This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Cerium oxide nanostructures have attracted considerable attention for their high oxidative catalytic activities and relatively large abundance in the earth's crust.1, 2 Cerium oxides have been applied in a wide range of important industrial applications including oxygen sensors, three-way catalytic converters, solid oxide fuel cells, the water gas shift reaction catalysts, and other oxidative reaction catalysis. Literature often attributes the catalytic activity of cerium oxide to its high oxygen storage capacity (OSC) which is largely due to the multi-valence nature of cerium. The shift between the cerium(III) (Ce3+) to cerium(IV) (Ce4+) states leads to a high oxygen mobility in ceria lattice that in turn leads to a strong catalytic potential.3-6 Recently, nanosized cerium oxide has been extensively studied for their enhanced oxidative and reductive catalytic activity because they possess large surface areas and increased surface energy when compared to the bulk materials. Many methods such as hydrothermal synthesis, precipitation, spray pyrolysis, and electrochemical methods to produce nanoscale cerium oxide have been explored to yield a wide variety of other materials morphologies such as nanowires, nanocubes, nanorods, and nanotubes.
Surface and sub-surface lattice defects of catalysts also play important roles in influencing their catalytic activity.7-9 Much of the catalytic ability of the cerium oxide is attributed to high mobility of oxygen in the lattice and to its large oxygen storage capacity.10-13 The high mobility of oxygen within the lattice combined with the ability of cerium to readily oxidize and reduce between Ce3+ and Ce4+ leads to the formation of oxygen vacancy defect (OVD) sites. OVDs have been studied as a major contributor to the catalytic activity of the cerium oxide because oxidation and reduction can be achieved easily at these sites. Specifically, the ratios of Ce3+ to Ce4+ in cerium oxides have been found to correlate to the number of oxygen vacancy defect sites present on the surface and just below the surface of the catalytic materials. This provides a convenient means to evaluate the OVD density in cerium oxide catalyst. Since the size, concentration, and type of OVDs can critically affect the enhanced catalytic activity of cerium oxide, these features are considered as possible catalyst design criteria. Nonetheless, most of the catalyst design research conducted for cerium oxides so far has revolved around the naturally occurring OVDs but not the intentional introduction of the defects.
Synthetic strategy to introduce oxygen lattice defects in cerium oxide catalysts essentially can be modulated to the control of Ce3+/Ce4+ ratio in a reproducible and energetically favorable manner. Cerium oxide naturally occurs as a mixture of Ce3+ and Ce4+, in the bulk form as O2- desorbs from the surface of the CeO2 the electrostatic balance is maintained by transferring one electron each to two neighboring Ce4+ reducing them to Ce3+ leaving behind CeO2-x as the reduced form.14 The ratio Ce3+ /Ce4+ at the bulk material surface is typically determined between 2 and 15 % primarily by x-ray photoelectron spectroscopy (XPS) analysis. Thus a significant increase in Ce3+ /Ce4+ ratio over 20 % can be applied an indirect indication of increased OVD in the cerium oxide catalyst.
Here we report our study of the oxidative catalytic activity of nanostructured cerium oxide catalyst with intentionally introduced of oxygen lattice defects. We found that the number of stable OVDs on cerium oxide nanoparticles and nanotubes could be increased through the use of elevated temperature annealing under vacuum in the post production activation process. The oxidative catalytic activities of these nanostructured catalysts were evaluated using the standard carbon monoxide (CO) oxidation reaction and were found to have much lower light-off temperatures when compared to the bulk counterpart. The chemical equilibrium reactions on the catalysts surface in vacuum were hypothesized to explain the unusual increase in the OVD density of the reported cerium oxide nanostructured catalysts.
All water used in this experiment was Ultrapure water of >18 Mâ„¦ resistivity and filtered through 0.22 nm pore-sized filters. All chemicals were used as purchased unless otherwise noted. Bulk cerium oxide powder with 300 mesh size (Sigma-Aldrich, St. Louis, MO) and cerium oxide nanopartices with 7-nm average diameter (Nanoscale, Manhattan, KS) were used for catalytic activity evaluations and comparisons with cerium oxide nanotubes. The cerium oxide nanoparticle samples have aggregate particle size â‰¤ 9 Âµm.
Synthesis of Cerium Oxide Nanotubes
Cerium oxide nanotubes were synthesized using a modified method developed by Zhou et al.3 (figure 1) Briefly, a sample of 0.5 g cerium(III) sulfate hydrate (Ce2(SO4)3Â· X H2O, Sigma-Aldrich, St. Louis, MO) was first dissolved into 40 mL of 10 M sodium hydroxide aqueous solution (NaOH (aq), Sigma-Aldrich, St. Louis, MO) solution. The solution was transferred to a 45 mL total volume Parr autoclave, and was allowed to react at 120 Â°C for 15 h. The cooled sample was filtered using 0.8 Âµm membranes (Millipore, Billerica, MA) and rinsed with 3 aliquots of 50 mL water. After rinsing, the sample was placed in a convection oven at 50 Â°C for 1 h. The samples were then gently powdered using a spatula and heated at 50 Â°C for an additional hour for this partial oxidation step. The resulting samples were mixed with 50 mL of water, and 50 mL of â‰ˆ15% hydrogen peroxide (H2O2, Sigma-Aldrich, St. Louis, MO), immediately followed by sonication for 30 min following sonication the samples were left in the H2O2 solution for an additional 60 minutes for their oxidative transformation into tubular structures. Lastly, the material was filtered using a 0.8 Âµm membrane rinsed with three aliquots of water and dried in a convection oven at 50 Â°C.
Activation treatment of cerium oxide samples
Typically, 100 mg of cerium oxide sample (nanoparticles, nanotubes or bulk materials) was activated by heating the sample in a 1" quartz tube furnace with a 100 SCCM flow of a nitrogen-oxygen mixture (80% N2 and 20% O2) for 1 hour at 350 Â°C under vacuum with an operating pressure of 0.1 Torr. Control samples were activated using similar experimental condition but with 1 atm operation pressure.
CO oxidation catalysis
The CO oxidation catalysis was carried out in a U-shaped quartz micro reaction chamber fitted with a coarse quartz frit sample platform. Typically, in each catalysis test, a mixture of 79% helium, 20 % oxygen and 1 % carbon monoxide was flowed continuously throughout the reaction at a rate of 30 SCCM through a 100-mg sample which completely covered the frit. The reaction chamber was warmed in 5 â°C increments from room temperature to 250 â°C. A 1 mL sample of process gas obtained at each reaction temperature was analyzed using a gas chromatography instrument (Gow-Mac, manufacturer information) equipped with an eight foot porapak Q column (Manufacturer information). The percentage of CO conversion was determined by quantifying the carbon dioxide concentration in the processed gas.
The structural morphology of the cerium oxide samples was examined with a field emission scanning electron microscope (FE-SEM, Hitachi S4700, Hitachi High Technologies America, Inc. Pleasanton, CA) operated at 15keV Detailed structures of nanomaterials were investigated by high resolution transmission electron microscopy (HRTEM) with a Tecnai G2 F20 S-Twin operated at 200keV (FEI, Hillsboro OR). Selected area electron diffraction (SAED) was used to determine the local structures of the materials. Each TEM sample was prepared by drop-casting a solution of sample sonnicated in methanol onto a holey carbon film on a copper grid support. X-ray diffraction (XRD) of the samples was performed with a Bruker AXS D8 Discover equipped with a GADDS area (Bruker AXS Inc. Madison, WI) to examine the crystallinity and crystal structure of the samples in bulk form. The weighted average wavelength of the Cu KÎ± x-ray source is 1.5417 Å.
The surface areas of the three types of cerium oxide catalysts (bulk powder, nanoparticles, and nanotubes) determined via the Braunauer Emmet Teller (BET) extension of the Langmuir isotherm with a Micromeritics ASAP 2010 (Micromeritics, City, ST). Nitrogen was used as the adsorbent gas in these experiments.
Chemical composition characterization
The elemental compositions of the cerium oxide samples were determined by energy dispersive x-ray spectroscopy (EDS) using an EDX detector (EDAX Inc, Mahwah, NJ) equipped with the HRTEM. The Ce3+/Ce4+ ratio for each sample was quantified by XPS (VersaProbeâ„¢ Scanning XPS Microprobe) Briefly, for each experimental spectrum, the XPSpeak program (publisher) was first applied to subtract a fitted base-line using the Shirley algorithm from the data. Semi-Voigt functions (convolved Gaussian-Lorentzian lineshapes) were then fitted to the resulting spectrum to determine the ten peak areas corresponding to the signals from Ce3+ and Ce4+ according to recent literature. No single peak is linearly proportional to the ratio of Ce3+ to Ce4+, therefore it was necessary to deconvolute and to calculate the areas of each of the ten peaks. The deconvolution of the XPS data to yield the Ce3+/Ce4+ ratios was accomplished by using the program XPSpeak15 to fit the baseline using the Shirley algorithm and the relevant XPS peaks corresponding to the cerium signals. Reported peak locations from the liturature were used; at least 80% Gaussian character was assumed. The peaks were fitted in a series of steps which allowed the areas and their full width at half maximum to vary throughout all steps. The percentage of Gaussian contribution for each lineshape was allowed to vary between 80 and 100% after the initial fit. The peak location was allowed to vary up to 0.2 eV due to the nanoscale nature of the materials during the last step. The ratio of the integrated peak area was then calculated by comparing the peak areas for Ce3+ to the total area calculated for both Ce3+ and Ce4+ according to the reported literature.
The physical properties of each of the tested cerium oxide materials demonstrated that all were consistent with previously reported values. The XRD revealed that all of the materials tested were in the CeO2 fluorite cubic structure (space group Fm3m (225)) with a lattice constant a = 5.411 Å (JCPDS 34-0394). BET surface area of the nanotubes, nanoparticles, and bulk cerium oxide was 113, 54 and 8 m2/g respectively. Field emission SEM studies showed that both of the nanoscale materials aggregated, the average size of the nanotube aggregate was â‰¤ 2Âµm and that of the nanoparticles was â‰¤ 9 Âµm the bulk material was a fine powder and the average particle size was 80 Âµm.
HRTEM of the nanotubes revealed that the majority were tube-like in structure with an average diameter of 20 nm with a 10 nm interior diameter; it also confirmed that the major lattice plane was  as the lattice fringes were 3.2 Å; some nanowires and nanoparticles were present with the nanotubes the nanowires were almost uniformly 5nm diameter, the average length of both nanotubes and nanowires was around 750 nm. The presence of nanowires and nanoparticles within our nanotubes was likely due to our intentionally harsh synthetic procedures. Sonication of particularly thin walled cerium oxide cerium hydroxide core shell structures caused damage to these individual nanorods leading to the lengthwise breakage of some tubes leading to the formation of nanowires. The rapid oxidation and change structure from the simple hexagonal of Ce(OH)3 to the cubic fluorite structure of CeO2-x led to the formation of many defect sites and small domains noted in the fast fourier transformation (FFT) of HRTEM images and selected area electron diffraction (SAED) shown in figure 3. SAED and FFT of the HRTEM images demonstrated that the materials were highly polycrystalline with very small domain sizes as expected based on the HRTEM data and the harsh synthetic conditions; the major lattice plane was . HRTEM of the nanoparticles revealed that the average size of the individual crystals was 5nm and that the major plane exposed was  with a lattice fringe spacing of 0.32 nm.
Once activated at 400 Â°C at 0.1 torr the appearance of numerous dark spots on the surface of both the nanotubes and nanoparticles was noted using HRTEM, while these dark spots were also apparent on the surface of the nanotubes and nanoparticles which were activated at standard pressures they were far less concentrated and the presence of linear oxygen vacancies and other types of clustered vacancies was almost insignificant. It was noted that at there were three distinct types of dark spots, linear, circular, and triangular when compared with scanning tunneling microscopy experiments these same three patterns can be shown to be specific types of OVDs; linear oxygen vacancies, circular surface-trimer oxygen vacancies, and triangular subsurface-trimer oxygen vacancies.13 The apparent shapes associated with each type of vacancy are distinct and are due to When samples were exposed to longer time under vacuum the number of visible vacancies increased up to an apparent maximum after about 6 hours under these conditions. With the bulk materials, there was no noticeable change in the HRTEM images after vacuum activation.
Chemically the cerium oxide materials were all of the formula CeO2-x where x accounts for the ratio of Ce3+ in the lattice. EDX for the nanotubes found no presence of elements other than cerium and oxygen this is consistent with the synthetic process used and with the reports from others utilizing the same method. By utilizing XPS the atomic ratio for the as synthesized nanotubes was found to be 16% Ce3+ while the ratio for the nanoparticles and bulk ceria were found to be 25 and 25 % respectively, the high percentage of Ce3+ found in the bulk material is likely accounted for by the micron scale of the powder. The atomic ratio of Ce3+ did not change for any of the samples when activated under standard pressure at 400 Â°C. However, when the samples were activated at 400 Â°C and 0.1 torr, the atomic ratio Ce3+ increased for both the nanotubes and the nanoparticles to 39 and 36 % respectively but did not change for the bulk material. The increase in Ce3+ is likely due to a change in the partial pressure of the system which encouraged the desorption of oxygen from the surface of the materials, in the bulk material the oxygen storage capacity of the cerium oxide was probably able to buffer the small amount of oxygen desorbing from the surface. Additionally the total surface area exposed to the low pressure is certain to be a factor in the amount of change in the Ce3+ ratio as demonstrated by the more significant change experienced by the nanotubes over the nanoparticles. There are two concerns with using XPS for the analysis of the ratio of Ce3+ to Ce4+. First, the x-ray source creates a core hole in the 3d band which in turn leads to a rearrangement of the electron energy leading to a change in the hybridization of the oxygen 2p and the cerium 4f making it appear that the cerium is in the 4f1 state rather than the 4f0 which adds to the complexity of the analysis of the peaks, however as long as all ten peaks are integrated this problem can be overcome. Secondly, XPS is by necessity done under high vacuum conditions. It has been shown that under these conditions CeO2 will reduce to form CeO2-x therefore if the spectrum are not recorded quickly the ratio of Ce3+ to Ce4+ will change. Careful studies have been done with good results being reported nanoparticles have been reported with natural ratios of 17 to 40 %, which corresponds well to the increased number of OVD sites due to increased surface energy. This of course leads to the question of stability of the materials when they are in an artificially high ratio of Ce3+ to Ce4+.
The catalyzed oxidation of CO was investigated and all materials were found to be catalytically active. The light off temperature (T50) was used as a comparison of the activity of the catalytic activity of each of the cerium oxide materials tested. For the control materials which were activated at 400 Â°C at ambient pressure the T50 for nanotubes, nanoparticles and bulk cerium oxide were found to be 205, 285, and 350 Â°C respectively. The light off temperature (T50) showed that both the nanotubes and the nanoparticles were more active after activation at 0.1 torr than when activated at ambient pressure, additionally the nanotubes were more active than the nanoparticles with T50 of 175 and 260 Â°C respectively. The turn over number (TON) at 250 Â°C for vacuum activated nanotubes, nanoparticles, and bulk cerium oxide were 1.77 Âµmol*g-1*sec-1, 0.22 Âµmol*g-1*sec-1, and 0.01 Âµmol*g-1*sec-1 respectively. To test the effect of surface area on the activity of these materials an amount of nanoparticles with an surface area equivalent 0.1 g of nanotubes was used; the T50 of these materials did not decrease significantly nor did the TON, an equivalent surface area bulk cerium oxide sample was not run as the amount needed would have exceeded the volume of our micro reactor.
Post catalytic XPS analysis of cerium oxide nanotubes which had been exposed to flowing CO/O2 gas mixture for at least 72 hours at T50 revealed that the Ce3+ ratio did not change from that of the activated cerium oxide, additionally there was no noticeable change in the number and types of OVD sites that were observed through HRTEM.
Previously the following mechanism has been suggested for the oxidation of CO on a pure cerium substrate.
In a stable Ce3+ rich environment it is possible that the adsorbed oxygen is made available to the CO more readily than in an environment where the cerium is attempting to return to the Ce4+ state. The fine balance between the energy of adsorption and the energy required to break the oxygen-oxygen bond is tipped in favor of the catalytic reaction and the rate of oxidation of the CO increases due to the presentation of a new energetically favourable pathway. Additionally it has been shown that the CO bond is strained when in contact with an OVD therefore it should be more reactive under these conditions.
During activation various vacancy cluster defects were added to the surface and subsurface of the materials. The addition of these defects to the ceria nanotubes led to enhanced catalytic activity of the nanotubes with respect to other nanoparticles synthesized in a similar fashion, or of similar structure. This increases our understanding of the importance of both types of defects for the studied materials.
The identification of vacancy clusters through the use of HRTEM proved promising as many of these types of defects were identifiable utilizing this method. The ability to identify VCs using HRTEM rather than STEM makes the identification of these defects accessible to more research groups. Step edge and grain boundary defects were introduced into the ceria nanotubes during synthesis.
The authors thank Dr. David Diercks and Dr. Nancy Bunce for their help with HRTEM and XPS data. We also thank Nebraska Research Initiative and Nebraska Center for Energy Science Research for financial support.