Removal of NO_x and PM from Non-Road Mobile Machinery by the Combination of Ozone Oxidation and Venturi Scrubbing

Abstract

Non-road mobile machinery (NRMM) emits a significant amount of NO_x and particulate matter (PM), yet there is still a lack of economically feasible purification technologies up to now. In response to this issue, and taking into account the relatively flexible installation conditions for exhaust purification systems in NRMM, this study proposes a novel technical approach for the simultaneous removal of NO_x and PM based on gas-phase O₃ oxidation combined with Venturi scrubbing. The effects of O₃ oxidation on the PM physicochemical properties, O₃ concentration, liquid-to-gas (L/G) ratio, and surfactant addition in the scrubbing liquid on the PM removal and simultaneous removal of PM and NO_x were investigated. The results showed that O₃ oxidation significantly promoted PM removal in the absence of NO, which was attributed to the increase in the hydrophilicity of PM resulting from O₃ oxidation. However, O₃ preferentially reacted with NO, thereby reducing the removal efficiency of PM under the conditions when PM and NO coexisted. Adding surfactants to the scrubbing liquid improved PM removal and increasing the liquid-to-gas ratio improved the removal of both PM and NO_x. When both the O₃/NO molar ratio and liquid-to-gas ratio were 1.5, the removal efficiencies of NO_x and PM reached 87% and 92%, respectively, and O₃ escape was also effectively controlled. These findings demonstrated that the combination of gas-phase O₃ oxidation and Venturi scrubbing is a promising purification technology for NRMM exhausts.

1. Introduction

NRMM typically uses diesel and heavy oil as fuel and emits various air pollutants, especially NOx and particulate matter (PM). According to the “2024 China Mobile Source Environmental Management Annual Report” released by the Ministry of Ecology and Environment, China [1], NRMM emitted 4.532 million tons of NOx and 224,000 tons of PM in 2023. NOx not only poses a direct health threat but also acts as one of the precursors of atmospheric haze and ozone (O₃) pollution. Although PM emissions from NRMM are not large relative to national total emissions, these emissions occur near the breathing zone and mainly consist of nanoscale soot, posing significant health and environmental risks. With the strengthening of atmospheric pollution control requirements, controlling NOx and PM emissions from NRMM has become one of the key tasks in China’s air pollution control efforts [2].

Although significant progress in controlling pollutant formation has been achieved through fuel optimization and in-cylinder combustion technology, these measures still remain insufficient to meet stringent emission standards. Therefore, exhaust post-treatment has become a rigid requirement. Diesel Oxidation Catalyst (DOC) + Diesel Particulate Filter (DPF) + Selective Catalytic Reduction (SCR) + Ammonia Slip Catalyst (ASC) has been widely applied as a mature exhaust gas purification technology for road diesel vehicles [3,4]. However, the fuel quality and engine technology of NRMM are generally inferior to those of road diesel vehicles. Moreover, NRMM operation modes change more frequently and significantly, with especially more extended periods of operation at low temperatures, which results in a wider fluctuation in NOx emissions and higher PM concentrations.

To adapt the NH3-SCR de-NOx technology to the wide exhaust temperature range of NRMM, extensive studies have been conducted on optimizing catalyst formulations [5,6,7,8,9]. However, tests have shown that the performance of NH3-SCR de-NOx systems for NRMM is still unsatisfactory, especially during cold starts and idle periods [10,11,12]. Although the filtration efficiency of DPF for PM from road diesel vehicle exhaust is generally above 90% [13], the frequent combustion regeneration of DPF leads to the decrease in both fuel economy and PM filtration efficiency during regeneration periods when using DPF to filter the high-concentration PM from NRMM exhausts [14,15,16]. For this reason, many studies have been conducted on particle catalytic oxidation (POC) for PM purification; the results show that the efficiency of POC in PM purification is generally less than 70% due to the limited solid–solid contact conditions between PM and catalyst [17,18,19,20]. Additionally, studies using conventional water scrubbing technology to purify PM from diesel engine exhaust have shown that the purification efficiency is only around 40% [21,22,23]. Recently, Kawakami et al. conducted an exploratory study on the removal of NOx and PM from diesel engine exhaust using non-thermal plasma. Their results showed that the removal efficiencies for PM and NOx could reach up to 98% and 70%, respectively, with NOx being converted into HNO₃. However, this process required a high energy input of 2400 J/L [24].

Considering the relatively flexible installation conditions for an NRMM exhaust purification system, this study puts forward a technical approach based on the combination of gas-phase O₃ oxidation and Venturi scrubbing for purifying both NOx and PM. The reliability of the efficient purification of NOx from industrial flue gas by the combination of O₃ oxidation and absorption has been verified through both experimental research and engineering practice. As long as the molar ratio of O₃/NO is larger than 1.5, NOx can be oxidized into higher-valence nitric oxides such as N₂O₅ and NO₃, which have good hydrophilicity and can be effectively removed by absorption [25,26,27,28,29,30,31]. Besides O₃ oxidation, the synergistic oxidation of vacuum ultraviolet (VUV)/Oxone/hydrogen peroxide (H₂O₂) coupled with absorption has also been studied and achieved a relatively good denitrification efficiency. However, overall, its comprehensive performance is inferior to that of ozone oxidation [32]. Because NOx concentration level and other characteristics of NRMM exhaust are similar to those of the industrial flue gas, it can be deduced that this technology is also feasible for the purification of NOx from NRMM exhaust. On the other hand, some studies have been conducted on the effect of O₃ oxidation on the physicochemical and toxicological properties of soot (or black carbon). The results show that the physical properties such as the particle size and soot morphology will change to varying extents after O₃ treatment [33,34]. Theoretically, O₃ oxidation can be utilized to modify the PM of NRMM exhaust, especially to enhance the hydrophilicity and thus promote the growth of PM particle size during Venturi scrubbing, thereby improving the separation effect.

However, the removal of both NOx and PM based on the combination of gas-phase O₃ oxidation and Venturi scrubbing remains unexplored up to now. The stringent regulations demanding deep reductions in NRMM emissions of NOx and PM pose a significant challenge, which makes the existing research gap particularly pressing. This study is designed precisely to address this shortcoming by examining the impact of O₃ oxidation on the physicochemical properties of PM and the efficacy of oxidation-coordinated scrubbing in achieving simultaneous NOx and PM removal.

2. Experimental Methodologies

2.1. Experimental System

A schematic diagram of the experimental setup is shown in Figure 1. The system comprises four units: the simulated NRMM exhaust gas unit, the O₃ generation and oxidation unit, the Venturi scrubbing and gas–liquid separation unit, and the detection unit. NRMM exhaust was simulated by mixing air with NO from a gas cylinder and flue gas containing a high concentration of black smoke (PM), produced from methane combustion under oxygen-deficient conditions. The ozone oxidation combined with a Venturi scrubbing system was positioned downstream of the Diesel Oxidation Catalyst (DOC). Therefore, it was not necessary to consider HC and CO in the simulated gas. Furthermore, it has been extensively reported that SO₂ is not preferentially oxidized by ozone over NOx and PM. Additionally, since diesel engines typically operate under excess air conditions, the use of air containing NOx and PM to simulate diesel exhaust is acceptable. The O₃ generator produced O₃ by high-voltage discharge in the O₂ atmosphere. The oxidation reaction tank is a stainless-steel cylinder with a diameter of 50 mm and a length of 80 mm. The Venturi scrubber, with a total length of 117 mm, a throat diameter of 3 mm, and a throat length of 10 mm, and a cyclonic liquid–gas separator were serially connected and fixed on an iron stand.

2.2. Experimental Methods

O₃ and the simulated exhaust gas were thoroughly mixed and then reacted in the oxidation tank. The gas after oxidation entered the Venturi scrubber, where the scrubbing liquid was atomized and accelerated by a high-speed gas flow at the throat. Theoretically, higher-valence NOx such as N₂O₅ and NO₃ are effectively absorbed by aqueous solution. Simultaneously, moisture condenses and adheres to the surface of the oxidized PM, which makes PM separated easier in the liquid–gas separator. A peristaltic pump was employed to circulate the scrubbing liquid. The purified exhaust gas entered the detection unit, where the O₃ concentration was analyzed using an ozone detector (Model 106-M, 2B Technologies, Inc., Broomfield, CO, USA), the NO and NO₂ concentrations were analyzed by a UV flue gas analyzer (MODEL 3080UV, Beijing SDL Technology Co., Ltd., Beijing, China), and the PM number concentration and number size distribution were detected by an ELPI^+TMparticle analyzer (Dekati, Kangasala, Finland). To ensure measurement accuracy, some necessary calibration of the relevant instruments was performed before the experiments commenced. All subsequent detection and measurement processes were carried out in strict compliance with the standardized operating procedures. Unless otherwise specified, the investigation was conducted under the conditions of an oxidation time of 0.3 s, a reaction temperature of 25 °C, an O₃ concentration of 300 ppm, a NO concentration of 200 ppm, a PM concentration of 3.65 (±0.25) × 10⁵/cm³, a liquid–gas ratio of 1.0 L/m³, and a total gas flow rate of 30 L/min.

NO_x removal efficiency is as follows:

$$\varphi ={\displaystyle \frac{C_{\mathrm{N}\mathrm{O},\mathrm{i}\mathrm{n}}-{{(C}_{\mathrm{N}\mathrm{O},\mathrm{o}\mathrm{u}\mathrm{t}}+C}_{\mathrm{N}{\mathrm{O}}_2,\mathrm{o}\mathrm{u}\mathrm{t}})}{C_{\mathrm{N}\mathrm{O},\mathrm{i}\mathrm{n}}}}\times 100$$

(1)

$\varphi$ is the NO_x removal efficiency, %; $C_{\mathrm{N}\mathrm{O},\mathrm{i}\mathrm{n}}$ and $C_{\mathrm{N}\mathrm{O},\mathrm{o}\mathrm{u}\mathrm{t}}$ are the NO concentrations before and after the oxidation–scrubbing–separation, ppm; and $C_{\mathrm{N}{\mathrm{O}}_2,\mathrm{o}\mathrm{u}\mathrm{t}}$ is the NO₂ concentration after the oxidation–scrubbing–separation, ppm.

PM removal efficiency is as follows:

$$\eta ={\displaystyle \frac{N_{\mathrm{P}\mathrm{M},\mathrm{i}\mathrm{n}}-N_{\mathrm{P}\mathrm{M},\mathrm{o}\mathrm{u}\mathrm{t}}}{N_{\mathrm{P}\mathrm{M},\mathrm{i}\mathrm{n}}}}\times 100$$

(2)

$\eta$ is the PM removal efficiency, %; N_PM,in and N_PM,out are the PM number concentrations before and after the oxidation–scrubbing–separation, particles/cm³, respectively.

In order to investigate the effects of O₃ oxidation on the physicochemical properties of PM, PM produced from methane combustion under oxygen-deficient conditions was collected and dried at 70 °C for 6 h. Then, PM oxidation was conducted under the conditions of O₃ concentration of 450 ppm. Subsequently, the number concentration and number size distribution, functional groups, electronegativity, and water contact angle of PM before and after oxidation were analyzed utilizing an ELPI^+TM particle analyzer (Dekati, Finlant), a FT-IR spectrometer (IR Tracer 100, Shimadzu, Kyoto, Japan), a Zeta potential meter (Zetasizer Nano, Malvern Panalytical, UK), and a contact angle meter (LSA100, Lauda, Germany).

3. Results and Discussion

3.1. Effects of O₃ Oxidation on Physicochemical Properties of PM

Figure 2 shows the number size distribution of PM before and after O₃ oxidation for 30 min. It can be seen that the O₃ oxidation resulted in a decrease in PM percentage with a large diameter (≥0.05 µm) and an increase in PM percentage with a small diameter (≤0.04 µm), which is similar to Liu’s research results [33]. This is because PM is not composed of dense, spherical particles. Instead, it consists of aggregates formed from thousands of nano-sized primary soot particles (approximately 20–50 nm in diameter) bound together by van der Waals forces and covalent bonds, creating complex, fluffy chain-like or cluster-like structures. The aerodynamic diameter of these aggregates can range up to several hundred nanometers or even micrometers. Ozone attacks the bridging sites that connect these primary particles. These junctions are typically defects or active sites within the carbon network, which are more vulnerable to oxidation than the crystalline structure itself. When these critical bridging points are oxidized and broken, the large aggregates disintegrate and fragment into several smaller aggregate fragments, or even individual primary particles. Consequently, the particle size distribution shifts, showing a decrease in the concentration of particles in the larger size ranges and a significant increase in the number of smaller particles.

The surface charge characterization indicated that after PM was oxidized for 30 min, the Zeta potential increased from −8.1 mV to −14.7 mV, which suggested that following O₃ oxidation, the acidic oxygen-containing functional groups formed on the PM surface and underwent weak acidic dissociation in suspension solution, thereby augmenting the negative charge density of the suspension solution.

Furthermore, the PM undergoing oxidation for 30 min was characterized by FTIR. As shown in Figure 3, the FTIR spectra of fresh and oxidized PM exhibited identical peak positions. However, the disparities existed in peak intensities. After oxidation, the absorption peaks corresponding to C=C stretching vibrations in aromatic rings or condensed rings at 1578 cm⁻¹, alkenes and alkynes at 890–1092 cm⁻¹, trisubstituted C-H aryl at 755 cm⁻¹, and tetrasubstituted C-H aryl at 843 cm⁻¹ weakened, whereas the absorption peak corresponding to C=O stretching vibrations in aldehydes, ketones, and carboxylic acids at 1629–1704 cm⁻¹ intensified. It was speculated that after O₃ oxidation, the condensation degree of the aromatic structure diminished. The aromatic rings and/or side chains underwent restructuring during oxidation, with a remarkable increase in highly active oxygen-containing functional groups. Unsaturated bonds were oxidized to aldehydes, ketones, and carboxylic acids. These outcomes indicated that ozone oxidation reduced the aromatic structure on the PM surface and significantly augmented the surface oxygen-containing functional groups, which corroborated the results of Li et al. [34].

The water contact angle (WCA) of PM before and after O₃ oxidation at different times was detected. As shown in Figure 4, the WCA of fresh PM was 132°, manifesting its strong hydrophobicity. After O₃ oxidation for 1 min, 5 min, and 15 min, the WCA decreased to 121°, 116°, and 64°, respectively, signifying that the PM transformed from hydrophobic to hydrophilic, which is consistent with the results reported by Liu et al. [33].

3.2. Removal of PM by the Combination of Ozone Oxidation and Scrubbing

3.2.1. The Influence of O₃ Concentration on PM Removal

Figure 5 shows the effect of O₃ concentration on PM removal by the combination of O₃ oxidation and scrubbing using water as the scrubbing liquid. It can be seen from Figure 5a that the total removal efficiency of PM gradually rose with the increase of O₃ concentration. When the O₃ concentration increased from 0 to 450 ppm, the PM removal efficiency increased from 54% to 80%. Figure 5b shows that the removal efficiency of PM in each particle size segment also increased with the increase of O₃ concentration and the removal efficiency of PM with a small size was higher than that of PM with a large size. The maximum and minimum PM removal efficiencies were obtained for 0.02 μm and 0.32 μm, respectively. The former was attributed to the most substantial diffusion effect, the latter was due to the relatively weak inertial collision, interception, and diffusion effects. On the other hand, the smaller the particle size, the higher the surface energy, and thus the stronger the hygroscopicity. Therefore, it was deduced that the increases in PM percentage with a small particle size (Figure 2), oxygen-containing functional groups (Figure 3), and hydrophilicity (Figure 4) after O₃ oxidation facilitated the wetting, water adhesion, and growth of PM in the scrubber, and thus promoted its removal.

3.2.2. The Influence of Scrubbing Conditions on PM Removal

(i)

The influence of liquid-to-gas ratio

The liquid-to-gas ratio (L/G ratio) is one of the key parameters affecting the purification performance of the Venturi scrubber. In this study, water was used as the scrubbing liquid, and the influence of the L/G ratio on PM removal was investigated. As shown in Figure 6a, with the increase in the L/G ratio, the total removal efficiency of PM significantly enhanced whether O₃ was added or not. This was attributed to the fact that when more liquid was supplied, more mist particles could be formed, thereby increasing the gas–liquid contact area and contact probability, and subsequently promoting the growth of particles. It can also be seen in Figure 6a that the promotion effect of O₃ oxidation weakened at higher L/G ratios, suggesting the dominating role of liquid supply. As shown in Figure 6b, the removal efficiencies of PM with different sizes all conspicuously increased with the increase in the L/G ratio.

(ii)

The influence of adding surfactants

Theoretically, surfactants can reduce the surface tension of scrubbing liquid, thereby enhancing the wetting properties and subsequently improving PM removal performance. In this study, the influence of surfactant addition on PM removal was investigated by adding three representative surfactants, namely LAS, CTAB, and AEO-9 (mass concentration of 0.25%). The results demonstrated that the addition of surfactants increased the PM removal efficiency. In particular, adding LAS made the PM removal efficiency increase from 56.3% to 73.3%. This improvement may be attributed to a large π bond contained in the LAS molecular structure, which exhibits a strong affinity with the hydrophobic groups of PM. Moreover, the electrostatic repulsion between PM particles reduced and the wettability of mist particles enhanced, and the formed micelles encapsulated PM particles, thereby facilitating PM growth.

3.3. Simultaneous Removal of NO_x and PM by the Combination of O₃ Oxidation and Scrubbing

3.3.1. The Influence of O₃ Concentration on Simultaneous Removal of NO_x and PM

The effects of O₃ concentration on the simultaneous removal of NO_x and PM were investigated using a 0.1% (w/w) Na₂CO₃ aqueous solution as the scrubbing liquid. As shown in Figure 7a, the NO_x removal efficiency initially increased linearly with the increase in the O₃ concentration. When the O₃ concentration reached 360 ppm, corresponding to an O₃/NO molar ratio of 1.8, NO_x removal efficiency reached 100%. Clearly, NO was oxidized to higher-valence nitric oxides with good water solubility, promoting the absorption removal of NO_x. On the other hand, with the increase of O₃ concentration, the increase in the PM removal efficiency was not as significant as that observed without NO in the simulated exhaust gas (Figure 5a). Cleary, a competitive relationship exists between O₃ reactions with NO and PM. O₃ preferentially oxidizes NO to form highly soluble, higher-valent nitrogen oxides. While this reaction significantly enhances NO_x removal in the subsequent scrubbing process, it also reduces the amount of O₃ available for direct PM oxidation, leading to a lower PM removal efficiency compared with conditions without NO in the simulated gas. Nevertheless, at higher O₃ concentrations, the residual O₃ and the generated higher-valent nitrogen oxides can still modify the PM surface, enhancing its hydrophilicity and thereby improving its wettability and removal during washing to a certain extent.

It can be seen from Figure 7b that the NO₂ concentration gradually decreased when the O₃ concentration increased from 180 ppm to 400 ppm, which indicated that NO₂ was the main oxidation product at the low O₃ concentration (O₃/NO ratio ≤ 1), the absorption of which was limited. At higher O₃ concentrations, NO₂ was further oxidized to nitric oxides such as N₂O₅ and NO₃, which can be easily absorbed. Figure 7b also showed that the O₃ concentration at the outlet of the reaction system increased with the concentration of O₃ at the inlet when it reached 300 ppm and above (O₃/NO ratio ≥ 1.5). Therefore, further research on optimizing the technical pathway and process parameters to prevent ozone escape is necessary.

3.3.2. The Influence of Scrubbing Conditions on Simultaneous Removal of NO_x and PM

(i)

The influence of scrubbing liquid compositions

Figure 8 shows the effects of the scrubbing liquid compositions on the removal of NO_x and PM. It can be seen that the removal efficiency of NO_x was relatively high (around 87%) when 0.1% (w/w) Na₂CO₃ was added into the scrubbing liquid, no matter whether 0.25% (w/w) LAS was added or not. On the other hand, the removal efficiency of PM was relatively high when 0.25% (w/w) LAS was added into the scrubbing liquid, especially when 0.1% (w/w) Na₂CO₃ was also added. These results suggest that the basicity of Na₂CO₃ is important for the absorption of NO_x while LAS, as a typical surfactant, mainly enhances the removal of PM. As expected, the highest removal efficiencies of NO_x and PM were obtained when both 0.1% (w/w) Na₂CO₃ and 0.25% (w/w) LAS were added into the scrubbing liquid (Figure 8).

(ii)

The influence of liquid-to-gas ratio

The effects of the L/G ratio on the removal efficiencies of NO_x and PM, as well as the escape concentration of O₃ were investigated using an aqueous solution containing 0.1% Na₂CO₃ and 0.25% LAS as the scrubbing liquid. As shown in Figure 9, the L/G ratio exerted a negligible influence on NO_x removal. However, the PM removal efficiency significantly increased as the L/G ratio rose. When the L/G ratio was 1.5 L/min, the removal efficiencies of PM and NO_x reached 92% and 87%, respectively. Notably, as the L/G ratio increased, the O₃ concentration at the outlet gradually diminished, indicating that the presence of more mist particles also alleviated O₃ escape.

3.4. Practical Challenges and Future Perspectives

While laboratory-scale results demonstrate the promising potential of O₃ oxidation-coordinated scrubbing technology for the efficient synergistic removal of NOx and PM, several challenges must be addressed for its practical engineering application. Firstly, the on-site generation of ozone involves certain energy consumption and equipment costs. However, the energy efficiency of ozone generators continues to improve with advancements in technology. More importantly, as this technology enables the synergistic control of multiple pollutants within a single system, it holds the potential to replace or simplify traditional multi-stage treatment devices, thereby offering cost competitiveness at the system level.

Secondly, safety concerns regarding ozone and system maintenance require careful consideration. The deployment of ozone leakage monitoring and interlock protection systems can effectively mitigate safety risks. Issues such as reactor clogging or corrosion can be addressed through modular design and regular predictive maintenance.

Future research will focus on developing intelligent ozone dosing control algorithms linked to engine operating conditions, aiming to optimize the balance between energy consumption and removal efficiency. Additionally, pilot-scale durability testing will be conducted to validate the long-term operational stability and economic feasibility of this technology under real-world conditions.

4. Conclusions

This study demonstrates that an integrated O₃ oxidation and Venturi scrubbing system achieves the efficient synergistic removal of NOx (87%) and PM (92%) under optimized conditions (O₃/NO = 1.5, L/G = 1.5), with effective control of O₃ escape. It was observed that an increased O₃ concentration enhances PM removal by elevating PM surface polarity and hydrophilicity; however, the competing reaction—where O₃ preferentially oxidizes NO over PM—resulted in a discernible decline in PM removal efficiency in the presence of NOx. Furthermore, both elevating the L/G ratio and incorporating surfactants into the scrubbing liquid were shown to improve PM capture. Raising the L/G ratio increases the number of mist droplets, thereby enhancing the probability of contact with PM, whereas surfactants reduce liquid surface tension and improve water dispersion—both mechanisms promote the adhesion of water to PM surfaces and facilitate particle growth. However, this research is limited to simulated exhaust under steady-state conditions and does not assess the energy consumption or potential secondary by-products of continuous O₃ generation. Future work should focus on dynamic engine testing, system durability, and techno-economic analyses to evaluate the technology’s feasibility for real-world NRMM applications.