Denitrification Mechanism of NaOH in the Presence of Carbon

Long Xu , Yanan Liu , Yanjun Li , Zhu Lin , Xiaoxun Ma , Yulong Zhang , Morris D. Argyle , Maohong Fan. Applied Catalysis A: General 2014 469, 387-397...
0 downloads 0 Views 203KB Size
Environ. Sci. Technol. 2005, 39, 9665-9668

Denitrification Mechanism of NaOH in the Presence of Carbon TIANJI LI, WATARU MINAMI, AND HEEJOON KIM* Department of Ecological Engineering, Toyohashi University of Technology, Tempaku-cho, Toyohashi, 441-8580, Japan

We tested a mixture of NaOH and carbon and found it to react rapidly with NO gas, thus demonstrating a promising new method of flue gas denitrification (FGD). We determined the reaction products: gas phases were analyzed by GCMS and NOx meter, and solid residues were analyzed by X-ray diffusion (XRD) and chemical analysis. The reaction process was clearly divided into two stages. In the first stage, the reaction proceeded extremely rapidly, with NO almost completely removed. The mechanism describing this stage consists of a series of reactions of NaOH and NO in the presence of carbon. The main conversion of sodium in this stage is proposed to be: NaOHfNaNO3 (and NaNO2)fNa2O2fNa2CO3. The pyrolysis of NaNO3 was examined and carbon was found to have the ability to reduce NOx emission during this process. In the second stage, NO reduction proceeds slowly and stabilizes at a constant value. The mechanism behind this stage is considered to be the reaction of carbon with NO in the presence of Na2CO3 as catalyst. Quantitative study of the reaction system demonstrated that the total amount of denitrified NO is proportional to the amount of carbon and that the denitrified NO in the first stage is proportional to the amount of NaOH.

briquette and NO/O2 mixtures at temperatures between 300 and 325 °C has also been investigated (13). In all of these reactions of carbon with nitric oxide, alkali is used as a catalyst in the reaction system, but the role of the alkali in the reaction has not been thoroughly explored. In the wet denitrification method, aqueous NaOH is used to absorb NO2. The chemical reaction of sodium compounds under high temperature has been studied for years, but the mechanisms are not fully understood (2-7). In our studies, we have found that the alkali hydroxides (NaOH and KOH) themselves react with nitric oxide violently in the presence of carbon at high temperatures (15). In this study, we tested and proposed a reaction mechanism for the reaction of NaOH and C with NO in the 500-900 °C temperature range. We discovered that the mixture of NaOH and carbon (NaOH + C) had very high reactivity with NO gas, suggesting the possibility of developing new denitrification methods based on this reaction mechanism.

Experimental Methods

Introduction Denitrification has been intensively studied for decades in the field of combustion technology. Some denitrification methods have been applied successfully in industry combustors. Low NOx burners with special structures have been designed to reduce NOx generation. Ammonia has been used as a denitrification agent, both in the combustion process and in flue gas denitrification. Recently, some new denitrification methods have been proposed, including dry denitrification by nonthermal plasma (1). Most of these successful methods were developed for large-scale boilers. In developing countries such as China and India, low-grade coal is used as fuel in domestic stoves and in small-capacity industrial boilers. In these countries, the currently available denitrification methods cannot be applied due to financial constraints. Therefore, a cost-effective denitrification technology is urgently needed for use in developing countries. Some researchers have focused their attention on the use of carbon for denitrification. Denitrification of NO using carbon at elevated temperatures (ranges of 350-700 °C (12) and 2447-3537 °C (9)) under different conditions (10, 11) has been reported. In particular, G. Badjai et al. (8) used char and activated carbon at 700 °C to adsorb up to 5000 ppm of NOx. The reaction of potassium-containing bituminous coal * Corresponding author e-mail: [email protected]. 10.1021/es050541u CCC: $30.25 Published on Web 11/11/2005

FIGURE 1. Schematic diagram of horizontally integrated experimental apparatus.

 2005 American Chemical Society

The experimental apparatus we used was a horizontally integrated electrical furnace, consisting of a quartz reactor, an electrical heater, and a temperature controller (Figure 1). It worked by heating the furnace to the predetermined experimental temperature and then placing the sample in the quartz reactor, just outside of the furnace (the quartz reactor is longer than the furnace). Next, the reaction or bulk gas was introduced into the reactor. And finally, the quartz reactor was moved horizontally, setting the sample at the center of the high-temperature furnace and starting the reaction. Exhaust gas was analyzed by a NOx meter and a Q-Mass spectrum. In the reaction of NaOH with NO, NaOH (96% pure) and carbon black (99% pure) were used as reactants. For qualitative experiments, a standard gas of 0.92% of NO with Ar base was used, and the exhaust gas was analyzed continuously by the Q-Mass spectrum system to monitor the components in gas phase. For quantitative experiments, the PC-200 gas analyzer (Horiba, Ltd.) was used to simultaneously measure the concentrations of CO, CO2, NO, NO2, and O2 in exhaust gas. Solid residue was analyzed by X-ray diffusion (XRD) to determine the components in solid phase. In the experiments of pyrolysis of NaNO3, Ar gas was used as bulk gas, and exhaust gas was analyzed by the Q-Mass spectrum. The solid-phase components in the residue were analyzed by XRD. VOL. 39, NO. 24, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9665

FIGURE 2. Time concentration history of NO for different reactions.

Results and Discussion Reaction Phenomenon of NaOH + Carbon with NO Gas. Utilization of active carbon or coal char (as a sorbent or reactant) to reduce nitric oxide has been reported by Badjai (8) and Gupta (12). In this study, we attempted to perform a direct reaction of pure carbon with NO/N2 gas (around 600 ppm) at 900 °C, however the reaction did not proceed under our experimental conditions (Figure 2, the dotted line). This result does not conform to the other researchers’ results (12), probably due to differences in the carbon samples used. Pure carbon black (99%) was used in our experiments, while the other researchers used bituminous or lignite coal char. It is possible that some substances, such as alkali and metal, were contained in their bituminous or lignite coal char and these substances acted as catalysts to promote the reaction. Our results showed that pure carbonswithout NaOHsdoes not react with NO gas effectively at a temperature under 900 °C. We then performed the reaction of NaOH with NO/N2 gas at 900 °C (Figure 2, the dashed line). In the first several minutes, NO concentration decreased to about 440 ppm, but at the conclusion of the reaction the NO concentration returned to the original value of 610 ppm. This result demonstrates that NaOH (alkali) without carbon reacts very poorly with NO. When we allowed the reaction of the mixture of NaOH and carbon (NaOH + C) with 610 ppm NO/N2 gas at 900 °C (Figure 2, the solid line), the NO concentration decreased rapidly (in approximately 2 min) from 610 to 0 ppm where it remained for about 25 min, and then increased rapidly to and stabilized at 470 ppm. This result shows that the reaction process clearly undergoes two stages. In the first stage, the reaction was rapid and NO was removed by 100 percent, and in the second stage the reaction was rather slow, the NO removal rate was constant, and the concentration of NO remained at 470 ppm for a very long period. To our knowledge, the phenomenon of the first stage, in which nitric oxide was removed at an extremely rapid rate, has not been reported by any other researchers. Quality Analysis by Q-Mass Spectrum and PC-200 Gas Analyzer. To elucidate the reaction mechanism of the first stage, NO/Ar gas was used and the exhaust gas was analyzed continuously during the reaction by a Q-Mass spectrum system. All compounds in gas phase that showed any change in concentration during the reaction are shown in Figure 3. Hydrogen was clearly detected. H2 (m/z ) 2) was emitted when NO started to reduce, reached its peak value when NO concentration decreased to 0 ppm, and stopped being emitted when the first reaction stage ended. Thus, H2 was emitted only in the first reaction stage. 9666

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 24, 2005

Nitrogen, N2 (m/z ) 28), was also detected in the first stage. Like H2, N2 emission corresponded with the reduction of NO in the first stage. At the end of the first stage, N2 concentration decreased to a value above baseline, which suggested that N2 continued to be emitted in the second reaction stage. N2 and CO have the same molecular weight of 28. Therefore, we needed to verify that the substance with a molecular weight of 28 that we detected by Q-Mass spectrum was N2 only. To do this, we measured the CO concentration using the PC-200 gas analyzer, which measures concentrations based on IR spectroscopy. Our results indicated that no CO was emitted in the first reaction stage. Interestingly, CO was detected in the second reaction stage. We had detected by Q-Mass spectrum a slight increase of the substance with molecular weight 28, as shown in Figure 3. This indicated the simultaneous emission of CO and N2 in the second stage. One point we had to consider was the presence of the N atom (m/z ) 14), as an increase of N2 concentration should cause a simultaneous increase of N atom concentration. However, as shown in Figure 3, the concentration of N atom actually decreased in the first stage. The concentration of N atom would be affected by both N2 concentration and NO concentration, and the effect of NO would be greater than that of N2 because NO is easier to decompose than N2 when ionized in the Q-Mass spectrum system. In the first reaction stage, as NO concentration was decreased from 9200 to 0 ppm, this would have a greater effect on N atom concentration than the N2 increase, and therefore the resultant concentration of N atom was decreased. XRD Analysis of the Residue. Residue from the reaction system was analyzed by XRD. The main peaks in our XRD resulted from Na2CO3 and Na2CO3‚H2O, although sodium mainly existed in the residue as Na2CO3 (Figure 4). These results indicate that sodium converted ultimately to Na2CO3 after the first reaction stage. Proposed Reaction Mechanism. On the basis of our results following gas-phase analyses (using the Q-Mass spectrum and the PC-200 gas analyzer) and solid-phase analysis (using the XRD), we propose the following reaction mechanism to describe the first stage conversion of sodium to Na2CO3: Carbon

NaOH + NO 98 NaNO2 + 1/2H2 Carbon

(1)

NaOH + 2NO 98 NaNO3 + 1/2H2 + N2

(2)

2NaNO3 f 2NaNO2+ O2

(3)

Carbon

2NaNO2 98 Na2O2 + N2 + O2

(4)

1 O + C f CO 2 2

(5)

O2 + C f CO2

(6)

1 NO + CO f N2 + CO2 2

(7)

Na2O2 + CO f Na2CO3

(8)

1 Na2O2 + CO2 f Na2CO3 + O2 2

(9)

The only possible reactions between NaOH and NO that can generate H2 are reactions 1 and 2, producing NaNO2 and NaNO3, respectively. Reactions 1 and 2 are considered to be present in our reaction system but with different rates. From our proposed reaction mechanism, it is obvious that, in the absence of carbon, the reaction of NaOH with

FIGURE 3. Concentration changes in gas phase during the reaction process.

FIGURE 4. X-ray diffusion of products of NaOH + carbon denitrification reaction. NO could not take place, which is what we showed in Figure 2 (the dashed line). We believe that carbon promotes reactions 1 and 2. In the presence of carbon in the reaction system, NaOH reacts with NO rapidly in the first stage, as shown in Figure 2 (the solid line). Pyrolysis Characteristics of NaNO3. The products NaNO2 and NaNO3 (reactions 1 and 2, respectively) are both unstable and should decompose under our experimental conditions (800-900 °C); sodium nitrite decomposes at 320 °C (14) and sodium nitrate decomposes at about 570 °C, according to our thermogravimetry analysis (TGA) experiment results. Thus, we investigated the pyrolysis characteristics of NaNO2 and NaNO3. NaNO3 first decomposes to NaNO2 and O2 and finally to Na2O2 at high temperatures (14). Since the decomposition of NaNO2 is contained within the decomposition reaction of NaNO3, we focused on determining the pyrolysis characteristics of NaNO3. According to the Chemistry Dictionary (14), Na2O decomposes to Na and Na2O2 at temperatures greater than 400 °C, thus Na2O2 is considered to be more stable than Na2O at our experimental temperatures (800-900 °C). Therefore, Na2O2 is considered to be the decomposition product of NaNO2 and NaNO3. The pyrolysis characteristics of NaNO3 were investigated by Q-Mass spectrum, and the decomposition results of NaNO3 without and with addition of carbon are shown in Figure 5 a and b, respectively. When carbon was not added, the main emission gases detected were O2 (m/z ) 32 and 16), NO (m/z ) 30), and N2 (m/z ) 28 and 14), as shown in Figure 5a. Nitrate ion is

FIGURE 5. Q-Mass spectrum results of NaNO3 decomposition at 800 °C (a) without the addition of carbon, and (b) with the addition of carbon. thought to convert mainly to NO, with some converting to N2. The main reaction is considered to be

2NaNO3 f Na2O2 + 2NO+O2 When carbon was added, the main emitted gases were CO2 (m/z ) 44) and N2 (m/z ) 28), as shown in Figure 5b. NO (m/z ) 30) was also detected, but the amount was greatly decreased. Almost no O2 (m/z ) 32) was detected. Nitrate radical is believed to convert mainly to N2 when carbon is present in the reaction system. O2 is considered to react with carbon, producing CO2, which was clearly detected, as shown in Figure 5b. The global reaction is considered to be Carbon

2NaNO3 + 2C 98 Na2O2 + N2 + 2CO2 The decomposition process of NaNO3 was shortened when carbon was present in the reaction system. In the reaction system without carbon, the decomposition process lasted for about 500 s, as shown in Figure 5a, and for the reaction system with carbon added, the decomposition process lasted for only 50 s, as shown in Figure 5b. Therefore, the NaNO3 decomposition process was accelerated 10 times when carbon was added to the reaction system, with carbon acting as a catalyst and changing the decomposition path, resulting in reduced emission of NO. In the reaction system of NaOH + C with NO, the emissions of CO and CO2 were not detected in the first reaction stage (Figure 3), however in the pyrolysis reaction of NaNO3, CO2 (m/z ) 44) was highly emitted upon addition of carbon (Figure 5b). According to the mechanism discussed VOL. 39, NO. 24, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9667

above, CO2 must react with Na2O2, producing Na2CO3, as described by reaction eq 9. The same reaction has been reported by Schofield Keith et al. (7). The rapid decomposition of NaNO3 (with carbon added) is considered to be the reason not all CO2 reacted with Na2O2. The above discussion of the pyrolysis characteristics of NaNO3 also supports the reaction mechanism described by reaction eqs 3-9 in the first stage. The second stage is considered to have the same reaction mechanism as that reported by Bedjai (8) and Gupta (12), in which carbon reacts with NO with sodium compound as catalyst: NaX

C + NO 98 CO + 1/2N2

(a)

We have tested the reaction of NaOH and C with NO in the 500-900 °C temperature range and inferred a logical reaction mechanism to describe the two stages of the reactions. We believe that we have uncovered a key to a promising new method of flue gas denitrification (FGD).

Literature Cited (1) Mizuno A.; Shimizu K.; Chakrabarti A.; Dascalescu L.; Furuta S. NO removal process using pulsed discharge plasma IEEE Trans. Ind. Appl. 1995, 31, 957-963. (2) McEwan, M. J.; Phillips, L. F. Dissociation energy of NaO2. Trans. Faraday Soc. 1966, 62, 1717-1725. (3) Carabetta, R.; Kaskan, W. E. Oxidation of Sodium, Potassium, and Cesium in Flames. J. Phys. Chem. 1968, 72, 2483-2489. (4) Kaskan, W. E. The reaction of alkali atoms in lean flames. In Tenth Symposium (International) on Combustion; Combustion Institute: Pittsburgh, PA, 1965; pp 41-46.

9668

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 24, 2005

(5) Hynes, A. J.; Steinberg, M.; Schofield, K. The chemical kinetics and thermodynamics of sodium species in oxygen-rich hydrogen flames. J. Chem. Phys. 1984, 80, 2585-2597. (6) Srinivasachar, S.; Helble, J. J.; Ham, D. O.; Domazetis, G. A kinetic description of vapor phase alkali transformations in combustion system. Prog. Energy Combust. Sci. 1990, 16, 303-309. (7) Schofield, K.; Steinberg, M. Sodium/sulfur chemical behavior in fuel-rich and lean flames. J. Phys. Chem. 1992, 96, 715-726. (8) Badjai, G.; Orbach, H. K.; Riensenfeld, F. C. Reaction of NO with activated carbon and hydrogen. Ind. Eng. Chem. 1958, 50, 1165. (9) Lindachers, D.; Burmeister, M.; Roth, P. Perturbation studies of high-temperature C and CH reactions with N2 and NO. In Twenty-Third Symposium (International) on Combustion; Combustion Institute: Pittsburgh, PA, 1990; pp 251-257. (10) Kirkpatrick, M.; Finney, W. C.; Locke, B. R. RVC electrode for gas phase pulsed corona reactors. IEEE Trans. Ind. Appl. 2000, 36 (2), 500-509. (11) Thomas, S. E.; Martin, A. R.; Raybone, D.; Shawcross, J. T.; Ng, K. L.; Beech, P.; Whitehead, J. C. Non thermal plasma after treatment of particulate - theoretical limits and impact on reactor design. CEC and SAE International, No. 01-1926; 2000; pp. 1-13. (12) Gupta, H.; Fan, L.-S. Reduction of Nitric Oxide from Combustion Flue Gas by Subbituminous Coal in the Presence of Oxygen. Ind. Eng. Chem. Res. 2003, 42, 2536-2543. (13) Garcia-Garcia, A.; Illan-Gomez, M. J.; Linares-Solano, A.; SalinasMartinez, C. L. NOx reduction by potassium-containing coal briquettes. Effect of NO2 concentration. Energy Fuels 1999, 13, 499-505. (14) Chemistry Dictionary; Tokyo Kagaku-Dojin: Tokyo, 1994 (Japanese). (15) Li, T. J.; Lu, G. Q.; Kim, H. J. Denitrification mechanism of biobriquette added NaOH in combustion. In SCEJ 66th Annual Meeting; 2001.

Received for review March 19, 2005. Revised manuscript received September 6, 2005. Accepted September 8, 2005. ES050541U