Photo-assisted Catalytic Removal of NO_x Over La_1–xPr_xCoO₃/Palygorskite Nanocomposites: Role of Pr Doping

Abstract

Photo-assisted selective catalytic reduction (photo-SCR) has been considered as a promising strategy for NO_x removal in recent decades. The purpose of the present work was to test the effectiveness of La_1–xPr_xCoO₃, supported on the surface of natural palygorskite (Pal) by a facile sol-gel method, as a photo-SCR for the removal of NO_x from wastewaters. The structure, acidity, and the redox property of the prepared La_1–xPr_xCoO₃/Pal nanocomposite were characterized using X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), UV-Visible diffuse reflectance spectroscopy (UV-Vis DRS), and X-ray photoelectron spectroscopy (XPS). Density functional theory (DFT) calculations were employed to determine the valence bands. The La_1–xPr_xCoO₃/Pal catalysts were then tested for SCR removal of NO_x with the assistance of photo-irradiation. The photo-SCR results revealed that the NO_x conversion and the N₂-selectivity were greatly improved by this method and reached >95% when carried out at the relatively low temperature of 200°C and with the Pr doping at x = 0.5. The improvements were attributed to the co-precipitation of a PrCoO₃ phase as in a solid solution forming a coherent heterojunction of PrCoO₃/La_0.5Pr_0.5CoO₃ on the Pal surface.

Introduction

Nitrogen oxide (NO _x ) is a significant pollutant and is a cause of a number of serious environmental problems including acid rain and atmospheric haze (Geng et al. 2017; Xiao et al. 2018). Developing a treatment for NO _x is, therefore, critically important. Selective catalytic reduction has been studied extensively and is used commonly for NO _x removal at high temperature (300–400°C); this process has comparatively high costs and undesirable by-products, however. To solve these problems, photo-assisted selective catalytic reduction (photo-SCR) of NO _x with NH₃ using semiconductor catalysts under the conditions of photo-illumination and low temperature is proposed as a possible alternative for NO _x reduction (Yu et al. 2017).

The structure of perovskite is ABO₃ where A is usually a lanthanide or alkaline earth metal ion, and B is a transition metal ion. Rare- or alkaline-earth elements provide mechanical resistance to the perovskite while transition metals improve the reactivity in redox processes (Cai et al. 2018). Perovskite has a narrow bandgap and good photoresponse to visible light. Due to the flexible alternation of A or B sites, perovskite has been studied widely in photocatalysis (Bhaskar et al. 2017; Humayun et al. 2018). As a typical perovskite, LaCoO₃ is easy to agglomerate because of its nano-size particles and large surface energy. Natural clay minerals are considered to offer potential as effective catalyst carriers because of their unique pore structure and layer structure (Zhou & Keeling 2013; Zhou et al. 2016; Chen et al. 2016; Zhu et al. 2019). Palygorskite (Pal), a naturally occurring, fibrous clay mineral, is an excellent example of such a supporting material in catalysts (Li et al. 2017; Liu et al. 2017; Wang et al. 2018) because of its large specific surface area, which inhibits the aggregation of particles, and good adsorption performance (Kadir et al. 2017; Lin et al. 2017]. The purpose of the present study was to measure the effectiveness of La_1–x Pr _x CoO₃/Pal nanocomposites, synthesized using a facile sol-gel approach, as a photo-assisted SCR of NO _x with NH₃, including an investigation of the effect of Pr doping on the conversion rate and the N₂ selectivity.

Experimental

Chemicals

La(NO₃)₃·6H₂O, C₆H₈O₇·H₂O, Pr(NO₃)₃·6H₂O, and Co(NO₃)₃·6H₂O were provided by Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Palygorskite was supplied by Jiangsu Nanda Zijin Technology Group Co., Ltd. (Changzhou, China). Citric acid and ethylene glycol were purchased from Shanghai Ling Feng Chemical Reagent Co., Ltd. (Shanghai, China). All reagents were of analytical grade and used without further purification.

Synthesis of La_1–xPr_xCoO₃/Pal

La_1–x Pr _x CoO₃/Pal nanocomposites were prepared by the facile sol-gel method which is summarized briefly as follows: 1.34 g of lanthanum nitrate, 0.12 g of cobalt nitrate, 1.50 g of praseodymium nitrate, and 1.80 g citric acid were mixed in a 250.0 mL beaker, giving x = 0.1. As these ratios were changed to yield varying values for x, the amount of nitrate also changed. The mixture was dissolved in 10.0 mL of deionized water under ultrasonication. The relative amounts of La and Pr, i.e. the value of x, varied from 0.1 to 0.9. The mixed solution was maintained at 80°C with stirring for 1 h, followed by the addition of 1.0 g of Pal. The composite catalyst was obtained after calcining at 650°C for 2 h.

Characterization

The TEM images were obtained using a transmission electron microscope (JEM–2100, JEOL, Tokyo, Japan), working at 200 kV. The XRD patterns were captured using a Rigaku D/Max-2500 X-ray diffractometer equipped with a Cu anode (Rigaku Corporation, Tokyo, Japan), running at 60 kV and 30 mA between the angles of 10 and 80°2θ at a scan rate of 0.05°2θ/s. The Fourier-transform infrared (FTIR) spectra were measured using a PerkinElmer Spectrum 100 FT-IR spectrometer (PerkinElmer, Shelton, Connecticut, USA).

The total acidity and acid distribution of the catalysts were measured by means of the temperature-programmed desorption (TPD) of NH₃ using a Micromeritics (Norcross, Georgia, USA) ASAP 2920 instrument equipped with a thermal conductivity detector (TCD). A sample of the La_1–x Pr _x CoO₃/Pal nanocomposite (sieved to 0.2–0.3 mm, 0.3 g) was added to a fixed-bed flow reactor using N₂ as the flow gas. The sample was degassed at 400°C for 30 min, and then cooled to room temperature (~25°C). A flow (30 mL/min) of NH₃ was introduced for 30 min followed by a purge with N₂ to remove the physically adsorbed NH₃ on the catalyst surface. TPD of chemically adsorbed NH₃ was then carried out under N₂ flow (50 mL/min) at 25–500°C with a heating rate of 10°C/min.

Temperature-programmed reduction by hydrogen (H₂-TPR) was performed using the same Micromeritics ASAP 2920 instrument as for TPD. Samples of ~50 mg were heated from ambient temperature to 700°C at 10°C/min in a reducing atmosphere of H₂ mixed (10 vol.%) with Ar (flow rate of 30 mL/min).

XPS measurements (Thermo Fisher Scientific, Waltham, Massachusetts, USA) were carried out using a Quantum 2000 Scanning ESCA Microprobe instrument using Al Kα. The C1s signal was set to a position of 284.6 eV.

Photo-SCR performance

The photo-SCR catalytic experiments were carried out in a fixed-bed reactor operated in a steady-state flow mode. A 500 W Xenon lamp was employed to provide a light source irradiation to simulate solar light. A UV-filter was placed at both ends of the window to cut off wavelengths <420 nm to guarantee only visible light irradiation. The catalyst temperature was measured through a thermocouple projecting into the center of the reactor. The reactant gas consisted of 1000 ppm NO, 1000 ppm NH₃, and 3 vol.% O₂, with the balance being N₂, which follows the conditions of “practical” SCR (Li et al. 2018). The total flow rate was adjusted to 1 L/min by the mass flow control corresponding to a GHSV (gas hourly space velocity) of 40,000 h^–1. A flue gas analyzer (KM9106, Kane International, Ltd., Welwyn Garden City, UK) was used to measure the inlet and outlet concentrations of NO _x . The concentration of gas was recorded once every 5 min during the reaction. The main photo-SCR reaction was

(1) $\mathrm{NO}+{\mathrm{NH}}_3+\frac{1}{4O_2}\to N_2+3/2H_{2}O$

The N₂ selectivity was calculated from

(2) $N_2\text{selectivity}=\left(1-\frac{2{\left(N_{2}O\right)}_{\text{outlet}}}{{\left({\mathrm{NO}}_x\right)}_{\text{inlet}}+{\left({\mathrm{NH}}_3\right)}_{\text{inlet}}-{\left({\mathrm{NO}}_x\right)}_{\text{outlet}}-{\left({\mathrm{NH}}_3\right)}_{\text{outlet}}}\right)\times 100\%$

Results and Discussions

XRD and BET Analysis

In the XRD patterns of pure Pal, LaCoO₃, and La_1–x Pr _x CoO₃/Pal (Fig. 1), the characteristic diffraction peak at 8.3°2θ corresponded to the (110) crystal plane of Pal. The intensity of this peak decreased gradually, indicating that rare-earth perovskite had been loaded successfully onto the surface of Pal (Takase et al. 2018). When x was 0, the diffraction peaks at 22.86, 32.51, 40.12, 46.70, 57.87, 68.08, and 77.39°2θ corresponded to the (012), (110), (014), (024), (214), (018), and (208) crystal planes of LaCoO₃, respectively (JCPDS Card, NO.48-123). The diffraction peaks at 23, 26, 33, 41, 48, 59, 70, and 80°2θ corresponded to the (200), (210), (220), (222), (400), (422), (440), and (620) crystal planes of PrCoO₃, respectively (JCPDS Card NO.25-1069). As the amount of Pr doping increased, the main peak (110) of perovskite was weakened gradually. When the doping amount was 0.5, PrCoO₃ was precipitated. With further increase in Pr doping, the characteristic peaks of PrCoO₃ at 32°2θ shifted progressively to a higher Bragg angle, which might be due to the fact that the Pr³⁺ radius is smaller than that of the La³⁺. Some Pr³⁺ ions entered into the lattice of LaCoO₃, leading to lattice distortion and, as a result, the tolerance factor value of the crystal structure decreased. Meanwhile, some PrCoO₃ might precipitate and accumulate on the surface of LaCoO₃. The main peak of PrCoO₃ coincided with that of partial Pal (the Pal part of PrCoO₃/Pal composites) which was not identified in the XRD spectrum. In addition, Pal had the largest specific surface area (S _BET = 127.74 m²/g) and the largest pore volume (V _t = 0.48 cm³/g) in all materials (La_1–x Pr _x CoO₃/Pal, x = ~0.1–0.9). After loading LaCoO₃ and La_0.5Pr_0.5CoO₃, the specific surface area (S _BET = 101.01 m²/g and 97.18 m²/g) and pore volume (V _t = 0.34 cm³/g and 0.21 cm³/g) of LaCoO₃/Pal and La_0.5Pr_0.5CoO₃/Pal decreased significantly. However, the pore size of LaCoO₃/Pal and La_0.5Pr_0.5CoO₃/Pal (d = 16.51 and 18.05 nm) was larger than that of Pal (d = 15.83 nm) because a large number of micropores and mesopores disappeared with the loading of LaCoO₃ and La_0.5Pr_0.5CoO₃. The pore size of the material generally appeared to increase. Therefore, in combination with the XRD results, La_1–x Pr _x CoO₃ was supported successfully on the surface of Pal.

Fig. 1 XRD patterns of LaCoO₃/Pal, PrCoO₃/Pal, and La_1–x Pr _x CoO₃/Pal (x = 0–1) Figure 2 blurry text inside the artwork.Dear editor, yes, it is.

SEM Analysis

A field emission scanning electron micrograph (FE-SEM) of the composite photocatalyst showed significant agglomeration at low resolution (Fig. 2a). Particles of various sizes and with many channels among the particles were observed and attributed to the accumulation of PrCoO₃ on the surface of Pal. The element mapping distribution of the La_1–x Pr _x CoO₃/Pal composites (Fig. 2b–2f) illustrated that the distribution of each element was relatively uniform, which was beneficial in that it improved the degradation of NO _x by the photo-SCR.

Fig. 2 SEM images and element distribution of La_0.5Pr_0.5CoO₃/Pal

TEM Analysis

Transmission electron microscopy images of La_1–x Pr _x CoO₃/Pal composites under various levels of doping (x = 0.1, 0.3, 0.5, and 0.7) depicted the process of formation of the PrCoO₃/La_1–x Pr _x CoO₃ heterostructure (Fig. 3). The diameter of Pal particles was 20–40 nm and the fiber length was ~800 nm (Fig. 3a,c). The nanoparticles of LaCoO₃ were scattered uniformly on the surface of Pal without significant agglomeration. The diameter of the LaCoO₃ nanoparticles was 10–25 nm (Fig. 3b,d). The lattice spacing of LaCoO₃ nanoparticles on the surface of Pal was 0.27 nm, corresponding to the (110) plane of LaCoO₃ (Wang et al. 2017). The nanoparticles of La_0.5Pr_0.5CoO₃ were distributed uniformly on the surface of Pal without obvious agglomeration and the average size of the particles was ~10 nm (Fig. 3e). The existence of heterojunctions in rare earth perovskite composites was clearly visible (Fig. 3f). The lattice spacing of the nanoparticles was 0.267 nm, corresponding to the (200) plane of PrCoO₃, which indicated that PrCoO₃ was precipitated. The particles were dispersed uniformly on the surface of Pal (Fig. 3g). A new phase of PrCoO₃ appeared, as evidenced by the (200) and (210) crystal planes (Fig. 3h), and was due to the conversion of the composite catalyst to pure PrCoO₃.

Fig. 3 TEM photomicrographs of La_1–x Pr _x CoO₃/Pal (x = 0.1, 0.3, 0.5, or 0.7)

FTIR Analysis

The FTIR spectra of LaCoO₃/Pal, PrCoO₃/Pal, and La_1–x Pr _x CoO₃/Pal (x = 0–1) showed that the absorption peaks around 787.14 and 3492.18 cm^–1 were the stretching vibrations of the coordination water from Pal (Fig. 4). The absorption peaks at ~1455.13 and ~1645.85 cm^–1 corresponded to the stretching vibration of Si-O-Al and zeolite water, respectively (Chen et al. 2018). The peaks at 1700 and 2800 cm^–1 in La_1–x Pr _x CoO₃/Pal represented the stretching vibrations of the La–O–Pr bond. Notably, the absorption peaks of La_1–x Pr _x CoO₃/Pal near 491.23, 1124.38, and 3492.18 cm^–1 became significantly weaker than those of pure Pal, indicating that La_1–x Pr _x CoO₃ was dispersed evenly on the Pal surface. The results above were consistent with the TEM results.

Fig. 4 FTIR spectra of La_1–x Pr _x CoO₃/Pal (x = 0–1)

NH₃-TPD Analysis

The temperature programmed adsorption (TPA) patterns of pure Pal were linear and independent of temperature (Fig. 5), indicating that the adsorption capacity of NH₃ gas for Pal was small compared with that of the La_1–x Pr _x CoO₃/Pal. However, weak peaks from LaCoO₃/Pal and PrCoO₃/Pal appeared at ~430°C, indicating that these two composites possessed minor acidic sites. With increased Pr doping, the intensity of the peak at ~430°C was enhanced gradually. Meanwhile, as the doping level of Pr increased, the peak shifted to higher temperatures and the peak intensity increased. When x was 0.5, the peak showed greatest intensity, indicating that La_0.5Pr_0.5CoO₃/Pal had the most acidic sites, which was beneficial to the SCR reaction. When the doping amount was >0.5, the peak intensity became weaker and the number of acid sites decreased.

Fig. 5 NH₃-TPD of La_1–x Pr _x CoO₃/Pal (x = 0.1, 0.3, 0.5, 0.7, or 0.9)

H₂-TPR Analysis

The H₂-TPR patterns of pure Pal and La_1–x Pr _x CoO₃/Pal were also acquired to evaluate the redox activity (Fig. 6). Peaks at 285 and 465°C represented chemisorbed oxygen and lattice oxygen of La_1–x Pr _x CoO₃, respectively. When x was 0.5, the two peaks of chemisorbed and lattice oxygen shifted to a lower temperature, indicating that La_0.5Pr_0.5CoO₃ was reduced more easily. The restoration process included Co³⁺→ Co²⁺ and Pr⁴⁺→ Pr³⁺ (Ayodele et al. 2017). According to the order and peak areas, the La_0.5Pr_0.5CoO₃ composite revealed the optimum redox activity compared with La_1–x Pr _x CoO₃/Pal composites (x = ~0.1–0.4, and ~0.6–0.9), which was consistent with the low-temperature photo-SCR activity.

Fig. 6 H₂-TPR of La_1–x Pr _x CoO₃/Pal (x = 0.1, 0.3, 0.5, 0.7, or 0.9)

XPS Analysis

XPS measurements of pure Pal and La_1–x Pr _x CoO₃/Pal were employed to obtain more information about elemental identification (Fig. 7). The full scan spectra of LaCoO₃/Pal and La_0.5Pr_0.5CoO₃/Pal confirmed the presence of Pr in the La_0.5Pr_0.5CoO₃/Pal (Fig. 7a). The high-resolution XPS spectra of La 3d, Co 2p, and Pr 3d binding energies (Fig. 7b, c, and d, respectively) revealed that the La 3d peak of La_0.5Pr_0.5CoO₃/Pal moved to higher binding energy compared with LaCoO₃/Pal (Fig. 7b), which might be due to the difference in ionic radii between Pr and La. The position of the Co 2p peak was virtually unchanged because the doping was in the A site but Co is in the B site (Fig. 7c). Pr 3d yielded two main peaks at binding energies of 935 eV and 955 eV (Fig. 7d), which was consistent with a previous study (Poggio-Fraccari et al. 2018).

Fig. 7 (a) XPS survey full scan spectra of LaCoO₃/Pal and La_0.5Pr_0.5CoO₃/Pal. XPS high-resolution scans of LaCoO₃/Pal and La_0.5Pr_0.5CoO₃/Pal binding energies of (b) La 3d, (c) Co 2p, and (d) Pr 3d

NO _x Conversion

Increasing the amount of Pr doping in the La_1–x Pr _x CoO₃/Pal photo-SCR nanocomposite greatly increased the denitration of NO_x (Fig. 8). The denitration activity of LaCoO₃/Pal at 100–250°C increased with increasing temperature. When x > 0.1, the denitration activity of La_1–x Pr _x CoO₃/Pal was greatly improved; in particular, the extent of NO _x elimination by La_0.5Pr_0.5CoO₃/Pal reached 95% at 200°C, a 20% increase over that of LaCoO₃/Pal. Some PrCoO₃ may have precipitated on the surface of perovskite to generate the heterojunction of PrCoO₃/La_1–x Pr _x CoO₃, contributing to separation of the photo-induced electron-hole pairs. The conversion of NO decreased with increasing doping levels of x > 0.5, and the conversion extent was less than with LaCoO₃/Pal. Under this circumstance, part of the Pr precipitated to form PrCoO₃ which adhered to the surface of rare-earth perovskite and Pal, and hindered the adsorption of NH₃ by the acidic sites of rare-earth perovskite, leading to a reduction of the conversion rate of NO, which was consistent with the results of XRD. From the perspective of denitration effect and economy, La_0.5Pr_0.5CoO₃/Pal composite had the best effect.

Fig. 8 Photo-SCR conversion of NO _x by LaCoO₃/Pal; PrCoO₃/Pal; and La_1–x Pr _x CoO₃/Pal (x = 0.1, 0.3, 0.5, 0.7, or 0.9)

N₂ Selectivity

The N₂ selectivity of LaCoO₃/Pal and PrCoO₃/Pal was ~60–70% (Fig. 9). When x > 0.1, the N₂ selectivity began to rise. When x = 0.5, the N₂ selectivity reached a maximum value of 99%, which was ascribed to the formation of PrCoO₃/La_1–x Pr _x CoO₃ heterojunction, accelerating the separation of photogenerated electron-hole pairs. However, when x > 0.5, the N₂ selectivity of La_1–x Pr _x CoO₃/Pal began to decrease, which was due to the excessive amount of PrCoO₃ precipitated on the surface of La_1–x Pr _x CoO₃/Pal, hindering further catalytic reaction.

Fig. 9 N₂ selectivity of LaCoO₃/Pal; PrCoO₃/Pal; and La_1–x Pr _x CoO₃/Pal (x = 0.1, 0.3, 0.5, 0.7, or 0.9)

DFT Calculations

DFT calculations of LaCoO₃, PrCoO₃, La_0.5Pr_0.5CoO₃, and the unit-cell diagram of La_0.5Pr_0.5CoO₃ (Fig. 10) revealed that the band gaps of LaCoO₃, PrCoO₃, and La_0.5Pr_0.5CoO₃ were 2.89, 2.94, and 3.07 eV, respectively. The Valence Band (VB) values were 2.59, 2.69, and 2.87 eV; and the Conduction Band (CB) values were –0.3, –0.25, and –0.2 eV. In addition, the occurrence of spin up and down should be considered when doing DFT calculations because LaCoO₃, PrCoO₃, and La_0.5Pr_0.5CoO₃ exhibited specific magnetic properties. From the view of the unit cell of La_0.5Pr_0.5CoO₃ (Fig. 10d), La_0.5Pr_0.5CoO₃ belongs to the orthorhombic system. The space group of La³⁺ and Pr³⁺ is Pnma, and the six oxygen atoms around Co³⁺ were arranged in the unit cell. The Co atoms at eight corners of the cuboid, and some La atoms were replaced by Pr atoms, demonstrating the Pr doping.

Fig. 10 DFT calculations in the presence of various scavengers for (a) LaCoO₃, (b) PrCoO₃, (c) La_0.5Pr_0.5CoO₃, and (d) the unit crystal cell of La_0.5Pr_0.5CoO₃. Atoms: green — La, blue — Pr, red — O, brown – Co

The Mechanism for the Photo-SCR of La_1-xPr_xCoO₃/Pal

On the basis of the above results, a mechanism for the photo-SCR of NO with NH₃ using La_1–x Pr _x CoO₃/Pal is proposed here (Fig. 11). The microspores of Pal had a physical-adsorption effect on gas enrichment (Yan et al. 2018). Under irradiation by visible light, La_1–x Pr _x CoO₃ absorbs light to generate electrons (e ^–) and holes (h ⁺). Simultaneously, the electrons shift from the VB to CB of La_1–x Pr _x CoO₃ and transferred to the surface of PrCoO₃, whereas the holes transferred from PrCoO₃ to La_1–x Pr _x CoO₃. The effective charge separation and the separated electrons, therefore, facilitated the reduction of NO on the Pal surface.

Fig. 11 Schematic illustration of photo-SCR mechanism for La_1–x Pr _x CoO₃/Pal

Conclusions

La_1–x Pr _x CoO₃/Pal nanocomposites with various levels of Pr doping were fabricated successfully by a facile sol-gel approach for the photo-SCR of NO _x with NH₃. Pal was an effective carrier with outstanding adsorption in the process of photo-SCR. Pr doping of LaCoO₃ produced an obvious increase in the NO conversion and on the N₂ selectivity. The performance of La_0.5Pr_0.5CoO₃/Pal as a catalyst was excellent, eliminating >95% of the NO _x over the low temperature range of 150–250°C. DFT calculations revealed that Pr incorporated into the LaCoO₃ lattice could modulate the band gap forming an intimate and staggered heterojunction of PrCoO₃/La_1–x Pr _x CoO₃, which enhanced the photo-absorption and facilitated the separation of electron-holes under visible light irradiation.