1 Introduction

With the development of global industrialization, air pollution is an increasing threat to the environment and human health. In particular, sulfur dioxide (SO2) and nitrogen oxide (NOX) are the main air pollutions emitted from fossil fuel consumption, which give rise to acid rain and photochemical smog. Besides, carbon dioxide (CO2) is the main greenhouse gas which causes global warming. Controlling pollutants of coal-fired flue gas becomes one of the keys to protect the atmospheric environment.

Flue gas desulfurization, selective catalytic reduction, and carbon capture and storage have been used to remove SO2, NO, and CO2, respectively. Comparing with individual removal of SO2, NO, and CO2, simultaneous removal has some advantages, such as the equipment and cost would be less in general. The study for simultaneous removal of SO2, NO, and CO2 by adsorption is few. Our previous studies showed that removal of SO2, NO, and CO2 with adsorption method is a promising way (Deng et al. 2012; Yi et al. 2012; Zhou et al. 2012). The literatures show that activated carbon is active for removal of SO2 (Bagreev and Bashkova 2002), NO (Lopez et al. 2007a; Lopez et al. 2007b; Izquierdo 2003), and CO2 (Zhang et al. 2010). In addition, CO2, SO2, and NOX are important industrial raw materials which can be made of chemical fertilizers (Dong et al. 2012). The adsorbed gases by active carbons can be further produced to be chemical fertilizers conveniently. It shows that the modified activated carbon usually improves the adsorption ability. Many kinds of metal oxides had been studied, such as Ce, Cu, Fe, and V (Gao et al. 2011; Li et al. 2010; Sumathi et al. 2010).

Thus, the paper studied the co-adsorption of SO2, NO, and CO2 from simulated fuel gas on metal-modified shell activated carbon (SAC). It included different metal-modified SACs for SO2, NO, and CO2 co-adsorption and effect of different feed gases on adsorption activity of Cu-SAC.

2 Experimental

2.1 Preparation of Sorbents

Coconut shell activated carbon (SAC) was supplied by Sensen Carbon Industry, Fujian, China. Prior to the impregnation, SAC was sieved to 40 ∼ 60 mesh, washed with deionized water to remove dirt and fines, and then dried at 383 K overnight. First, the SAC was treated with metal nitrate of an appropriate concentration by incipient wetness impregnation. Copper nitrate (Cu(NO3)2 · 3H2O), calcium nitrate (Ca(NO3)2 · 4H2O), magnesium nitrate (Mg(NO3)2 · 6H2O), and zinc nitrate (Zn(NO3)2 · 6H2O) were used as the metal precursors in this study. The solution concentrations of metal nitrates were all 0.3 mol/L. Then, the mixtures of SAC and solutions of metal nitrate were being impregnated with ultrasonic for 30 min at 303 K. After that, the samples were washed with appropriate amount distilled water and dried at 383 K for 12 h. Finally, the prepared samples were heat-treated at 573 K for 2 h to form the reduced sorbents. The sorbents were signed as X-SAC, where X was the metal used. The metal content of modified SACs was listed in Table 1.

Table 1 The metal content of different sorbents
Full size table

2.2 Activity Test

The schematic diagram was described elsewhere (Yi et al. 2012). The co-adsorption experiment was carried out in a fixed-bed reactor. The reactor was 18 mm in diameter and 24 cm in length. The feed gas mixture contained 2,000 ppm SO2, 1,000 ppm NO, 10 % CO2, 5 % O2, and balance N2. The feed flow through the reactor was controlled at 400 mL min−1. The gas hourly space velocity was 7,000 h−1. The reaction temperature was preset to 323 K. The inlet and outlet concentration of gases was measured using a fuel gas analyzer (Kane KM9106). The detection limit for SO2, NO, and CO2 is 1 ppm, 1 ppm, and 0.1 %, respectively.

2.3 Characterization of Samples

The Brunauer-Emmett-Teller (BET) specific surface areas and pore size distribution for the samples were determined by N2 adsorption using an Autosorb-1-C instrument. The samples were first outgassed at 573 K for more than 12 h before adsorption isotherms were generated by dosing nitrogen (at 77 K) on the catalysts.

X-ray photoelectron spectroscopy (XPS) was performed on ESCALab220i-XL electron spectrometer from VG Scientific using Al Kα radiation. The power was 300 W, with a pass energy of 50.0 eV and a step size of 0.1 eV. The data were analyzed with Xpspeak 4.1 software.

3 Results and Discussion

3.1 Different Metal-Modified SACs for SO2, NO, and CO2 Co-adsorption

The performance of co-adsorption for SO2, NO, and CO2 on unmodified SAC and metal-modified SACs are shown in Figs. 1, 2, and 3. It shows that supported metals all improve the adsorption of SO2, NO, and CO2. However, it is hard to determine the best sorbent directly because of their different adsorption efficiencies of SO2, NO, and CO2.

Fig. 1
figure 1

Co-adsorption breakthrough curves for SO2 on sorbents

Full size image
Fig. 2
figure 2

Co-adsorption breakthrough curves for NO on sorbents

Full size image
Fig. 3
figure 3

Co-adsorption breakthrough curves for CO2 on sorbents

Full size image

According to Fig. 1, Ca-SAC exhibits the longest breakthrough (100 % removal efficiency) time for SO2. The effect of sorbents on adsorption of SO2 increased from SAC-Zn to SAC-Ca in the following order: Zn-SAC < Cu-SAC < Mg-SAC < Ca-SAC. The longer the sorbent could maintain 100 % removal efficiency, the better the prepared sorbent is. Zn-SAC has the lowest activity for SO2 among the four sorbents. However, Zn-SAC shows the highest activity of adsorption of NO as shown in Fig. 2. Instead, the increasing activity for NO is in the following order: Ca-SAC < Mg-SAC < Cu-SAC < Zn-SAC. In addition, Fig. 3 shows that Cu-SAC obtains the best adsorption of CO2 under the operation condition. The adsorption performance of these adsorbents for SO2, NO, and CO2 are not in agreement.

In order to select the best sorbent under the condition, the adsorption capacities were calculated. Because we mainly studied the co-adsorption interaction of SO2, NO, and CO2, we select the deadline when CO2 reached saturation. C(SO2), C(NO), and C(CO2) represent adsorption capacities of SO2, NO, and CO2 at the time that CO2 reached saturation. The adsorption capacities of SO2, NO, and CO2 were all measured by dynamic adsorption curves. The adsorption capacities of sorbents were summarized in Table 2.

Table 2 Adsorption capacities of sorbents obtained from the co-adsorption curves
Full size table

The C(SO2) of all sorbents is the same (0.357 mmol g−1) as shown in Table 2. The C(NO) of Zn-SAC and the C(CO2) of Cu-SAC is the largest among the four sorbents. In addition, the C(NO) of Cu-SAC is the second largest. Concerning more adsorption capacities of SO2, NO, and CO2 simultaneously, copper seemed to be most favorable. Thus, Cu-SAC was the best sorbent for the co-adsorption of SO2, NO, and CO2.

The porous structure parameters of SAC and Cu-SAC were measured by BET as shown in Table 3. The surface area and pore volume of Cu-SAC become slightly smaller than those of SAC, which may result from that impregnated process giving rise to some pore blocked by the metal particles. It indicates that the preparation method used in this paper allows the copper particles to be located in the internal part of the pores. And supported copper may be formed into active ingredients on the SAC so that Cu-SAC is more active for the co-adsorption of SO2, NO, and CO2.

Table 3 Porous structure parameters of SAC and Cu-SAC
Full size table

3.2 Effect of Different Feed Gases on Adsorption Activity of Cu-SAC

In order to investigate the different effects of SO2, NO, and CO2 adsorption, the experiment of different feed gases was designed. The adsorption capacities were shown in Table 4, and the dynamic adsorption curves were described in Fig. 4a–d.

Table 4 Adsorption capacities of Cu-SAC for SO2, NO, and CO2 in different feed gases
Full size table
Fig. 4
figure 4

Co-adsorption breakthrough curves on Cu-SAC in different feed gases. a Pure component. b Double components. c Triple components. d Simulated flue gas

Full size image

In the pure component experiment, the adsorption performance for SO2 is the highest, while NO and CO2 penetrated the bed quickly. The adsorption capacity of NO and SO2 both increased in the double-component experiment. It is indicated that NO promotes the adsorption of SO2 and vice versa. In the three-component experiment, the curves of SO2 and NO have a little difference with that in double-component experiment. The adsorption capacity of SO2 and NO decreases a little in the three components. It is indicated that CO2 is adverse to the adsorption of SO2 and NO but it was limited. In addition, O2 improved the adsorption of SO2 and NO obviously. Thus, O2 is the key to co-adsorption as its good oxidizing ability. The detail S 2p and N 1s XPS spectral of different sorbents in co-adsorption experiment are shown in Figs. 5a–c and 6a–c, respectively. Comparing with fresh Cu-SAC, there are two new peaks which fit in with N 1s in the experiment with 5 % O2 and there is just one peak in the experiment without O2. There are two new peaks which fit in with S 2 s both in the experiment with O2 or not. It is indicated that sulfur compounds and nitrogen compounds are formed in the Cu-SAC during the adsorption process. O2 is important to adsorption of SO2 and NO, and O2 is more significant to adsorption of NO than SO2. Without sufficient O2, NO may be oxidized to NO2 difficultly and SO2 could be oxidized by less O2 which is from the active carbon.

Fig. 5
figure 5

Detail S 2p XPS spectral of different sorbents. a Fresh Cu-SAC. b Spent Cu-SAC in experiment without O2. c Spent Cu-SAC in experiment with 5 % O2

Full size image
Fig. 6
figure 6

Detail N 1s XPS spectral of different sorbents. a Fresh Cu-SAC. b Spent Cu-SAC in experiment without O2. c Spent Cu-SAC in experiment with 5 % O2

Full size image

For all sorbents, the adsorption time of SO2 was much longer compared with NO and CO2 as shown in Figs. 1, 2, and 3. CO2 component is much larger than SO2 and NO in the flue gas, so CO2 reaches saturation rapidly. The difference may be due to the characteristic nature and the reaction between sorbent and sorbate. The activated carbon adsorbs SO2 more easily than NO, because SO2 has higher boiling point than NO (Zhu et al. 2005a; Tang et al. 2005; Lisovskii et al. 1997). It is known that the higher the boiling point of molecule is, the more Van der Waals force is created. The molecule with lower boiling point has weak intermolecular interactions. Hence, SO2 may replace NO which is physically adsorbed (Tang et al. 2005). Not only SO2 can influence adsorbed NOX physically due to the higher point of SO2 than NO, but also chemically due to the better bond force of SO2 as its stronger dipole moment. The adsorbed NOX and SO2 can interact with each other so as to form intermediate species, which facilitates SO2 and NO oxidized to SO3 and NO2, respectively. The reaction can be shown as follows (Raymundo-Pinero et al. 2001; Zhu et al. 2005b):

$$ {\mathrm{SO}}_2+*\to {{\mathrm{SO}}_2}^{*} $$
(1)
$$ \mathrm{NO}+{\mathrm{O}}_2+*\to {{\mathrm{NO}}_2}^{*} $$
(2)
$$ {\mathrm{O}}_2+*\to {{\mathrm{O}}_2}^{*} $$
(3)
$$ \mathrm{NO}+{{\mathrm{O}}_2}^{*}+{{\mathrm{SO}}_2}^{*}\to \left[\left({\mathrm{NO}}_2\right){\left({\mathrm{SO}}_3\right)}^{*}\right]+* $$
(4)
$$ {{\mathrm{NO}}_2}^{*}+{{\mathrm{SO}}_2}^{*}\to {{\mathrm{SO}}_3}^{*}+\mathrm{NO}+* $$
(5)

4 Conclusion

Several metal nitrates can be loaded on activated carbon by impregnation method. Co-adsorption of SO2, NO, and CO2 can be put into effect by SACs impregnated with metal nitrate. In co-adsorption experiments, SACs impregnated respectively with Ca, Zn, Cu showed good activity for SO2, NO, and CO2. Cu-SAC showed the best adsorption among the four metal-modified SACs. There is an interaction effect between SO2 and NO, and CO2 is reduced to inferior position in the co-adsorption system. In the presence of O2, SO2 and NO can be formed intermediate species, which facilitates SO2 and NO oxidized to SO3 and NO2, respectively. Cu-SAC could be a promising sorbent for co-adsorption of SO2, NO, and CO2.