Mechanistic Study on the Removal of NO2 from Flue Gas Using Novel Ethylene Glycol-tetrabutylammonium Bromide Deep Eutectic Solvents.

The removal of NO x (approximately 90% of which is NO) from flue gas is a crucial process for clean power generation from coal combustion. Oxidation of NO to NO2 followed by NO2 absorption using sorbents is considered to be a promising technology alternative to selective catalytic reduction (SCR). This study investigated the absorption of NO2 in flue gas by ethylene glycol (EG)-tetrabutylammonium bromide (TBAB) deep eutectic solvents (DESs) under a range of experimental conditions. The effects of experimental conditions including molar ratio of EG to TBAB, operating temperature, residence time, and the O2 and steam partial pressure in the flue gas on the denitrification performance of EG-TBAB DESs were systematically analyzed. The concentrations of NO2 in the inlet and outlet were evaluated using a flue gas analyzer. The chemical structure changes of DESs after denitrification were characterized using Fourier transform infrared (FT-IR) spectroscopy. The obtained analysis signified that maximum denitrification efficiency and capacity were achieved at a EG/TBAB molar ratio of 5:1, 50 °C, and 6 s residence time. EG-TBAB DESs were able to maintain a stable denitrification performance after five absorption-desorption cycles. The results of quantum chemical calculation and 1H NMR spectra of EG-TBAB DES show that bromide anions in the EG-TBAB DES maintained strong interactions with NO2 via hydrogen bonding, leading to increased NO2 adsorption. The presence of O2 and steam in the flue gas improved the absorption of NO2 in EG-TBAB DESs due to chemical reactions and formation of nitrate.

Introduction

Elimination of nitrogen oxin class="Chemical">des (NO) that are present in the flue gas is critical as NO emissions raise significant environmental concerns such as acid rain, ozone depletion, and photochemical smog.[1−3] NO is the primary component of atmospheric NO (∼90%); it is inert and has extremely low solubility in common solvents. Current industrial technologies for NO reduction include selective catalytic reduction and selective noncatalytic reduction (SCR[4] and SNCR,[5] respectively). However, these technologies require a sizable initial investment, high operating costs, and generate secondary pollution, which restrict their commercial utilization.[6,7] Therefore, developing environmentally friendly, high-performing, and low-cost NO sorbents for simple NO removal systems is of paramount importance.

Deep eutectic solvents (DESs) are mixtures of different compounds that have lower melting points compared to the individual components of mixure.[8−10] Recently, n class="Chemical">DESs have been considered as promising alternatives to traditional gas sorbents due to their excellent properties such as “tailormade” characteristics, better structural integrity, ideal solubility, and extraordinary thermal and chemical stabilities.[11−13] Furthermore, DESs are produced from cheap and nontoxic natural sources. Hence, they exhibit great potential in large-scale gas sequestration and sustainable applications.[14−17]

Current studies on gas absorption using n class="Chemical">DESs focus on gases such as CO2, SO2, and H2S.[14,18−20] However, there are limited studies on using DESs as sorbents for NO.[21−23] Duan et al. found solubility differences in DESs with differing component ratios and operating temperatures by investigating the solubility of pure nitrogen oxide and nitrogen dioxide in various caprolactam-based DESs.[23] To the best of our knowledge, a few reports involving the selective removal of NO from flue gas using DESs have been published;[22] these reports indicated a 100% NO removal efficiency within 40 min. However, the denitrification capacity was significantly lower compared to active acidic gases such as NO2. Over the past 20 years, many denitrification methods have incorporated an oxidation step to oxidize NO to higher-valence NO (e.g., NO2 and N2O3) to improve denitrification efficiency in an absorption tower.[24−26] These oxidation methods included using a strong oxidant, catalytic oxidation, plasma, and others.[27,28] Thus, it is feasible to reduce NO emissions via catalytic oxidation of NO followed by NO2 removal.

Limited literature data show that DESs significantly absorb pure n class="Chemical">NO2.[23] However, the absorption mechanism is still unclear. Meanwhile, little research has been carried out on the absorption of NO2 by DES at low partial pressures of NO2. In this study, we addressed the unexploited research by studying the selective absorption of NO2 from simulated flue gas as a function of DES composition, operation temperature, residence time, and flue gas composition. Current literature supports the understanding of interaction mechanisms between DES components and NO2 molecules, which facilitate the advancement of high-performing and low-priced NO sorbents.

Results and Discussion

Influence of Operation Temperature

The effect of operating temperature on the removal of NO2 was analyzed using an n class="Chemical">ethylene glycol (EG)-tetrabutylammonium bromide (TBAB) molar ratio of 5:1 and 6 s residence time. As shown in Figure a, the fresh EG-TBAB DES was colorless. However, as it absorbed NO2, it turned brown after absorbing NO2 at 30 °C, dark brown at 50 °C, and light brown at 70 °C. Since pure NO2 is a red-orange colored gas, the color change before and after NO2 absorption suggested that NO2 was successfully absorbed in the EG-TBAB DESs solution. The darkest color at 50 °C indicated the highest NO2 absorption concentration.

Figure 1

Color change of EG-TBAB DESs before and after NO2 absorption (a), NO2 removal efficiency (b), and denitrification capacity (c) (at an EG-to-TBAB molar ratio of 5:1, residence time of 6 s, and 0.15% NO2).

Color change of EG-TBAB DESs before and after n class="Chemical">NO2 absorption (a), NO2 removal efficiency (b), and denitrification capacity (c) (at an EG-to-TBAB molar ratio of 5:1, residence time of 6 s, and 0.15% NO2).

NO2 removal efficiency at different operation temperatures was computed using eq , and the calculated results were plotted against the absorption time, as illustrated in Figure b. Initially, the n class="Chemical">NO2 removal efficiency consistently surpassed 95% at all three operation temperatures. However, the breakthrough time (when removal efficiency drops below 90%) changed significantly with operating temperature. The breakthrough time initially increased from ∼600 to ∼800 min as the temperature increased from 30 to 50 °C, and then dropped to ∼110 min at 70 °C. As shown in Figure c, the denitrification capacity (calculated using eq ) peaked at 50 °C, which was consistent with the longest breakthrough time and dark-brown appearance.

NO2 removal efficiency and denitrification capacity as a function of temperature were similar to those of NO, as previously reported.[22] The longest breakthrough time and the maximum denitrification capacity were detected at 50 °C for both n class="Gene">gases in EG-TBAB DESs. The physical absorption was more favorable at low temperatures, whereas elevated temperatures favored chemical absorption. The optimal operating temperature at 50 °C implied that EG-TBAB DESs adsorbed NO2 both physically and chemically, similar to NO. However, the breakthrough period and the denitrification capability of NO2 were more than an order of magnitude higher than those of NO. EG-TBAB showed a higher adsorption efficiency for NO2. Therefore, removal of NO2 via NO oxidation is a promising alternative for removing NO in flue gases.

Figure S1 displays the Fourier transform infrared (FT-IR) spectra of fresh and used EG-TBAB DES before after n class="Chemical">NO2 absorption at different operating temperatures. No changes in chemical functionality were observed for fresh or used EG-TBAB DESs; FT-IR peaks of ranges 1545–1570, 700–900, ∼1610, and 1560–1490 cm–1 attributed to the aromatic ring C=C stretching vibrations, out-of-plane vibration of aromatic C–H groups, aromatic carbon, and stretched aromatic rings, respectively, were observed for all samples.[29,30] These results indicate that NO2 absorption is primarily physical for EG-TBAB. However, differences such as the C=C stretching vibration of the aromatic ring were observed in the intensity and fine structure of absorption wavelengths between fresh and used EG-TBAB DESs; these could be due to weak chemical absorptions between NO2 and EG-TBAB.

Effect of Residence Time

EG-TBAB n class="Chemical">DESs NO2 removal efficiency and denitrification capacity with respect to residence time are shown in Figure a,b, respectively. For a 6 s residence time, the NO2 removal efficiency remained at 100% for the first 550 min and subsequently dropped to the breakthrough point at which the removal efficiency decreased to 90% after ∼800 min. When the residence time was reduced to 4.5 s, NO2 removal efficiency remained at 100% for the first 400 min, and then dropped sharply to 90% for 455 min. When the residence time was lowered to 3 s, the NO2 removal efficiency remained at 88% for the first 200 min and then decreased rapidly. The denitrification capacity decreased with shorter residence times, similar to the removal efficiency. These results indicated that the solubility of NO2 in DESs did not reach saturation when the residence time was between 3 and 4.5 s. Increasing the residence time of NO2 gas in the EG-TBAB DESs promoted the interactions and retention of NO2.

Figure 2

Denitrification efficiency (a) and denitrification capacity (b) of EG-TBAB DESs at different residence times (operating temperature, 50 °C; EG-to-TBAB molar ratio, 5:1; and 0.15% NO2).

Denitrification efficiency (a) and denitrification capacity (b) of EG-TBAB DESs at different residence times (operating temperature, 50 °C; n class="Chemical">EG-to-TBAB molar ratio, 5:1; and 0.15% NO2).

Effect of EG/TBAB Molar Ratio

Figure shows the EG-TBAB n class="Chemical">DES denitrification capacity and the removal efficiency of NO2 as a function of EG-to-TBAB ratio at 50 °C, for 6 s residence time. The breakthrough time increased as the molar ratio increased from 3:1 to 5:1, but decreased with a further increase of the molar ratio from 5:1 to 50:1 (cf. Figure a). With an EG-TBAB of 50:1 and pure EG, the NO2 removal efficiency was very low and did not reach the breakthrough point at any stage throughout the process. The denitrification capacity followed the same trend of the removal efficiency, i.e., initially increasing as the EG content reached 5:1 but subsequently decreasing significantly (cf. Figure b). NO2 interacted strongly with EG-TBAB DESs via H-bonding, which is detailed in the following section. DESs contain a large number of asymmetric ions, which in turn leads to a lower lattice energy. The delocalization of charges in hydrogen bonds between halogen ions and HBD is the primary reason for the reduction of the melting point of DES. In addition to the expected BrHN hydrogen bonds, EG also forms hydrogen bonds with the bromine cation (e.g., OHNBr), resulting in both Ch-EG and Br-EG clusters. At the optimal molar ratio of 5:1, the original crystal structures may have been destroyed, and extensive H-bond networks were formed. This observation was in agreement with our previous study,[22] which speculated that there was no melting point and the lowest glass-transition temperature was detected at the 5:1 molar ratio. These results differed from previous work on NO removal by EG-TBAB DESs,[22] where the optimum molar ratio was 50:1. This result can be attributed to the weak H-bonding between NO and EG-TBAB.

Figure 3

NO2 removal efficiency (a) and denitrification capacity (b) of EG-TBAB DESs at different molar ratios (operating temperature, 50 °C; residence time, 6 s; and 0.15% NO2).

Figure 4

NO2 removal efficiency (a) and denitrification capacity (b) for different absorption–desorption cycles (operating temperature, 50 °C; EG-to-TBAB molar ratio, 5:1; residence time, 6 s; and 0.15% NO2).

NO2 removal efficiency (a) and denitrification capacity (b) of n class="Chemical">EG-TBAB DESs at different molar ratios (operating temperature, 50 °C; residence time, 6 s; and 0.15% NO2).

NO2 removal efficiency (a) and denitrification capacity (b) for different absorption–n class="Chemical">desorption cycles (operating temperature, 50 °C; EG-to-TBAB molar ratio, 5:1; residence time, 6 s; and 0.15% NO2).

Regeneration Capability of EG-TBAB DESs

To investigate the regeneration performance and n class="Chemical">desorption of NO2, N2 was bubbled through the NO2-EG-TBAB DESs at 90 °C for 3 h under a similar experimental setup. NO2 removal efficiency and denitrification capacities of the regenerated EG-TBAB DES samples were analyzed at 6 s flue gas residence time at 50 °C, as shown in Figure . No observed drop in removal efficiency and denitrification capacity was detected after five cycles of absorption–desorption runs, e.g., the absorption capacities (g NO2/100 g DES) of DES are 48.94, 48.15, 47.31, 47.29, and 47.19, which indicated an excellent regeneration ability of EG-TBAB DESs. The reversibility and stability of NO2 absorption–desorption for EG-TBAB DESs further proved that the NO2 molecules were absorbed by EG-TBAB DESs.

Quantum Chemical Calculation on the Interaction of EG-TBAB DESs with NO2

The structures of EG-TBAB DESs with n class="Chemical">NO2 were optimized and calculated using DFT-DMol.[3] Each preliminary configuration was considered, and the lowest-energy complex was considered with optimized structure (Figure ). The structure variables for the optimized geometries are listed in Table . The organization of gaseous molecules around the anion is correlated to the absorption. Figure a shows that the Br anion interacted strongly with the N atom in NO2 at 5.36 Å. The positively charged H in EG interacted strongly with an O atom from NO2 at 2.24 Å. Due to these strong interactions, the average O–N–O angle bends from 179.98 to 134.01° and the N–O bond length elongates by 7% to 1.19 Å. These angle and bond length changes were more pronounced for NO2 in an EG-TBAB-NO2 complex than for isolated NO2. Figure c shows that the Br anion in EG-TBAB DES was the primary active site for NO2 absorption. Hydrogen bonding played a critical role in NO2 absorption. The interaction energy and absorption enthalpy are essential to affect the gas absorption capacity. The binding energy and absorption enthalpy for the EG-TBAB-NO2 complex (−36.97 kJ/mol) were larger than those for EG-TBAB DES (−34.91 kJ/mol) and NO2 (−2.05 kJ/mol), indicating that NO2 absorption preferably occurs in EG-TBAB DES.[23] In addition, absorption resulted in a change in charge distribution. The net charge (−0.65e) transferred from DES to NO2 on the EG-TBAB DES–NO2 complexes, which allowed the Br anions to maintain strong interactions with NO2 that facilitate NO2 adsorption.

Figure 5

Optimized structure of complexes (a) EG-TBAB DES, (b) NO2, and (c) EG-TBAB DESs with NO2.

Table 1

Structural Parameters of EG-TBAB DESs before and after NO2 Absorption

structural parameters	DES	NO2	DES-NO2	Δ
C–O–H (deg)	108.19		108.01	0.18
O–H (Å)	0.97		0.97	0
O–N (Å)		1.11	1.19	0.08
O–N–O (deg)		179.98	134.01	45.97

Optimized structure of complexes (a) EG-TBAB DES, (b) n class="Chemical">NO2, and (c) EG-TBAB DESs with NO2.

A comparison between the 1H NMR spectra of n class="Chemical">EG-TBAB DES before and after NO2 absorption is shown in Figure . It can be observed that the standard peaks of H atom in TBAB at 1.3 and 3.35 ppm, which are assigned to the chemical shift of the H atom connected to C(2) and C(4,5), respectively, shifted down to 1.40 and 3.40 ppm. The peak of −OH in EG shifted up from 3.85 to 3.95 ppm. These results demonstrate that the EG and Br– can form a hydrogen bond and an intermolecular force with NO2, which in turn may enhance the NO2 absorption.[19,31] Also, it can be seen that the intensity of the peaks experienced a slight change in intensity. This phenomenon could be caused by the reaction of NO2 and OH···Br–, which further indicates that the positive charge is not evenly distributed over the TBAB. This finding was also consistent with the density functional theory (DFT) results of the EG-TBAB DES NO2 absorption.

Figure 6

1H NMR spectra of EG-TBAB DES before and after NO2 absorption.

1H NMR spectra of n class="Chemical">EG-TBAB DES before and after NO2 absorption.

Effects of Steam on NO2 Absorption

Figure shows the EG-TBAB n class="Chemical">DESs NO2 removal efficiency and denitrification capacity for steam compositions of 0 and 5% at 50 °C. In the absence of steam, the NO2 removal efficiency of EG-TBAB DESs remained at 100% for the first 500 min but dropped significantly thereafter. However, with 5% steam, the NO2 removal efficiency remained at 100% for the duration of the experiment (1000 min), with a concentration of 50 ppm NO after 200 min (Figure a). The denitrification capability of the EG-TBAB DESs increased by 30% in the presence of steam (cf. Figure b). These results signified that the existence of steam in the flue gas enhanced the removal of NO2 by EG-TBAB DESs and may be due to interactions between polar H2O and NO2 molecules. NO2 could react with H2O to form HNO3 at low temperatures, i.e., 3NO2 + H2O = 2HNO3 + NO. EG-TBAB had a high adsorption capacity for HNO3 and NO.[22] Only the generated NO (50 ppm) was detected.

Figure 7

NO2 absorption efficiency and NO formation (a) and denitrification capacity (b) of EG-TBAB DES at steam concentrations of 0 and 5% (operating temperature, 50 °C; EG-to-TBAB molar ratio, 5:1; residence time, 6 s; NO2 concentration, 0.15%).

NO2 absorption efficiency and NO formation (a) and denitrification capacity (b) of n class="Chemical">EG-TBAB DES at steam concentrations of 0 and 5% (operating temperature, 50 °C; EG-to-TBAB molar ratio, 5:1; residence time, 6 s; NO2 concentration, 0.15%).

Figure S2 displays the FT-IR spectra of EG-TBAB DESs before and after n class="Chemical">NO2 removal in the presence and absence of steam. Peaks ascribed to aromatic carbon (∼1610 cm–1), stretched aromatic rings (1560–1490 cm–1), and out-of-plane aromatic C–H vibrations (700–900 cm–1) were observed for all samples. The peak at 1650 cm–1 corresponds to −N–O vibration bands from NO3–. When 5% steam was added to the flue gas, NO3– was formed after absorption of NO2, suggesting that steam improved NO2 absorption. This result further confirmed the aforementioned mechanism of NO2 absorption by EG-TBAB DESs.

Influence of Steam and Oxygen on NO2 Removal

Oxygen plays a critical role in n class="Chemical">NO2 removal by EG-TBAB DES because it oxidizes NO to NO2 at low temperatures and thus enhances NO2 removal efficiency. Figure a shows the changes of NO2 removal efficiency and NO concentration in the outlet flow as a function of flue gas O2 concentration with 5% steam. The NO2 removal efficiency was 100% for all three conditions investigated. However, the outlet concentration of NO decreased as the O2 partial pressure increased and dropped below the detection limit with 10% O2 in the flue gas. The denitrification capacity also increased with O2 partial pressure (cf. Figure b).

Figure 8

NO2 absorption efficiency and outlet concentration of NO (a) and denitrification capacity (b) at different oxygen partial pressures in the presence of 5% steam (operating temperature, 50 °C; EG-to-TBAB molar ratio, 5:1; residence time, 6 s; 0.15% NO2).

NO2 absorption efficiency and outlet concentration of NO (a) and denitrification capacity (b) at different n class="Chemical">oxygen partial pressures in the presence of 5% steam (operating temperature, 50 °C; EG-to-TBAB molar ratio, 5:1; residence time, 6 s; 0.15% NO2).

These results indicated that the NO2 removal efficiency and denitrification ability of the n class="Chemical">EG-TBAB DESs can be promoted by the coexistence of oxygen and steam. One possible reason for this synergic effect of O2 and steam is that absorbed O2 in EG-TBAB could dissociate and convert to active lattice oxygen and thus catalytically oxidize NO2 to HNO3. As shown in Figure S3, peaks corresponding to vibration bands of −N–O from NO3– (1650 cm–1) were noted in both the presence and absence of O2, but higher peak intensities were observed at higher O2 partial pressures; this indicated an increased concentration of NO3– and greater removal of NO2. NO2 removal efficiency and denitrification capacity were higher than those for NO in EG-TBAB DESs, further indicating the feasibility of oxidizing NO into NO2 in flue gases, with subsequent removal using EG-TBAB DES in the presence of steam and oxygen without the formation of secondary products. The outcome provides a potential route for designing future industrial denitrification technologies.

Conclusions

EG-TBAB DESs have demonstrated excellent performance in the removal of n class="Chemical">NO2 from flue gas. The peak denitrification efficiency and capacity were attained at a 5:1 EG-to-TBAB molar ratio, 50 °C, and 6 s residence time. The presence of steam and oxygen in the flue gas promoted NO2 absorption by EG-TBAB DESs due to the formation of nitrate via reactions with O2 and H2O. Hydrogen bonding played a significant role during NO2 absorption. Br anions maintained a strong interaction with NO2, resulting in high denitrification capacity. EG-TBAB DESs displayed an excellent denitrification consistency after five cycles of absorption and desorption, suggesting that the desorption of NO2 was relatively easy and required low energy input for regeneration. Therefore, DESs have potential as high-performing denitrification sorbents with applications in industrial denitrification technologies.

Materials and Methods

Ethylene Glycol-tetrabutylammonium Bromide DES Preparation

Pure analytical ethylene glycol (n class="Chemical">EG) and tetrabutylammonium bromide (TBAB) reagents utilized in this study were from Sinopharm Chemical Reagent Co. Ltd., Shanghai, China. To understand the effect of molar ratio of EG and TBAB, different ratios (4:2, 6:2, 8:2, 10:2, 20:2, and 100:2) of these compounds were prepared and stored based on the previously reported methods.[22] The Karl Fischer coulometric titration method was used to analyze the H2O content of the desiccated DES solutions, which was found to be less than 0.5 wt %.

Classification of Ethylene Glycol-tetrabutylammonium Bromide DESs

The characterization methods for the EG-TBAB DESs are n class="Chemical">described in detail from our previous study.[22] In brief, a Thermo Fisher Nicolet PerkinElmer device with universal attenuated total reflectance (UATR) was used to detect the change in the ATR-FT-IR spectra. The difference in the 1H NMR spectra of EG-TBAB DES samples before and after the absorption of NO2 was recorded using a 400 MHz spectrometer (Bruker Avance III, Switzerland) with heavy water (D2O) as a solvent and tetramethylsilane (TMS) as the internal standard.

DFT Analysis of DES Structure

Density functional theory (DFT) was used to determine quantum chemical interactions between DESs and n class="Chemical">NO2. Detailed specifications of the DFT analysis method/procedure are found elsewhere. The Brillouin-zone integrations were achieved by means of a 4 × 2 × 1 Monkhorst Pack grid. All of the computations were calculated by the DMol[3] program.[22]

NO2 Absorption Experiments

NO2 absorption experiments were conducted in a bubble column reactor and were illustrated and n class="Chemical">described in detail in our previous study.[22] The experiments were carried out in a glass container immersed in a H2O bath, which had a temperature accuracy of ±0.10 °C. The gas distributor consisted of a 2 cm disk base, and the magnitude of the bubbles formed was about 2 mm. The effects of DES residence time, operational temperature, and molar ratio on the removal efficiency of NO2 were examined using a simulation of the flue gas that contained 0.15 vol % NO2 and then stabilized with the addition of nitrogen gas. In addition, 0–5 vol % steam and 0–10 vol % oxygen were injected to the simulated flue gas to investigate the effects of oxygen and steam on the removal of NO2 with EG-TBAB DESs. The concentration of NO2 in the outlet gas was continuously monitored by a German-made MRU MGA5 online flue gas analyzer. The online flue gas analyzer had a detection accuracy of ±2 ppm, and the experimental runs were repeated for at least three times for result accuracy. Residence time ranging from 3 to 6 s was used in this study.

The NO2 removal efficiency (η) was computed by eq where Cin and Cout are the concentrations of NO2 (ppmv) in the inlet and outlet of the flue gas, respectively.

The quantity of the absorbed NO2 (denitrification capacity) was calculated by eq (22)where mNO is the mass absorption capability of NO2 in EG-TBAB DES, t1 and t2 are the initial and saturation times during absorption, respectively, Q (mL/min) is the flow rate of the gas stream, and mDES and ρNO stand for the mass of DES that was used during absorption and the density of NO2, respectively.