Introduction

Nitrogen oxides (NOx) are harmful gases that give rise to a variety of environmental problems, such as acid rain, photochemical smog, and ozone depletion, which threaten human health to a great extent1,2,3. In the pursuit of a better living environment and with increasingly stringent environmental regulations, NOx emissions have become a research hotspot in the field of environmental science and engineering. NOx storage and reduction (NSR) plays an important role in controlling NOx emissions from automobile sources while permitting operation under predominantly lean-burn conditions4,5,6. So far, the most efficient technique to control NOx emissions from coal-fired power plants and automobiles is the selective catalytic reduction (SCR) of NO (4NO + 4NH3 + O2 = 4N2 + 6H2O)7,8,9,10,11 and NO2 (NO + 2NH3 + NO2 = 2N2 + 3H2O) with NH3 using various catalysts12,13. For SCR of NO by NH3, it has been found that NO and NH3 are introduced into the reaction vessel at a ratio of 1:1, and a certain amount of O2 is essential for the reaction. However, the excessive use of NH3 can cause air pollution and the corrosive nature of NH3 is also harmful to the experimental apparatus, which may cause secondary pollution. Therefore, it is necessary to reduce the use of NH3 for the elimination of NO. Also, the removal of NOx is generally carried out at high temperatures (>600 °C), where many catalysts may lose their activity14,15,16,17,18, leading to reduced elimination efficiency of NOx. Within this context, it is extremely urgent to find more suitable active species or to develop an efficient, green approach to the elimination of NOx at high temperatures.

Transition metal phosphides (e.g., ZrP, FeP, Ni2P, etc.) are known for their hardness and chemical inertness even at elevated temperatures19,20,21. In the past few years, substantial efforts have been devoted to the study of the catalytic properties of transition metal carbides and nitrides22,23. Notably, transition metal phosphides possess similar physical properties and even better catalytic activity and selectivity in many reactions such as hydrogen evolution reaction (HER) and oxygen evolution reaction (OER), and have become a new research focus in the field of catalytic materials and water oxidation catalysis. For instance, Ni2P has been used as an excellent catalyst precursor for water oxidation catalysts24, and FeP as an efficient catalyst for HER25,26. Also, there are many studies on the catalytic performance of zirconium phosphate (ZrP2O7)27,28,29,30,31. In contrast, whereas ZrP has been synthesized by a variety of methods in the past decades, such as the reactions of zirconium metal with red phosphorus (elemental combination method)32,33, zirconium or zirconium tetrachloride with Ca3P2 20, zirconium with PH3 33, or sodium co-reduction of ZrCl4 and PCl3 34, there are few reports on the chemical properties of ZrP and, up to date, there is no study of the activity of ZrP towards NO elimination.

In this study, ZrP was synthesized by elemental combination method via the direct reaction of zirconium powders with red phosphorus in a quartz tube. Since the phosphorus atoms in the ZrP are negatively charged (P3−) and the nitrogen atoms in the NO are positively charged (N2+), ZrP might be used to reduce NO by the reaction, ZrP + 2NO → ZrO2 + N2 + P. The results showed that the efficiency of NO elimination by ZrP increased with increasing temperature. In addition, it was found that when a small amount of NH3 was added to the reaction system, the following reaction could occur, 8ZrP + 22NO + 4NH3 → 8ZrO2 + 13N2 + 8P + 6H2O, leading to enhanced efficiency of NO elimination. The experimental results showed that at the ratio of the NH3 to NO concentration of 2:11, 0.5 g of ZrP powders was sufficient for the complete reduction of 500 ppm NO gas for up to 14 h at 750 °C. As for the products of the elimination reaction, ZrO2 powders can be used as a biocompatible material35,36 and catalysts/catalyst support, red phosphorus can be used to produce safety matches and organophosphorous pesticide, and the produced N2 may be collected and used as a protective gas or be converted into liquid nitrogen for other purposes. In this study, we not only identified an active species for NO elimination, but also reduced the amount of NH3 for NO reduction. The elimination process was of high efficiency and all reaction products could be used for other purposes.

Results

Structures of ZrP

The structures of the obtained ZrP were first characterized by X-ray diffraction (XRD) measurements. As shown in the Fig. 1a, XRD measurements of the samples prepared at 800–1000 °C all show a series of distinct diffraction peaks at 27.86°, 28.78°, 31.36°, 35.55°, 40.49°, 46.21°, 49.58°, 58.24°, 59.26° and 62.55°, which can be indexed to the ZrP (100), (101), (102), (103), (104), (105), (110), (201), (107) and (203) lattice planes of the hexagonal phase of ZrP (space group P63/mmc, No. 194). The observed unit cell parameters are a = b = 3.6840 Å, c = 12.5540 Å and α = β = 90°, γ = 49.58°, in good agreement with data in JCPDS No. 016–0034, with the crystal structure of ZrP shown in the Fig. 1b. In addition, from Fig. 1a, it can seen that at higher temperatures, the diffraction peaks become sharper, suggesting increasing crystallinity of the obtained ZrP samples.

Figure 1
figure 1

XRD patterns and Crystal structure of the ZrP samples. (a) XRD patterns of the ZrP prepared at different temperatures (b) Crystal structure of ZrP. The red and green balls represent phosphorus and zirconium atoms, respectively.

SEM characterization

The microstructures of ZrP were then investigated by SEM measurements. From the SEM images (Fig. 2a,b), it can be seen that ZrP was an irregular bulk material.

Figure 2
figure 2

SEM and TEM analyses of the ZrP sample. (a) and (b) SEM images of the ZrP sample (c) TEM images of the as-prepared ZrP (d) HRTEM images of the as-prepared ZrP. Inset to (d) is the electron-diffraction patterns.

TEM characterization

The microstructures of ZrP were also investigated by TEM measurements (Fig. 2c). From the HRTEM image (Fig. 2d), several lattice fringes of 3.2 Å and 2.5 Å can be observed clearly, in good agreement with the crystal planes (100) and (103) of ZrP. In addition, the corresponding SAED patterns (inset to Fig. 2d) can also be assigned to single-crystalline ZrP42,43,44.

XPS characterization

In view of the same diffractions of XRD for ZrP samples prepared at different temperatures (Fig. 1a), only the typical ZrP samples synthesized at 800 °C was selected for XPS, XRF characterization. XPS analyses of the ZrP samples are shown in Fig. 3. The survey spectrum (Fig. 3a) shows that the sample surface consists of zirconium and phosphorus, and based on the integrated peak areas, the mole ratio of Zr/P is estimated to be 1.3:1 (Fig. 3b,c), which is close to that of ZrP. The oxygen may come from surface adsorption. Carbon is also found, which may be from carbon dioxide adsorption.

Figure 3
figure 3

XPS analyses of the ZrP sample. (a) survey spectrum (b) Zr3d region (c) P2p region.

XRF characterization

XRF analysis was then carried out to further analyze the composition of the ZrP samples, and the results are summarized in Table 1. It can be seen that for the as-prepared sample, the mole ratio of Zr/P is estimated to be 1.4:1, which is close to 1:1 (when oxygen adsorption on the sample surface was excluded). Yet, after reacting with NO at 700 °C and 750 °C, the relative content of phosphorus in the reaction products diminished with the calculated mole ratio of Zr/P increased from 2.1:1 to 3.6:1. This is consistent with the results shown in Supporting Figure S1.

Table 1 Summary of XRF analyses.

Elimination of NO

Figure 4 showed the reaction of ZrP with NO at different temperatures. Specifically, Fig. 4a shows that when the reaction temperature between ZrP and NO was controlled at 650 °C, it was worth noticing that ZrP samples (prepared at 800 °C, 900 °C, 1000 °C, respectively) would eliminate almost the same amount of NO from three almost identical curves, which indicated that the temperature of synthesis for ZrP had little influence on the elimination of NO at 650 °C. As shown in Fig. 4b–d, when reaction temperatures between ZrP and NO were increased to 700 °C, 750 °C and 800 °C, the elimination amount of NO over ZrP prepared at different temperatures still had no obviously difference. So ZrP samples synthesized at 800 °C was typically selected to the next study of elimination for NO. Figure 5a shows the effect of reaction temperature for the elimination of NO in the temperature range from 650 °C to 800 °C. In the course of experiment, a mixture of gases containing 500 ppm NO balanced with N2 was introduced into the reactor at the 174th min. It can be seen that at the controlled temperature of 650 °C, after the introduction of NO, the concentration of NO at the outlet increased rapidly and equal to the concentration of NO in the inlet, which means that there was no reaction between NO and ZrP. Yet, when the temperature was raised to 800 °C, no signal of NO was detected at the outlet for 1080 min (18 h), indicating that NO was completely reduced by ZrP in the reactor within this period. At lower temperatures (700 and 750 °C), the period of time where no NO was detected was shorter, suggesting that temperature was an important factor for the elimination of NO. Such a disparity of the reaction activity can also be manifested in XRD measurements of the solids after reaction. From Supporting Figure S1, we can see that when the reaction temperature was controlled at 650 °C, the diffraction patterns of the solids after reaction were consistent with those of the ZrP, as there was no reaction with NO at this temperature. When the reaction temperature was raised to 700 °C, the solids were found to consist of a mixture of ZrP and ZrO2, indicating that part of the ZrP was oxidized into ZrO2 due to reduction of NO to N2. At the even higher temperature of 800 °C, no diffraction pattern of ZrP was detected and only ZrO2 diffraction patterns were observed, most likely due to complete consumption of ZrP in the reduction of NO, (ZrP + 2NO = ZrO2 + N2 + P).

Figure 4
figure 4

Reductions of NO at 650 °C (a), 700 °C (b), 750 °C (c) and 800 °C (d) (three curves in each Figure represented the ZrP samples synthesized at different temperatures, i.e. 800, 900 and 1000 °C, respectively).

Figure 5
figure 5

Reaction of NO with ZrP at different temperatures (a), Effect of NH3 concentration on the elimination of NO (b).

The tests were also carried out to elucidate the influence of NH3 addition in NO elimination. As depicted in Fig. 5b, prior to the addition of NH3, 0.5 g of ZrP powders can reduce 500 ppm NO gas for 5 h at 750 °C. With the addition of NH3 into the reactor where the molar ratio of NH3:NO increased from 1:3 to 2:3, the time for the complete conversion of NO into N2 was increased from 6 h to 8 h. Yet, too much NH3 is not conductive to the elimination of NO. For instance at the mole ratio of NH3:NO = 3:3, the time for the complete conversion of NO by ZrP actually decreased to only 3 h, even shorter than that without any ammonia addition at the same condition (~5 h). The optimal NH3:NO ratio was identified at 2:11, where 0.5 g of ZrP powders were found to reduce 500 ppm NO gas for up to 14 h at 750 °C. This may be attributed to the following reaction, 8ZrP + 22NO + 4NH3 → 8ZrO2 + 13N2 + 8 P + 6H2O. From this reaction, we can see that at a fixed amount of ZrP, the addition of a small amount of NH3 could facilitate the reduction of NO. In fact, this elimination process not only reduces the use of NH3, but also greatly improves the elimination efficiency of NO as compared to SCR of NO by NH3.

Verification of red phosphorus

After elimination reactions of ZrP with NO, in addition to the product of ZrO2, some red brown powders were found adhering to the inner wall at the end of the quartz tube (Supporting Figure S2a), which were collected, dried, and subject to a combustion test. The results showed that the powders burned violently, accompanied by the generation of white smoke (Supporting Figure S2b) and irritating smell, forming a white solid product. These observations were consistent with the combustion of red phosphorus, suggesting red phosphorus as the part of the elimination reaction of NO by ZrP.

Discussion

ZrP powders were prepared by thermal treatment of zirconium and red phosphorus in an argon atmosphere at controlled temperatures, which exhibited apparent activity in the reductive elimination of NO to N2. It was found that the reaction temperature and concentration of NH3 were important factors that affected the elimination efficiency of NO by ZrP. The products of the elimination process included ZrO2, N2, H2O (if NH3 was added) and red phosphorus. In summary, a new method based on ZrP was developed for the reductive elimination of NO, where the reaction products might be collected and used for other applications.

Methods

Material Preparation

All chemicals were purchased from Alfa Aesar and used as received without further purification. To prepare ZrP, zirconium powders were placed in a quartz boat at one end of a quartz tube and the required quantity of 99.99% pure red phosphorus powders were placed at the other end under an atmosphere of purified argon. The temperature was slowly raised (2 °C/min) from room temperature to 800 °C~1000 °C and kept for 6 h37,38.

Material Characterizations

The phase structures and morphologies of the ZrP powders were characterized by X-ray powder diffraction (XRD, Bruker D8 with Cu Kα radiation, λ = 1.54 Å), Scanning electron microscope (SEM, HITACHI, S4800, 15 kV), X-ray photoelectron spectroscopy (XPS, ESCALAB 250 with Al Kα radiation) and X-ray fluorescence (Shimadzu, XRF-1800) measurements. The lattice fringes of the obtained samples and the corresponding selected-area electron diffraction (SAED) patterns were examined by using a high-resolution transmission electron microscope (HRTEM JEOL 2010F, 200 kV).

Elimination of NO

The elimination of NO was carried out in a fixed-bed quartz tube reactor with an internal diameter of 6 mm39. 0.5 g of ZrP powders were sieved with a 40–60 mesh and placed on the quartz wool held in the reactor, and the reactor was heated by a vertical electrical furnace. The total flow rate was 198 mL·min−1 (room temperature), the mass of ZrP was 500 mg, and the corresponding gas hourly space velocity (GHSV) was 6 × 104 cm3·g−1·h−1, which was evaluated by the equation (1):

$$GHSV=\frac{{q}_{V}}{\pi {{\rm{hr}}}^{{\rm{2}}}},$$
(1)

where qv is the total flow rate, h is the height of the reactant in the reactor and r is the radius of the reactor40. The feed contained 500 ppm of NO, 500 ppm of NH3 (when used), and balance of N2. The concentration of NO was continuously detected by a gas chromatographic analyzer equipped with a flame photometric detector (Beijing Beifen-Ruili 3420 A). The NO conversion was calculated according to the equation (2):

$$N{O}_{conversion}=\frac{{[NO]}_{in}-{[NO]}_{{\rm{out}}}}{{[NO]}_{in}}\times 100 \% ,$$
(2)

where [NO]in indicates the inlet NO concentration at steady state and [NO]out indicates the sum concentration of outlet NO41.

Data Availability

All data generated or analysed during this study are included in this article (and its Supplementary Information files).