Effects of high-temperature CeO₂ calcination on the activity of Pt/CeO₂ catalysts for oxidation of unburned hydrocarbon fuels

Abstract

A CeO₂-supported Pt catalyst (denoted Pt/CeO₂(800)) was prepared by pre-calcining a commercial CeO₂ support at high temperature (800 °C) before loading Pt via the incipient wetness impregnation method. Pt/CeO₂(800) exhibits enhanced redox and hydrocarbon (HC) oxidation activity, in comparison to a catalyst with untreated CeO₂ support (denoted Pt/CeO₂(UT)). A combination of multiple characterization techniques, including X-ray diffraction (XRD), transmission electron microscopy (TEM), CO temperature-programed reduction (CO-TPR), and diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS), shows that the promoted redox activity of Pt/CeO₂(800) is independent of the morphology of Pt clusters. Instead, it is associated with the enhanced mobility of surface lattice oxygen on the high temperature pretreated CeO₂ support. The catalysts were evaluated for the catalytic oxidation of various HC fuel blendstocks with different functionalities (C=C, C–O, and C=O bonds) under different simplified and simulated exhaust conditions (with 0.74% and 10% O₂). Under simulated exhaust conditions, fuel molecules, regardless of functionalities, are always more active under lean (high O₂) conditions than under rich (low O₂) conditions although their oxidation reactivity is suppressed by the competitive adsorption of NO and CO. The high redox activity of Pt/CeO₂(800) facilitates the HC oxidation by accelerating the rate-limiting O₂ activation step and broadening the operation window for an oxygen-dominant Pt surface.

---

Introduction

The Co-optimization of Fuels and Engines (Co-Optima) initiative by the U.S. Department of Energy aims to simultaneously develop highly efficient advanced engines along with high-performance fuels to reduce petroleum consumption. Emerging multimode engine technologies offer higher efficiency by combining spark ignition (SI) at high engine loads with advanced compression ignition (ACI) at low loads. To date, researchers have identified a number of potential fuel blendstocks that could provide the desired fuel properties.¹ These blendstocks include a broad range of functionalized hydrocarbons, such as alkenes, ketones, alcohols, and furans, which could be derived from biomass, further reducing overall petroleum consumption and greenhouse gas emissions. Newly developed engines and fuels bring new challenges for exhaust aftertreatment catalysts. On the one hand, varieties of blendstocks complicate hydrocarbon components in exhaust. On the other hand, the catalysts need to operate under both stoichiometric and lean-burn exhaust conditions as the engine switches between SI and ACI operation modes.

Platinum (Pt) has been widely studied for hydrocarbon and VOC (volatile organic compound) oxidation, because of its high activity.² Ceria (CeO₂) possesses high oxygen storage capacity (OSC) and redox properties due to the cycle of Ce⁴⁺/Ce³⁺ redox pairs, making it an active support for oxidation catalysts. Combining these two materials, Pt/CeO₂-based catalyst systems have been widely applied as oxidation catalysts for automotive emission control.³ The OSC of CeO₂-based catalysts is especially important for three-way catalysts (TWCs) which need to operate under rich-burn or stoichiometric conditions (low O₂). In addition, a strong metal–support interaction endows CeO₂-supported noble metal systems with unique redox activities. These features make Pt/CeO₂-based catalyst systems good candidates for addressing unburned hydrocarbon fuel exhaust for multimode engines.

Oxygen vacancy on CeO₂ plays a critical role in the oxidation activity of CeO₂-supported noble metal catalysts, as extensively reported for both CO^4–6 and hydrocarbon oxidation.^7–9 Specifically for HC oxidation on Pt/CeO₂ catalysts, Peng et al.⁸ studied the effects of CeO₂ morphology on the activity of Pt/CeO₂ for toluene oxidation. Three Pt/CeO₂ catalysts were synthesized using CeO₂ nanorods, nanoparticles, and nanocubes, which preferentially exposed the (110), (111), and (100) facets, respectively. The Pt/CeO₂-nanorods exhibit the best toluene oxidation activity, due to their highest concentration of surface oxygen vacancies. In addition, the reaction rates on the three catalysts were found to be proportional to the oxygen vacancy concentrations but independent of Pt dispersion, suggesting that the supply of active oxygen controls the reaction rate. The surface oxygen vacancies play a role in replenishing active oxygen species from the gaseous phase and accelerating the oxygen cycle for toluene oxidation. Wang and co-workers⁹ modified a Pt/CeO₂-nanorod catalyst with dielectric barrier discharge plasma and significantly improved its activity for toluene oxidation (with T₉₀ lowered by 79 °C). It was proposed that the improved activity is associated with the higher oxygen vacancy concentrations on the plasma-treated Pt/CeO₂ sample.

The redox activity of CeO₂ depends not only on the oxygen vacancy concentration but also on the structure of oxygen vacancies. Liu et al.¹⁰ compared the CO oxidation activity of two types of CeO₂ nanorods with predominant exposure of the (100) + (111) and (100) + (110) facets, respectively. They found that the former containing large size oxygen vacancy clusters is more active than the latter predominated with isolated oxygen vacancies. Clusters of more than two vacancies, such as linear surface oxygen vacancies, proved to be favorable for migration of oxygen.¹¹

Lee et al.¹² reported a simple approach for promoting the CO oxidation activity of Pt/CeO₂ by annealing the CeO₂ support at a high temperature of 800 °C before loading Pt. The high temperature pretreatment reduces the surface defects on CeO₂.^12,13 A low defect concentration leads to a weaker metal–support interaction on the Pt/CeO₂ catalyst and facilitates the formation of PtO₂ species, which is essential for the high CO oxidation ability. In this work, we applied a similar CeO₂ pre-annealing approach to synthesize a Pt/CeO₂ catalyst for hydrocarbon (HC) oxidation. We found that the HC oxidation activity is only related to the mobility of the surface lattice oxygen and not impacted by the morphology of Pt clusters.

Other than the activity of the catalysts, the identities of the target pollutants and the composition of the exhaust also significantly impact the performance of the after-treatment system. Previous work by ORNL^14,15 has showed how functionalities of fuel blendstocks might impact the performance of TWC systems in vehicles powered by spark-ignition (SI) engines under stoichiometric conditions. More information about the effects of fuel chemistry under lean-burn conditions is still lacking.

Therefore, in this work, we also investigated the effects of functional groups on the oxidation reactivity of hydrocarbon fuels on Pt/CeO₂ catalysts under various exhaust conditions. Under simulated exhaust conditions, where HC oxidation is suppressed by the competitive adsorption of NO and CO, fuel molecules, regardless of functionalities, are always more active under lean (high O₂) conditions than under rich (low O₂) conditions. The improved redox activity of Pt/CeO₂ benefits the HC oxidation by accelerating the rate-limiting O₂ activation step and broadening the operation window of an O dominant Pt surface. This simple approach for tuning the surface properties of CeO₂ can potentially be applied in other CeO₂-based systems for improved catalytic activities.

Experimental

Catalyst preparation

CeO₂-supported Pt catalysts were prepared by the incipient wetness impregnation method. Commercial CeO₂ (99.99%, 10 nm, US Research Nanomaterials, Inc.) was used as the support, and [Pt(NH₃)₄](NO₃)₂ (99.995%, Sigma Aldrich) as the Pt precursor. CeO₂ was either used as received or pre-calcined at 800 °C for 10 h in flowing air (zero grade) before loading the Pt precursor. The pre-calcined CeO₂ support is denoted CeO₂(800), whereas the untreated CeO₂ is labelled CeO₂(UT). An aqueous solution of [Pt(NH₃)₄](NO₃)₂ was impregnated on the respective CeO₂ supports to achieve a 1.5 wt% Pt loading. The Pt loaded CeO₂ samples were dried at 70 °C for 4 h and then calcined at 500 °C for 4 h in flowing air to obtain the Pt/CeO₂ catalysts. The catalyst samples are denoted as Pt/CeO₂(UT) and Pt/CeO₂(800), respectively.

Catalyst characterization

Powder X-ray diffraction (XRD) measurements were performed on a Philips PW3040/00 X'Pert powder X-ray diffractometer with CuKα radiation (λ = 1.5406 Å). Data were collected with 2θ ranging from 2° to 100° using a step size of 0.01°.

Transmission electron microscopy (TEM) measurements were conducted with an FEI Titan 80-300 operated at 300 kV. The observations were performed in scanning transmission electron micrscopy (STEM) mode using HAADF detector, as well as conventional TEM mode. All TEM images were digitally recorded using a charge-coupled device camera. The TEM specimens were prepared by dispersing the catalyst samples in ethanol and depositing the suspension onto a lacey carbon-coated copper grid.

Temperature-programed CO reduction (CO-TPR) was performed on a Micromeritics AutoChem II analyzer combined with an MSK Cirrus 2 mass spectrometer. About 70 mg of catalysts was loaded in the reactor for measurement. The samples were first pretreated in 10% O₂/He (50 cm³ min⁻¹) at 250 °C for 30 min and then purged with He (50 cm³ min⁻¹) at room temperature for 30 min. The first cycle of CO-TPR was performed in 1% CO/He (50 cm³ min⁻¹) with a heating rate of 10 °C min⁻¹ to 250 °C and held for 30 min. The second cycle of CO-TPR was performed with the same procedure and conditions, except the TPR temperature was increased to 500 °C.

Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) studies of CO adsorption were performed on a Nicolet iS50R FT-IR spectrometer (Thermo Scientific) equipped with a liquid-nitrogen-cooled MCT detector and a Praying Mantis diffuse reflection accessory (Harrick Scientific Products Inc.) operated at a resolution of 4 cm⁻¹. The background spectra were collected with KBr powder in flowing He at 30 °C. Before measurements, the catalyst powders placed in the sample cell were pretreated by heating to 400 °C in 10% O₂/He (20 cm³ min⁻¹) and holding for 30 min, followed by cooling down to 30 °C in flowing He (20 cm³ min⁻¹). CO adsorption was performed with 1% CO/He (20 cm³ min⁻¹). The samples were heated with a heating rate of 10 °C min⁻¹ to 250 °C and held at 30, 50, 100, 150, 200, and 250 °C for 1 min for spectrum collection.

X-ray photoelectron spectroscopy (XPS) measurements of the pristine Pt/CeO₂(UT) and Pt/CeO₂(800) catalysts were performed on a Physical Electronics Quantera SXM scanning X-ray microprobe, with a focused monochromatic Al Kα X-ray (1486.7 eV) source for excitation and a spherical section analyzer. High energy resolution spectra were collected using a pass-energy of 69.0 eV with a step size of 0.125 eV. The binding energy (BE) scale was calibrated using the Cu 2p_3/2 feature at 932.62 ± 0.05 eV and Au 4f_7/2 at 83.96 ± 0.05 eV.

Catalyst activity evaluation

The catalyst activities for the oxidation of various hydrocarbon (HC) fuel blendstocks were evaluated by temperature-programed reaction using a plug flow microreactor system, with the reactor effluent composition quantified using an FTIR gas analyzer (MKS, Multigas 2030). The catalyst powders were pressed, crushed, and sieved (180–250 μm) for the tests. 150 mg of catalysts were mixed with 400 mg of SiC diluent and loaded in a quartz reactor (inner diameter ∼10 mm) above a porous frit. The catalysts were tested under either fully simulated exhaust (containing HCs, O₂, H₂O, CO₂, CO, NO, H₂, and N₂ balance) or simplified conditions (containing HCs, O₂, and N₂ balance), as listed in Table 1, with a space velocity of 200 L g⁻¹ h⁻¹. The simulated exhaust conditions are based on the U.S. DRIVE protocols LTC-G (low-temperature combustion with gasoline) and S-GDI (stoichiometric gasoline direct injection).¹⁶ The concentrations of CO, NO, and H₂ were adjusted from the original protocols due to the limitation by the available gas mixture source (5000 ppm CO/2000 ppm NO/1000 ppm H₂). The liquid HC fuels and H₂O were introduced into vaporization zones located at the upstream of the reactor using a syringe infusion pump and an HPLC pump, respectively. The gas line was kept above 200 °C. In the light-off test, the reactor was heated with flowing air to 100 °C before switching to the testing gas feed and held for 10 min. Then, the temperature was ramped from 100 to 500 °C with a heating rate of 5 °C min⁻¹.

Table 1 Gas composition for the catalyst tests

Gas components	Concentration
	LTC-G	S-GDI	High O2	Low O2
O2	10%	0.74%	10%	0.74%
H2O	6%	13%	—	—
CO2	6%	13%	—	—
CO	2000 ppm	3500 ppm	—	—
NO	800 ppm	1400 ppm	—	—
H2	400 ppm	700 ppm	—	—
HCs	3000 ppm C1	3000 ppm C1	3000 ppm C1	3000 ppm C1

---

Kinetic measurements of steady state propene oxidation were performed using a plug flow microreactor system. 8–16 mg catalyst pellets (180–250 μm) were mixed with 400 mg of SiC and loaded in a quartz reactor (inner diameter ∼10 mm) above a porous frit. The reaction was conducted isothermally at 220 °C. The reactant mixture consists of 500–3500 ppm propene and 2.5–15 vol% O₂ (balanced by He and N₂). The propene concentrations in the reactant feed and reactor effluent were quantified using a gas chromatograph (Agilent, 7890A) equipped with a capillary column (Agilent HP-1, 19091Z-433, 30 m, 0.25 mm ID, 0.25 μm film) connected to a flame ionization detector (FID). The rate of propene oxidation was determined based on the propene concentration difference between the reactant feed and reactor effluent.

Results and discussion

Influence of CeO₂ calcination on the structure of Pt/CeO₂ catalysts

Fig. 1 shows the XRD patterns of the CeO₂ supports and Pt/CeO₂ catalysts. The high temperature calcination caused crystalline growth (Table 2) and thus the sharpening of the CeO₂ diffraction peaks. Correspondingly, the surface area of the support decreases from 82 m² g⁻¹ on the untreated CeO₂(UT) to 18 m² g⁻¹ on the 800 °C-calcined CeO₂(800) (Table 2). No phase change was observed. Pt clusters were formed on both Pt/CeO₂ catalysts, as indicated by the presence of a weak Pt (111) peak at 40°.

Fig. 1 XRD profiles of the CeO₂ supports and Pt/CeO₂ catalysts.

Table 2 Properties of the CeO₂ supports and Pt/CeO₂ catalysts

Sample	Surface area^a (m^2 g^−1)	Pt loading^b (wt%)	CeO2 crystalline size^c (nm)
a Measured by the BET method. b Measured by ICP. c Determined by XRD.
CeO2(UT)	82	—	7.7
CeO2(800)	27	—	23.0
Pt/CeO2(UT)	82	1.54	8.0
Pt/CeO2(800)	30	1.66	23.4

---

The CeO₂ calcination impacted the morphology of the loaded Pt clusters. Fig. 2 shows the TEM images and the Pt nanoparticle size distributions of Pt/CeO₂(UT) and Pt/CeO₂(800). On the untreated CeO₂(UT), Pt formed nanoparticles with sizes of around 1–2 nm (Fig. 2a). On the high temperature treated CeO₂(800), although the XRD spectrum indicates the existence of Pt nanoparticles (Fig. 1), TEM shows that Pt particles were predominantly highly dispersed as two-dimensional monolayer clusters (Fig. 2b). However, these monolayer Pt clusters were not stable. Upon reductive treatment with CO at 250 °C, Pt aggregated to form small nanoparticles with a size of 1–1.5 nm (Fig. 2c). The Pt nanoparticles on Pt/CeO₂(800), however, show much more active redox properties than those on Pt/CeO₂(UT).

Fig. 2 STEM HAADF images and Pt particle size distributions of (a) Pt/CeO₂(UT), (b) Pt/CeO₂(800), and (c) reduced Pt/CeO₂(800).

Influence of CeO₂ calcination on the redox activity of Pt/CeO₂ catalysts

The redox properties of the Pt/CeO₂ catalysts were examined with CO-TPR. As shown in Fig. 3a, in the first cycle of CO-TPR to 250 °C, the reduction of the pristine Pt/CeO₂(UT) initiates at around 100 °C and reached maximum at 125 °C. This low temperature reduction peak is associated with the CeO₂ surface lattice oxygen activated by Pt nanoparticles. In the second cycle of CO-TPR to 400 °C, the as-reduced Pt/CeO₂(UT) shows a lower reduction peak temperature at 105 °C. A second reduction peak appears at 250 °C, which is assigned to CeO₂ surface lattice oxygen not activated by Pt,^10,17 as bare CeO₂(UT) support also shows a weak reduction peak at the same temperature (Fig. 3a).

Fig. 3 CO-TPR profiles of the pristine and reduced (a) Pt/CeO₂(UT) and (b) Pt/CeO₂(800) (after the first CO-TPR of the pristine catalysts to 250 °C in 1% CO/He and being held for 30 min; the reduced catalysts were re-oxidized at 250 °C in 10% O₂/He for 30 min before the second cycle of CO-TPR).

The 800 °C pretreatment of the CeO₂ support promoted the redox properties of the Pt/CeO₂ catalyst. As shown in Fig. 3b, in the first CO-TPR cycle, the reduction of the pristine Pt/CeO₂(800) initiates below 50 °C, with the first reduction peak centered at 79 °C. Again, this low temperature reduction is assigned to surface CeO₂ lattice oxygen activated by Pt clusters. Similar high redox properties of Pt/CeO₂ have previously been reported by Lee and co-workers,¹² who also reported that a Pt/CeO₂ catalyst using a 800 °C-pretreated CeO₂ support showed a low temperature reduction peak at 60 °C in H₂-TPR, in comparison to the 165 °C reduction peak for a Pt/CeO₂ sample with untreated CeO₂ support. The pristine Pt/CeO₂(800) has a second reduction peak at 120 °C, which is ascribed to lattice oxygen activated by atomically dispersed Pt atoms.⁶ In the second CO-TPR cycle to 400 °C, the reduced Pt/CeO₂(800) only showed one low temperature reduction peak at 81 °C associated with Pt clusters, as the Pt single-atoms had aggregated to Pt clusters in the first CO-TPR cycle, as confirmed by TEM (Fig. 2c). Both the pristine Pt/CeO₂(800) with 2D Pt rafts and the reduced Pt/CeO₂(800) with 3D Pt particles show a similar low temperature reduction peak (79–81 °C), indicating that the morphology of the Pt clusters only has minor effects on the high redox activity of Pt/CeO₂(800) reflected by CO-TPR.

It is noted that the Pt/CeO₂(800) sample does not show any additional reduction peak at around 250 °C related to unactivated surface lattice oxygen as observed on Pt/CeO₂(UT), although the bare CeO₂(800) support does have a weak reduction peak at around 250 °C. We postulate that all the reducible surface lattice oxygen species on CeO₂(800) can easily migrate to the Pt clusters to react with CO within the first reduction peak at around 80 °C.

The state of the Pt species on the Pt/CeO₂ samples was further investigated by in situ DRIFTS of the CO adsorption during and after CO-TPR. As shown in Fig. 4a, during CO-TPR starting at 30 °C, the as-calcined Pt/CeO₂(UT) sample shows a peak at 2096 cm⁻¹ for CO adsorption on ionic Pt^δ+,^6,18 as the Pt cluster surfaces are either covered with PtO_x or chemisorbed O. As the temperature increases to 100 °C, a band at 2061 cm⁻¹ and a broad tail with lower frequencies (2060–1950 cm⁻¹) appeared, which are assigned to linearly adsorbed CO on metallic Pt⁰ atoms with different coordination numbers,^6,18 indicating the initiated reduction of the Pt species by CO at this temperature. Therefore, the reduction peak of the pristine Pt/CeO₂(UT) sample in Fig. 3 at 130 °C is accompanied by the reduction of the Pt clusters. As the temperature further increases, the ionic Pt^δ+ (2091–2096 cm⁻¹) gradually disappears accompanied by the growth of metallic Pt (2061 cm⁻¹). After the re-oxidation treatment at 250 °C and cooling to 30 °C, CO adsorption on the catalyst shows a peak at 2080 cm⁻¹, which is assigned to the co-adsorption of CO and oxygen on the Pt clusters.¹⁸

Fig. 4 DRIFTS spectra of CO adsorption on (a) Pt/CeO₂(UT) and (b) Pt/CeO₂(800) during and after CO-TPR to 250 °C.

For Pt/CeO₂(800), the as-calcined sample also shows a peak at 2098 cm⁻¹ for the CO adsorption on ionic Pt^δ+ (ref. 6 and 18) at room temperature. These ionic Pt^δ+ species remain stable during CO-TPR up to 100 °C, and metallic Pt species appear at 150 °C, 50 °C higher than for Pt/CeO₂(UT) (Fig. 4a). Since no reduction of Pt^δ+ species occurs on pristine Pt/CeO₂(800) at 100 °C, the reduction peak at 70 °C must be exclusively associated with the consumption of lattice oxygen from the CeO₂(800) support. As the reduction temperature increases above 150 °C, the appearance of metallic Pt species (2077, 2060, and 2020–2009 cm⁻¹) is associated with the aggregation of two-dimensional Pt rafts (Fig. 2b) to three-dimensional Pt nanoclusters (Fig. 2c). After the re-oxidation at 250 °C and cooling to 30 °C for CO adsorption, the Pt/CeO₂(800) sample presents two peaks at 2077 and 2093 cm⁻¹, respectively. Although the former is associated with the co-adsorption of CO and oxygen on the Pt clusters, similar to that on Pt/CeO₂(UT), the latter is assigned to the ionic Pt^δ+ species, suggesting that Pt is easier to be oxidized on Pt/CeO₂(800). The lower reduction temperature of CeO₂ surface lattice oxygen than that of PtO_x, together with the ease of PtO_x formation, suggests a strong metal–support interaction on this catalyst. In addition, the pristine and reduced Pt/CeO₂(800) have different Pt oxidation states (Fig. 4b) but similar redox activity (Fig. 3b), indicating that the oxidation states of the Pt clusters (except Pt single atoms) do not influence the redox activity of Pt/CeO₂(800). Therefore, we postulated that the redox activity was related to the nature of the CeO₂ support.

The TEM images (Fig. 2b and c) show that the CeO₂(800) support consists of larger CeO₂ particles with the well-defined (111) facet exposed. In contrast, the untreated CeO₂(UT) support (Fig. 2a) has a less defined morphology. Another difference between these two CeO₂ supports is the surface carbonate species. Fig. 5 shows the DRIFTS spectra of the carbonate/bicarbonate species on the two supports CeO₂(500) (prepared by calcining CeO₂(UT) at 500 °C in flowing air for 4 h) and CeO₂(800) and two catalysts Pt/CeO₂(UT) and Pt/CeO₂(800). The surfaces of CeO₂(500) and Pt/CeO₂(UT) contain much more carbonate species than those of CeO₂(800) and Pt/CeO₂(800), including monodentate (1464 cm⁻¹), polydentate (1395 cm⁻¹), and bidentate (1301 cm⁻¹) carbonate species.^19–22 These carbonate species are likely formed by the adsorption of CO₂ on the defect sites on the CeO₂ surface. The carbonate species on CeO₂(UT) are very stable, as they cannot be removed by 400 °C pretreatment in 10% O₂/He. The CeO₂(800) and Pt/CeO₂(800) surfaces mainly contain less stable bicarbonate species (3609 and 1607 cm⁻¹).^19–22 There are two possible reasons for the much lower surface carbonate concentration on CeO₂(800): (1) the high temperature calcination (800 °C for 10 h) decreases the amount of surface defects; (2) CeO₂(800) predominantly exposes the (111) facet which is much less favorable for CO₂ adsorption in comparison to the (100) and (110) facets.²³ The lower surface defect density on the high temperature treated sample was further confirmed by XPS. Fig. 6 shows the O1s XPS spectra of the pristine Pt/CeO₂(UT) and Pt/CeO₂(800) catalysts. The peaks at 529.3, 530.6, and 532.8 eV are associated with CeO₂ lattice oxygen, CeO₂ surface defects/PtO_x oxygen, and adsorbed water/other adsorbed species, respectively.²⁴ Given a similar amount of PtO_x species on these two catalysts, a smaller peak of Pt/CeO₂(800) at 530.6 eV than that of Pt/CeO₂(UT) indicates a lower amount of surface defects on the former.

Fig. 5 DRIFTS spectra of carbonated/bicarbonate species on the two supports CeO₂(500) (prepared by calcining CeO₂(UT) at 500 °C in flowing air for 4 h) and CeO₂(800) and two catalysts Pt/CeO₂(UT) and Pt/CeO₂(800) at 80 °C (all the samples were pretreated with 10% O₂/He at 400 °C for 30 min before cooling down to 80 °C for spectrum collection).

Fig. 6 O1s XPS spectra of the pristine Pt/CeO₂(UT) and Pt/CeO₂(800) catalysts.

The high redox activity of Pt/CeO₂(800) is likely related to much lower carbonate concentrations on the CeO₂(800) support. It's been widely accepted that the redox properties of CeO₂ is associated with the surface oxygen vacancies.^10,11,25 The migration of oxygen in CeO₂ takes place via a vacancy hopping mechanism.^10,26 We postulated that the formation of stable carbonates on the surface defects likely hinders the hopping of oxygen vacancies along the CeO₂ surface. Therefore, oxygen migration is more favorable on the CeO₂(800) surface which contains much less stable carbonate species. CO-TPR (Fig. 3) cannot reflect the difference between the bare CeO₂(UT) and CeO₂(800) supports, simply because the CeO₂ lattice oxygen requires Pt species as active sites to react with CO.

Catalytic activity for HC oxidation

The catalytic performance of Pt/CeO₂(UT) and Pt/CeO₂(800) for HC oxidation was evaluated with selected fuel blendstock components with different functional groups. Fig. 7a shows the T₅₀ and T₉₀ (temperatures of 50% and 90% conversion, respectively) values for the light-off of 8 fuel components on Pt/CeO₂(800) under different simplified and simulated test conditions (as listed in Table 1). Under simplified conditions with 0.74% or 10% O₂, the functional group of the fuel molecule largely impacts its oxidation reactivity. The unfunctionalized alkanes (isooctane and n-heptane) are more reactive (with lower T₅₀ and T₉₀) in low O₂ (0.74%) than in high O₂ (10%), whereas all the functionalized HCs with C=C, C–O, or C=O show a reverse trend, being more reactive under a high O₂ condition. This is associated with the competitive adsorption between O₂ and fuel molecules on the Pt catalysts. The unfunctionalized alkanes adsorb relatively weakly on Pt. Therefore, under high O₂ pressure (10%), the Pt surface is predominated by oxygen species with low coverage of alkanes, resulting in a low reaction rate. On this O-dominated Pt surface, reducing the O₂ pressure increases the alkane coverage and promotes the reaction rate. In contrast, the functionalized HC molecules (with C=C, C–O, or C=O) adsorb strongly on Pt, leading to an HC-dominated surface with low oxygen coverage, so the oxidation reaction is limited by the availability of active O species. In this case, increasing the O₂ pressure promotes the reaction rate.

Fig. 7 (a) T₅₀ and T₉₀ (temperatures of 50% and 90% conversion) for the oxidation of selected fuel components on the Pt/CeO₂(800) catalyst under different simulated or simplified exhaust conditions; (b) comparison of T₅₀ and T₉₀ for isooctane and toluene oxidation on the Pt/CeO₂(800) and Pt/CeO₂(UT) catalysts under different conditions (space velocity = 200 L g⁻¹ h⁻¹; gas compositions are summarized in Table 1; all the catalysts were reduced by 1% CO/N₂ at 250 °C before the test).

Under the simulated stoichiometric (S-GDI) and lean (LTC-G) conditions, the T₅₀ and T₉₀ values of all the fuel components are significantly increased, in comparison to the corresponding simplified conditions (0.74% and 10% O₂, respectively), as shown in Fig. 7a. This is the well-known inhibition effects of other exhaust components (CO, NO, and H₂O) on HC oxidation over Pt catalysts, associated with adsorption competition.^27,28

We selected isooctane and toluene as representative fuel components with weak and strong adsorption, respectively, to compare the catalytic performance of Pt/CeO₂(UT) and Pt/CeO₂(800). As shown in Fig. 7b, for both isooctane and toluene, Pt/CeO₂(800) shows lower T₅₀ and T₉₀ than Pt/CeO₂(UT) under all the conditions.

We rule out Pt dispersion or Pt morphology to be the reason for the high activity of Pt/CeO₂(800). As shown in Fig. 8, the light-off curves for the isooctane oxidation on the pristine Pt/CeO₂(800) with 2D Pt rafts (Fig. 2b) and the reduced Pt/CeO₂(800) with 3D Pt particles (Fig. 2c) almost overlap, indicating that the morphology of the Pt clusters does not influence the HC oxidation activity. On the other hand, Pt/CeO₂(UT) (Fig. 2a) and Pt/CeO₂(800) (Fig. 2c) have close Pt particle sizes (1.5–2 nm) and thus similar Pt dispersion. Therefore, the high HC oxidation activity of Pt/CeO₂(800) is likely related to its high redox activity. Next, we applied more rigorous kinetic measurements of propene oxidation to gain more insights into the superior activity of Pt/CeO₂(800) for HC oxidation.

Fig. 8 Light-off curves for the isooctane oxidation on the pristine and reduced Pt/CeO₂(800) catalysts (light-off condition: S-GDI; catalyst reduction treatment conditions: 1% CO/He, 250 °C, 30 min).

Kinetic insights into the influence of redox activity on HC oxidation on Pt/CeO₂

To better understand the promoted HC oxidation activity of Pt/CeO₂(800) in comparison to Pt/CeO₂(UT), we examined the kinetic dependence of propene (C₃H₆) oxidation on these two catalysts. We fixed the O₂ partial pressure and varied the C₃H₆ pressure. As shown in Fig. 9, under a specific O₂ pressure, as the C₃H₆ pressure (or C₃H₆/O₂ ratio) increases, the rate of C₃H₆ oxidation first increases and then decreases. Such kinetic dependence transition is due to the change of relative coverage of oxygen and HC species on the Pt surface, as previously reported for HC (e.g., methane, propane) oxidation on Pt/Al₂O₃ catalysts.^29,30 With a low C₃H₆/O₂ ratio, the Pt surface is predominated by O species, and the total oxidation rate is limited by the availability of activated C₃H₆. Thus, increasing the C₃H₆/O₂ ratio and C₃H₆ coverage will increase the oxidation rate. Under a high C₃H₆/O₂ ratio, the Pt surface is predominantly covered by C₃H₆ and the oxidation rate is limited by the availability of activated oxygen species, so further increasing C₃H₆/O₂ decreases the oxidation rate.

Fig. 9 Rates of propene oxidation on (a) Pt/CeO₂(UT) and (b) Pt/CeO₂(800) as a function of C₃H₆-to-O₂ ratio (220 °C).

It is noted that, regardless of the O₂ pressure, the C₃H₆/O₂ ratio for the kinetic regime transition remains constant (0.01 on Pt/CeO₂(UT) and 0.025 on Pt/CeO₂(800), respectively), suggesting that under this condition the Pt surface is fully covered by oxygen and C₃H₆ species and is free of bare Pt sites. In other words, under a fixed C₃H₆/O₂ ratio (e.g., 0.01 for Pt/CeO₂(UT)), increasing the O₂ pressure does not change the coverage of C₃H₆ or oxygen species on the Pt surface. However, the C₃H₆ oxidation rate still increases with increasing O₂ pressure. Such a positive reaction order on O₂ pressure under constant Pt surface species coverage indicates that the rate of C₃H₆ oxidation is not limited by the Langmuir–Hinshelwood reaction between the activated C₃H₆ and oxygen species on the Pt surface. Instead, the kinetically relevant step appears to proceed either via the Eley–Rideal mechanism occurring between a C₃H₆-derived intermediate on the Pt surface and gas phase oxygen or via the Mars–van Krevelen mechanism between a C₃H₆-derived intermediate and active oxygen from the CeO₂ surface (lattice oxygen). Both the mechanisms show a positive reaction order under O₂ pressure.³¹ This phenomenon is in line with previous discovery by Peng and co-workers,⁸ who demonstrated that oxygen species for toluene oxidation on Pt/CeO₂ catalysts is supplied by the surface lattice oxygen of the CeO₂ support and the rate-limiting step for toluene oxidation on Pt/CeO₂ catalysts is the activation of oxygen species at the oxygen vacancies of the CeO₂ surface, as they observed that the rates of toluene oxidation over three Pt/CeO₂ catalysts with different CeO₂ morphologies are not relevant to the Pt dispersion but proportional to the oxygen vacancy concentrations on CeO₂. In this work, the high HC oxidation activity of Pt/CeO₂(800) is likely associated with its high mobility of surface lattice oxygen as demonstrated by CO-TPR (Fig. 3b).

It is also noted that the C₃H₆/O₂ ratio for the transition from the oxygen-dominated surface to the HC-dominated surface is higher on Pt/CeO₂(800) than on Pt/CeO₂(UT) (0.025 vs. 0.01). This indicates that the high lattice oxygen mobility of Pt/CeO₂(800) increases the chemical potential of oxygen species on the Pt surface and broadens the operation window of the oxygen-dominated surface, which is beneficial for the catalyst operation under rich conditions.

Conclusions

High temperature (800 °C) pretreatment of CeO₂ support before loading Pt significantly promotes the redox activity of Pt/CeO₂ catalysts. The improved redox activity is not related to the structure or oxidation state of Pt clusters. Instead, it is likely associated with the lower amount of surface defects on the 800 °C-treated CeO₂ support. The low-defect surface favors the mobility of lattice oxygen and thus improves the redox activity of Pt/CeO₂ catalysts.

The Pt/CeO₂ catalysts were evaluated in the oxidation of a series of hydrocarbon fuel molecules under various conditions. The functional groups (C=C, C–O, and C=O bonds) of the fuel molecules affect their adsorption strengths and thus the oxidation reactivities under different conditions. Under simplified conditions with only O₂ and fuel molecules, the unfunctionalized alkanes, which adsorb relatively weakly on Pt catalysts, are more reactive under stoichiometric (low O₂) conditions than under lean (high O₂) conditions, whereas all the functionalized (by C=C, C–O, and C=O) fuel molecules with stronger adsorption show a reverse trend, being more reactive under lean conditions. Under simulated conditions, the presence of NO and CO suppresses the oxidation reactivities of all fuel molecules by competitive adsorption. In this case, all fuel molecules are more reactive under lean (LTC-G, high O₂) conditions than under stoichiometric (S-GDI, low O₂) conditions.

The Pt/CeO₂ catalyst with 800 °C-treated CeO₂ support presents promoted catalytic activity for HC oxidation, for both weakly-adsorbed alkanes (represented by isooctane) and strongly-adsorbed aromatics (represented by toluene), under both lean and stoichiometric conditions. Such improved HC oxidation activity is associated with the promoted redox activity. Kinetic measurement indicates that the promoted redox activity benefits the HC oxidation reaction in two ways: (1) accelerating the kinetically-relevant O₂ activation step and (2) increasing the oxygen chemical potential on Pt clusters so as to broaden the operation window of an oxygen-dominant Pt surface.

Author contributions

Fan Lin: methodology, investigation, formal analysis, and writing – original draft. Kenneth Rappé: investigation. Libor Kovarik: investigation. Miao Song: investigation. Xiaohong Shari Li: investigation. Mark Engelhard: investigation. Yong Wang: funding acquisition, project administration, conceptualization, methodology, and writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was sponsored by the US Department of Energy (DOE), Vehicle Technologies Office under the DOE Co-optimization of Fuels and Engines Initiative. A part of the research described in this paper was performed in the Environmental Molecular Sciences Laboratory (EMSL), a national scientific user facility sponsored by the DOE's Office of Biological and Environmental Research and located at PNNL. PNNL is operated for the US DOE by Battelle.

References

1. D. J. Gaspar, B. H. West, D. Ruddy, T. J. Wilke, E. Polikarpov, T. L. Alleman, A. George, E. Monroe, R. W. Davis, D. Vardon, A. D. Sutton, C. M. Moore, P. T. Benavides, J. Dunn, M. J. Biddy, S. B. Jones, M. D. Kass, J. A. Pihl, M. M. Debusk, M. Sjoberg, J. Szybist, C. S. Sluder, G. Fioroni and W. J. Pitz, Top Ten Blendstocks Derived From Biomass For Turbocharged Spark Ignition Engines: Bio-blendstocks With Potential for Highest Engine Efficiency, United States, 2019.
2. L. F. Liotta, Appl. Catal., B, 2010, 100, 403–412.
3. M. A. Salaev, A. A. Salaeva, T. S. Kharlamova and G. V. Mamontov, Appl. Catal., B, 2021, 295, 120286.
4. S. Chang, M. Li, Q. Hua, L. Zhang, Y. Ma, B. Ye and W. Huang, J. Catal., 2012, 293, 195–204.
5. N. Singhania, E. A. Anumol, N. Ravishankar and G. Madras, Dalton Trans., 2013, 42, 15343–15354.
6. X. I. Pereira-Hernández, A. DeLaRiva, V. Muravev, D. Kunwar, H. Xiong, B. Sudduth, M. Engelhard, L. Kovarik, E. J. M. Hensen, Y. Wang and A. K. Datye, Nat. Commun., 2019, 10, 1358.
7. H. Huang, Q. Dai and X. Wang, Appl. Catal., B, 2014, 158-159, 96–105.
8. R. Peng, X. Sun, S. Li, L. Chen, M. Fu, J. Wu and D. Ye, Chem. Eng. J., 2016, 306, 1234–1246.
9. B. Wang, B. Chen, Y. Sun, H. Xiao, X. Xu, M. Fu, J. Wu, L. Chen and D. Ye, Appl. Catal., B, 2018, 238, 328–338.
10. X. Liu, K. Zhou, L. Wang, B. Wang and Y. Li, J. Am. Chem. Soc., 2009, 131, 3140–3141.
11. F. Esch, S. Fabris, L. Zhou, T. Montini, C. Africh, P. Fornasiero, G. Comelli and R. Rosei, Science, 2005, 309, 752–755.
12. J. Lee, Y. Ryou, J. Kim, X. Chan, T. J. Kim and D. H. Kim, J. Phys. Chem. C, 2018, 122, 4972–4983.
13. Q. Fu, H. Saltsburg and M. Flytzani-Stephanopoulos, Science, 2003, 301, 935–938.
14. S. Sinha Majumdar, J. A. Pihl and T. J. Toops, Appl. Energy, 2019, 255, 113640.
15. S. Sinha Majumdar and J. A. Pihl, Energy Fuels, 2020, 34, 12900–12910.
16. K. G. Rappé, C. DiMaggio, J. A. Pihl, J. R. Theis, S. H. Oh, G. B. Fisher, J. Parks, V. G. Easterling, M. Yang, M. L. Stewart and K. C. Howden, Emiss. Control Sci. Technol., 2019, 5, 183–214.
17. A. A. Voskanyan, C.-K. J. Tsui and K.-Y. Chan, J. Catal., 2020, 382, 155–164.
18. P. Bazin, O. Saur, J. C. Lavalley, M. Daturi and G. Blanchard, Phys. Chem. Chem. Phys., 2005, 7, 187–194.
19. Z. Wu, A. K. P. Mann, M. Li and S. H. Overbury, J. Phys. Chem. C, 2015, 119, 7340–7350.
20. M. Li, U. Tumuluri, Z. Wu and S. Dai, ChemSusChem, 2015, 8, 3651–3660.
21. U. Tumuluri, G. Rother and Z. Wu, Ind. Eng. Chem. Res., 2016, 55, 3909–3919.
22. C. Li, Y. Sakata, T. Arai, K. Domen, K.-I. Maruya and T. Onishi, J. Chem. Soc., Faraday Trans. 1, 1989, 85, 929–943.
23. A. R. Symington, R. M. Harker, M. T. Storr, M. Molinari and S. C. Parker, J. Phys. Chem. C, 2020, 124, 23210–23220.
24. A. I. Stadnichenko, V. V. Muravev, S. V. Koscheev, V. I. Zaikovskii, H. A. Aleksandrov, K. M. Neyman and A. I. Boronin, Surf. Sci., 2019, 679, 273–283.
25. C. T. Campbell and C. H. F. Peden, Science, 2005, 309, 713–714.
26. C. R. A. Catlow, J. Chem. Soc., Faraday Trans., 1990, 86, 1167–1176.
27. K. Daneshvar, R. Krishna Dadi, D. Luss, V. Balakotaiah, S. B. Kang, C. M. Kalamaras and W. S. Epling, Chem. Eng. J., 2017, 323, 347–360.
28. S. E. Voltz, C. R. Morgan, D. Liederman and S. M. Jacob, Product R&D, 1973, 12, 294–301.
29. C. P. O'Brien, G. R. Jenness, H. Dong, D. G. Vlachos and I. C. Lee, J. Catal., 2016, 337, 122–132.
30. Y.-H. Chin, C. Buda, M. Neurock and E. Iglesia, J. Am. Chem. Soc., 2011, 133, 15958–15978.
31. J. R. H. Ross, in Contemporary Catalysis, ed. J. R. H. Ross, Elsevier, Amsterdam, 2019, pp. 161–186,  DOI:10.1016/B978-0-444-63474-0.00007-2.

---