The Decomposition of Dilute 1-Butene in Tubular Multilayer Dielectric Barrier Discharge Reactor: Performance, By-Products and Reaction Mechanism

Abstract

Butene is a typical component of exhaust gas in the petrochemical industry, the emission of which into the atmosphere would lead to air pollution. In this study, a tubular multilayer dielectric barrier discharge (TM-DBD) reactor was developed to decompose 1-butene at ambient pressure. The experimental results show that a decomposition efficiency of more than 99% and CO_x selectivity of at least 43% could be obtained at a specific energy density of 100 J/L with an inlet concentration of 1-butene ranging from 100 to 400 ppm. Increasing the volume ratio of O₂/N₂ from 0 to 20% and the specific energy density from 33 to 132 J/L were beneficial for 1-butene destruction and mineralization. Based on organic byproduct analysis, it was inferred that the nitrogenous organic compounds were the main products in N₂ atmosphere, while alcohol, aldehyde, ketone, acid and oxirane were detected in the presence of O₂. In addition, the contents of formaldehyde, acetaldehyde, ethyl alcohol, acetic acid and propionic acid increased with an increase in specific energy density, but the contents of propionaldehyde, ethyl oxirane, butyraldehyde and formic acid decreased. Three main pathways of 1-butene destruction were proposed involving Criegee intermediates and ozonolysis of the olefins, and the following degradation could be the dominant pathways rather than epoxidation. Overall, the developed TM-DBD system paved the way for scaling up the applications of plasma technology for gaseous pollutant decomposition.

1. Introduction

Butenes are important raw materials for synthetizing masses of industrial chemicals (such as 1,3-butadiene [1] and acetic acid [2]) and polymers (such as linear low density polyethylene and atactic polypropylene [3,4]). Meanwhile, butenes are the typical gaseous pollutants in petrochemical waste gas, which are unwanted byproducts in dehydrogenation and unreacted monomers in polyreactions. Although butenes have low toxicity and exert little effect on human health, their emission into the atmosphere should be controlled to avoid secondary pollution mediated by photochemical reactions [5,6,7]. Therefore, it is imperative to develop appropriate technology to remove butenes before being emitted into the atmosphere.

At present, there are many conventional methods to remove 1-butene, including adsorption, thermal combustion and catalytic oxidation. However, these techniques have some limitations in VOC treatment. For example, the adsorption method suffers from secondary pollution (such as spent adsorbent and collected organics). Thermal combustion has high natural gas consumption and produces extra CO₂. The catalytic oxidation can work well only at a high temperature, and catalyst poisoning is also a major problem [8]. In comparison, non-thermal plasma (NTP) technology has been widely used for odor and exhaust gas treatment due to its good decomposition capacity and low operating cost [9,10], especially for low concentration VOCs (<1000 ppm) [11,12,13]. Additionally, NTP technology also has some specific advantages, such as simple operation, quick switch of equipment, and simultaneous treatment of multiple pollutants. There are many ways to generate NTP, such as glow discharge, corona discharge and dielectric barrier discharge (DBD). Among these methods, DBD can prevent the occurrence of arc discharge and avoid the corrosion of metal electrodes by covering the electrodes with barrier layers [14,15]. Therefore, DBD has been recognized as the most commonly industrial NTP. In the discharge space, the molecules of VOCs can be decomposed and oxidized by highly reactive species, e.g., O(¹D), O(¹S), O(³P), O⁺, O₂⁺, O₃, N₂(A₃${\mathsf{\Sigma }}_{\mathrm{u}}^{+}$) and N₂(B₃${\mathsf{\Pi }}_g$), which are produced by collisions between energetic electrons and background gas molecules, such as N₂ and O₂ [16,17,18]. Thus far, only a handful of studies have been concerned with alkene decomposition by NTP, and most of them focus on the removal of ethylene. Mok et al. [19] reported that alkene and substituted alkene have much higher decomposition rate constants than aromatics and substituted alkane compounds. Aerts et al. [20] studied ethylene decomposition by the global (0D) model and found that atomic oxygen is the dominant destruction species at a low specific energy density (SED) and low inlet concentration, whereas the metastables dominated the destruction process at a high concentration and high SED. In the discharge space, O₂ promotes the mineralization of VOCs but also produces some hazardous organics, e.g., alcohol, aldehyde, ketone, acid, etc. [21,22,23,24]. In addition, SED and O₂ concertations also affect the formed CO_x and byproducts [25,26]. The decomposition and mineralization of ethylene can be improved with plasma-catalytic reactors developed by Mok et al. [27,28,29], and some unwanted byproducts, including O₃ and NO_x, were inhibited at a low ethylene concertation. In conclusion, these researchers offered a rich diversity of methodologies to investigate 1-butene decomposition in the plasma system. However, as an important waste gas during industrial processes, 1-butene decomposition by DBD plasma is still unknown. The comprehensive analysis of the 1-butene decomposition mechanism in NTP would be beneficial to promote the application of NTP technology in industrial VOC management.

In this work, a tubular multilayer dielectric barrier discharge (TM-DBD) reactor was developed for the effective decomposition of 1-butene at ambient pressure. The industrial operation parameters, such as inlet 1-butene concentration, SED and O₂/(N₂+O₂) ratio on 1-butene decomposition, were studied in a DBD reactor. The distribution of organic byproducts was detected by GC-MS analysis under different SED and N₂/air conditions. The decomposition mechanism of 1-butene was proposed in terms of byproduct analysis results.

2. Experimental Setup and Method

2.1. Experimental Setup

As shown in Figure 1, the experimental setup consisted primarily of a gas feeding system, a DBD reactor, a power supply, and electric and gaseous analytical systems. All continuous flow gases were supplied by gas cylinders and were adjusted by mass flow controllers (MFC). The 1-butene was mixed and diluted with N₂ and O₂ in a flask and then fed into the DBD reactor. The gas flow rate was selected at 40 L/min, 60 L/min and 80 L/min, and the initial 1-butene concentration was set at 100 ppm, 200 ppm and 400 ppm, respectively. The O₂ concentration in the fixed flow rate was adjusted by changing the proportions of O₂ and N₂. A tubular multilayer dielectric barrier discharge (TM-DBD) reactor was designed based on a study by Zhang et al. [30]. The TM-DBD reactor was energized by a high-frequency and high-voltage AC power supply (CTP-2000K, China) with sine wave output, which could supply voltage varying from 0 to 30 kV and frequency varying from 1 to 100 kHz. The applied voltage and current were monitored with a digital oscilloscope (Tektronix DPO3054, Beaverton, OR, USA), a high voltage probe (Tektronix P6015A) with an attenuation ratio of 1000:1, and a passive probe (Tektronix TPP0500). Gas samples were collected at the outlet of the DBD reactor. The 1-butene concentration was measured on a gas chromatograph (Agilent 8890B, Santa Clara, CA, USA) equipped with a flame ionization detector (FID). Organic intermediates were qualitatively analyzed on gas chromatography mass spectrometry (GC-MS) (Agilent 7890B-5973N, America). The concentrations of CO and CO₂ were quantified using a Fourier transform infrared (FTIR) spectrometer (Thermo Fisher, Antaris IGS, Waltham, MA, USA) fitted with a 20 cm long gas cell. The O₃ concentration was monitored by a O₃ analyzer (uSafe 3000, Shenzhen, China). The NO₂ concentration was determined online using a flue gas analyzer (Testo 350, Titisee-Neustadt, Germany).

2.2. DBD Reactor

The TM-DBD reactor is shown in Figure 2a and the pinouts were led out from the endpoints of iron powder for connecting high voltage power supply or ground. Each hollow quartz dielectric tube had a wall thickness of 1.5 mm, outer diameter of 10 mm and length of 80 mm. All electrodes were divided into 2 layers arranged in parallel. Adjacent electrodes were placed in every layer evenly with a gap distance of 4 mm. One layer consisting of 4 electrodes was connected with a high voltage power supply, acting as the discharge electrodes. The other layer, consisting of 5 electrodes, was connected with the earth, serving as the ground electrodes. The gap distance between the adjacent discharge electrode and the ground electrode was 3.8 mm. The gas could be treated through the “V” shape of the discharge spaces, which were located between the discharge electrodes and two adjacent ground electrodes. The cross section of gas through the reactor had a height of 80 mm and a width of 60 mm. Figure 2c shows the actual discharge conditions of the TM-DBD reactor in this experiment.

2.3. Measurement Methods

The energy consumed in the TM-DBD reactor was calculated using the voltage-charge (V-Q) Lissajous method [31]. The specific energy density (SED) was defined as the discharge power (W) divided by the total gas flow rate (L/min), which was calculated using the following relations:

$$\mathrm{S}\mathrm{E}\mathrm{D}(\mathrm{J}/\mathrm{L})=\frac{\mathrm{D}\mathrm{i}\mathrm{s}\mathrm{c}\mathrm{h}\mathrm{a}\mathrm{r}\mathrm{g}\mathrm{e} \mathrm{p}\mathrm{o}\mathrm{w}\mathrm{e}\mathrm{r}}{\mathrm{G}\mathrm{a}\mathrm{s} \mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w} \mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}}\times 60$$

(1)

The percentage of decomposition efficiency (DE) and selectivities toward CO, CO₂ and CO_x were obtained based on the following equations:

$$\mathrm{D}\mathrm{E}(\mathrm{\%})=\frac{{\left[{\mathrm{C}}_4{\mathrm{H}}_8\right]}_{\mathrm{i}}-{\left[{\mathrm{C}}_4{\mathrm{H}}_8\right]}_{\mathrm{o}}}{{\left[{\mathrm{C}}_4{\mathrm{H}}_8\right]}_{\mathrm{i}}}\times 100$$

(2)

$$\mathrm{C}\mathrm{O} \mathrm{s}\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y}(\mathrm{\%})=\frac{{\left[\mathrm{C}\mathrm{O}\right]}_{\mathrm{o}}}{4\times \left({\left[{\mathrm{C}}_4{\mathrm{H}}_8\right]}_{\mathrm{i}}-{\left[{\mathrm{C}}_4{\mathrm{H}}_8\right]}_{\mathrm{o}}\right)}\times 100$$

(3)

$${\mathrm{C}\mathrm{O}}_2 \mathrm{s}\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y}(\mathrm{\%})=\frac{{\left[{\mathrm{C}\mathrm{O}}_2\right]}_{\mathrm{o}}}{4\times \left({\left[{\mathrm{C}}_4{\mathrm{H}}_8\right]}_{\mathrm{i}}-{\left[{\mathrm{C}}_4{\mathrm{H}}_8\right]}_{\mathrm{o}}\right)}\times 100$$

(4)

$${\mathrm{C}\mathrm{O}}_{\mathrm{x}} \mathrm{s}\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y}\left(\mathrm{\%}\right)=\mathrm{C}\mathrm{O} \mathrm{s}\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y}+{\mathrm{C}\mathrm{O}}_2 \mathrm{s}\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y}$$

(5)

where [C₄H₈]_o and [C₄H₈]_i are the concentrations of 1-butene at the inlet and outlet of the TM-DBD reactor, respectively.

3. Results and Discussion

3.1. 1-Butene Decomposition

The capacity of the TM-DBD reactor toward the destruction of 1-butene was evaluated at different initial 1-butene concentrations and specific energy density (SED). In this series of experiments, a fixed O₂/(N₂+O₂) volume ratio of all the carrier gases was set at 20% for simulating air. As shown in Figure 3a, the decomposition efficiency of 1-butene increased with increasing SED at any inlet concentration until complete degradation. Specifically, 1-butene decomposition efficiency increased from 91.2% to nearly 100% with elevating SED from 33 to 66 J/L at an inlet 1-butene concentration of 200 ppm. The SED for attaining complete 1-butene degradation required at least 33 J/L, 66 J/L and 101 J/L when inlet concentrations were 100 ppm, 200 ppm and 400 ppm, respectively. The CO_x selectivity could directly reflect the mineralization extent of 1-butene because CO and CO₂ were the main decomposition products of 1-butene after plasma treatment. Figure 3b shows that CO selectivity was significantly higher than CO₂ selectivity. Notably, it was found that the CO₂ selectivity increased remarkably with the increase of SED, but the increase of CO selectivity was not evident under the same condition. For example, when the inlet 1-butene concentration was 400 ppm and SED increased from 33 to 203 J/L, the CO₂ selectivity rose rapidly from 4% to 13%; however, the CO selectivity rose slowly from 33% to 35%, which was higher than CO₂ selectivity. These results could be attributed to the fact that CO was much more easily produced than CO₂ in air plasma at atmospheric pressure, and these two oxidation products were generated through different paths (discussed in Section 3.4). Even though both oxidation processes could be enhanced with the increase of SED, CO selectivity increased slightly while an evident increase of CO₂ selectivity was observed. The reason might be that an increase of SED promoted the production of CO; however, most of the generated CO was subsequently consumed to synthesize CO₂ in the presence of extensive active particles.

In addition, Figure 3a,b indicates that lower inlet 1-butene concentration gained higher decomposition efficiency (under the state of incomplete degradation) and low CO/CO₂ selectivity. Taking the SED of 51 J/L and inlet 1-butene concentration of 100 ppm as an example, the decomposition efficiency, CO selectivity and CO₂ selectivity were 100%, 37% and 10%, respectively, while the results decreased to 68%, 34% and 5%, respectively, at an inlet 1-butene concentration of 400 ppm. The reason for this phenomenon was more possibilities for less organic molecules to react with highly reactive species in the same discharge space [32].

O₃ and NO_x were the main inorganic byproducts during the plasma treatment process, which should also be measured in the VOC decomposition process. In the air plasma discharge system, these two gaseous compounds were generated via dissociation, excitation and ionization. First, the excited-state atomic nitrogen N(²D) and N(²P), as well as excited-state molecular nitrogen N₂(A³${\mathsf{\Sigma }}_{\mathrm{u}}^{+}$) were produced by electron collisions with N₂ in Equations (6)–(8) [33,34]. These reactive species reacted with O₂ to generate NO and atomic O, as shown in Equations (9)–(11). In addition, the electron collisions with O₂ also produced the atomic O in Equation (12). O₃ was generated from atomic O via reactions with O₂ and M in Equation (13), where M could be either O₂ or N₂ [35,36]. NO₂ was detected as a major NO_x species, but no NO was detected in the gas treated by plasma because of the oxidization of O₃ and atomic O, as shown in Equations (14) and (15) [37]. As the equations shown, O₃ and NO_x are both involved in the generation and consumption of “useful” active species, such as atomic O and O₂, which could directly oxidize VOC compounds.

$$\mathrm{e}+{\mathrm{N}}_2\to \mathrm{e}+\mathrm{N}\left({}^2\mathrm{D}\right)+\mathrm{N}\left({}^2\mathrm{D}\right)$$

(6)

$$\mathrm{e}+{\mathrm{N}}_2\to \mathrm{e}+\mathrm{N}\left({}^2\mathrm{P}\right)+\mathrm{N}\left({}^2\mathrm{P}\right)$$

(7)

$$\mathrm{e}+{\mathrm{N}}_2\to \mathrm{e}+{\mathrm{N}}_2\left({\mathrm{A}}_3{\mathsf{\Sigma }}_{\mathrm{u}}^{+}\right)$$

(8)

$$\mathrm{N}\left({}^2\mathrm{D}\right)+{\mathrm{O}}_2\to \mathrm{N}\mathrm{O}+\mathrm{O}\cdot$$

(9)

$$\mathrm{N}\left({}^2\mathrm{P}\right)+{\mathrm{O}}_2\to \mathrm{N}\mathrm{O}+\mathrm{O}\cdot$$

(10)

$${\mathrm{N}}_2\left({\mathrm{A}}_3{\mathsf{\Sigma }}_{\mathrm{u}}^{+}\right)+{\mathrm{O}}_2\to \mathrm{O}\cdot +\mathrm{O}\cdot +{\mathrm{N}}_2$$

(11)

$$\mathrm{e}+{\mathrm{O}}_2\to \mathrm{e}+\mathrm{O}\cdot +\mathrm{O}\cdot$$

(12)

$$\mathrm{O}\cdot +{\mathrm{O}}_2+\mathrm{M}\to {\mathrm{O}}_3+\mathrm{M}$$

(13)

$${\mathrm{O}}_3+\mathrm{N}\mathrm{O}\to \mathrm{N}{\mathrm{O}}_2+{\mathrm{O}}_2$$

(14)

$$\mathrm{O}\cdot +\mathrm{N}\mathrm{O}+\mathrm{M}\to \mathrm{N}{\mathrm{O}}_2+\mathrm{M}$$

(15)

Figure 3c,d shows that the concentration of O₃ and NO₂ increased with the increase of SED. When the inlet 1-butene concentration was 100 ppm and SED increased from 33 to 203 J/L, the concentrations of O₃ and NO₂ increased from 514 to 1325 ppm and from 286 to 620 ppm, respectively. However, the increase of inlet 1-butene concentration led to an inhibition of the formation of O₃ and NO₂. For example, the O₃ concentration was reduced from 1021 to 533 ppm at an SED of 101 J/L as the inlet 1-butene concentration increased from 100 to 400 ppm. Meanwhile, the NO₂ concentration decreased from 520 to 367 ppm accordingly. The reason could be attributed to the consumption of energetic electrons and reactive species (e.g., N species and atomic O) during the process of 1-butene degradation [38]. In consideration of the significant formation of O₃ and NO₂ produced in the plasma process, many studies have reported that packing catalysts downstream of the DBD reactor could consume these byproducts effectively [39,40].

3.2. Effect of O₂/(N₂+O₂) Ratio

O₂ often plays an important role in VOC decomposition by affecting the electric field and the active species (Equations (9)–(15)). Three effects of O₂/(N₂+O₂) volume ratio on VOC decomposition performance based on the literature are shown below. First, the naphthalene decomposition efficiencies increased with the O₂/(N₂+O₂) volume ratio by the DBD plasma, indicating that the reactive species derived from O₂ mainly contributed to pollutant decomposition [41]. Next, the decomposition efficiencies of chlorodifluoromethane (CHF₂Cl) decreased with the increase of the O₂/(N₂+O₂) ratio because the reactive species from N₂ primarily contributed to the destruction of CHF₂Cl rather than those from O₂ [42]. Finally, the decomposition efficiencies of benzene and toluene reached a maximum value at 3–5% O₂ and then decreased with increasing O₂/(N₂+O₂) volume ratio. This result might be caused by the fact that an enhanced generation of O atoms with the increase of O₂ concentration generally led to a higher degradation efficiency. However, with the further increase in O₂ concentration, more excited-state nitrogen species, O₃ and O atoms, were consumed through collisions between active particles (Equations (9)–(11), (14) and (15)), which would otherwise be used for destroying benzene [43,44]. In order to understand the influence of O₂ on the degradation of 1-butene, a series of experiments were conducted in which the O₂ volume ratio was varied from 0 to 20% at different SED, and the experimental results are illustrated in Figure 4. In this series of experiments, the inlet 1-butene concentration was maintained at around 200 ppm.

As shown in Figure 4, the decomposition efficiency, CO_x selectivity, and the yield of O₃ and NO₂ increased rapidly with the O₂ concentration. Taking the SED of 66 J/L as an example, in the absence of O₂ in the gas, the 1-butene decomposition efficiency was only 11%, and CO_x, O₃ and NO₂ were not detected. When the O₂ volume ratio was raised to 20%, the 1-butene decomposition efficiency and CO_x selectivity increased to 99% and 45%, respectively, and the concentrations of O₃ and NO₂ were 588 ppm and 301 ppm, respectively. The increase of O₂/(N₂+O₂) ratio in the gas was beneficial for 1-butene mineralization and oxidation product formation by facilitating the production of atomic O, which was concordant with the previous literature [41,43,44]. In addition, SED could enhance the decomposition efficiency and CO_x selectivity even at a low O₂/(N₂+O₂) ratio. For example, when the O₂/(N₂+O₂) volume ratio was 5% and SED was increased from 33 to 132 J/L, the decomposition efficiency increased from 31% to 99% and CO_x selectivity increased from 15% to 38%. The results suggest that raising the SED was one of the most effective ways to improve 1-butene degradation, especially at a low O₂/(N₂+O₂) ratio.

3.3. Organic Byproduct Analysis

In order to illustrate the reaction mechanism of 1-butene destruction in N₂ and air, the gas compounds in the outlet of the TM-DBD reactor were analyzed by GC-MS at an SED of 51 and 203 J/L, respectively. The inlet concentration was maintained at around 300 ppm. The detected organic byproducts and their relative abundance under four different experimental conditions are presented in Figure 5 and

Table 1. The organic byproducts detected in four experimental conditions.

Peak Numbers	Time (min)	Organic Compound	Structure	Relative Abundance
				SED = 51 J/L in N2	SED = 51 J/L in Air	SED = 203 J/L in N2	SED = 203 J/L in Air
1	6.839	1-butene	CH2 = CHCH2CH3	5,094,566	3,544,865	4,561,915	n.d
2	6.034	Formaldehyde	CH2O	n.d	169,841	n.d	177,739
3	7.126	Acetaldehyde	CH3CHO	n.d	215,131	n.d	263,062
4	8.686	Ethyl alcohol	CH3CH2OH	n.d	639,90	n.d	148,173
5	9.382	Propionaldehyde	CH3CH2CHO	n.d	1,231,359	n.d	147,682
6	12.017	Ethyl oxirane	CH3CH2CH(O)CH2	n.d	840,904	n.d	704,603
7	12.091	Butyraldehyde	CH3CH2CH2CHO	n.d	371,202	n.d	82,059
8	12.225	Formic acid	HCOOH	n.d	67,179	n.d	n.d
9	12.344	2-butanone	CH3COCH2CH3	n.d	77,893	n.d	79,384
10	13.322	Acetic acid	CH3COOH	n.d	n.d	n.d	63,755
11	15.705	Propionic acid	CH3CH2COOH	n.d	n.d	n.d	93,962

n.d: not detected.

. For 1-butene decomposition in N₂ plasma, hydrocarbon and oxy-organics were not found in the outlet gas even though the SED was increased. Moreover, CO_x has not been found either under similar experimental conditions, as shown in Figure 4b. These results indicate that the decomposition products of 1-butene mediated by the N reactive species may be in the main forms of nitrogenous organic compounds (such as amine, nitrile and nitrogen heterocyclic) [41]. Unfortunately, these nitrogenous organic compounds cannot be detected by GC-MS analysis. In comparison, abundant organic byproducts of 1-butene decomposition were found in air plasma, which included alcohol, aldehyde, ketone, acid and oxirane. With the SED raised from 51 to 203 J/L, the contents of formaldehyde, acetaldehyde, ethyl alcohol, acetic acid and propionic acid increased, while the contents of propionaldehyde, ethyl oxirane, butyraldehyde and formic acid decreased.

3.4. Proposed Reaction Mechanism

Based on the intermediates detected by GC-MS analysis, three possible 1-butene decomposition pathways were proposed during the air plasma treatment. As shown in Figure 3, 200–1500 ppm O₃ were detected during the air plasma discharge procedure at a 1-butene concentration of 100–400 ppm. The cleavage of the double bonds of 1-butene (1) and related isomer 2-butene could easily take place by ozonolysis via 1,3-dipolar cycloaddition to give a carbonyl compound and a carbonyl oxide commonly called Criegee intermediates [45]. For Path A, due to the asymmetric structure of 1-butene (1), two zwitterion species, 5-O and 2-O, could be formed along with formaldehyde (2) and propanal (5), respectively. Criegee intermediates 5-O and 2-O could be trapped by nucleophiles, such as N-oxides, in air plasma. The generated species subsequently decomposed into related reduction products, propanal (5) and formaldehyde (2). With the aid of the active oxidants (i.e., ^•OH, O^•), aldehyde compounds could proceed through further oxidation to produce the related acid propionic acid (11) and formic acid (8) [46], which could go through a direct decarbonylation process to furnish CO and H₂O. Additionally, after further oxidation of formic acid to carbonic acid (12), a decarboxylation reaction could take place to eliminate a carboxyl group and with the release of CO₂. As shown in Figure 3b, it suggests that CO and CO₂ originated in two independent ways, and it was easier to obtain CO than CO₂. It should be noted that other possible pathways from acid compounds to generate CO and CO₂ could not be completely excluded, especially in surroundings full of high-energy electrons. For Path B, in the discharge atmosphere, the charged 1-butene (1-H) was capable of isomerizing with the resonance structure charged 2-butene (1-H’). Owing to the symmetrical structure of 2-butene, only one zwitterion specie 3-O was obtained and then reduced to acetaldehyde (3). The acetaldehyde (3) could not only be sequentially reduced to alcohol (4) but also be oxidized to acetic acid (10). For Path C, 1-butene could also react with Criegee intermediates to generate the epoxidized compound (6). According to the breaking method of C-O bonds, routes a and b were proposed in Figure 6. The n-butylaldehyde (7) and 2-butanone (9) could be determined, accompanying the rupture of the correspondent C₂-O (route a) and C₁-O bonds (route b), respectively.

In addition, the amounts of the above organic byproducts were examined under conditions of 51 and 203 J/L, respectively. In Path A, as shown in Table 1, with the increase of SED, extra propionic acid (11) was found with the decrease of propionaldehyde (5), while formaldehyde (2) was enriched, and formic acid (8) totally vanished to form CO_x. (i.e., CO and CO₂). In detail, there was a slight increase in the amount of CO₂, while the amount of CO remained the same (Figure 3b). Active oxygen species were required for the formations of the CO_x precursors, i.e., formic acid and CO₂. The O₂ concentration plays a positive role in raising CO_x selectivity, which can be supported by the experimental results shown in Figure 4b. As for Path B, acetic acid (10) appeared along with both acetaldehyde (3) and ethyl alcohol (4). However, unlike Paths A and B, ethyl oxirane (6) and butyraldehyde (7) in Path C both declined, while 2-butanone (9) was well maintained. On the basis of the above results, ozonolysis of the olefins and the subsequent degradation reaction may be the dominant pathways rather than epoxidation.

4. Conclusions

In this study, we developed a TM-DBD reactor for the decomposition of 1-butene at ambient pressure, which is a typical gaseous pollutant in petrochemical exhaust gas. The results indicate that DBD plasma was an effective technology for removing a low 1-butene concentration of 100–400 ppm from the exhaust gas. First, the decomposition efficiency, CO_x selectivity and by-product yield were systematically studied with operating parameters (e.g., inlet concentration, SED). When SED reached 100 J/L, a decomposition efficiency of more than 99% and CO_x selectivity of at least 43% could be obtained; meanwhile, the production of O₃ and NO₂ were lower than 1021 ppm and 520 ppm, respectively. Second, the effect of O₂/(N₂+O₂) ratio was also examined. The increasing O₂/(N₂+O₂) ratio and SED were beneficial in improving the 1-butene decomposition and CO_x selectivity. Finally, the degradation mechanism of 1-butene was proposed according to the analysis of organic byproducts. The main organic byproducts of 1-butene decomposition in air plasma consisted of alcohol, aldehyde, ketone and organic acid, as well as oxirane. Based on the above experimental results from the GC-MS analysis, three main pathways were proposed involving Criegee intermediates. In particular, ozonolysis of the olefins and the following reduction (Paths A and B) could be the dominant pathways rather than epoxidation (Path C), which was consistent with the GC-MS analysis results. In general, this study did not only offer an effective TM-DBD reactor for 1-butene decomposition but also illustrated the potential 1-butene degradation pathways, facilitating the industrial application of plasma technology for gaseous pollutant control. More importantly, these original data played a constructive role in promoting plasma industrialization, especially in the treatment of petrochemical waste gas.