English

• In 2021, global CO₂ emissions from energy combustion and industrial processes reached the highest annual level ever recorded, 36.3 billion tons [1]. To achieve the IPCC (Intergovernmental Panel on Climate Change) goal of limiting global warming to 1.5 ℃ by 2100, global greenhouse gas emissions, especially CO₂, must be significantly reduced. Using CO₂ as a feedstock to produce value-added chemicals is a more economically valuable way to reduce carbon than storing it deep underground. Currently, the reaction routes for converting CO₂ into chemicals can be divided into reductive and non-reductive reactions. Products that can be obtained by reductive reactions include CO, methane, methanol, formaldehyde, formic acid, carbon-rich hydrocarbons, oxygenated compounds, etc. Since CO₂ has a very low Gibbs free energy and carbon is in its highest oxidation state, it is chemically inert in most cases. Therefore, reduction of CO₂ requires external energy or the use of high-energy molecules (e.g., H₂). Non-reductive reactions are another route for CO₂ conversion that require less external energy input because the oxidation state of the carbon in the target product is the same as in CO₂. This route can produce urea, carbamates, organic carbonates, etc. Among them, direct synthesis of organic carbonates from CO₂ and alcohols is an attractive process [2,3]. For example, using CO₂ and glycerol as feedstock to produce GC (Eq. 1) cannot only help reduce CO₂ emissions, but also consume glycerol, a byproduct generated in biodiesel production. GC can be used as an environmentally friendly solvent, additive, chemical intermediate, polymer monomer, etc. [4,5]. Especially in recent years, with the explosive development of the new energy vehicle industry, the demand for organic carbonates as the electrolyte component of lithium-ion battery grows rapidly.

  (1)

  The atomic utilization (atom economy) of this process is 87% and the only byproduct is H₂O, which is a typical reaction in accordance with the principles of "green chemistry" [6]. However, this process has not yet been commercialized due to the lack of highly active and long-lived catalysts, in addition to the low yield caused by the thermodynamic equilibrium limitation of the reaction, but this limitation can be overcome to some extent by innovating the reaction process [7], using dehydrating agents to remove the water produced during the reaction [2,3], or applying an external field [8]. Organic tin compounds were first used as homogeneous catalysts to catalyze this reaction, mainly ⁿBu₂SnO or ⁿBu₂Sn(OMe)₂ [9,10]. However, there have been few follow-up studies for this catalyst due to its strong toxicity and difficulty in recycling. Heterogeneous catalysts such as inorganic bases, Cu/La₂O₃ [11], La₂O₂CO₃-ZnO [12], Zn/Al/La and Zn/Al/La/M (M = Li, Mg, Zr) [13] also have been studied for this reaction, which have the advantage to be easily separated.

  Recently, various metal oxide catalysts (including CeO₂, ZnO, TiO₂, ZrO₂, Al₂O₃, Ho₂O₃, Pr₆O₁₁, Nd₂O₃, Sm₂O₃, Bi₂O₃, Er₂O₃, GeO₂, SrO, Gd₂O₃, and Mo₂O₃) were tested for the direct synthesis of CO₂ and glycerol to GC and found that CeO₂ showed the best catalytic performance [14,15]. However, to our knowledge, no experimental study with CeO₂ model catalysts (single crystals or shaped nanocrystals) and no theoretical study on the catalytic mechanism of the direct GC formation from CO₂ and glycerol on CeO₂ (111), (110) and (100) surfaces have been reported. In the present study, we synthesized single-crystalline CeO₂ nano-octahedra, nanorods, and nanocubes as model catalysts, which predominantly expose {111}, {110}, and {100} facets, respectively. The nanocrystals and reaction system were characterized by scanning and transmission electron microscopy (SEM, TEM), X-ray diffraction (XRD), temperature programmed reduction of H₂ (H₂-TPR), temperature programmed desorption of CO₂ and NH₃ (CO₂-TPD, NH₃-TPD), Raman spectroscopy, and attenuated total reflectance infrared spectroscopy (ATR-IR). We also systematically investigated the mechanism of adsorption, activation, and reaction of CO₂ and glycerol on the surface of CeO₂ using first-principles calculations. This study for the first time determined the relationship between the surface structure, acid-base properties of CeO₂ and its catalytic performance, provided deeper insights into the catalytic mechanism at the molecular level and thus will support the development of CeO₂-based catalysts for direct GC synthesis.

  Figs. 1a−h show the SEM and TEM images of the as-synthesized CeO₂ nanocrystals with different morphologies. The CeO₂ nano-octahedra have a size of 200−500 nm and predominantly expose {111} facets. Both nanorods have a diameter of 10−15 nm and the length varying from 150 to 500 nm with the {110} (or {100}) facets preferentially exposed. The nanocubes reveal a regular shape with edge lengths in the range of 150−200 nm, and the surface is dominated by {100} facets.

  Figure 1

  

  Figure 1.  SEM and TEM images of CeO₂ nano-octahedra (a, e), nano cubes (b, f), nanorods from Ce(NO₃)₃ (c, g), nanorods from CeCl₃ (d, h).

  DownLoad: Full-Size Img PowerPoint

  Fig. S1 presents the XRD patterns collected for the CeO₂ nanocrystals. All these samples display diffraction peaks at 2θ of 28.7°, 33.1°, 47.5°, 56.4°, and 59.1°, which are in line with the (111), (200), (220), (311), and (222) planes of pure fluorite cubic structure of CeO₂ (JCPDS 34-0394). The XRD peaks of CeO₂ nanorods are relatively broader compared to other CeO₂ nanocrystals, since CeO₂ nanorods have smaller size in some dimensions. The Raman spectra of the CeO₂ nanocrystals are collected as shown in Fig. S2. To visualize and compare the intensity difference between the different vibrational bands, all Raman spectra in Fig. S2 have been normalized to the F_2g band at 462 cm⁻¹. The band at about 258 cm⁻¹ is assigned to the second-order transverse acoustic (2TA) mode [16]. The band at ca. 598 cm⁻¹ has been attributed to the defect-induced (D) mode [16], which is very weak for all CeO₂ nanocrystals as synthesized indicating that the amount of oxygen vacancies on the sample surface can be neglected.

  Table 1 summarizes the average particle sizes determined from SEM and TEM images, the specific surface areas measured by Brunauer−Emmett−Teller (BET) analysis, and the density of basic and acid sites obtained from TPD-QMS. The specific area of CeO₂ nano-octahedra and nanocubes is comparable, but an order of magnitude smaller than that of nanorods. The acid-base properties will be discussed in detail below.

  Table 1

  

  Table 1.  Physicochemical properties of the CeO₂ nanocrystals. rod-N and rod-C refer to nanorods synthesized with Ce(NO₃)₃ and CeCl₃, respectively.

  DownLoad: CSV

  Fig. 2 shows the H₂-TPR of CeO₂ nanocrystals. It is known that TPR gives indications of the reducibility of surface and bulk of ceria. The reduction of conventional CeO₂ powders is characterized by a bimodal shape with a peak (or peaks) at low temperature characteristic of surface reduction and a large peak at high temperature typical of bulk reduction. In agreement with previous studies, the bulk reduction of all samples in this study shows a major peak in the temperature range from 650 ℃ to 850 ℃, which is nearly independent of the facets exposed by the nanocrystals. However, CeO₂ nanocrystals of different morphologies show a strongly different behavior for the low temperature peaks. The starting temperature of surface reduction of the CeO₂ nanocrystals follows the order of rods (200 ℃) < cubes (320 ℃) < octahedra (390 ℃). This indicates that the surface reducibility follows the order (110) < (100) < (111), which are predominantly exposed by CeO₂ nanorods, nanocubes, and nano-octahedra, respectively. Since the reduction of the surface leads to the formation of surface oxygen vacancies, the formation energy of surface oxygen vacancies should play a decisive role in the reducibility of the CeO₂ surface. Previous DFT calculations show that the oxygen vacancy formation energy of low-indexed CeO₂ surfaces follows the order (110) < (100) < (111) [17,18], which is thus consistent with our experimental observations. Two ceria nanorods synthesized with Ce(NO₃)₃ and CeCl₃ exhibit almost the same TPR profiles, with the surface reduction even showing a relatively higher peak intensity than the peak of bulk reduction at high temperature. The surface reduction of the CeO₂ nanocubes leads to two peaks at 400 and 560 ℃, with the second peak being quite close to the surface reduction peak of the CeO₂ nano-octahedra (545 ℃). This is because above 500 ℃ the nanocubes begin to change their morphology, resulting in the development of extended flat {111} facets (corner-truncated) and {110} planes (edge-truncated) [19–21]. And the edge-truncated {110} planes are further fully {111}-faceted, exposing a characteristic sawtooth-like structure [21,22].

  Figure 2

  

  Figure 2.  H₂-TPR-QMS profiles of the CeO₂ nanocrystals.

  DownLoad: Full-Size Img PowerPoint

  Based on previous studies on the synthesis of organic carbonates from alcohols and CO₂, acid-base properties are crucial for the catalyst performance since acidic sites usually facilitate the adsorption of methanol and basic sites activate CO₂. To investigate the acid-base properties of the catalysts, we measured the temperature programmed desorption (TPD) of NH₃ and CO₂. The chemisorption analyzer equipped with a downstream quadruple mass spectrometer (QMS) as detector can provide very high sensitivity. The NH₃-TPD-QMS profiles of all catalysts used to study the amount and strength of acid sites are shown in Fig. 3a. As expected, the NH₃-TPD-QMS profiles of CeO₂ nanocrystals of different morphologies are very different. The desorption peaks at 100–300 ℃ and 430–600 ℃ are attributed to the weak and strong acidic sites, respectively. It can be observed that two CeO₂ nanorods show nearly the same desorption behavior, a large peak for NH₃ desorbed from weak acid sites and a small peak desorbed from strong acid sites. The CeO₂ nano-octahedra exhibit a relatively small peak for NH₃ desorbed from weak acid sites. Surprisingly, the CeO₂ nanocubes show nearly no visible NH₃ desorption peak, indicating the absence of acidic sites on the (100) surface.

  Figure 3

  

  Figure 3.  NH₃-TPD-QMS (a) and CO₂-TPD-QMS (b) profiles of CeO₂ nanocrystals with different morphologies.

  DownLoad: Full-Size Img PowerPoint

  The CO₂-TPD-QMS profiles of all catalysts used to study the amount and strength of basic sites are shown in Fig. 3b. The CO₂-TPD-QMS profiles suggest that the CeO₂ nanocrystals contain weak, moderate, and strong basic sites with peak maxima of < 200, 200−400, and > 450 ℃, respectively. Again, the two CeO₂ nanorods exhibit almost the same CO₂-TPD-QMS profiles, indicating the quite similar surface properties of these two CeO₂ nanorods synthesized with different cerium salt precursors. The CeO₂ nano-octahedra and nanocubes exhibit similar desorption peaks, namely a large peak at about 100 ℃ for CO₂ desorbed from weak basic sites and a broad peak with peak maxima above 450 ℃ for CO₂ desorbed from strong basic sites. In addition, the nano-octahedra also exhibit a small peak at about 260 ℃ for CO₂ desorbed from moderate basic sites. The CO₂-TPD-QMS profiles over CeO₂ nano-octahedra and nanocubes agree well with those over well-defined CeO₂ (111) and (100) single crystalline films [23–25], suggesting that the CeO₂ nano-octahedra and nanocubes predominantly expose {111} and {100} facets.

  Fig. 4a shows the rate of GC formation per unit surface area with varying reaction time over CeO₂ nano-octahedra, nanorods, and nanocubes at 160 ℃ and 40 bar of CO₂ pressure. The rate of GC formation was high at the initial phase and decreased after 10 h. According to previous studies, a longer reaction time leads to the decomposition of GC, i.e., the reverse reaction, and thus to a lower instantaneous GC formation rate of the catalysts. At 24 h, the CeO₂ nano-octahedra have the cumulative formation rate of 1.548 mmol/m², while the two CeO₂ nanorods show much lower cumulative formation rate of 0.155 mmol/m² and 0.134 mmol/m², with the formation rate over CeO₂ nanocubes being in between, 0.988 mmol/m². CeO₂ nanorods synthesized with Ce(NO₃)₃ and CeCl₃ have almost the same activity, indicating that the residual NO₃⁻ and Cl⁻ anions have a similar influence on the activity or the influence can be neglected. Fig. S3 shows the comparison of ATR-IR spectra for glycerol carbonate and products obtained after 16 h of reaction with CeO₂ nanocrystal catalysts. Glycerol carbonate exhibits a C−O vibrational band at 1760 cm⁻¹, which appears for the products of all CeO₂ nanocrystal catalysts, indicating the formation of GC. Figs. S4 and S5 show the SEM and XRD data after catalytic reactions of the CeO₂ nanocrystals, which retained their original morphology and structure, indicating the high stability as reported in previous study [14].

  Figure 4

  

  Figure 4.  (a) GC formation rate per unit surface area at different reaction time over CeO₂ nano-octahedra, nanorods, and nanocubes. Reaction conditions: 160 ℃, 40 bar CO₂, 4.68 g glycerol, 8.515 g 2-cyanopyridine, 0.52 g CeO₂, and 24.44 g DMF. (b) Correlation between the catalytic activity (GC formation rate at reaction time of 5 h) and the basicity. The density of basic sites can be obtained by dividing the peak area value of basic sites (Fig. 3b) by the specific surface area.

  DownLoad: Full-Size Img PowerPoint

  It is commonly reported that for the synthesis of organic carbonates from CO₂ and alcohol, both acidic and basic sites must be present on the catalyst surfaces. In many reactions catalyzed by CeO₂, the catalytic reactivity is attributed to the redox properties of CeO₂, which are related to the oxygen vacancies or cerium cations of multiple oxidation states (Ce³⁺/Ce⁴⁺) [25–31]. Their importance in the present reaction cannot be completely ruled out. However, it should be noted that this reaction occurs under mild and non-reducing conditions, so it is unlikely that oxygen vacancies and Ce³⁺ cations are involved. The reaction is probably catalyzed only by acid and base sites.

  According to Fig. 4b, we found that the catalytic performance (GC formation rate) is not strictly linearly dependent on the acidity and the ratio of acidity to basicity, but linearly dependent on the density of basic sites. The oxygen anions on the CeO₂ surface are Lewis basic sites, which are relevant to the adsorption and activation of CO₂. It is quite plausible that CO₂ activation is more important for a three-phase reaction in which a very inert gaseous reactant, CO₂, and a high-velocity liquid reactant, glycerol, are coupling up over the surfaces of a solid catalyst, CeO₂. This correlation is also mentioned by many previous studies [12,13,32,33].

  To depict a more microscopic reaction mechanism underlying the formation of GC from CO₂ and glycerol on all three CeO₂ (111), (110), and (100) surfaces, we also performed a comprehensive first principles calculations.

  First, various placements and geometric configurations of pre-adsorbed CO₂ and glycerol are examined. After full geometry optimization, the adsorption energies of CO₂ and glycerol on the stoichiometric CeO₂ (111), (110) and (100) surfaces are computed. Note that the CeO₂ (100) is a polar surface and must be restructured to compensate the surface polarity. Although the structure of the CeO₂ (100) surface reconstruction is still in debate currently, herein we adapted the more accepted O-terminated (2 × 2) model by removing 50% of surface oxygen atoms [21].

  As shown in Fig. 5a, on all three surfaces, the adsorbed CO₂ can be in either linear or bent configuration. In the linear configuration, the CO₂ molecule retains the linear geometry as in the gas phase, and no C–O bond is formed between CO₂ and CeO₂ surface. In the bent configuration, the CO₂ molecule forms a C–O bond (denoted as C–O_latt) with a lattice O on the CeO₂ surface. Due to the formation of the C–O_latt bond, the CO₂ molecule is negatively charged with the C–O double bond being elongated, resulting in a carbonate species. In such a bent configuration, the carbonate species forms a 'Y' shape with the original two oxygen atoms of CO₂ dangling toward the vacuum. Linear adsorption of CO₂ on CeO₂ (111), (110) and (100) surfaces entail comparable adsorption energies of −0.23, −0.25 and −0.21 eV, respectively. Stronger adsorption is expected in the bent configuration, which is true for (110) and (100) surfaces with more negative adsorption energy of −0.32 and −0.57 eV, respectively. Unexpectedly, the bent configuration of CO₂ adsorption on the CeO₂ (111) surface is a bit weaker than that for the linear configuration, entailing an adsorption energy of −0.19 eV; this is due to a larger distortion of surface lattice upon CO₂ adsorption on the CeO₂ (111) surface.

  Figure 5

  

  Figure 5.  Computed adsorption energies and optimized geometries of CO₂ (a) and glycerol (b) adsorbed on the CeO₂ (111), (110) and (100) surface. The black, red, white, and gray balls represent C, O, H, and Ce atoms, respectively. "dis" and "mol" refer to dissociative and molecular adsorption, respectively.

  DownLoad: Full-Size Img PowerPoint

  Fig. 5b shows the adsorption configurations of glycerol on the CeO₂ (111), (110), and (100) surfaces. The glycerol molecules can be adsorbed molecularly and dissociatively. In the molecular adsorption configuration, two glycerol O bind to the surface Ce cations and the −CH₂OH group dangles toward the vacuum. The molecular adsorption configurations of glycerol on CeO₂ (110) and (100) entail close adsorption energies, −1.30 and −1.51 eV, respectively, which are more negative than that of the molecularly adsorbed glycerol on the CeO₂ (111) surface (−0.94 eV). That sugg-ests the stronger binding of glycerol on the CeO₂ (110) and (100) surfaces. The dissociative adsorption configurations with one H dissociated from the −OH group of glycerol on CeO₂ (110) and (100) surfaces have more negative adsorption energies, −2.00 and −1.55 eV, respectively, than the molecular adsorption counterpart. Again, the dissociative adsorption of glycerol on the CeO₂ (111) surface is slightly weaker than the molecular adsorption and it has an adsorption energy of −0.82 eV.

  Upon the adsorption of two reactants, CO₂ and glycerol, three successive steps are identified towards the formation of GC on the CeO₂ surface, as shown by the calculation: the interaction of glycerol with the CeO₂ surface to form an adsorbed glyceroxide, which then attacks the CO₂ molecule to form a linear carbonate intermediate. Subsequently, the secondary hydroxyl of glycerol attacks the linear carbonate group to form the cyclic carbonate. Fig. 6 shows the most favorable energy profile for the formation of a GC from a CO₂ and a glycerol molecule on the CeO₂ (111), (110) and (100) surfaces. The activation energy for the first step on the CeO₂ (111), (110) and (100) surfaces is comparable and is 0.54, 0.50, and 0.55 eV, respectively.

  Figure 6

  

  Figure 6.  Energy profiles of the GC formation on the CeO₂ (111), (110) and (100) surfaces.

  DownLoad: Full-Size Img PowerPoint

  On the CeO₂ (111) surface (black dotted curve in Fig. 6), the linear carbonate intermediate can be formed from a slightly endothermic reaction between the adsorbed CO₂ (bent structure) and a co-adsorbed glycerol molecule. The formation of the linear carbonate intermediate from the bent CO₂ entails an activation energy barrier of 0.5 eV. Note that the formation of the intermediate from the linear CO₂ is also considered and both CO₂ configurations exhibit comparable activation energy barrier, which is also true for CeO₂ (110) and (100) surfaces. The subsequent attack of the secondary hydroxyl of glycerol to the carbonate group leads to the formation of GC. Such a step must overcome a high activation energy barrier (1.24 eV). On the CeO₂ (111) surface, the annulation step is the rate-limiting step for GC formation.

  The formation of GC from CO₂ and glycerol on the CeO₂ (110) surface proceeds via a three-step mechanism similar to that on the CeO₂ (111) surface. As shown in Fig. 6 by the red dotted curve, the co-adsorbed CO₂ and glycerol react with each other to form the linear carbonate intermediate with a higher activation energy barrier of 1.03 eV compared to the CeO₂ (111) surface. This process is slightly endothermic with an energy input of 0.48 eV. The subsequent annulation step entails a very high activation energy barrier of 2.61 eV, leading to the formation of the GC. As with the CeO₂ (111) surface, the annulation step is the rate-limiting step with a high activation energy barrier of 2.61 eV, which is 1.37 eV higher than that on the CeO₂ (111) surface.

  Next, we investigate the formation of GC on the CeO₂ (100) surface. The CeO₂ (100) surface is a polar surface and has a higher surface energy compared to the CeO₂ (111) and (100) surfaces. As shown in Fig. 6 by blue dotted curve, the GC formation on the CeO₂ (100) surface also follows the three-step mechanism as on the CeO₂ (111) and (110) surface. On the CeO₂ (100) surface the co-adsorbed CO₂ and glycerol react with each other and form the linear carbonate intermediate with an activation energy barrier of 1.28 eV. On the CeO₂ (100) surface, this step is associated with a higher activation energy barrier than on the CeO₂ (111) and (110) surfaces. In the following step, the annulation reaction to form GC is the rate-limiting step with the activation energy barrier of 1.45 eV. Passing the transition state leads to the formation of GC and hydration of the CeO₂ surface, and the activation energy is higher than for CeO₂ (111) but lower than for CeO₂ (110).

  From the optimized energy profile, the overall activation energy of GC formation on the CeO₂ surface follows the order of (111) < (100) < (110), which agrees well with the evaluation of the performance of CeO₂ nanocrystal catalysts, i.e., the CeO₂ nano-octahedra, predominantly exposing {111} facets, exhibit the highest GC formation rate and the CeO₂ (110) nanorods, predominantly exposing {110} facets, show the lowest GC formation rate.

  In summary, we synthesized CeO₂ nanocrystals with well-defined surface structure as model catalysts to investigate the reaction mechanism for the direct synthesis of GC from glycerol and CO₂. Under the same reaction conditions, CeO₂ nano-octahedra, predominantly exposing {111} facets, showed the highest GC formation rate than CeO₂ nanorods and nanocubes, which predominantly expose {110} and {100} facets, respectively. Using high-sensitivity TPD with quadrupole mass spectrometry as detector, we characterized the acid-base properties of CeO₂ nanocrystals. The catalytic performance (GC formation rate) was found to be strictly linearly dependent on the density of basic sites, suggesting that the density of surface basicity is crucial for the high catalytic performance, which is relevant to the adsorption and activation of CO₂.

  We have also performed the first principles calculations for adsorption, activation, and reaction of CO₂ and glycerol over CeO₂ (111), (110), and (100) surfaces and obtained deeper insights into the catalytic mechanism at the molecular level. It is found that CO₂ can be adsorbed on CeO₂ surfaces in either a linear or bent configuration, and glycerol can be adsorbed on CeO₂ surfaces molecularly and dissociatively. A three-step Langmuir–Hinshelwood (LH) mechanism was identified in which the annulation reaction is the rate-limiting step. The CeO₂ (111) surface exhibits the lowest overall activation energy, which agrees well with the catalytic performance that the CeO₂ nano-octahedra, predominantly exposing {111} facets, have the highest GC formation rate. The CeO₂ (110) surface exhibits the highest overall activation energy, which agrees well with the catalytic reaction performance that the CeO₂ nanorods, predominantly exposing {110} facets, have the lowest GC formation rate. Clarification of the catalytic reaction mechanism will help to develop CeO₂-based catalysts with better performance.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (No. 21902005); Beihang University; and State Key Laboratory of Advanced Technology for Materials Synthesis and Processing (Wuhan University of Technology).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.108395.

  
  
• 1. [1]

     IEA (2021), Global energy review 2021, IEA, Paris https://www.iea.org/reports/global-energy-review-2021, License: CC BY 4.0

  2. [2]

     D. Procopio, M.L. Di Gioia, Catalysts 12 (2022) 50. doi: 10.3390/catal12010050

  3. [3]

     S. Christy, A. Noschese, M. Lomelí-Rodriguez, N. Greeves, J.A. Lopez-Sanchez, Curr. Opin. Green Sustain. Chem. 14 (2018) 99–107. doi: 10.1016/j.cogsc.2018.09.003

  4. [4]

     M. Szőri, B.R. Giri, Z. Wang, A.E. Dawood, B. Viskolcz, Sustain. Energy Fuels 2 (2018) 2171–2178. doi: 10.1039/c8se00207j

  5. [5]

     M.O. Sonnati, S. Amigoni, E.P. Taffin de Givenchy, et al., Green Chem. 15 (2013) 283–306. doi: 10.1039/C2GC36525A

  6. [6]

     P. Anastas, N. Eghbali, Chem. Soc. Rev. 39 (2010) 301–312. doi: 10.1039/B918763B

  7. [7]

     D.C. Shi, S. Heyte, M. Capron, S. Paul, Green Chem. 24 (2022) 1067–1089. doi: 10.1039/d1gc04093f

  8. [8]

     J.X. Liu, Y.J. Li, H.M. Liu, D.H. He, Appl. Catal. B: Environ. 244 (2019) 836–843. doi: 10.1016/j.apcatb.2018.12.018

  9. [9]

     J. George, Y. Patel, S.M. Pillai, P. Munshi, J. Mol. Catal. Chem. 304 (2009) 1–7. doi: 10.1016/j.molcata.2009.01.010

  10. [10]

      M. Aresta, A. Dibenedetto, F. Nocito, C. Pastore, J. Mol. Catal. Chem. 257 (2006) 149–153. doi: 10.1016/j.molcata.2006.05.021

  11. [11]

      J. Zhang, D.H. He, J. Colloid Interface Sci. 419 (2014) 31–38. doi: 10.1016/j.jcis.2013.12.049

  12. [12]

      H.G. Li, D.Z. Gao, P. Gao, et al., Catal. Sci. Technol. 3 (2013) 2801–2809. doi: 10.1039/c3cy00335c

  13. [13]

      H.G. Li, X. Jiao, L. Li, et al., Catal. Sci. Technol. 5 (2015) 989–1005. doi: 10.1039/C4CY01237B

  14. [14]

      J.X. Liu, Y.M. Li, J. Zhang, D.H. He, Appl. Catal. A Gen. 513 (2016) 9–18. doi: 10.1016/j.apcata.2015.12.030

  15. [15]

      X.L.N. Su, W.W. Lin, H.Y. Cheng, et al., Green Chem. 19 (2017) 1775–1781. doi: 10.1039/C7GC00260B

  16. [16]

      Z.L. Wu, M.J. Li, J. Howe, H.M. Meyer III, S.H. Overbury, Langmuir 26 (2010) 16595–16606. doi: 10.1021/la101723w

  17. [17]

      M. Nolan, S.C. Parker, G.W. Watson, Surf. Sci. 595 (2005) 223–232. doi: 10.1016/j.susc.2005.08.015

  18. [18]

      J. Paier, C. Penschke, J. Sauer, Chem. Rev. 113 (2013) 3949–3985. doi: 10.1021/cr3004949

  19. [19]

      E. Aneggi, D. Wiater, C. de Leitenburg, J. Llorca, A. Trovarelli, ACS Catal. 4 (2014) 172–181. doi: 10.1021/cs400850r

  20. [20]

      Z.L. Wu, M.J. Li, S.H. Overbury, J. Catal. 285 (2012) 61–73. doi: 10.1016/j.jcat.2011.09.011

  21. [21]

      C.W. Yang, M. Capdevila-Cortada, C.Y. Dong, et al., J. Phys. Chem. Lett. 11 (2020) 7925–7931. doi: 10.1021/acs.jpclett.0c02409

  22. [22]

      C.W. Yang, X.J. Yu, S. Heißler, et al., Angew. Chem. Int. Ed. 56 (2017) 375–379. doi: 10.1002/anie.201609179

  23. [23]

      S.D. Senanayake, D.R. Mullins, J. Phys. Chem. C 112 (2008) 9744–9752. doi: 10.1021/jp8016425

  24. [24]

      P.M. Albrecht, D.E. Jiang, D.R. Mullins, J. Phys. Chem. C 118 (2014) 9042–9050. doi: 10.1021/jp501201b

  25. [25]

      D.C. Yang, Y.C. Xu, K. Pan, et al., Chem. Phys. Lett. 33 (2022) 378–384.

  26. [26]

      J.Y. Liu, X.Q. Gong, Chin. Chem. Lett. 32 (2021) 1127–1130. doi: 10.1016/j.cclet.2020.08.033

  27. [27]

      W.J. Yang, J.J. Li, X.Y. Cui, et al., Chin. Chem. Lett. 32 (2021) 2489–2494. doi: 10.1016/j.cclet.2020.12.057

  28. [28]

      L.N. Wang, F.M. Meng, K.K. Li, F. Lu, Appl. Surf. Sci. 286 (2013) 269–274. doi: 10.1016/j.apsusc.2013.09.067

  29. [29]

      F.M. Meng, Z.H. Fan, C. Zhang, et al., J. Mater. Sci. Technol. 33 (2017) 444–451. doi: 10.1016/j.jmst.2016.06.018

  30. [30]

      L.N. Wang, F.M. Meng, Mater. Res. Bull., 48 (2013) 3492–3498. doi: 10.1016/j.materresbull.2013.05.036

  31. [31]

      F.M. Meng, L.N. Wang, J.B. Cui, J. Alloy. Compd. 556 (2013) 102–108. doi: 10.1016/j.jallcom.2012.12.096

  32. [32]

      D. Stoian, F. Medina, A. Urakawa, ACS Catal. 8 (2018) 3181–3193. doi: 10.1021/acscatal.7b04198

  33. [33]

      J. Zhang, D.H. He, J. Chem. Technol. Biotechnol. 90 (2015) 1077–1085. doi: 10.1002/jctb.4414

• 

  Figure 1  SEM and TEM images of CeO₂ nano-octahedra (a, e), nano cubes (b, f), nanorods from Ce(NO₃)₃ (c, g), nanorods from CeCl₃ (d, h).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  H₂-TPR-QMS profiles of the CeO₂ nanocrystals.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  NH₃-TPD-QMS (a) and CO₂-TPD-QMS (b) profiles of CeO₂ nanocrystals with different morphologies.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) GC formation rate per unit surface area at different reaction time over CeO₂ nano-octahedra, nanorods, and nanocubes. Reaction conditions: 160 ℃, 40 bar CO₂, 4.68 g glycerol, 8.515 g 2-cyanopyridine, 0.52 g CeO₂, and 24.44 g DMF. (b) Correlation between the catalytic activity (GC formation rate at reaction time of 5 h) and the basicity. The density of basic sites can be obtained by dividing the peak area value of basic sites (Fig. 3b) by the specific surface area.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Computed adsorption energies and optimized geometries of CO₂ (a) and glycerol (b) adsorbed on the CeO₂ (111), (110) and (100) surface. The black, red, white, and gray balls represent C, O, H, and Ce atoms, respectively. "dis" and "mol" refer to dissociative and molecular adsorption, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  Energy profiles of the GC formation on the CeO₂ (111), (110) and (100) surfaces.

  下载: 全尺寸图片 幻灯片

  Table 1.  Physicochemical properties of the CeO₂ nanocrystals. rod-N and rod-C refer to nanorods synthesized with Ce(NO₃)₃ and CeCl₃, respectively.

  下载: 导出CSV
•