Improving the NO_x decomposition and storage activity through co-incorporating ammonium and copper ions into Mg/Al hydrotalcites

Abstract

We report a 40% and 70% improvement of NO_x decomposition and storage rate based on synthetic Mg/Al/Cu/NH₄⁺ hydrotalcites (HT), compared with Mg/Al/NH₃ and Mg/Al/Cu HTs. TGA and DTG show that ammonium has been released from HT below 160 °C. A combined characterization through WAXS (Wide-Angle X-ray Scattering), SEM, sXAS (soft X-ray Absorption Spectroscopy), CHNS elemental analysis and N₂ absorption and desorption reveals the structural and physical properties of this co-incorporated system. While Mg/Al/Cu/NH₄⁺ HT retains the typical structure of hydrotalcite-like compounds, its crystallinity weakened. WAXS shows that the layer–layer spacing of Mg/Al/Cu and Mg/Al/Cu/NH₄⁺ HTs decreases due to the Jahn–Teller effect. The particle size of Mg/Al/Cu/NH₄⁺ decreases about 10 nm compared with that of Mg/Al/Cu HT. sXAS and CHNS elemental analysis reveal that ammonium was successfully introduced into Mg/Al/Cu/NH₄⁺ HT by analyzing the N-K edge and CHN content, but not into HT without Cu²⁺, indicating the importance of the co-incorporation of ammonium and Cu²⁺. Additionally, the test of NO_x decomposition and storage activity revealed that the introduction of ammonium into the system should be responsible for the performance improvement.

---

Introduction

Pollution is one of the central concerns when considering the impact of human behavior on our environment. Nitrogen oxides (NO_x) are major pollutants that contribute heavily to the production of smog, depletion of tropospheric ozone and the inception of acid rain.¹ It has become an issue of utmost urgency to reduce the total amount of NO_x emission because of the severity of the present smog proliferation, especially in developing countries. Coal-powered plants and nitric acid factories are the main stationary sources of NO_x emissions and automobiles account for another half of man-made NO_x emissions.² Among the technologies being used to curb NO_x emission, catalytic reduction through various reducing agents has attracted considerable attention.³

Hydrotalcites (HTs) are layered double hydroxides with the general formula [M_1−x²⁺M_x³⁺(OH)₂]^x+A_x/m^m−·zH₂O. They are typically positively charged layers of brucite-like Mg(OH)₂, with trivalent cations substituting divalent cations in the octahedral sites.⁴ A wide range of derivatives with various combinations of M²⁺, M³⁺ and Aⁿ⁻ ions can be synthesized. Three important characteristics make HTs promise for practical applications. First, they have a good anion-exchange capacity, so they can be used as ion exchangers, adsorbents or sensors. Second, they operate as solid bases depending on the composition. Therefore, they have been broadly studied and successfully employed as basic catalysts for several reactions.^5,6 Third, since they present a high surface area and basic character, calcined HTs such as Mg–Al mixed oxides are potentially useful as catalysts and/or catalyst precursors.

HTs combined with metals can store NO_x as nitrates on lean (oxidizing) conditions and reduce the stored nitrates under rich (reducing) condition.⁵ Yexin Zhang and coworkers synthesized Pd and K co-supported Mg–Al mixed oxides to stimulate the reduction of pre-absorbed NO_x with CO.³ Ja Hun Kwak and coworkers studied the system of Pt–BaO/MgAl₂O₄ supported catalysts and demonstrated its reduction to NO_x.⁷ Xinyan Zhang and coworkers investigated the selective catalytic reduction of N₂O by NH₃ over an Fe–mordenite. They found that some NO may form NO₂, which reacts with NH₄⁺ to produce NH₄NO₂, and further decomposes to N₂ and H₂O.⁸ Zhongpeng Wang synthesized Cu/Mg/Al HT-derived metal oxides and found that the Cu substitution content strongly influenced the crystal phases, porous structures and redox properties of the catalysts.⁹ Corma and coworkers studied the simultaneous catalytic removal of SO_x and NO_x with HT-derived oxides containing copper. Their purpose was to ascertain the possibility of usage in FCC units. They suggested that Cu-containing hydrotalcites might be an applicable active additive to FCC catalysts for NO removal in the regenerator unit. When the NO concentration is low, 100% conversion of NO should be expected. However, when the O₂ concentration increases, the additive becomes inactive for NO decomposition.¹⁰ A. E. Palomares and coworkers found that the addition of 1 wt% of transition metals with redox properties such as Pt, Pd, V and Ru to the HTs increases the activity. This is the result of a combination of the redox properties of these metals and the acid–base properties of the hydrotalcite. However, these noble metals will increase the cost of the catalysts. Wang Haitao and coworkers studied the emission characteristics of gasoline-fueled taxicabs during cold and hot starts. They found that NO_x emission accounts for 76% during cold starts and 85% during hot starts, and it increases less than 6% during the process of driving. However, most of the catalysts in use can only carry out catalytic action above 300 °C.¹¹ So, it's necessary to develop catalysts that are efficient below 300 °C.

Ammonia has been used as a traditional reducing agent for NO_x removal for many years in industrial boilers and has recently been employed for removing NO_x from vehicle exhaust. However, in present technologies, it is supplied by independent equipment and leads to increasing the cost of NO_x removal. On the other hand, the additional pollution resulting from ammonia itself is another serious problem. In this work, we report an efficient NO_x storage-decomposition hydrotalcite based on Mg/Al/Cu/NH₄⁺ HT for low-temperature NO_x removal activity. The microscopic structure, electronic structure and thermal stability were investigated through combined lab-based and synchrotron-based tools. The decomposition and storage activity for removing NO_x is reported and the possible incorporation mechanism is discussed.

Experimental section

HT preparation

The Mg/Al HT co-incorporated with Cu²⁺ and NH₄⁺ (Mg/Al/Cu/NH₄⁺ HT) was prepared via co-precipitation. 0.11 M aluminum nitrate nonahydrate, 0.33 M magnesium nitrate hexahydrate, 0.22 M copper nitrate hexahydrate and 0.11 M NH₄Cl were dissolved in DI water to form solution 1. 0.22 M sodium hydroxide and 0.11 M sodium carbonate were dissolved in DI water to form solution 2. Solution 1 and solution 2 were slowly added dropwise into a flask containing 50 mL of water during vigorous stirring. The pH was controlled at 9–10 by varying their addition rate at 60 °C. The slurry was stirred for an additional 1 h and aged quiescently at 65 °C for 18 h. The obtained precipitate was filtered, washed with distilled water until the pH was 7 and then vacuum dried at 80 °C for 12 h into Mg/Al/Cu/NH₄⁺ HT in the powder form. Mg/Al HT (Mg²⁺ : Al³⁺ ratio of 3 : 1), Mg/Al/NH₃ HT (Mg²⁺ : Al³⁺ : NH₃·H₂O ratio of 3 : 1 : 1) and Mg/Al/Cu HT (Mg²⁺ : Al³⁺ : Cu²⁺ ratio of 3 : 1 : 2) reference samples were prepared with the same method as a reference. All of the chemical reagents (AR) above were obtained from Shanghai Sinopharm Chemical Reagent Co., Ltd.

WAXS measurements

Wide-angle X-ray scattering (WAXS) was performed on beamline 7.3.3 of the Advanced Light Source (ALS) at the Lawrence Berkeley National Laboratory (LBNL). The beamline was configured with an X-ray energy of 10 keV (wavelength 0.124 nm) and a beam size of ∼300 μm by 700 μm. Two-dimensional scattering patterns were collected on a Pilatus 1M detector (Dectris). The distance between the Pilatus detector and the samples was 0.29 m.^12–14 The resultant isotropic scattering pattern was radially averaged using NIKA, an Igor Pro based package, to produce a 1D scattering profile.¹⁵

SEM morphology

The scanning electron microscopy (SEM) measurements were carried out using a HITACHI SU8010, magnification of 30 800 000. The images were taken with an emission current = 10 μA and an accelerator voltage = 5 kV. The samples were secured onto brass stubs with carbon conductive tape and sputter coated with gold.¹⁶

X-ray absorption spectroscopy

Synchrotron-based soft X-ray Absorption Spectroscopy (sXAS) was carried out using undulator beamline 8.0.1 at the ALS, LBNL. The spherical grating monochromator supplied a linearly polarized photon beam with resolving power up to 6000. Powder samples were pressed into carbon tape and loaded into an ultrahigh vacuum chamber. Experiments were performed at room temperature and checked at the low temperature of 80 K with linear polarization of the incident beam inclined 45° to the sample surface. sXAS spectra were collected by recording the total electron yield (TEY) and total fluorescence yield (TFY) with a probe depth of about 10 nm and 100 nm, respectively. All data shown was normalized to the incident photon flux and monitored by a clean gold mesh. The energy values of the spectra were calibrated by measuring reference samples of Cu and N. The overall resolution of sXAS was better than 0.15 eV.^17,18

Elemental analysis

An Elementar Vario EL III CHNS elemental analyzer was used to determine the CHN content of the samples.⁴

Thermal analysis

Thermogravimetric analysis (TGA) and DTG were carried out on a SII Nano TG-DTG instrument. Analysis was done from 20 °C to 700 °C at a heating rate of 10 °C min⁻¹ under nitrogen (50 mL min⁻¹).⁶

N₂ absorption–desorption

The specific surface area, pore volume and pore diameter of the samples were measured with a liquid N₂ absorption/desorption isotherm on a Micro Active for ASAP 2460 and the samples were degassed at 38 °C prior to measurement.¹⁹

Catalytic activity tests

The experiments were designed to obtain information on the catalytic activity of the samples at atmospheric pressure: nitrogen with a flow rate of 54 to 60 L h⁻¹ and oxygen with a flow rate of 5.4 to 5.5 L h⁻¹ were fed into the reactor as carrier gasses. An approximately 300 mg sample was placed in the center of a quartz reactor tube (Fig. 1) with a mass space-velocity of 252 per h.

Fig. 1 Catalytic test system.

The sample that had been placed for 24 h in the drier was held in a small instrument covered with a quartz cloth and a K-type grounded thermocouple was placed in the center of the tube for temperature measurements. The reactor was heated via a furnace and the temperature was controlled by a thermocouple connected with a temperature control device at room temperature and 60% humidity.^20,21 10% NO with a flow rate of 2.6 to 2.8 mL min⁻¹ was fed into the reactor. The inlet concentration of NO and NO_x was about 220 to 250 ppm and 270 to 290 ppm, respectively. Also, the reactor's exit was connected to a German MRU MGA infrared gas analyzer that was used to analyze the inlet and outlet NO_x concentrations. NO_x concentration was the averaged value of multiple measurements in the span of 5 to 20 min.

The following equation was used to calculate the decomposition and storage rate of NO_x.

η = (NO_xi − NO_xo)/NO_xi × 100%

NO_xi: the inlet NO_x concentration; NO_xo: the outlet NO_x concentration.

Results and discussion

WAXS characterization

Wide angle X-ray scattering results for the as-synthesized samples (Mg/Al, Mg/Al/NH₃, Mg/Al/Cu and Mg/Al/Cu/NH₄⁺ HTs) are shown in Fig. 2. The resulting data is comprised of the characteristic features for HTs. Specifically, the interlayer correlation is present in the form of intense (00L) Bragg peaks that are usually indexed as (003) and (006) for a tri-layered stacking sequence.¹² The diffraction peaks of Mg/Al and Mg/Al/NH₃ HTs were relatively narrow and symmetrical, which indicates that the crystallizations of these two HTs are complete and there is a single crystal phase. Compared with Mg/Al and Mg/Al/NH₃ HTs, the diffraction peaks of Mg/Al/Cu HT became weaker, but was of similar full width at half maxima (Table 1), while that of Mg/Al/Cu/NH₄⁺ HT became weaker and broader, indicating that their crystallization strongly decreased. In addition, a distorted octahedral salt complex forms prior to the double layered HTs because of the special electronic structure of Cu²⁺.²²

Fig. 2 WAXS spectra. (a) Mg/Al HT, (b) Mg/Al/NH₃ HT, (c) Mg/Al/Cu HT, (d) Mg/Al/Cu/NH₄⁺ HT.

Table 1 WAXS data

Samples (HT)	2θ (degrees)	β (degrees)	d (nm)	t (nm)
Mg/Al	11.45	0.42	0.77	15.3
Mg/Al/NH3	11.40	0.33	0.78	19.1
Mg/Al/Cu	11.53	0.31	0.77	20.6
Mg/Al/Cu/NH4^+	11.50	0.52	0.77	12.5

---

The (003) peaks from Mg/Al, Mg/Al/NH₃, Mg/Al/Cu and Mg/Al/Cu/NH₄⁺ HTs are located at 0.0814, 0.0808, 0.0824 and 0.0814 nm, respectively, which corresponds to the 2θ Bragg diffraction angles of 11.45°, 11.40°, 11.53° and 11.50°, respectively. Layer–layer d-spacing was calculated from d = 2π/q and the crystallite size, t, from the Debye–Scherrer formula t = 0.89λ/β cos θ, where λ is the wavelength of the radiation used and β is the line broadening at half the maximum intensity in radians.²³ The results in Table 1 use the (003) peak. The layer–layer spacing of Mg/Al/NH₃ HT is 0.78 nm and those of Mg/Al/Cu and Mg/Al/Cu/NH₄⁺ HTs are 0.77 nm and 0.77 nm, respectively.

The above results suggest that the effect of ammonium and copper ions to the layer–layer spacing is opposite and changes little. The layer–layer spacing of Mg/Al/NH₃ HT is similar with Mg/Al HT, but that of Mg/Al/Cu and Mg/Al/Cu/NH₄ HTs becomes smaller due to the copper ions. According to the ion-exchange theory, cations will be exchanged on the layers and anions will move into the interlayers of HTs. Usually, when magnesium ions are substituted with divalent metal ions with bigger radii, the layer–layer spacing will increase, but copper ions are very special. Just as Ts. Stanimirova mentioned, the Cu²⁺ oxysalt minerals are not generally isostructural with non-copper ion analogues, and as such, they are often considered somewhat of an enigma in mineralogy. Owing to the Jahn–Teller effect, the structures of copper hydroxysalts are described as being built from the edge-shared or corner-shared octahedral in which Cu²⁺ coordination is strongly distorted from the ideal octahedral environment.²⁴ The particle size comparison results indicate that the crystallite size of the Mg/Al/Cu/NH₄⁺ HT decreases to 12.50 nm, which is much smaller than that of Mg/Al/Cu HT, and implies that the introduction of ammonium into Mg/Al/Cu/NH₄⁺ HT greatly affects its structure.

SEM morphology analysis

Representative SEM images obtained for the samples are shown in Fig. 2. Pure Mg/Al HT shows a clear and structured crystalline perfection [Fig. 2(a)] and lamellae crystals. Mg/Al/NH₃ HT [Fig. 2(b)] and Mg/Al/Cu HT [Fig. 2(c)] exhibit layer structures with crystal dispersion. However, because of the nano effect, all of them appear to have partial agglomeration and gather into small balls, and the micrographs of Cu containing samples are of poor quality due to the conducting nature of the samples (charging effect). Mg/Al/Cu/NH₄⁺ HT has obvious crystal structures and the size of the particles becomes smaller. It can be seen clearly that the crystallinity is weakened due to the incorporation of ammonium with copper ions (Fig. 3).

Fig. 3 SEM morphologies (×100 000). (a) Mg/Al HT, (b) Mg/Al/NH₃ HT, (c) Mg/Al/Cu HT, (d) Mg/Al/Cu/NH₄⁺ HT.

sXAS spectroscopic study

sXAS directly probes the 3d unoccupied valence states of Cu in the HTs through the dipole-allowed 2p–3d transition. Both surface-sensitive TEY (∼10 nm probe depth) and bulk-sensitive TFY (∼100 nm probe depth) were collected. Fig. 4(i) shows the Cu-L edge sXAS spectra of Mg/Al/Cu HT (c) and Mg/Al/Cu/NH₄⁺ HT (d). The spin–orbit interaction splits 2p core states into 2p_1/2 and 2p_3/2, which correspond to the well-separated L₂ and L₃ absorption features. In those two samples, the energy position, as well as the overall spectral lineshape, of the L₃ and L₂ peaks suggests that the copper oxidation state in the HTs is Cu²⁺.¹⁶ Introducing NH₃, from the NH₄Cl reaction with NaOH, into the HTs does not lead to any change on the Cu-L spectral lineshape. However, a downward energy shift of almost 0.4 eV could be seen with ammonium intercalation, as shown in Fig. 4(ii). Since the unchanged lineshape indicates that the formal valence of Cu remains the same at Cu²⁺,¹⁸ the obvious energy shift stems from the overall chemical potential drop when NH₃ is introduced into the HT and coordinates to Cu²⁺. This likely delocalizes the Cu-3d states, leading to the drop in chemical potential. The unchanged Cu valence is overall consistent with the WAXS results that show no new crystalline phase (Fig. 2). The energy level of the chemical potential is known to be critical to catalytic behavior.²⁵ The lower chemical potential through incorporating NH₃ generally improves the electron mobility of the system and adjusts the relative energy position of the transition-metal d states versus the Fermi level, which is likely the main factor responsible for the improved catalytic activity.

Fig. 4 (i) Cu L-edge TEY and TFY spectra of HTs, (ii) Cu L₃ edge TEY and TFY spectra of HTs. (c) Mg/Al/Cu HT and (d) Mg/Al/Cu/NH₄⁺ HT.

Fig. 5 shows the N-K edge sXAS spectra of Mg/Al/Cu/NH₄⁺ HT (d) and Mg/Al/NH₃ HT (b). In the N-K edge spectrum, there are three absorption peaks in the 1s → π* region, two prominent peaks (A₁ and A₃) and one small peak (A₂), and a broad feature centered at around 407 eV in the 1s → σ* region.²⁶ As shown in Fig. 5, only the Mg/Al/Cu/NH₄⁺ HTs (d) shows N-K features in the bulk-sensitive TFY spectrum. The very weak N-K features in the TEY spectrum indicate that NH₄OH (NH₃·H₂O) is depleted from the surface after the samples are vigorously pumped in our ultra-high vacuum experimental chamber. There is no absorption feature in the N-K sXAS spectra of the Mg/Al/NH₃ HTs in either TEY or TFY modes (b), suggesting that NH₃ could not be effectively introduced into the HTs without Cu²⁺. These straightforward results suggest the critical role of retaining ammonium by copper ions into HT systems. Cu²⁺ stabilizes the NH₃ incorporation while incorporating ammonium lowers the chemical potential of the system.

Fig. 5 N-K edge TEY and TFY spectra of the HTs. (b) Mg/Al/NH₃ HT, (d) Mg/Al/Cu/NH₄⁺ HT.

CHNS elemental analysis

Table 2 summarizes the results of CHNS elemental analysis. The Mg/Al/Cu/NH₄⁺ HT containing copper ions has 0.39% nitrogen content. On the contrary, almost no nitrogen is present in the Mg/Al/NH₃ HT, Mg/Al HT and Mg/Al/Cu HT. This is coherent with the results obtained from sXAS in which the importance of the copper ions for NH₃ incorporation in Mg/Al/Cu/NH₄⁺ HT was shown. The content of hydrogen and carbon is much higher than that of nitrogen and indicates that the amount of NH₃ incorporated into the Mg/Al/Cu/NH₄⁺ HT is lower due to the easy volatilization of NH₃.

Table 2 Elemental analysis data

Samples (HT)	C (%)	N (%)	H (%)
Mg/Al	3.83	≤0.05	3.99
Mg/Al/NH3	3.95	≤0.05	4.26
Mg/Al/Cu	2.18	≤0.05	2.80
Mg/Al/Cu/NH4^+	2.22	0.39	3.09

---

Thermal analysis

TGA and DTG were carried out under nitrogen atmosphere in the 25–700 °C temperature range. Fig. 6 and Table 3 describe the degradation process and data of the samples. The samples present two stages of weight loss. The first stage, which occurs at a temperature below 200 °C, is associated with the removal of small molecules weakly adsorbed in the interlayer. The next stage of the thermal decomposition is observed over 300 °C.⁶ Mg/Al/Cu/NH₄⁺ HT displays very different degradation curves, the two small peaks in the DTG below 159 °C is attributed to the removal of two kinds of small molecules weakly adsorbed to metal ions. The sXAS and CHNS elemental analysis have proved that nitrogen only exists in the Mg/Al/Cu/NH₄⁺ HT. The two adsorption peaks below 159 °C should be assigned to the release of NH₃ incorporated with copper ions and water intercalated in the interlayers. However, only one degradation peak at about 180 °C is assigned to water adsorbed in the interlayer for each of the other three HTs and the small peak at 45 °C maybe assigned to NH₃ adsorbed on the surface of HT because the samples weren't dried in the ageing box. The mass loss is assigned to the decomposition and removal of hydroxyl groups in the brucite-like layers as well as the OH⁻, CO₃²⁻ and other interlayer anions that decompose at over 200 °C. In the case of the Mg/Al/Cu and Mg/Al/Cu/NH₄⁺ HTs, the adsorption peak appears at about 320 °C lower than that of the other two samples without copper ions, as previously suggested by Xie and coworkers (2008).²⁷

Fig. 6 (i) TGA and (ii) DTG trace. (a) Mg/Al HT, (b) Mg/Al/NH₃ HT, (c) Mg/Al/Cu HT, (d) Mg/Al/Cu/NH₄⁺ HT.

Table 3 Mass loss rates and decomposition temperatures

Samples	The first mass loss stage		The second mass loss stage
	Mass loss (%)	Decomposition temperature (°C)	Mass loss (%)	Decomposition temperature (°C)
Mg/Al HT	15.0	187	46.1	399
Mg/Al/NH3 HT	15.0	191	43.9	379
Mg/Al/Cu HT	11.0	191	25.9	328 378
Mg/Al/Cu/NH4^+ HT	13.2	47 126	38.2	319

---

N₂ adsorption–desorption

The BET specific surface area, BJH adsorption pore distribution, pore volume and pore diameters of the samples were measured with the liquid N₂ adsorption technique as shown in Fig. 7 and Table 4. Inspection of the N₂ adsorption–desorption results indicate that all samples present IUPAC type IV isotherms, confirming that the thermal decomposition of all samples are indicative of mesoporous solids. However, the copper content partially blocked its porous structure, significantly influencing the pores diameter and pore volume. The surface areas are 75.8 and 63.7 m² g⁻¹ for Mg/Al and Mg/Al/NH₃, respectively. The surface areas are 43.6 and 51.5 m² g⁻¹ for Mg/Al/Cu and Mg/Al/Cu/NH₄⁺ HTs, respectively, and there are observable changes in comparison with those of the Mg/Al and Mg/Al/NH₃ HTs, which indicates that a quantity of Cu²⁺ will partially block the porous structure. It is interesting to find a small surface area increase from NH₃ incorporation into copper ions. As can be seen in Table 4, the pores volume and pore diameter of Mg/Al/Cu/NH₄⁺ HT are smaller than those of the other HTs, which agrees with the result from WAXS.

Fig. 7 N₂ adsorption–desorption isotherms. Inset: BJH pore size distributions obtained from the corresponding adsorption isotherms.

Table 4 Adsorption–desorption data

Sample (HT)	BET surface area (m^2 g^−1)	Pore volume (cm^3 g^−1)	Pore diameter (nm)
Mg/Al	75.8	0.59	27.9
Mg/Al/NH3	63.7	0.49	25.2
Mg/Al/Cu	43.6	0.44	32.0
Mg/Al/Cu/NH4^+	51.5	0.27	16.5

---

Catalytic activity test

The aforementioned variation of the CHN elemental analysis, the drop of overall chemical potential of Cu²⁺ and the N-K edge TEY and TFY spectra of Mg/Al/Cu/NH₄⁺ HT from the sXAS data provide direct evidence for the incorporation of NH₃ into the Mg/Al/Cu/NH₄⁺ HT. Otherwise, the catalytic performance of Mg/Al/Cu HT over 300 °C has been mentioned in many papers.^6,9 It has been shown clearly that the decomposition temperature of Mg/Al/Cu/NH₄⁺ HT is 159 °C from the TGA results herein. Thus, NO_x removal performance of different samples at 160 °C will be the focus.

Fig. 8(i) and (ii) clearly shows that the changes of NO_x emission concentration, average decomposition and storage rates with removal time when simulated gas NO_x flows through different HTs, separately. The results of curve (i) indicate the emission concentrations of the four HTs are lower than that of “No catalyst”, and the emission concentration of the Mg/Al/Cu/NH₄⁺ HT is much lower than those of the other three within 450 seconds. In the first 75 seconds, no NO_x was detected due to the storage of NH₃ adsorbed on the HT surface to NO_x, which is coherent with the explanation about the small peak on its DTG. With prolonged NO_x staying with HTs times, the NO_x emission concentration increases and reaches a balance in 20 minutes. The Mg/Al/Cu/NH₄⁺ HT has the same storage effects with Mg/Al and Mg/Al/NH₃ HTs after 10 minutes, which indicates NH₃ was completely released, but it still shows catalytic and decomposition performance to NO_x. It can be seen from graph (ii) that the average decomposition and storage rate of Mg/Al/Cu/NH₄⁺ HT within 5 minutes is 84.56% but that of Mg/Al, Mg/Al/NH₃ and Mg/Al/Cu HTs are 43.11%, 37.11% and 10.01%, respectively. After 20 minutes, the average decomposition and storage rate of Mg/Al/Cu/NH₄⁺ HT decreases to 34.49%, but it is also higher than that of the other three HTs. The results of N₂ absorption and desorption have proved that the specific area, pore volume and pore diameter of Mg/Al/Cu/NH₄⁺ HTs are not as big as that of the Mg/Al or Mg/Al/NH₃ HTs. Otherwise, the higher decomposition and storage performance of the Mg/Al/Cu/NH₄⁺ HT cannot be assigned to its specific area, pore volume and pore diameter, but can be attributed to the incorporated ammonium reacting with NO_x below 160 °C effectively. In contrast, the NO_x decomposition and storage rate of Mg/Al/NH₃ HT displays similar NO_x catalytic performance with Mg/Al HT, and Mg/Al/Cu HT has little NO_x decomposition and storage performance at this temperature, which agrees with the result of the N-K edge sXAS and CHNS elemental analyses.

Fig. 8 (i) NO_x emission concentration with removal time. (ii) Average decomposition and storage rate of NO_x with removal time. (a) Mg/Al HT, (b) Mg/Al/NH₃ HT, (c) Mg/Al/Cu HT, (d) Mg/Al/Cu/NH₄⁺ HT.

Conclusion

The Mg/Al/Cu/NH₄⁺ HT is prepared by co-precipitation, and the NO_x removal catalytic mechanism is discussed by comparing Mg/Al, Mg/Al/NH₃ and Mg/Al/Cu HTs. Based on the sXAS results obtained from the Cu-L edge, the N-K edge and CHNS elemental analysis, we determine that ammonium can only be effectively introduced into HTs with existing Cu²⁺. Ammonium incorporation does not change the valence of Cu ions, but leads to a decrease in the chemical potential. TGA and DTG analysis indicate Mg/Al/Cu/NH₄⁺ HT begins to lose ammonium and water below 159 °C, which agrees with its effective decomposition and storage activity to NO_x at this temperature. From WAXS and SEM characterization, Mg/Al/Cu/NH₄⁺ HT has the typical structure of a hydrotalcite-like material, but its crystallinity decreases due to the Jahn–Teller effect. The N₂ absorption and desorption result indicates that the specific surface area, pore volume and pore diameter are also smaller than that of Mg/Al and Mg/Al/NH₃ HTs. However, it is interesting to find the NO_x decomposition and storage rate of Mg/Al/Cu/NH₄⁺ HT is about 40% higher than that of Mg/Al and Mg/Al/NH₃ HTs, and 70% higher than that of Mg/Al/Cu HT at 160 °C, which confirms that the significant decomposition and storage activity can be attributed to the reaction of ammonium with NO_x.

Acknowledgements

We acknowledge the financial support received from Shanghai municipal education commission with “Twelfth Five” scientific connotation construction project (number: nhky-2012-05), foreign visiting scholar fellowship program (number: B-8938-12-0406). The author gratefully thanks beamline 7.3.3 and 8.0.1 at the Advanced Light Source of Lawrence Berkeley National Lab, supported by the Director of the Office of Science, Office of Basic Energy Sciences, of the U.S. Department of Energy under Contract No. De-AC02-05CH11231.

References

1. J. L. Shi, Chem. Rev., 2013, 113, 2139–2181.
2. J. Perez-Ramirez, J. M. Garcia-Cortes, F. Kapteijin, M. J. Illan-Gomez, A. Ribera, D. L. C. Salinas-Martinez and J. A. Moulijn, Appl. Catal., B, 2000, 25, 191–203.
3. Y. X. Zhang, X. Wang, Z. P. Wang, Q. Li, Z. L. Zhang and L. M. Zhou, Environ. Sci. Technol., 2012, 46, 9614–9619.
4. E. Palomares, A. Uzcategui, C. Franch and A. Corma, Appl. Catal., B, 2013, 142–143, 795–800.
5. P. K. Sahu, P. K. Sahu, S. K. Gupta and D. D. Agarwal, Catal. Sci. Technol., 2013, 3, 1520–1530.
6. C. M. Jinesh, V. Rives, D. Carriazo, A. Churchil and S. K. Antonyraj, Catal. Lett., 2010, 134, 337–342.
7. J. H. Kwak, D. H. Kim, J. Szanyi, S. J. Cho and C. H. F. Peden, Top. Catal., 2012, 55, 70–77.
8. X. Y. Zhang, Q. Shen, C. He, C. Y. Ma, J. Chen, L. D. Li and Z. P. Hao, ACS Catal., 2012, 2, 512–520.
9. Z. P. Wang, Q. Li, L. G. Wang and W. F. Shangguan, Appl. Clay Sci., 2012, 55, 125–130.
10. A. Corma, A. E. Palomares, F. Rey and F. Marquez, J. Catal., 1997, 170, 140–149.
11. H. T. Wang, J. N. Hu, X. F. Bao, Y. Wu, S. X. Xie and Z. H. Li, J. Automot. Saf. Energy, 2011, 1(2), 146–151.
12. X. Wang, R. Deng, S. A. Kulkarni, X. Y. Wang, S. S. Pramana, C. C. Wong, M. Gratzel and S. Uchida, J. Mater. Chem. A, 2013, 1, 4345–4351.
13. H. W. P. Carvalho, S. H. Pulcielli, C. V. Santilli, F. Leroux, F. Meneau and V. Briois, Chem. Mater., 2013, 25, 2855–2867.
14. B. D. Olsen, X. Gu, A. Hexemer, E. Gann and R. A. Segalman, Macromolecules, 2010, 43, 6531–6534.
15. A. Kusoglu, A. Hexemer, R. C. Jiang, C. S. Gittleman and A. Z. Weber, J. Membr. Sci., 2012, 421–422, 283–291.
16. L. Chmielarz, M. Rutkowska, B. Dudek, R. Dziembaj and M. Michalik, J. Therm. Anal. Calorim., 2011, 105, 161–170.
17. Q. G. He, X. F. Yang, R. H. He, A. Bueno-Lopez, H. Miller, X. M. Ren, W. L. Yang and B. E. Koel, J. Power Sources, 2012, 213, 169–179.
18. P. Malet, J. A. Odriozola, F. M. Labajos, V. Rives and M. A. Vlibarri, Nucl. Instrum. Methods Phys. Res., Sect. B, 1995, 97, 16–19.
19. R. Y. Jiang, H. H. Shan, J. L. Zhang and C. H. Yang, Influence of the Copper Content of De-SO_x Catalyst on Performance, ASCE, 2014, ISSN, 0733–9402/04014038, pp. 1–5.
20. C. E. Stere, W. Adress, R. Burch, S. Chansai, A. Goguet, W. G. Graham, F. D. Rosa, V. Palma and C. Hardacre, ACS Catal., 2014, 4, 666–673.
21. Y. X. Zhang, D. S. Liu, M. Meng, Z. Jiang and S. Zhang, Ind. Eng. Chem. Res., 2014, 53, 8416–8425.
22. Z. M. Ni and J. L. Xue, Chem. Res. Chin. Univ., 2013, 34, 503–508.
23. S. Kannan and C. S. Swamy, J. Mater. Sci. Lett., 1992, 11, 1585–1587.
24. T. Stanimirova, S. Dencheva and G. Kirov, Clay Miner., 2013, 48, 21–36.
25. J. K. Nørskov, J. Rossmeisl, A. Logadottir, L. Lindqvist, J. R. Kitchin, T. Bligaard and H. Jonsson, J. Phys. Chem. B, 2004, 108, 17886–17892.
26. H. Niwa, K. Horiba, Y. Harada, M. Oshima, T. Ikeda, K. Terakura, J. I. Ozaki and S. Miyata, J. Power Sources, 2009, 187, 93–97.
27. X. M. Xie, K. Yan, Q. X. Hu, J. L. Song and Z. Z. Wang, Chin. J. Inorg. Chem., 2008, 24(1), 32–36.

---