CeO2 and CeO2-based oxides have been studied intensively as catalyst, metal support and promoter [1, 2, 3, 4, 5]. CeO2 is widely used in three-way catalysts because of its unique oxygen storage capacity (OSC). Depending on the nature and temperature of the pretreatment, CeO2 has versatile acid-base properties and can be applied in many organic reactions [6, 7]. Many synthesis approaches have been employed in the synthesis of ceria-based nano-materials with well-defined crystal planes. As early as 1988, Hsu et al. [8] reported the preparation of polycrystalline CeO2 spheres with a uniform size by forced hydrolysis of cerium salt in sulfuric acid followed by aging at a critical temperature. Mai et al. [9] reported the synthesis of shape- selective CeO2 nanorods, nanopolyhedrals and nanocubes by an alkali hydrothermal method. In 2011, it was reported for the first time that CeO2 nanosheets exposing (110) facets were successfully prepared by a facile hydrothermal method [10].
Density functional theory (DFT) studies have shown that the (111) facet of CeO2 has the lowest surface energy, followed by the (110) facet, and the (100) facet has the highest surface energy. Less energy would be required to form oxygen vacancies on a facet with a higher surface energy [11, 12, 13, 14]. Thus the selective exposure of high energy facets would facilitate the formation of oxygen vacancies, which in turn would improve the OSC of ceria. In addition, oxygen vacancy clusters promotes the reducibility and activity of ceria nanorods [15]. Clusters such as linear surface oxygen vacancies had proved helpful for the migration of oxygen from the bulk to the surface, and thus would make CeO2 more reducible [16].
Nowadays, much attention is paid to global warming, and it is desirable to design materials to store CO2 or develop process to convert CO2 into valuable chemicals. CO2 can be activated by reduced CeO2. One mechanism proposed for CO2 activation is that CO2 is reduced directly into CO on reduced CeO2 [17, 18]. However, Lykhach et al. [19, 20] reported that CO2 activation on reduced CeO2 led to the rapid formation of a carbonate and surface carboxylates. A DFT study of CO2 activation on reduced CeO2 (110) also proved that CO2 activation on reduced CeO2 (110) proceeded with CO2 adsorbed parallel to the surface by forming stable carbonate species [21]. Yang and coworkers reported that a Sm-doped CeO2 nanomaterial showed excellent CO2 capture ability. They attributed the CO2 adsorption ability to the abundant oxygen vacancies of Sm-doped CeO2 [22].
Glycol is a very useful organic solvent in metal and metal oxide preparation because of its strong reducibility. CeO2 prepared in its presence would have lots of oxygen defects on the surface, which would improve its OSC and gas capture ability. Here, an improved CeO2 nanomaterial was prepared by a glycol solvothermal method and its OSC property was studied. Its CO2 storage property was explored by a CO2 pulse technique and CO2 in situ diffuse reflectance infrared spectroscopy (DRIFTS).
CeO2 nanomaterials were prepared by three different methods with commercial reagents. Analytical grade Ce(NO3)3∙6H2O, NaOH, ethylene glycol and citric acid were purchased from Sinopharm Chemical Reagent Co., Ltd and used without further purification.
For the preparation of CeO2-GST, 5 mmol of Ce(NO3)3∙6H2O was added into 60 mL ethylene glycol with stirring for 20 min under ambient temperature. This was transferred into a 100 mL Teflon-lined stainless steel autoclave for crystallization at 180 °C for 24 h. The solid product was separated by centrifugation and washed with acetone three times and then dried at 50 °C overnight. This sample was marked as CeO2-GST. In most cases, the sample was calcined at 600 °C for 3 h with a heating rate for 1 °C/min.
The CeO2-nanorod was synthesized by a traditional hydrothermal method given in a previous report [9]. 5 mmol of Ce (NO3)3∙6H2O and 0.8 mol of NaOH were dissolved in 80 mL deionized water and stirred for 30 min at ambient temperature. The mixture was transferred into a Teflon-lined stainless steel autoclave for crystallization at 120 °C for 48 h. The product was filtered and washed with distilled water until it was neutral and then dried at 50 °C. This sample was denoted as CeO2-nanorod. In most cases, the sample was also calcined at 600 °C for 3 h with a heating rate for 1 °C/min.
For CeO2-CA, 5 mmol of Ce (NO3)3∙6H2O and 15 mmol of citric acid were dissolved in 80 mL deionized water. The solution was heated to 80 °C under stirring and kept for 12 hours at 80 °C in a closed environment to form a gel. After that, the gel was dried at 100 °C. The sample was calcined at 600 °C for 3 h with a heating rate for 1 °C/min and denoted as CeO2-CA.
X-ray diffraction (XRD) patterns were obtained on a Bruker D8 Focus diffractometer (Cu Kα radiation, l = 1.5406 Å) operated at 40 kV and 40 mA. BET specific surface areas were measured at -196 °C on a Quantachrome NOVA 4200e instrument. The sample was outgassed at 200 °C for 12 h before measurement. High resolution transmission electron microscopy (HR-TEM) images of the synthesized ceria were taken on a JEOL-2010 field emission transmission electron microscope. X-ray photoelectron spectroscopy (XPS) spectra were measured using a Thermo ESCA LAB-250 spectrometer with monochromatic Al Kα radiation. The binding energies were calibrated internally by the carbon deposit C 1s binding energy (BE) at 284.8 eV. The deconvolution of the XPS spectra used Gaussian functions.
H2 temperature-programmed reduction (H2-TPR) of the CeO2 samples (100 mg) was investigated by heating a CeO2 sample in a 5% H2/N2 flow (40 mL/min) at the heating rate of 10 °C/min from 30 to 600 °C. The hydrogen consumption was monitored by a thermal conductivity detector (TCD) using a zeolite for water adsorption. After the H2-TPR analysis, the same sample was cooled to 100 °C, and the gas flow was switched to 3% O2/He and the remaining H2/N2 was purged for 10 min. Then, the sample was re-oxidized by the O2/He mixture gas at 500 °C for 2 h and cooled to ambient temperature. Afterwards, the gas was switched to the H2/N2 flow again, which completed one H2 reduction-O2 oxidation cycle. Several cycles were repeated
CO pulse adsorption was investigated on Auto Chem Ⅱ2920 Micrometrics instrument using a mass spectrometer (MS) to detect the gas product. For each of the three samples, 50 mg of CeO2 was placed at the bottom of a U-shaped quartz tube which was first pretreated in an 5% O2/He flow (purity 99.999%, 30 mL/min) at 400 °C for 30 min. Then the sample was flushed with pure He. When the baseline was stable, the CO pulse experiment was carried out using 5% CO/He (30 mL/min) as the pulsed gas.
CO2 pulse experiments were also carried out on the Auto Chem Ⅱ2920 Micrometrics instrument using both MS and TCD as detectors. For each of the three samples, 25 mg of CeO2 was first pre-reduced by 10% H2/Ar flow (50 mL/min) at 600 °C for 30 min. Then the sample was cooled to 50 °C and flushed with pure He. After the baseline was stable, the CO2 pulse experiment was conducted using 5% CO2/He (30 mL/min) as the pulsed gas.
CO2 in situ DRIFTS was performed using a Thermo Nicolet Nexus 670 spectrometer equipped with a MCT detector and a high temperature reaction chamber with ZnSe windows. Prior to each CO2 adsorption experiment, ceria was first pretreated with 10% H2/N2 (50 mL/min) at 500 °C for 30 min, cooled down to 300 °C, and then finally flushed with He (80 mL/min) for 30 min. After that, the sample was cooled to 50 °C and the background spectra were recorded in flowing He for the subtraction from the sample spectrum for each measurement. The CO2 adsorption experiment was conducted at 50 °C using 5% CO2/He (80 mL/min). As soon as CO2 was flowed into the sample cell, the spectrum was collected in the range of 4000-650 cm-1 by accumulating 64 scans at a 4 cm-1 resolution, DRIFTS spectra were recorded every 2 min. After the sample was exposed to CO2 flow for 20 min, He flow was switched in to flush for 20 min. During the He flushing, spectra were also collected every two minutes.
Figure 1 shows the XRD patterns of the calcined ceria prepared by the three different methods. CeO2-GST, CeO2-nanorod, and CeO2-CA were all in the pure cubic phase (fluorite structure, JCPDS 34-0394, space group Fm-3m). The BET surface areas of CeO2-GST, CeO2-nanorod, and CeO2-CA were 79.0, 78.0, and 49.0 m2/g, respectively.
Figure 2 shows the HR-TEM images of the calcined ceria samples prepared by the different methods. For the sample prepared by the glycol solvothermal method (Fig. 2(a) and (b)), a hexahedral morphology was found for most of the CeO2 particles. The particle size was uniformly distributed at 10-15 nm and the particle was mainly enclosed by (111) facets. The morphology of CeO2 was not totally in accordance with that in Liang’s report, where it was almost sphere-like and could be tuned to be cubic by adding some water into the original solution [23]. In our experiment, Ce(NO3)3∙6H2O was dissolved directly in glycol and no water was added. The morphology of CeO2-GST would be affected and it finally formed the hexahedral shape. The CeO2-nanorods (Fig. 2(c) and (d)) showed a rod-like shape (20 nm across, 60-300 nm long) with a preferentially growth direction along the [110] direction, which exposed (100) facets. The CeO2-CA prepared by the sol-gel method (see Fig. 2 (e) and (f)) showed an irregular morphology and exposed (111) facets, with particle sizes that ranged from 5 to 20 nm.
XPS was conducted to investigate the surface oxidation state of CeO2-GST before and after calcination. The CeO2-nanorod and CeO2-CA samples were also tested for comparison. As presented in Fig. 3, the Ce 3d spectra were composed of eight peaks corresponding to four pairs of spin-orbit doublets. According to a work reported previously [24, 25, 26, 27], the Ce4+ spectrum was labelled with v, v", v''' and u, u", u''', which were the 3d5/2 and 3d3/2 photoelectron peaks for 4f2, 4f1, and 4f0 occupation, respectively; v' and u' represented the lower binding energy peaks of the 3d5/2 and 3d3/2 electrons for the 4f1/4f2 configuration in Ce3+. The peaks marked as u (900.6-901.0 eV), u" (907.5-907.7 eV), and u''' (916.6-916.9 eV) were from Ce4+ 3d3/2, while the peaks labelled as v (882.2-882.6 eV), v" (889.1-889.3 eV), and v''' (898.2-898.5 eV) were from Ce4+ 3d5/2. The couples corresponding to one of the two possible electron configuration of the final state of the Ce3+ species were labelled as u' (903.5-904.2 eV) and v' (885.1-885.8 eV). The proportion of Ce3+ cation ratio is summarized in Table 1, which was calculated using the equation in reference [24].
From the XPS results, uncalcined CeO2-GST showed the largest Ce3+ concentration of 63% (Table 1). The relative intensity of u' and v' attributed to Ce3+ was much higher than with the other samples and the u''' peak for Ce4+ could be barely seen. After calcination, surface Ce3+ was still preserved and present as 11.2% of the surface cations. The intensity of the u' and v' (Fig. 3(b) peaks decreased and the u''' peak attributed to Ce4+ increased correspondingly. For the three samples (Fig. 3(c) and (d)), the CeO2-nanorod showed the lowest surface Ce3+ ratio of 5% and CeO2-CA, which could be reduced by the reducing gas generated in the decomposition of citric acid, showed the highest surface Ce3+ ratio of 14.6%. CeO2-GST contained more Ce3+ than the sample from the traditional hydrothermal process because of the reduction of glycol by the valence change Ce4+ → Ce3+ [28, 29]. This in turn indicated that there were more oxygen vacancies on the surface of CeO2-GST than on the CeO2-nanorod.
The H2 reduction-O2 oxidation cycle experiment was designed to investigate the reversible redox behavior of the CeO2 samples. The oxygen release property of the samples was qualitatively analyzed by the CO pulse method using a MS as detector.
Figure 4 displayed the H2 reduction-O2 oxidation cycles of the three samples. The H2 consumptions for all three samples are listed in Table 2. For the first cycle, the CeO2-nanorod displayed the highest H2 consumption (470 μmol/g), followed by CeO2-GST (380 μmol/g) and CeO2-CA (369 μmol/g). After that, the H2 uptake for all three samples decreased. CeO2-CA showed the weakest ability for oxygen re-storage and the H2 uptake decreased from 369 to 118 μmol/g in just three cycles. The H2 consumption of CeO2-nanorod and CeO2-GST were similar from the second cycle. However, the rod-like structure was broken after five cycles (Fig. 5(a) and (b)), while CeO2-GST still remained intact even after cycling 10 times (Fig. 5(c) and (d)). The BET measurement showed that after the cycling, the surface area of the nanorod decreased from 78 to 6 m2/g, which was in agreement with the HR-TEM results. It should be pointed out that CeO2-GST presented the most stable reversible redox ability of the three samples. The oxygen re-storage activity began to be constant after the sixth cycle (Fig. 4(a) and Table 2). The oxygen vacancies would favor the migration of oxygen and thus continuously supply oxygen from the bulk to the surface, which in turn stabilized the H2 uptake capacity of CeO2-GST during the cycling. In addition, the glycol solvothermal condition improved structural stability of CeO2-GST, and the morphology was preserved after 10 cycles.
Figure 6 shows the MS signal of the CO pulse experiments for the different CeO2 samples at 400 °C. Each test with CO pulses lasted 16 pulses. It was seen that the CO peak for all of the three samples showed a linear growth in the first 10 pulses. After 13 CO pulses, the CO peak intensity showed no more growth for all the three samples. The integration of the peak areas represented the remaining CO amount. A linear relationship between the peak area and pulse number was obtained by fitting (Fig. 7). The intercept k is relevant to the oxygen release rate in our experiment and a larger value of k corresponded to a lower oxygen release rate. The k values of the CeO2 samples are shown in Table 1. The linear correlation for k is displayed in Fig. 7. The CeO2-nanorod exhibited the highest oxygen release activity, followed by the CeO2-CA and CeO2-GST.
According to first principle studies, the exposure of the more reactive (100) and (110) surfaces facilitate the formation of oxygen vacancies on the CeO2 surface [24, 25, 26]. This was verified by our results that the CeO2-nanorod had the highest oxygen release ability and H2 consumption (470 μmol/g) due to the exposure of more reactive (100) facets. Both CeO2-GST and CeO2-CA mainly exposed the most stable (111) facet (Fig. 2(a) and (c)). By analyzing the other two samples, it was found that the release rate for CeO2-GST was smaller than that of CeO2-CA, as was the surface Ce3+. Thus it can be speculated that the surface defects generated by the formation of Ce3+ on the CeO2 surface contributed to the OSC performance. A similar experiment conducted by Liu and co-workers made the same conclusion [15]. Thus, it can be deduced that the controlled synthesis of a CeO2 nanomaterial that exposed abundant vacancies would improve the oxygen release ability, thus enhancing the catalytic activity of CeO2.
Generally, CO2 capture is measured using an amine solvent, a membranes, or an ionic liquid [30, 31]. The biggest challenge for all traditional CO2 adsorption methods is the high regeneration energy for the adsorbent and high capital cost. Since CeO2 is a Lewis base, it is thermodynamically desirable for CO2 adsorption [30, 32]. In addition, CeO2 with abundant exposed oxygen vacancies is a potent surface for gas adsorption [33]. In this section, the CO2 capture ability of the CeO2 samples is reported. The CO2 adsorption state was also investigated by CO2 in situ DRIFTS.
To test the CO2 capture ability of the CeO2 samples at 50 °C, a CO2 pulse method was designed using TCD and MS to detect the products. It was found that no CO product was generated at 50 °C except for a small signal of 28 fragment peaks generated by CO2. Figure 8 shows the MS signal of CO2. The amount of adsorbed CO2 was calculated from the TCD signals. CeO2-GST showed the highest CO2 adsorption of 149 μmol/g and the CeO2-nanorod and CeO2-CA showed similar values, which were 86 and 90 μmol/g, respectively.
Prior to the CO2 in situ DRIFTS measurement, the sample was pretreated under a flow of 10% H2/N2 (50 mL/min) for 120 min at 500 °C and then switched to pure helium flow for 30 min at 300 °C (80 mL/min). After that, the temperature was cooled down to 50 °C and the background spectrum was recorded at this temperature. Then a gas mixture of 5% CO2/He (80 mL/min) was fed into the reactor chamber to flow through the sample and the spectra were recorded. The assignments for the peaks were based on previous studies and listed in Table 3 [34, 35, 36, 37, 38, 39].
Figure 9 shows the infrared spectra of CeO2 samples at 50 °C. For CeO2-GST, six distinct peaks at 856, 1021, 1217, 1289, 1397, and 1584 cm-1 in the region 800-2000 cm-1 were detected (shown in Fig. 9(a)) after CO2 was introduced into the sample cell. The peak increased when the sample was treated with CO2 for a longer time and all the peaks became constant after 20 min of CO2 exposure (Fig. 9(a)), which was regarded as that the sample had reached its saturation for CO2 adsorption. CeO2-CA showed a similar behavior to CeO2-GST except that all the peak intensities were weaker (Fig. 9(c)). The CeO2-nanorod showed quite different peaks in comparison with the other two samples. The bands at 1214 and 1392 cm-1 were much stronger and there were several new bands at 1458, 1542, and 1617 cm-1 (Fig. 9(b)). The peaks observed on CeO2-GST and CeO2-CA at 856, 1028, 1289, and 1584 cm-1 were assigned to the bidentate carbonate. The bands at 1612, 1392, and 1214 cm-1 were assigned to bicarbonate, the bands at 1217 (1214) and 1397 (1392) cm-1 were assigned to bridge carbonate, 1458 cm-1 was assigned to unidentate carbonate, and 1542 cm-1 was assigned to formate [35, 36].
After the samples were flushed with pure He for 10 min, some changes were observed. The peak at 1217 cm-1 (1214 cm-1 for CeO2-nanorod) almost disappeared for all three samples. This was attributed to the instability of the bridge carbonate on the CeO2 surface and the adsorbed bridge carbonate was easily removed by the He flush [34]. For the CeO2-nanorod, the peak at 1584 cm-1 did not appear and there was a new peak at 1563 cm-1 (Fig. 9(b)), which indicated that the reduced CeO2 was re-oxidized by CO2, making the highest wavenumber due to the v(OCO) mode of bidentate carbonate shift from 1584 cm-1 to 1563 cm-1 [36]. The peaks attributed to bicarbonate, unidentate carbonate and formate on nanorod were all preserved after helium flushing. It is worth noting that the intensity of the 1217 cm-1 peak on the nanorod was stronger than on the other two samples, which was due to the easier formation of bridge carbonate on the reduced (100) facet. For CeO2-GST, the band at 1584 cm-1 was changed slightly (Fig. 9(a)) and the peak at 1563 cm-1 was not found during the experiment. This indicated that there was no bicarbonate species on the surfaces of CeO2-GST and CeO2-CA.
From the stable species after the pure He gas flush, it was found that CO2 was activated by CeO2-nanorod forming more stable species such as unidentate carbonate, bicarbonate and formate. Since there was no signal of CO detected in the MS signal and no adsorbed CO IR peak at 2000-2100 cm-1. It can be deduced that CO2 re-oxidized the reduced CeO2-nanorod by forming stable carbonate species, which was in agreement with the DFT study on the reduced CeO2(110) facet [21]. Thus, the CeO2-nanorod would react CO2 rather than capture it. The IR spectra also suggested that CO2 was preferentially adsorbed on the reduced (111) facet in the form of the less stable bidentate carbonate. The differences in the band intensities between CeO2-GST and CeO2-CA were distinct, which was attributed to their different BET surface areas. Since the active site for CO2 adsorption is Ce3+, reduced CeO2(100) was so active that it would activate CO2 by oxidizing Ce3+ into Ce4+, which would limit CO2 capture. It was considered that CeO2-GST was more efficient for CO2 adsorption because of abundant oxygen vacancy clusters on its surface and the CeO2(111) facets also contributed to the CO2 capture ability of CeO2-GST by stabilizing the bidentate carbonate species.
A CeO2-GST sample prepared by a glycol solvothermal method had abundant surface oxygen vacancy clusters and exhibited excellent reversible redox ability. The oxygen vacancies generated in the solvothermal process were preserved even after calcination and they improved the reducibility of CeO2, thus providing a continuous oxygen supply from the bulk to the surface, which made this CeO2 stable during H2- consumption in the H2 reduction-O2 oxidation cycles. In contrast, the CeO2-nanorod displayed rather poor reversible redox performance Moreover, the CO2 storage measurements with the reduced CeO2 nanomaterials (CeO2-GST, CeO2-nanorod, CeO2- CA) showed that CeO2-GST had the highest CO2 storage capacity (149 μmol/g) of the three samples. CO2 adsorption on the CeO2 nanomaterials was also different, which was mainly due to the exposure of different planes on the different samples. The bidentate carbonate was preferentially adsorbed on the reduced (111) facet, while the reduced (100) facet favored the stable bridge carbonate. CeO2 with more vacancy clusters on the surface and with a larger BET surface area gave good CO2 adsorption capacity and thus is an excellent material for CO2 capture.
作为稀土氧化物中重要的一员, CeO2以及Ce基复合氧化物已经被人们广泛研究[1, 2, 3, 4, 5]. CeO2是一个很好的负载型催化剂载体, 同时也具有一些非常优异的催化性能. 例如, CeO2因其独特的储放氧性能而被用于汽车尾气净化器中的三效催化剂中; 通过预处理后的氧化铈表现出一定强度的碱性, 应用于催化有机缩合反应中[6, 7].
CeO2材料可通过多种成熟的方法合成. 早在1988年, Hsu等[8]在强酸体系中水解铈盐的前驱体合成了尺寸十分均一的多晶CeO2纳米球; Mai等[9]在强碱性溶液中采用水热晶化的方式成功合成了CeO2纳米棒、CeO2纳米立方以及CeO2多面体等不同形貌的CeO2纳米材料; 最近有人报道, 通过简便的水热法即可成功制备暴露(110)活性面的CeO2纳米片[10].
人们通过密度泛函数理论对CeO2的低密勒指数晶面模拟计算发现, CeO2(111)面具有最低的表面能, 因而在(111)面上需要较高的能量才能形成氧空穴, 而(100)晶面具有最高的表面能, 有利于表面氧空穴的产生[11, 12, 13, 14]. 因此选择性地暴露出活性面将有利于CeO2表面氧空穴的形成, 可提高其储放氧性能. 此外, 氧空穴簇能够增加CeO2纳米棒的还原性和活性[15], 这是因为氧空穴簇以线式结构存在于CeO2表面, 促进体相氧向表面的迁移, 从而利于CeO2的还原[16].
如今温室效应越来越受到人们的关注. 通过设计优良的存储材料将温室气体CO2存储或者通过一系列催化反应将其转化为高价值的精细化学品有利于减小CO2温室气体的排放. 还原的CeO2能够吸附CO2, 其中有研究者认为CO2在吸附过程中被活化生成CO [17, 18]; Lykhach等[19, 20]认为CO2在CeO2表面快速生成了碳酸盐物种而不会直接生成CO. 通过还原CeO2(110)晶面的CO2吸附理论计算结果证明, CO2在CeO2表面是以碳酸盐形式吸附的[21]. 最近, 研究发现Sm掺杂的CeO2纳米材料在50 oC下具有非常优异的CO2吸附性能, 表面丰富的氧空穴有利于CO2的吸附[22].
乙二醇具有很好的还原性能, 常被用于金属单质和金属氧化物的制备过程中. 在乙二醇溶剂热条件下合成CeO2纳米材料, 有望使CeO2表面暴露出更多的氧空穴, 从而能够提高其可逆氧化还原性、储放氧性能和CO2的捕获能力. 本文采用乙二醇溶剂热法制备了CeO2-GST纳米材料, 并测试了其可逆氧化还原性、储放氧性能及CO2捕获性能. 此外还通过CO2原位红外漫反射(DRIFTS)实验进一步研究了CO2在CeO2表面的吸附状态.
分析纯的Ce(NO3)3∙6H2O, NaOH, 乙二醇和柠檬酸从国药集团化学试剂有限公司购买, 所有试剂在使用前均未经过进一步纯化.
采用三种不同的方法制备CeO2材料.
量取60 mL的乙二醇于聚四氟乙烯内衬中, 在搅拌条件下, 加入5 mmol Ce(NO3)3∙6H2O, 继续搅拌20 min. 在Ce(NO3)3全部溶解后, 将聚四氟内衬转移到晶化釜内, 180 oC晶化24 h. 反应结束后, 离心, 并用丙酮洗涤样品数次, 50 oC干燥过夜. 600 oC焙烧 3 h, 升温速率1 oC/min. 所得样品标记为CeO2-GST.
将5 mmol Ce(NO3)3∙6H2O和0.8 mol的NaOH分别溶解在80 mL去离子水中, 然后在搅拌条件下, 将NaOH水溶液逐滴加入到硝酸铈溶液中, 滴完后继续搅拌30 min. 再转移到水热晶化釜中, 拧紧, 120 oC晶化48 h. 反应结束后取出晶化釜, 过滤, 并用去离子水多次洗涤至中性, 50 oC烘箱中干燥过夜. 600 oC焙烧3 h, 升温速率为1 oC/min. 所得样品标记为CeO2-nanorod.
称取5 mmol Ce(NO3)3∙6H2O和15 mmol柠檬酸溶解在80 mL去离子水中, 将所得混合溶液转移到80 oC水浴中, 静置12 h使其成胶, 并陈化过夜. 之后把所得到的溶胶放入100 oC的烘箱中, 烘干剩余水分. 将所得淡黄色片状干胶研碎, 600 oC焙烧3 h, 升温速率为1 oC/min. 得到的样品标记为CeO2-CA.
样品的晶体结构以及有关的晶胞参数测试均使用Bruker D8 FOCUS X射线衍射(XRD)仪. 仪器参数为: Cu靶Kα射线(λ = 1.5406 Å), Ni单色器, 管电压为40 kV, 管电流为40 mA. 样品的比表面积采用Quantachrome NOVA 4200e型吸附仪进行分析, 实验前样品于200 oC脱气12 h. 样品的形貌以及晶格条纹采用日本电子公司的JEOL-2010型场发射高分辨透射电子显微镜(HR-TEM)进行分析. X射线光电子能谱(XPS)表征在Thermo ESCA LAB-250型光谱仪上测得, 真空度为1 x 10-9-2 x 10-10 torr, 用150 W单色Al Kα射线(hv = 1486.6 eV)为激发源, 能量分析器固定透过能为20 eV, 结合能通过表面积炭(C 1s = 284.8 eV)进行电荷校正.
对于H2还原-O2氧化循环测试, 称取100 mg的样品于U形石英管中, 在5% H2/N2 (45 mL/min)气氛下, 以10 oC/min的升温速率程序升温到600 oC, 采用TCD检测. 在程序升温完成后, 降温至100 oC, 切换气体至3% O2/He, 吹扫10 min后升温到500 oC, 保持120 min. 氧化结束后, 降至室温, 再将气体切换为H2/N2气氛, 再次程序升温到600 o C进行第二次循环测试. 每次循环均包括H2还原和O2氧化两个过程.
CO以及CO2脉冲测试在Auto Chem Ⅱ2920 Micrometrics化学吸附仪上进行, 采用TCD检测器及质谱仪在线分析产物.
对于CO脉冲测试, 称取50 mg样品置于U形管中. 首先, 样品在5% O2/He (50 mL/min)气氛中400 oC预处理30 min, 然后He气吹扫, 待基线走平后开始脉冲测试, 测试条件为400 oC, 5% CO/He (30 mL/min)作为脉冲气.
对于CO2脉冲测试, 称取25 mg样品置于U形管中. 样品首先还原预处理, 还原条件为10% H2/Ar (50 mL/min)混合气中, 10 oC/min程序升温到600 oC保持30 min, 之后He气吹扫30 min. 降温到室温后, 再次升温到50 oC, 以He气为载气, 5% CO2/He (30 mL/min)为脉冲气开始脉冲实验, 每个脉冲相邻的间隔为2.5 min.
二氧化碳原位红外漫反射(CO2 in-situ DRIFS)在Thermo Nicolet Nexus 670型红外仪上进行, 采用MCT检测器, 高温样品池为ZnSe窗口. 样品先经过预处理再进行红外测试, 预处理条件为: 在500 oC用10% H2/N2 (60 mL/min)气氛下还原30 min, 之后降温到300 oC He气预处理20 min. 预处理结束后, 降温到50 oC, He (99.99%, 80 mL/min)吹扫下采集背景, 之后通入5% CO2/He (80 mL/min)气, 同时开始采集数据. 红外光谱的扫描范围为650-4000 cm-1, 分辨率为4 cm-1,Q 97;扫描次数为64次. DRIFTS的采点频率为每隔2 min采集一次. 通CO2/He气体20 min后, 切换为He气吹扫, 同时每隔2 min采集数据一次, He吹扫时间为20 min.
图1为采用三种不同方法制备得到的CeO2的XRD谱, 所有样品都在600 oC下焙烧3 h (1 oC/min). 从图中可见, CeO2样品均为立方萤石结构(JCPDS 34-0394), Fm-3m空间群. 通过N2吸附曲线测得样品CeO2-GST,CeO2-nanorod和CeO2-CA的比表面积分别为79, 78和49 m2/g.
图2为三种样品的TEM和HR-TEM照片. 从图2(a)和(b)可见, CeO2-GST为纳米六面体颗粒, 粒径在10-15 nm左右, 高分辨透射电镜表明所暴露的晶面为稳定的(111)面. 该形貌与文献报道的有所不同, 因为在制备过程中水的引入会对形貌有很大影响, 不同水的引入量会导致其形貌从球形变化为立方形, 本实验中将铈源直接溶于乙二醇中未引入水, 从而最终形成六面体形状[23]. 图2 (c)和(d)为CeO2-nanorod的透射电镜和高分辨透射电镜图. 从图中可以看到, CeO2-nanorod直径为20 nm左右, 纳米棒的长度在100-300 nm不等. 纳米棒主要沿着[110]轴生长, 暴露出(100)面. 对于柠檬酸法制备的CeO2-CA样品, 其没有规则形状, 主要暴露出(111)面, 粒径大小为5-20 nm.
采用XPS研究了焙烧前后CeO2样品的表面性质(铈的价态以及表面氧空穴), 结果见图3. 通过分峰软件, 对XPS谱图进行分峰拟合, 并根据文献对拟合后的各个峰进行归属[24, 25, 26, 27]. Ce 3d的光电子能谱分为u和v两个系列的自旋-轨道多重谱. 其中对应于Ce4+的峰有6个, 其中v (882.2-882.6 eV), v" (889.1-889.3 eV)和v''' (898.2-898.5 eV)归属于v系列Ce 3d5/2的三个特征峰; u (900.6-901.0 eV), u" (907.5-907.7 eV)和u''' (916.6-916.9 eV)对应于u系列Ce 3d3/2的三个特征峰. u' (903.5-904.2 eV)和v' (885.1-885.8 eV)分别归属于Ce3+的Ce 3d3/2和Ce 3d5/2的特征峰. 其中Ce3+含量见表1, 其含量计算公式来自文献[24].
由图3(a)可以看出, 对于CeO2-GST样品, 焙烧前其Ce3+的含量为63%(表1), 归属于Ce3+的v'和u'出现强吸收, 而归属于Ce4+的u'''峰强度很低. 焙烧之后, 样品表面仍有丰富的Ce3+, 含量为11.2%, 且v'和u'的峰强度减弱, 同时Ce4+的特征峰u'''出现(图3(b)). 通过对比发现, CeO2-nanorod表面的v'和u'峰强度比其他两个样品弱很多, 表面存在最少的Ce3+含量. 对拟合后的峰进行面积积分, 计算出各个样品的Ce3+含量(表1), 其中CeO2-nanorod表面Ce3+含量最低, 只有5%, 柠檬酸法制备的CeO2-CA样品表面的Ce3+含量最高, 为14.6%, 这是由于柠檬酸在贫氧条件下分解出还原性气体, 将CeO2表面还原. CeO2-GST在焙烧之后, 其表面Ce3+含量仍能够保留为11.2%, 是CeO2-nanorod的两倍. 从XPS结果可以证明, 在乙二醇还 原条件下制备得到的CeO2-GST比传统的水热法制备的CeO2样品具有更丰富的表面Ce3+, 从而可以推断, CeO2-GST表面具有更丰富的氧空位[28, 29].
为了测试所制备CeO2样品可逆氧化还原的稳定性, 我们设计了H2还原-O2氧化的循环实验. 样品储放氧速率通过CO-脉冲进行评价,采用MS为检测器. 图4为三种样品的H2还原-O2氧化循环图, 其耗氢量见表2. 由第一次H2-TPR结果可见, CeO2-nanorod样品表现出最高的氢消耗量470 μmol/g, 其次为CeO2-GST (380 μmol/g), CeO2-CA最低, 为369 μmol/g. 在后续的循环中, 三个样品的耗氢量均逐渐减小. 最终结果表明, CeO2-CA的稳定性最差, 在3次循环后, 其耗氢量降低到至118 μmol/g; CeO2-nanorod与CeO2-GST从第二次循环开始表现出相似的下降规律, 但是对循环测试后的CeO2-nanorod样品进行TEM分析, 发现其棒状结构被破坏(图5(a)和(b)), 同时产生严重的烧结现象, 比表面积降低至6 m2/g. 但是10次循环后的CeO2-GST样品形貌没有发生太大变化(图5(c)和(d)). 因此CeO2-GST样品在H2还原-O2氧化的循环评价实验中表现出最优异的可逆氧化还原性能(图4a), 同时其结构在循环过程中也最稳定. 此外, 对于CeO2-GST样品, 在第6次循环之后, 其耗氢量不再降低. CeO2表面丰富的氧空穴将有利于体相氧向表面的迁移, 从而使得耗氢量在循环过程中趋于稳定. 此外通过溶剂热法制备得到的CeO2-GST能够提升材料结构的稳定性, 在10次循环后其结构仍能够保留.
图6为三个样品的CO-脉冲的MS谱图. 每个样品在16次脉冲内均达到饱和, 不再消耗CO. 从图6可以看出, 前10次脉冲3个样品的CO剩余量(峰面积)呈线性增加趋势, 在13次脉冲后CO均不再消耗. 对前10次脉冲的CO剩余量进行面积积分并对脉冲数做图, 其表现出很好的线性递增关系(图7(a), (b)和(c)). 在本试验中直线的斜率k应与样品的氧释放速率有关. 由于峰面积代表剩余的CO量, 因此样品对CO的消耗应该呈现相反的趋势, 即k越大, 样品的氧释放速率越慢. 三个样品的k值见表1, 其中CeO2-nanorod的k值最小, 表明其氧释放速率最快, 而CeO2-GST样品氧释放速率最慢.
根据第一性原理, 选择性暴露出(110)和(100)面有利于CeO2表面形成氧空穴[24, 25, 26]. 由HR-TEM结果可以发现, CeO2-nanorod选择性暴露出活性(100)晶面, 具有最快的氧迁移速率, 这一结论也与理论计算一致. CeO2-GST与CeO2-CA均暴露出稳定的(111)面, 具有相对更慢的氧迁移速率. 进一步对后两种样品对比发现, 表面Ce3+含量和氧迁移速率为CeO2-GST < CeO2-CA, CeO2表面具有更丰富的氧空穴, 同样能够提高其表面晶格氧的释放速率. Liu等[15]通过研究表面富含氧空穴的纳米棒的CO催化活性也得出相同的结论. 因此可以推断, 通过控制合成条件制备富含表面氧空穴的CeO2纳米材料将能够提高其氧的释放速率乃至催化活性.
对于CO2的捕获一般采用有机胺溶液、薄膜吸附以及离子液体等[30, 31], 但其再生需要较高的能量, 而且生产成本高, 因而开发新的CO2捕获材料具有十分重要的意义. CeO2是十分优良的路易斯碱, 从热力学上看, 将CeO2用于CO2的吸附是可行的[30, 32]. 此外, CeO2表面暴露出丰富的氧空穴时将有利于CO2的吸附[33]. 本文通过设计CO2-脉冲实验来测试三种CeO2材料吸附CO2的性能, 同时采用原位DRIFTS来研究CO2在CeO2表面的吸附状态.
50 oC下CO2的脉冲实验是通过TCD检测器来分析经过样品后剩余的CO2含量的, 同时对尾气进行质谱分析, 确定实验过程中是否有CO产生. 质谱分析结果表明, 在50 oC下没有CO特征信号出现, 只有很小的CO2在28处的碎片峰. 图8为CO2的MS信号. 可以发现在13个脉冲内所有样品对CO2的吸附都达到饱和. 同时通过对CO2信号的峰面积进行积分发现, CeO2-GST表现出最大的CO2吸附量(149 μmol/g), ;CeO2-nanorod和CeO2-CA表现出相近的CO2吸附活性, 分别为86和90 μmol/g.
通过CO2原位DRIFTS对CO2在CeO2样品表面的吸附状态进行研究. 在DRIFTS实验前, 先对样品进行预处理, 预处理条件为在500 oC通入10% H2/N2混合气对样品进行预还原, 之后降温到300 oC通入He吹扫30 min, 再降温到50 oC并走基线以及采集背景. 之后通入5%CO2/He进行实验. 根据文献[34, 35, 36, 37, 38, 39], 对所得谱图中的各个峰进行归属, 其结果汇总于表3中.
从图9(a)可以发现, 当CO2进入样品池时, 在800-2000 cm-1观察到6个强吸收峰, 分别位于856, 1021, 1217, 1289, 1397和1584 cm-1, 随着CO2通入时间的延长, 各个峰的峰强度增加, 但没有新峰出现. 20 min后, 所有吸收峰强度都不再增加, 表明此时样品吸附饱和. CeO2-CA样品也表现出类似的吸附行为, 但是其峰强度要低于CeO2-GST (图9(c)). CeO2-nanorod (图9(b))则表现出和前两个样品不同的红外吸收谱图, 首先在1214和1392 cm-1处出现更强的吸收特征峰, 而且出现了一系列新的吸收特征峰, 分别位于1458, 1542和1617 cm-1. 对于CeO2-GST和CeO2-CA样品, 856, 1028, 1289和1584 cm-1处峰归属于双齿碳酸盐物种, 1217和1397 cm-1归属于桥连碳酸盐物种. 对于CeO2-nanorod, 桥连碳酸盐物种的吸收峰分别红移至1214和1392 cm-1, 此外, 1612 cm-1处的吸收峰归属于重碳酸盐物种, 1458 cm-1处的吸收峰归属于单齿碳酸盐物种, 1542 cm-1处的吸收峰归属于甲酸盐物种[35,36].
当再次通入He对样品表面吹扫10 min后, 样品的红外谱图出现明显的变化. 首先三个样品位于1217 cm-1处归属于桥连碳酸盐的吸收峰明显减弱甚至消失, 这是由于桥连碳酸盐物种在CeO2表面的吸附较弱, 很容易从表面脱附. 对于CeO2-nanorod样品, 位于1584 cm-1的吸收峰消失, 同时在1563 cm-1出现一个新的吸收峰. 有文献报道, 双齿碳酸盐在+3价Ce和+4价Ce表面的吸附峰存在差异, 其最强吸收峰v(OCO)在Ce3+上为1584 cm-1, 而在Ce4+上则会向低波数偏移至1563 cm-1 [36], 因此可以推断出, CeO2-nanorod表面Ce3+被CO2氧化生成Ce4+. 此外, CeO2-nanorod样品归属于重碳酸盐、单齿碳酸盐和甲酸盐的特征峰在He吹扫下仍然保留. 同时通过对比桥连碳酸盐在三个样品上的吸附情况可以发现, 桥连碳酸盐物种在CeO2-nanorod表面的吸附最强, 其原因可能是高活性的(100)面更有利于桥连碳酸盐物种的稳定吸附. 对于CeO2-GST和CeO2-CA样品, 研究表明, CO2主要以双齿碳酸盐形式吸附, 没有观察到重碳酸盐物种.
通过对He吹扫后样品表面保留的吸附物种分析, CeO2-nanorod由于暴露出活性的(100)面, 能够活化CO2生成稳定的单齿碳酸盐、重碳酸盐和甲酸物种, 同时其表面的Ce3+被CO2氧化. 由于在该过程中没有发现CO的线式吸附峰, 同时结合在CO2脉冲过程中同样没有跟踪到CO的特征峰, 可以认为CO2主要以碳酸盐的形式在CeO2-nanorod表面被活化, 并不会直接生成CO. 同时红外光谱同样证明, CO2通过以不稳定的双齿碳酸盐物种吸附在CeO2(111)表面. 通过对比CeO2-GST和CeO2-CA的红外信号强度发现, CeO2-GST具有更高的吸收强度, 这也与其相对更高的CO2吸附量对应, 归结为CeO2-GST具有相对更大的比表面积. CO2吸附过程中Ce3+为活性位, 因此具有更高比表面积和丰富的表面Ce3+的CeO2-GST更利于CO2的捕获.
通过溶剂热法制备了CeO2-GST纳米材料, 其表面具有丰富的氧空穴, 且具有很好的可逆氧化还原性能. 这些丰富的氧空穴在高温条件下仍然能够得到保留,从而促进了CeO2的还原, 使得其在H2-TPR循环过程中表现出更好的循环稳定性. CeO2-nanorod表现出相对较低的循环稳定性. CO2脉冲测试发现, CeO2-GST表现出最好的CO2吸附性能(149 μmol/g). 此外, CO2在不同样品表面具有不同的吸附形式, 这与样品所暴露的晶面有一定关系. 双齿碳酸盐物种优先吸附在还原的(111)面; 而(100)面更利于桥式碳酸盐的吸附. 此外, 表面富含更多的氧空位和相对更大的比表面积更有利于CO2的捕获.
2014, Vol. 35 Issue (8): 1364-1375
PDF (1583KB)
2014, Vol. 35 Issue (8): 1364-1375
PDF (1583KB)