks from one type of bicarbonate (1593 and 1376 cm-1) also increased rapidly up to 60 min. At the oxidation stage shown in Fig. 6(d), the relative intensities of the surface carbonate species peaks (1543 and 1346 cm-1) were found to change, although no new peaks were generated. The hydroxyl species (3373 cm-1) and the formate species (1593 and 1376 cm-1) initially increased in intensity and then decreased on going from 40 to 160 ℃, while the peaks due to DOM decreased continuously. At 200 ℃, the KHCO3-Co spectrum indicates the presence of only a few OH groups (3373 cm-1), as well as DOM and carbonate species.
At lower temperatures (80 ℃), few DOM species could be converted to formate species due to absence of hydroxyl groups on the NH4HCO3-Co. In contrast, at higher temperatures (> 80 ℃), the decomposition of formate species became the key step. In the case of the KHCO3-Co, the DOM species generated via the adsorption of HCHO could be converted to formate species in the presence of K2CO3, and so the key step would be the oxidation of formate species to carbonate species. It is thought that hydroxyl groups are formed from the hydrolysis of surface K2CO3, and are then consumed during the reaction. Subsequently, these hydroxyl groups are replaced by the H2O generated by the reaction. Thus, surface alkaline hydrolysis is continued and accelerates the HCHO oxidation process.
Co3O4 catalysts prepared using various precipitants were studied with regard to low-temperature HCHO oxidation. It was found that the KHCO3 precipitated material (KHCO3-Co) exhibits the best catalytic efficiency, and is able to provide 100% conversion of HCHO to CO2 at 90 ℃. The AAS and XPS data indicate that the KHCO3-Co surface holds residual K. The structures of the Co3O4 catalysts obviously vary with different types of precipitants, such that those catalysts precipitated using reagents containing CO32- or HCO3- ions have higher BET surface areas and BJH pore sizes. H2-TPR and XPS analyses show that the KHCO3-Co and PC/AHC-Co have higher oxidizing abilities due to their relatively high surface Co3+/Co2+ molar ratios. In situ DRIFTS results suggest that few DOM species are converted to formate species due to the absence of OH groups, which can be regenerated by K+ and CO32- loaded on the surface of the Co3O4.
2016,
Vol. 37 Issue (6): 947-954 
PDF (1018 KB)
2016,
Vol. 37 Issue (6): 947-954 PDF (1018 KB)