NOx Removal from Flue Gas Streams by Solid Adsorbents: A

Aug 4, 2015 - Liu , Y.; Bisson , T. M.; Yang , H.; Xu , Z. Fuel Process. Technol. 2010, 91 (10) 1175– 1197 DOI: 10.1016/j.fuproc.2010.04.015. [Cross...
1 downloads 0 Views 6MB Size
Review pubs.acs.org/EF

SOx/NOx Removal from Flue Gas Streams by Solid Adsorbents: A Review of Current Challenges and Future Directions Fateme Rezaei,*,† Ali A. Rownaghi,† Saman Monjezi,† Ryan P. Lively,‡ and Christopher W. Jones*,‡ †

Department of Chemical and Biochemical Engineering, Missouri University of Science and Technology, 1101 North State Street, Rolla, Missouri 65409, United States ‡ School of Chemical & Biomolecular Engineering, Georgia Institute of Technology, 311 Ferst Drive, Atlanta, Georgia 30332, United States ABSTRACT: One of the main challenges in the power and chemical industries is to remove generated toxic or environmentally harmful gases before atmospheric emission. To comply with stringent environmental and pollutant emissions control regulations, coal-fired power plants must be equipped with new technologies that are efficient and less energy-intensive than status quo technologies for flue gas cleanup. While conventional sulfur oxide (SOx) and nitrogen oxide (NOx) removal technologies benefit from their large-scale implementation and maturity, they are quite energy-intensive. In view of this, the development of lowercost, less energy-intensive technologies could offer an advantage. Significant energy and cost savings can potentially be realized by using advanced adsorbent materials. One of the major barriers to the development of such technologies remains the development of materials that are efficient and productive in removing flue gas contaminants. In this review, adsorption-based removal of SOx/ NOx impurities from flue gas is discussed, with a focus on important attributes of the solid adsorbent materials as well as implementation of the materials in conventional and emerging acid gas removal technologies. The requirements for effective adsorbents are noted with respect to their performance, key limitations, and suggested future research directions. The final section includes some key areas for future research and provides a possible roadmap for the development of technologies for the removal of flue gas impurities that are more efficient and cost-effective than status quo approaches.

1. INTRODUCTION AND MOTIVATION

next 30 years, which will keep coal as a significant fuel for power generation for decades to come. SOx and NOx impurities are emitted into the atmosphere from both stationary and mobile combustion sources, including fossil fuel combustion in heating and thermal power plants, petroleum refineries, and on-road transportation vehicles. Flue gas streams from fossil-fuel-fired power plants alone are responsible for 87% of SOx and 67% of NOx emissions.1,2 Among various sulfur and nitrogen oxide species, SO2, NO, and NO2 are considered the most toxic and harmful gases emitted into the atmosphere. These acidic gases are primary sources for atmospheric pollution and are believed to be causing increasingly serious environmental problems, mainly through the formation of acid rain and photochemical smog as well as ozone layer destruction. Furthermore, high concentrations of these undesirable contaminants in air pose serious health threats to human beings by contributing to a broad range of health issues, including respiratory diseases such as asthma, bronchitis, emphysema, and throat inflammation, among others. In particular, these acid gases highly contribute to the formation of secondary organic aerosols (SOAs), which have an impact on the climate as well.3,4 These emissions stem from various sources, including the combustion of fossil fuels in power plants and petroleum refineries as well as in transportation vehicles.

The reliance of the energy sector on fossil fuels for electricity generation is still growing to address rising electricity demands. According to the U.S. Energy Information Administration (EIA),1 coal-fired power generation is expected to increase by an average of 0.2% per year from 2011 through 2040, and coal remains the largest source of electricity despite the growing interest in natural gas as a fuel, as can be seen from Figure 1. This is primarily the case because natural gas prices are projected to increase more rapidly than coal prices over the

Received: June 9, 2015 Revised: August 4, 2015 Published: August 4, 2015

Figure 1. Electricity generation by fuel for the time period of 1990− 2040 (1012 kwh). Reprinted from ref 1. © 2015 American Chemical Society

5467

DOI: 10.1021/acs.energyfuels.5b01286 Energy Fuels 2015, 29, 5467−5486

Review

Energy & Fuels

alkaline chemical reagent such as limestone (CaCO3), lime (CaO), ammonia (NH3), or sodium hydroxide (NaOH) is used as a water suspension or solution to convert SO2 to either liquid or solid waste byproducts. The process can operate at both low temperatures (typically below 150 °C) and high temperatures (700 to 1200 °C). The acidic nature of SO2 promotes the chemical reaction between this gas and basic surfaces of metal oxides or with ammonia. This reaction results in the retention of SO2 in the liquid phase and thus the purification of the gas phase. According to the U.S. Annual Energy Outlook published in 2013,2 SO2 emissions from U.S. electric power plants are projected to drop from 3.3 million tons in 2012 to 1.3 million tons in 2016 (64%) followed by an annual increase of 0.9%/ year from 2016 to 2040, reaching 1.6 million tons, as shown in Figure 2. These reductions are based on the assumption that all coal-fired power plants will be equipped with either FGD or DSI systems with full fabric filters.

The adverse environmental and health effects caused by SOx/NOx emissions have catalyzed research and development on the purification of flue gas streams containing these acid gases over the past two decades. Indeed, various methods and technologies have been proposed/practiced to curtail the emission of SOx and NOx. Such methods are divided into two main categories: (i) nonregenerable and (ii) regenerable. These two approaches are distinguished on the basis of the way that the sorbent (either ab- or ad-) is treated after sorption of the SOx/NOx. Technologies within the first classification include wet flue gas desulfurization (FGD), dry sorbent injection (DSI), and selective noncatalytic reduction (SNCR).5 The second classification includes the regenerable sorbent-based methods, including selective catalytic reduction (SCR). Among them, FGD and SCR are the most widely used technologies for the removal of SOx and NOx, respectively, from flue gas streams. In many respects, the regenerable-based technologies could offer substantial advantages over their nonregenerable counterparts in terms of cost and waste production. The first grouping of technologies is wellestablished and has been practiced on a large scale for decades; however, the second class is still in the early stages of development. Another important aspect of SOx/NOx removal is the way that these impurities influence downstream CO2 capture systems. Although the removal of SOx/NOx impurities alone is of crucial importance from both environmental and health points of view, the significant impact of these undesirable acid gases on CO2 capture systems is another aspect that needs to be carefully considered. This review is organized as follows. The first section focuses primarily on current methods applied to decrease the level of SOx/NOx gases in gas streams. In particular, the challenges associated with current technologies are discussed in detail along with projections about the future of these technologies. The second part provides an in-depth discussion of solid adsorbent materials that have been studied for the removal of SOx and NOx via single- or binary-component gas removal. In addition, multicomponent SOx/NOx/CO2 removal is discussed for each class of materials. The third section discusses mechanisms of interactions of SOx and NOx with solid adsorbents during adsorption by evaluating the effects of temperature, oxygen, and humidity on the reaction pathways and formed species. Finally, key areas for future research are discussed.

Figure 2. Sulfur dioxide emissions from electricity generation for the time period of 1990−2040 (million tons). Reprinted from ref 2.

Figure 3 shows an overview of a typical wet FGD process. In wet scrubbing systems, the flue gas normally passes first through a fly ash removal device, either an electrostatic precipitator or a wet scrubber (not shown here), and then into

2. CURRENT TECHNOLOGIES FOR SOX AND NOX REMOVAL Among various widely used large-scale emission control processes available today, FGD and SCR by NH3 are the most effective methods for removal of SO2 and NOx, respectively. These technology options rely mostly on purification of a single pollutant in flue gas, although other multipollutant removal processes have also gained significant attention recently. Indeed, the simultaneous removal of all flue gas impurities in a single unit has the potential to lower the capital and operating costs of the process. Along with FGD and SCR processes, other well-established technologies such as DSI and SNCR are discussed below as well. 2.1. Wet Flue Gas Desulfurization (FGD). Currently, the most widely used technology for the abatement of SO2 from industrial effluents is wet flue gas desulfurization because of its high efficiency and stability. In this process, a solution of an

Figure 3. Overview of a typical wet FGD process. 5468

DOI: 10.1021/acs.energyfuels.5b01286 Energy Fuels 2015, 29, 5467−5486

Review

Energy & Fuels

Rule (CAIR), as shown in Figure 4.2 This limited change in emissions is primarily a result of stringent clean air regulations

the SO2 spray tower absorber, where it reacts with an alkaline reagent. The solid waste byproducts may be used in gypsum or fertilizer processes, or they may be regenerated. Approximately 80% of conventional FGD systems use lime or limestone as a reagent. There are a number of control variables that dictate SO2 absorber operation. Such parameters include the pH of the slurry, the liquid-to-gas ratio, the SO2 inlet concentration, the allowable SO2 outlet concentration, and the slurry residence time. Of these metrics, the slurry pH is the most important control variable, as it determines the amount of alkaline reagent. The allowable SO2 outlet emissions are based on the maximum outlet level requirement; this is dictated by either environmental regulations or the overall system efficiency. Despite its widespread use and high operational efficiency, this process suffers from many drawbacks, including high capital cost, complexity of equipment, large occupation of land area, large consumption of fresh water, and formation of secondary pollutants. 2.2. Dry Sorbent Injection (DSI). Dry sorbent injection processes based on lime or limestone are generally used for flue gas cleanup in power systems that use low-sulfur coal. The design is similar to that of a wet system except that the lime or limestone is first mixed with water and then enters the dryer. Adsorption of SOx impurities on carbonaceous materials such as activated carbon or coke under dry conditions provides an alternative to the conventional DSI process using lime or limestone. In contrast to the wet method, carbon-based DSI technology offers a series of advantages, including a low environmental footprint (due to the limited production of waste), the capability to recover sulfur, low water consumption (40% less than wet FGD), and lower capital cost. Because of the advantages offered by this technology, there has been an explosion of interest in the improvement of DSI systems by focusing on optimization of the sorbent characteristics as well as the desulfurization and regeneration conditions. 2.3. Selective Catalytic NOx Reduction (SCR). As with SOx, NOx control strategies adopt either prevention or treatment approaches. The former is done by modifying the operating combustion conditions or changing the fuel source so that less NOx is formed, whereas the later approach relates to the removal of NOx that has been formed from flue gas streams. Selective catalytic reduction is the most practiced and wellestablished method and results in the formation of environmentally nonpoisonous reaction products (i.e., nitrogen and water). It has become a technologically important process over the last two decades, with many processes in practice (in particular, the automotive industry relies heavily on this technology for NOx removal). In this method, NO is reduced to N2 and H2O using ammonia, hydrogen, or hydrocarbons as reducing agents over a supported catalyst. Numerous catalyst materials have been evaluated for this process, including zeolites, transition-metal oxides, noble metals, pillared clays, and perovskite-type oxides, among others. However, the catalyst of choice over which the SCR process is performed both industrially and in vehicles is a vanadium oxide (V2O5)supported catalyst at temperatures above 300 °C. According to the U.S. Annual Energy Outlook published in 2013,2 the annual NOx emissions from electricity generation in 2040 are expected to experience little change compared to the 2011 level, ranging between 1.6 and 2.1 million short tons from 2011 to 2040, as opposed to the 47% drop from 2005 to 2011 as a result of the implementation of the Clean Air Interstate

Figure 4. Nitrogen dioxide emissions from electricity generation for the time period of 1990−2040 (million tons). Reprinted from ref 2.

and is based on the assumption that 15.4 GW of coal-fired capacity will be retrofitted with NOx controls from 2011 to 2040. Despite significant progress in the abatement of NOx emissions by SCR, some major issues still remain unsolved, all of which contribute to technological limitations of this approach. For example, the fate of unreacted species involved in the reduction reaction (i.e., ammonia and hydrocarbons), the complex multistep operation, and high cost are among disadvantages of this method. Furthermore, the temperatures required for the reduction reactions are often inconveniently high (above 300 °C). An overview of a typical SCR unit is shown in Figure 5. Although the SCR process operates at relatively high temperatures, many applications exist for which lower reaction temperatures and possibly cheaper catalysts would be preferable, e.g., the removal of nitric oxide after the flue gas desulfurization step. There have been numerous efforts to improve the SCR technology by finding more energy-efficient approaches such as the development of low-temperature catalyst/sorbents for SCR.6 2.4. Selective Noncatalytic NOx Reduction (SNCR). Selective noncatalytic reduction of NOx to nitrogen and water is another well-established technology in which ammonia (NH3) or urea (NH2CONH2) are injected into flue gas to react with NO at temperatures between 800 and 1100 °C according to the following gas-phase reactions: NH 2CONH 2 + H 2O → 2NH3 + CO2

(1)

4NO + 4NH3 + O2 → 4N2 + 6H 2O

(2)

Such technologies were first commercialized in Japan in the late 1970s. NOx emissions can be reduced by 30−50% using this method. A typical SNCR system consists of reagent storage, multilevel reagent-injection equipment, and associated control instrumentation. The SNCR reagent storage and handling systems are similar to those for SCR systems. However, because of higher stoichiometric ratios, ammonia and urea SNCR processes require 3−4 times as much reagent as SCR systems to achieve similar NOx reductions. The formation of nitrous oxide (N2O), which contributes to the greenhouse effect, and 5469

DOI: 10.1021/acs.energyfuels.5b01286 Energy Fuels 2015, 29, 5467−5486

Review

Energy & Fuels

Figure 5. Overview of a typical wet SCR process.

For this molecular sieve, the SO2 adsorption equilibrium constant was found to be 167.8 at 250 °C. The experimental data were then used to estimate the heat of adsorption (ca. −12.40 kcal/mol). In another study performed by the same group,10 the adsorption performance of SO2 on zeolites 4A and 5A over the temperature range of 250−445 °C was investigated, and it was shown that zeolite 4A gives rise to higher SO2 capacity than its 5A counterpart. Gupta et al.11 evaluated the dynamic adsorption of trace SO2 ( NO > N2 on both zeolites. In addition, it was shown that zeolite NaY displays the minimum mass transfer resistance as a result of its large pore size.22 3.1.2. Binary-Component SOx/NOx Removal. Investigating the synergy between the coadsorption of SOx and NOx is another important consideration, as these impurities coexist in flue gas streams and may influence the adsorption of each other. Coadsorption of SO2 and NO on zeolites has rarely been reported in the literature. Recently, by performing dynamic breakthrough experiments on NaX, Yi et al.22 showed that after initial breakthrough the NO concentration descends to a minimum and then gradually ascends, while the breakthrough profile of SO2 is not influenced that much. In another study by the same research group,23 it was shown that SO2 and NO positively influence the adsorption of each other on ionexchanged zeolites. In another study, Deng and co-workers21 showed that zeolite 5A exhibits better adsorption efficiency than zeolite 13X for coadsorption of SO2 and NO flue gas impurities under equilibrium conditions. 3.1.3. Multicomponent SOx/NO x/CO 2 Removal. The presence of SOx/NOx impurity gases greatly influences CO2 capture processes by negatively impacting the lifetime of the adsorbents. Despite the great promise of adsorbent materials in practical CO2 capture applications, their stability in the face of these impurities remains a significant challenge. Indeed, these contaminants have the potential to permanently adsorb within the adsorbent, effectively reducing the available number of active sites and potentially shutting down transport through the

Figure 6. Structure of the cobalt−NO complex in zeolite A, as revealed by single-crystal X-ray diffraction. Reprinted from ref 14. Copyright 2006 American Chemical Society.

40 °C. The authors noted that in the presence of oxygen, zeolite 13X exhibits a large capacity for NOx whereas zeolite Hβ displays outstanding performance for the removal of NO when oxygen is not present in the feed gas. Furthermore, copper-exchanged ZSM-5 was shown to have a high adsorption affinity toward NO2 over a wide temperature range using both dry and humid feeds. However, this zeolite does not adsorb NO appreciably, and it has been hypothesized that for NO to adsorb on this zeolite, it should be first oxidized to NO2.16−19 More recently, in a study conducted by Chang et al.20 it was demonstrated that zeolite H-β modified with calcium oxide is highly effective at capturing NOx: a dramatic increase in adsorption capacity (from 0.1 to 221 μmol/g) compared with pure CaO or other modified zeolites (such as NaY and H-ZSM5) with the same oxide material was observed. The measurements were carried out at 40 °C using a feed gas containing 400 ppm NO, 1−20% O2, and Ar. It was proposed that there are two distinct sites for adsorption of NOx on the modified zeolite

Figure 7. CO2 coadsorption breakthrough curves for SO2, NO, and CO2 on K-NaX zeolite for (a) SO2 (2000 ppm)/NO (1000 ppm)/CO2 (10%)/ balance N2 and (b) SO2 (2000 ppm)/NO (1000 ppm)/CO2 (10%)/O2 (5%)/balance N2. Reprinted from ref 23. Copyright 2013 American Chemical Society. 5471

DOI: 10.1021/acs.energyfuels.5b01286 Energy Fuels 2015, 29, 5467−5486

Review

Energy & Fuels

removal of CO2 and SO2 under both atmospheric- and highpressure conditions while the inlet concentrations of SO2 and O2 were maintained at 2900 ppm and 3.0 vol %, respectively. SO2 was found to impede cyclic CO2 capture because of pore blockage by sulfate products, resulting primarily from direct sulfation during the later stage of each cycle. Moreover, the adsorbents showed similar patterns during cocapture. The results revealed that direct sulfation becomes dominant after completion of an initial fast stage of carbonation, filling larger pores by sulfation from the outside through the formation of a shell, enveloping the adsorbents with this impermeable shell, and thus inhibiting further carbonation and impeding subsequent calcination. For the atmospheric-pressure tests, prehydration improved the cocapture reversibility. Moreover, steam helped cocapture, but increasing the molar Ca/C ratio and increasing the CO2 mole fraction from 80% to 93% had little effect on CO2 and SO2 cocapture. Although changing the total pressure did not change the adsorbent reversibility appreciably, increasing the CO2 partial pressure was helpful. The typical technique to reactivate partially sulfated calcium adsorbents is called dry hydration, in which steam is employed to crack the sulfate outer shell. Partially sulfated limestone can also be reactivated by CO2 via a limestone calcination/ sulfation/carbonation cycle, as shown schematically in Figure 8.30,31

porous network. In addition, they act as a strong competitor with CO2 for adsorption sites, reducing the adsorbent efficacy toward CO2 removal from gas streams. Therefore, conventional adsorption-based CO2 capture processes rely heavily on using a pretreatment stage to remove SOx and NOx, which adds considerably to the overall cost. The higher acidity of SO2 results in its preferential removal from flue gas streams over CO2, while the preferential adsorption of NO2 over CO2 is attributed to its higher polarity and adsorption strength. The evaluation of the effect of flue gas impurities on zeolitebased CO2 adsorbents has been the subject of several studies. Webley and co-workers24 studied the effect of flue gas impurities such as O2, SOx, NOx, and water on the CO2 capture performance of alumina and NaX zeolite by vacuumswing adsorption. They reported a very slow SO2 desorption rate and attributed this behavior to its large capacity and chemical adsorption, whereas NO desorption was shown to occur fairly quickly. In another study conducted by Deng and co-workers,23 the interactive effect for simultaneous removal of SO2, NO, and CO2 from flue gas on ion-exchanged NaX zeolite was studied by carrying out dynamic breakthrough experiments using a feed stream containing the pure components and binary and tertiary mixtures. It was found that K-NaX zeolite is capable of removing the contaminants simultaneously. It was also shown that in the presence of SO2 and NO the CO2 uptake drops dramatically from 0.51 mmol/g (in the case of the pure component) to 0.28 mmol/g (in the case of multiple components) and that adding O2 to the feed gas reduces the capacity even further to 0.13 mmol/g. On the other hand, the presence of CO2 was shown to have a trivial impact on the adsorption of SO2 and NO. The corresponding breakthrough profiles for SO2, NO, and CO2 on K-NaX zeolite are presented in Figure 7. To further understand the mechanism of coadsorption, the authors conducted thermodynamic simulations (i.e., by considering oxidation reactions) and showed that SO2 is oxidized to form SO42− on the surface of the zeolite, while other species were not observed. 3.2. Calcium-Based Adsorbents. Calcium-based adsorbents such as lime, limestone, and dolomite are widely used for the removal of sulfur oxides from flue gases derived from combustion of fossil fuels. These naturally occurring minerals are abundant sources that are available at low cost, making them very attractive for dry-based approaches to sulfur removal. In addition, these materials are very popular in calcium looping cycles for high-temperature removal of CO2 in a two-step carbonation/calcination process (typically operated at 700 °C). However, the irreversible sulfation reaction of CaO adversely impacts the CO2 adsorption capacity and decreases the activity of the adsorbent for CO2.25−27 3.2.1. Single-Component SOx/NOx Removal. Lu and Smirniotis28 studied the behavior of different dopants on the performance of CaO as an adsorbent for capture of CO2 in the presence of SO2. They investigated carbonation and sulfation both separately and simultaneously and showed that SO2 reduced the capability of the adsorbent to capture CO2 because of the competition between the carbonation and sulfation reactions. CaCO3 and CaSO4 were formed upon carbonation and sulfation, respectively. Their results indicated that the carbonation is reversible, while this is not the case with the sulfation. Anthony and co-workers27,29 studied the in situ removal of CO2 by calcium-based adsorbents in the presence of SO2. Seven calcium-based adsorbents were tested for simultaneous

Figure 8. Schematic sequence of limestone calcination, sulfation, and carbonation reactions. Reprinted from ref 31. Copyright 2005 American Chemical Society.

3.2.2. Binary-Component SOx/CO2 Removal. Several research groups have investigated the impact of SO2 on the CO2 adsorption capacity of CaO during postcombustion CO2 capture by calcium looping technology.27,29,31−35 On the basis of their experimental findings, they reported a negative impact of SO2 on the CO2 capacity, as indicated by a significant decrease in capacity during cycling. For example, the simultaneous CO2/SO2 capture characteristics of three limestones in a pilot-scale fluidized-bed reactor were investigated by Ryu et al.,32 who showed that when the adsorbent is exposed to a mixture of CO2 and SO2, some portions of it adsorb CO2 while others adsorb SO2, causing the CO2 capture capacity to decrease with increasing SO2 concentration and number of cycles. As can be seen from Figure 9, for simultaneous CO2/ 5472

DOI: 10.1021/acs.energyfuels.5b01286 Energy Fuels 2015, 29, 5467−5486

Review

Energy & Fuels

Figure 10. Schematic representation of the proposed process to integrate CO2 capture and SO2 retention with adsorbent reactivation. Reprinted from ref 35. Copyright 2007 American Chemical Society.

ash with little or no unreacted CaCO3 and high S/Ca molar ratio, clean flue gas with less than 5% CO2 and “zero” SO2, and a practically pure CO2 stream (>95%). The advantages of this integrated process are that it can eliminate the problem of ash disposal and reduces the amount of solid waste by half. Ridha et al.27 reported a dramatic decrease in CO2 capacity after two or three consecutive cycles for limestone after exposure to a gas mixture containing 15% CO2, 3% O2, and 0.45% SO2 with the balance N2 at 650 °C. This capacity loss is demonstrated in Figure 11. As can be seen in the figure, the Figure 9. CO2 concentration vs time during simultaneous capture of CO2 and SO2 at 700 °C. Reprinted from ref 32. Copyright 2006 American Chemical Society.

SO2 capture, the duration of capture decreased with the number of cycles, indicating rapid deactivation of the adsorbent in the presence of SO2. On the other hand, the SO2 capture increased with both the number of cycles and the SO2 concentration. The total calcium utilization decreased as the number of cycles increased, but the effect of SO2 concentration on the total calcium utilization was shown to be strongly dependent on the sulfation pattern of the limestone. For one limestone (with sulfation of only the outer core), the total calcium utilization decreased with increasing SO2 concentration. However, for the other two limestones (with uniformtype sulfation), the total calcium utilization was almost independent of the SO2 concentration over the range investigated. The results show that SO2 reduces the CO2 capture capacity of limestone and that the sulfation pattern drastically impacts the CO2 capture capacity. Manovic and Anthony35 attributed the loss in CO2 capacity to the irreversible reaction between SO2 and CaO, resulting in the sulfation of calcium-based adsorbents and the formation of CaSO4, which is thermodynamically stable during adsorbent regeneration at temperatures well above 900 °C. While the formation of calcium sulfate results in the loss of active CaO adsorbent in consecutive cycles, the formed CaSO4 covers the adsorbent surface, hence obstructing the adsorption of CO2 molecules on the surface of CaO and causing pore blockage, giving rise to a significant drop in CO2 adsorption capacity. As a result of the efficient performance of CaO-based adsorbents in carbonation and sulfation steps, a new process based on the utilization of these materials for enhanced CO2 capture and SO2 retention was conceptualized, as shown schematically in Figure 10. According to this flow scheme, the final products are

Figure 11. Capture capacities of CaO-based adsorbent pellets (formed from powdered acetified limestone with a binder) at 650 °C in 15% CO2/3% O2/0.45% SO2/balance N2. Reprinted with permission from ref 27. Copyright 2011 Elsevier Ltd.

CO2 capacity decreased from 4.55 mmol/g in cycle 1 to essentially zero in cycle 5, while the SO2 capacity experienced a dramatic increase from 1.56 to 5.47 mmol/g, which clearly justifies the loss in CO2 capacity. The simultaneous removal of CO2 and SO2 from air under both dry and humid conditions was studied by Czyzewski et al.36 using a CaO/MgO−carbon material. They observed that while both contaminants are simultaneously removed from air, humidity favors removal of CO2 and greatly enhances the CO2 uptake (i.e., 7-fold). However, it diminishes retention of SO2 and decreases the SO2 removal by ca. 20%. Therefore, it appears that calcium-based adsorbents are not capable of efficiently removing SOx impurities in the presence of water. In general, very few studies have considered the impact of humidity on the adsorptive behavior of calcium-based adsorbents, which is unfortunate considering the humid nature of flue gas streams. Future work in this area must be conducted 5473

DOI: 10.1021/acs.energyfuels.5b01286 Energy Fuels 2015, 29, 5467−5486

Review

Energy & Fuels

indicate there is synergy in coadsorption of these gases over activated carbon surfaces.44−49 As previously shown by different research groups, activated carbon behaves differently when exposed to pure or mixed gases. Generally, while NOx promotes SO2 adsorption, the activated carbon exhibits greater affinity for SOx than NOx. This is largely due to the fact that SO2 has a higher permanent dipole moment and polarizability than NOx, resulting in stronger dispersion and electrostatic interactions with the surface. In a recent work by Guo et al.,50 the simultaneous adsorption of SO2 and NO on activated carbon at 120 °C was studied by evaluating the impacts of various SO2 and NO concentrations on the adsorption capacity. It was shown that NO enhances the adsorption of SO2 when the NO concentration is relatively high (i.e., >200 ppm), primarily by promoting the chemisorption of SO2 over the surface of the activated carbon. On the contrary, the NO adsorption capacity decreases as the SO2 concentration increases (i.e., >700 ppm), mainly because of blocking of adsorption sites available for the conversion of NO to NO2. The authors also demonstrated that the maximum total adsorption capacity for the gases was obtained at the SO2/ NO ratio of 1.7. At an initial concentration of 1000 ppm, the SO2 adsorption capacity of activated carbon at 120 °C was found to be 0.48 mmol/g, with almost all of the adsorbed SO2 desorbed above 207 °C. This interest in simultaneous adsorptive removal of SOx and NOx stems from the relative ease of retrofitting existing power plants with low-temperature capture systems. If bare activated carbon (or activated carbon fiber) is used as the adsorbent of choice, then the SOx/NOx removal should be performed by a two-step process in which SOx is first captured in the initial stage of the reactor, and then in the following step, NOx removal is performed as a second stage. This technology can be simplified to a single-step process by using hybrid catalyst− adsorbent systems such as composite metal oxide-activated carbon materials.47 In that regard, various metal oxides such as V2O5, CuO, Fe2O3, MnO2, Cr2O3, and CeO2 have been incorporated into activated carbon supports for this separation process. Of the supported catalyst−adsorbent systems studied, V2O5-activated carbon was found to be a promising candidate, showing higher removal efficiency at low temperatures in comparison with the bare carbon support.47,51,52 The metal oxide here acts as a catalyst, and its role is to oxidize NO and/ or reduce it to N2. This hybrid catalyst−adsorbent system shows promising performance in terms of removal efficiency and the ability to capture both SOx and NOx at relatively low temperatures (i.e., the stack temperature). On the basis of the experimental results published by Wey and co-workers,47 the V2O5- and CuO-activated carbons showed better performance, being more effective in catalytic removal of SO2 and NO, with removal efficiencies of 88% and 74%, respectively, compared with 60% for alumina-based hybrid catalysts. The catalytic reactions were performed at 485−547 °C at SO2 and NO concentrations of 1500−2000 and 150−250 ppm, respectively. Liu and co-workers53,54 used other types of supports such as an activated-carbon honeycomb to incorporate metal oxides, and their results indicated that the removal efficiency may be even better than that of the granular-type hybrid materials. In another study,6 the selective catalytic reduction of NO with NH3 was studied using transition-metal oxides (i.e., Fe2O3, Cr2O3, and CuO) loaded into active carbon in the presence and absence of oxygen at reaction temperatures between 140 and 340 °C. The composite materials proved to be efficient catalysts

to properly evaluate the impact of water on the performance of calcium-based adsorbents for SOx/NOx removal. Moreover, a review of the literature found no reports focusing on the use of calcium-based adsorbents for multicomponent adsorption of SOx/NOx/CO2. 3.3. Activated Carbon. Activated carbon is a microcrystalline, non-graphitic form of carbon with very high porosity and surface area, which positions it as an ideal adsorbent candidate for the removal of impurities from air, liquids, and soil. As with metal oxide and zeolite materials, the adsorption of SOx and NOx on carbon-based adsorbents has been extensively studied in the past. This is mainly due to the favorable characteristics of this class of adsorbents, such as high adsorption capacity, fast adsorption kinetics, high hydrophobicity, regenerability, low cost, and thermal and mechanical stability. The textural and surface characteristics of activated carbon significantly influence the capture efficiency. 3.3.1. Single-Component SOx/NOx Removal. It has been shown that SO2 can be oxidized to SO3 over activated carbon, followed by the formation of sulfuric acid (H2SO4) due to subsequent reaction with water. This results in the acidification of the surface, and therefore, activated carbons with a basic surface are preferred for effective removal of SO2.37,38 Lignitebased activated carbon was used by Karatepe et al.39 for adsorptive removal of SO2 at a concentration of 4800 ppm at 25 °C. SO2 adsorption values as high as 5.55 mmol/g were reported, and it was suggested that carboneous adsorbents with higher micropore volume and not specifically surface area could be considered the best candidates for SO2 removal. Moreover, these researchers observed that phenolic and lactone groups present in the structure of activated carbon play an important role in SO2 adsorption. Zhang et al.40 studied NO adsorption at a concentration of 500 ppm over activated carbon at temperatures below 100 °C. Their results indicated that there is almost negligible NO capture when O2 is absent, while the presence of oxygen (up to about 10%) helps oxidize NO and hence improves the NO removal efficiency. Activated carbon acts as a catalyst to convert NO to NO2 in the presence of O2, and then NO2 adsorbs on the carbon surface. On the basis of their findings, the authors suggested that the NO to NO2 conversion is independent of the surface area and is correlated to a greater extent with the narrow micropores. Other studies have also emphasized the importance of O2 for NO adsorption on activated carbons.41,42 This is mainly because NO is readily converted to NO2 in the presence of O2 at relatively low temperatures over activated carbon. Because NO2 can be readily removed by water, the catalytic oxidation of NO to NO2 may be considered as a potentially practical way of removing NOx from flue gas, potentially replacing the conventional ammonia-based SCR technology. The incorporation of silver nanoparticles on biomass-based carbonaceous materials was shown to be helpful in adsorption of NO2, its conversion to NO, and the retention of NO, which can later be further reduced to N2.43 The deposited silver nanoparticles enhance the reduction capability of the carbon surface and increase the surface interaction with NO. Breakthrough tests were conducted at room temperature using dry air with 0.1% NO2 as the feed gas, and a NO2 breakthrough capacity of 0.40 mmol/g was reported. 3.3.2. Binary-Component SOx/NOx Removal. The simultaneous removal of SOx and NOx by carbon-based materials (activated carbon, activated coke, or activated carbon fibers) has recently gained attention, and the outcome of such studies 5474

DOI: 10.1021/acs.energyfuels.5b01286 Energy Fuels 2015, 29, 5467−5486

Review

Energy & Fuels

10% O2, and balance N2, the silica-based metal oxide adsorbent exhibited a high adsorption capacity of 2.0 mmol of SO2/g. Recently, Bandosz and co-workers60 employed copper oxide/ SBA-15 silica composites for NO2 adsorption. Their results indicated that silanol groups (Si−OH) on the surface of silica play an important role in NO2 adsorption and NO retention at room temperature. The same group61 evaluated the adsorption of NO2 on copper-modified activated carbon at ambient temperature. At a SO2 concentration of 1000 ppm in N2 at room temperature, a removal capacity as high as 3.2 mmol of NO2/g was obtained for copper-modified carbon. The authors linked this high capacity to the presence of copper metal and its high dispersion on the surface of the carbon support. In similar studies conducted by the same group,62,63 the interaction of NO2 with SBA-15-supported cerium−zirconium mixed oxides was investigated at room temperature. Compared with unsupported cerium−zirconium mixed oxides, the silicasupported adsorbents showed a significant enhancement of the NO2 adsorption capacity, ranging from 0.3 mmol of NO2/g for bare SBA-15 to 5.0 mmol of NO2/g for supported adsorbents upon exposure to 1000 ppm NO2 in nitrogen. It was also reported that the structure of the adsorbent remained stable after exposure to SO2. The coadsorption of SOx and NOx on metal oxides has not been reported. 3.5. Solid Amine Adsorbents. Supported amine adsorbents, as an emerging class of hybrid organic−inorganic adsorbents, have gained increased attention for flue gas capture because of their inherent high adsorption capacities toward acidic gases, particularly CO2, similar to those of their pure liquid amine counterparts.64 However, these solid-supported amines are susceptible to poisoning by flue gas contaminants such as SOx and NOx impurities, just as the liquid amine media are adversely affected. SOx and NOx have been shown to bind irreversibly to most amine groups.65−70 Typical amine hybrid materials are made by either impregnating a mesoporous support with aminopolymers such as poly(ethylenimine) (PEI) or by covalent tethering of amine groups such as aminosilanes to the surface.64 3.5.1. Single-Component SOx/NOx Removal. Beckman and co-workers studied the adsorption of weak acidic gases such as CO2, SO2, NO2 and NO onto amine-containing polymeric adsorbents and investigated the effect of amine structure on acid gas adsorption and thermal reversibility of the adsorption− desorption reactions.66,67 They found that the thermal reversibility of the capture process decreased in the order CO2 > SOx > NOx. They also showed that although tertiaryamine-functionalized copolymers exhibited rather poor CO2 adsorption capacity, they were found to have a high affinity for Lewis acidic gases stronger than CO2 (e.g., SO2). In a study performed by Chuang and co-workers,71 the SO2 adsorption capacity of amine-grafted SBA-15 silica and its influence on the CO2 adsorption capacity were evaluated, and the results revealed that SO2 adsorbed irreversibly onto the material and blocked the active amine sites toward subsequent CO2 adsorption, thus reducing the CO2 adsorption capacity. Belmabkhout and Sayari72 studied the SO2 desorption performance of triamine-functionalized pore-expanded MCM41 silica by exposing the material to pure SO2 at a pressure of 0.23 mbar. Their results indicated that only 85% of the SO2 desorbed during regeneration at 100 °C under vacuum. 3.5.2. Binary-Component SOx/NOx Removal. Song and coworkers65 probed the adsorption separation performance and the stability of a “molecular basket” adsorbent made of PEI-

for SCR of NO with NH3, especially in the presence of O2, exhibiting high catalytic activity and achieving complete conversion of NO to N2 (at an initial concentration of 960 ppm) with high selectivity (i.e., nearly 100%) for the desired main product, N2. One of the drawbacks of the simultaneous removal of multiple impurities on activated-carbon-supported metal oxides is that the hybrid materials are prone to SOx deactivation at low temperatures and poisoning in the presence of SOx as a result of the formation of sulfate salts and sulfation of the carbon support (although they exhibit high activity for NOx reduction in the low-temperature range). Therefore, the development of carbon-based hybrid materials capable of simultaneous removal of SOx/NOx impurities at low temperatures in a one-step process remains a challenge for future research. Activated carbon adsorbents for multicomponent adsorption of SOx/ NOx/CO2 have not yet been reported. 3.4. Other Metal Oxides. In addition to calcium oxides, the use of other metal oxides in SOx abatement applications has been extensively studied in the past. However, despite promising performance in some cases, their use is limited as a result of the formation of stable salt compounds after exposure to SOx, which causes rapid deactivation that further decreases the SOx removal capacity. In a review published by Mathieu et al.,55 the use of metal oxide adsorbents for SOx removal is discussed in detail by classifying the materials into three categories: single oxides, mixed oxides, and supported oxides. The authors conclude that silica-based adsorbents are promising candidates for removal of SOx impurities, mainly because of their high surface area and the inherent inert characteristics of silica toward SOx. By contrast, only a few studies have focused on the use of metal oxides as NOxremoving solid materials. 3.4.1. Single-Component SOx/NOx Removal. Peterson and co-workers56 investigated commercial zirconium hydroxide, Zr(OH)4, for the removal of SO2 from air at room temperature. Breakthrough tests using a humid feed containing 765 ppm SO2 at 15% relative humidity (RH) gave rise to a removal capacity of 1.30 mmol of SO2/g. In a similar study, Seredych and Bandosz57 reported the use of zirconium hydroxide and its composites (hydrous zirconia with graphite oxide) for SO2 adsorption under both dry and humid conditions. The measurements were performed at room temperature using dry (