Article pubs.acs.org/est
Combination of Plasma with a Honeycomb-Structured Catalyst for Automobile Exhaust Treatment Woo Seok Kang,* Dae Hoon Lee, Jae-Ok Lee, Min Hur, and Young-Hoon Song Korea Institute of Machinery and Materials (KIMM), 156 Gajungbukno, Yuseong-gu, Daejeon 305-343, Republic of Korea ABSTRACT: To activate a catalyst efficiently at low temperature by plasma for environmental control, we developed a hybrid reactor that combines plasma with a honeycombstructured catalyst in a practical manner. The reactor developed generated stable cold plasma at atmospheric pressure because of the dielectric and conductive nature of the honeycomb catalyst by consuming low amounts of power. In this reactor, the applied voltage and temperature determined the balance between the oxidation and adsorption by the plasma and catalyst. The synergistic reaction of the plasma and catalyst was more effective at low temperatures, resulting in a reduction in a lowered light-off temperature.
■
INTRODUCTION As emerging technical solutions to global environmental and energy issues, plasma and/or catalyst research have been given attention.1 It is noteworthy that, while the catalysts are highly selective and active only at high temperatures, plasma is reactive even at low temperatures, albeit not selective.1−4 Hence, the complementary function of plasma on a catalyst can provide a solution for the low-temperature activation of a catalyst, and previous studies have shown the synergistic effect of using a catalyst−plasma combination in removal of volatile organic compounds (VOCs),5,6 automobile exhaust treatment,1,7,8 carbon dioxide reduction,2,3 and methane reforming4,9,10 on a lab scale. Especially in the automobile industry, the plasma−catalyst combination can suggest a solution for the treatment of exhaust emissions in next-generation diesel engines. The trends in diesel engine development are toward low-temperature combustion (LTC), such as homogeneous charge compression ignition (HCCI) engines. Newly arising critical issues are that low-temperature exhaust gas emissions contain more hydrocarbons (HCs). First, the highly fuel-efficient engine produces lower temperature exhaust gas; this cannot sufficiently activate the exhaust catalyst, which might therefore affect the catalytic decomposition of exhaust gases.11−13 In addition, this can exacerbate the problem of cold starts, during which most emissions are released, while the catalyst is insufficiently hot to be activated. The second issue is that the composition of exhaust gas differs in a LTC-based diesel engine. Although LTC emits less NOx and particulate matter (PM) than previous engines, the emission of HC becomes severe.12,13 As a common configuration, a plasma reactor with beadshaped catalysts, which is called packed-bed discharge,1,4−6 has been widely used because stable plasma can be generated in the void between pellets that are packed in the reaction volume. However, this configuration is difficult to be commercialized or © 2013 American Chemical Society
scaled-up, especially for automobile exhaust treatment or other environmental applications, because the bead-shaped catalysts affect pressure drop in a reactor and are vulnerable to mechanical vibrations of a system. To overcome these problems, recent studies to combine the plasma and catalyst have focused on using industrially proven honeycomb-structured catalysts, such as three-way catalysts (TWCs) or diesel-oxidation catalysts (DOCs), that are already widely used in automobile exhaust treatment. For instance, Kirkpatrick et al. proposed a two-stage reactor that consisted of a cylindrical plasma reactor with a honeycomb DOC located in the downstream of the flow gas that undergoes a plasma reaction.8 Son et al. reported direct discharge in the photocatalytic honeycomb monolith in NOx treatment.7 Hensel et al. demonstrated the direct current−alternating current (DC−AC) dual-power operating concept to generate plasmas inside the honeycomb catalyst or ceramic form.14,15 Despite prior studies demonstrating the feasibility of combining a honeycomb catalyst with plasma generation, scaling up of a reactor still requires excessive applied voltage with high energy consumption. We propose an effective way to generate plasma over a honeycomb-structured catalyst using a dielectric and conductive nature of the catalyst. On the basis of the plasma− catalyst concept developed, this work presents a plasma− catalytic synergistic effect with the feasibility of controlling next-generation diesel automobile exhaust gas. Received: Revised: Accepted: Published: 11358
June 14, 2013 August 14, 2013 August 30, 2013 August 30, 2013 dx.doi.org/10.1021/es402477a | Environ. Sci. Technol. 2013, 47, 11358−11362
Environmental Science & Technology
■
Article
EXPERIMENTAL SECTION Schematic diagram of the plasma−catalyst reactor developed in this study is shown in Figure 1. We placed a platinum-coated
discharge occurs. However, in the reactor that we developed, which consisted of a catalyst between metallic electrodes, the plasma did not transit to a spark and sustained stable plasmas within the air gap between the high-voltage electrode and the honeycomb monolith. The discharge voltage and current shapes in Figure 2 show that most of the discharge current
Figure 1. Schematic diagram of the plasma−catalyst reactor.
honeycomb monolith catalyst (Al2O3 + 0.507% Pt, with a catalyst thickness of 20 mm and diameter of 40 mm) between high-voltage and ground electrodes that are perforated. The spacing between the high-voltage electrode and the catalyst was 2 mm, while the ground electrode was in contact with the bottom of the catalyst. In this configuration, the plasma was produced within the air gap between the high-voltage electrode and the catalyst via moderate levels of applied voltage. For plasma generation, external power was introduced into the reactor at the voltage of 8−12 kV at the frequency of 2 kHz by a high-voltage amplifier (20/20C, TREK, Inc.) synchronized with a function generator. We measured the discharge current and the voltage of the reactor using voltage and current probes equipped with a digital oscilloscope (P6015A and TCP0030 with DPO5054, Tektronix, Inc.). The discharge power was estimated by a Q−V Lissajous plot.3,16 Because the next-generation diesel engine minimizes the emission of NOx and PM, the reduction of HC rather than NO x is more focused on future automobile emission regulations, although interaction of NOx with HC and HC fragments has been a topic of importance.11−13,17−19 We therefore assumed a simplified simulant gas from the diesel automobile exhaust, comprising a mixture of 900 part per million carbon (ppmC) of propylene (C3H6), 5 standard liter per minute (slpm) of nitrogen, and 500 standard cubic centimeters per minute (sccm) of oxygen. The simulant gas was preheated to a desired temperature in a furnace and was subsequently allowed to flow into the reactor through the perforated electrode at the high-voltage side. The gas flowed through the air gap and the catalyst serially; i.e., the flowing gas reacted with the plasma within the air gap first and then reacted with the catalyst. The space velocity was about 1200 h−1. We used a flame ionization analyzer (MEXA-1170HFID, Horiba, Ltd.) to measure the total hydrocarbon (THC) at the rate of one sample per second in ppmC. For the average gas monitoring, Fourier transform infrared (FTIR) spectroscopy (Tensor-27, Bruker Co.) was used. To avoid interference from electric noise during catalyst temperature measurement in the course of plasma operation, using an electrically insulated thermocouple, we monitored only the temperature of the bottom of the catalyst, at which the ground electrode was located.
Figure 2. Current and voltage characteristics of the reactor at 140− 170 °C.
existed during a positive rising phase of the applied voltage (from −250 to −100 μs), exhibiting a discharge current pattern with an active period of discharge similar to a dielectric barrier discharge (DBD). The dielectric barrier in a DBD reactor is known to prevent spark transition by suppressing a discharge voltage increase by accumulated surface charges over the surface.20,21 The measured discharge current pattern supports that the alumina monolith is also believed to play a role as a dielectric material, such as that in the case of DBD in the reactor that we developed. The plasma was initiated at over 8 kV between the two electrodes at over 140 °C, while the plasma was not generated at room temperature. From the breakdown voltage criteria of dried air (∼35 kV cm−1), over 7 kV is applied within the air gap (of 2 mm) and less than 1 kV was consumed within the 20 mm thick catalyst. A high ambient temperature is believed to change the electrical properties of a dielectric material. Because the catalyst presently used is based on a dielectric material, polarization may be the reason for the small potential drop within the catalyst. This assumption is supported by results reported by other researchers in which temperature over the Curie point was assumed to affect the dielectric constant.22 However, the small voltage drop within the catalyst cannot be explained by the polarization of the dielectric material alone. Because the catalyst is not a pure solid dielectric material and is a conductor (platinum)-coated monolith, the conductive nature may also be a contributing factor. This hypothesis is supported by Figure 2. When the ambient temperature was varied from 140 to 170 °C while sustaining the applied voltage at a particular value, discharge current amplitudes increased with an increase in the ambient temperature. This explains that the potential difference within the air gap increased with a decrease in the potential within the catalyst because the catalyst became more conductive at high temperatures. The increase in the gas temperatures is also believed to play a role in enhancing the conductivity of the catalyst. Over 14 kV and at 160 °C, the plasma turned to a spark between the metallic electrodes through the catalyst because the conductive catalyst enabled surface discharge propagation through holes14,15 and the spark-transiting voltage limit decreased with an increase in the ambient temperature. To
■
RESULTS AND DISCUSSION Generally, when high voltage is applied between metallic electrodes at atmospheric pressure, unstable localized spark 11359
dx.doi.org/10.1021/es402477a | Environ. Sci. Technol. 2013, 47, 11358−11362
Environmental Science & Technology
Article
temperature of the catalyst in the other side might have been higher. Destruction and removal efficiency (DRE) varied from 59.2 to 69.2% after 10 min of reaction with HC. At the moment when external voltage was applied at event B in Figure 3a, THC dropped instantaneously from 406 to 360 ppmC (difference Δ of 46 ppmC), after which the THC rapidly increased. Figure 3b describes the characteristics of the variation in THC in detail. The THC finally decreased gradually to 265 ppmC (DRE of 70.6%). When the plasma was in operation, the catalyst temperature increased gradually and reached steady state, when the THC was saturated. When the plasma was turned off at event C, as shown in panels a and c of Figure 3, the THC increased gradually, peaking at 328 ppmC, after which the THC gradually increased further, reaching a higher value, in comparison to the previous peak (328 ppmC). The increase in THC indicates that the catalyst became less active, during which the catalyst temperature decreased gradually. While comparing the gas sampled by the reaction with the catalyst alone and that sampled by the reaction in the presence of the catalyst and plasma (from phase B to C), intense FTIR peaks indicating the presence of N2, O2, and C3H6 appeared. Additional peaks corresponding to the presence of CO2, water vapor, and a negligible presence of NOx were also present, as shown in Figure 4. Because of the oxygen-rich condition, C3H6
avoid catalyst damage by the spark, we operated the plasma under the voltage limit in conjunction with the ambient temperature change, and the maximum voltage was set at below 12 kV, which is within the stable operating window. By changing the applied voltage from 8 to 12 kV, the discharge power varied in the range of 26−60 W (specific energy density of 240−740 J/L). We operated the system with the following procedure. First, the heated N2/O2 was switched on. (A) Then, the HC supply was switched on, after which the (B) plasma was operated. Then, the (C) plasma was switched off, and the (D) HC supply was switched off. Figure 3 shows the temporal variation of THC and the catalyst temperature at various distinct reaction phases.
Figure 4. Comparison of the FTIR spectra of the gas decomposed by the catalyst alone and the influence of the catalyst and plasma.
decomposed to CO2 in the absence of lower order HC byproducts. Despite the gas being composed mostly of nitrogen, less quantities of NOx were obtained at temperatures under 200 °C. While the ambient temperature was varied from 140 to 170 °C, the THC variations are similar; the THC dropped instantaneously when the plasma was generated, and the THC gradually decreased when the plasma was sustained, as shown in Figure 5. The THC values increased and attained previous levels of ppmC when the plasma was switched off. At 140 and 150 °C, when the catalytic reaction was less active, THC varied without any continued desorption and the gradual decrease in THC (and the increase in the temperature) was observed, while the plasma power was maintained. Desorption of THC was noticed only when the temperature was greater than 160 °C. Focusing on the variation of THC (events B and C shown in Figure 4), Figure 6 compares the DRE of three cases, i.e., the value observed under the influence of catalyst alone, the value obtained instantaneously under the influence of the catalyst with the plasma, and the value obtained under the influence of the catalyst and plasma after 10 min of plasma operation. Figure 6 indicates that the developed plasma−catalyst system showed
Figure 3. Temporal variations of the THC and catalyst temperature (12 kV at 2 kHz and 160 °C) in a (a) wide range, (b) during plasma operation, and (c) when the plasma was not operated.
During the transition from phase A to B, as shown in Figure 3a, the catalyst solely reacted with the HC. The preheated simulant gas, which contained 900 ppmC of HC, increases the temperature of the catalyst, and the activated catalyst reacted with HC. While the catalyst began to react with the HC, an exothermic reaction induced an increase in the catalyst temperature. After HC injection for 10 min, the temperature of the catalyst at the bottom side reached 166 °C and the 11360
dx.doi.org/10.1021/es402477a | Environ. Sci. Technol. 2013, 47, 11358−11362
Environmental Science & Technology
Article
explains the abrupt drop in THC without time delay at the instance of the introduction of the plasma. After the rapid reaction caused by the plasma, the catalytic reaction occurred, accompanied by the occurrence of two plasma-related catalyst phenomena, i.e., HC desorption and catalyst temperature change. The HC desorption occurred after the initial THC drop at the moment that the plasma power was switched on (from phase B to C in Figure 3). Although previous researchers have observed plasma-induced desorption from a catalyst,5,24 the mechanism of desorption and the role of plasma is unclear. Some species that originate during plasma generation may have driven the adsorbed propylene on the active site over the catalyst surface. Low-order hydrocarbon fragments,25 which exhibit short lifetimes under the presently employed FTIR detection limits, can replace the adsorbed propylene on the catalyst. Atomic oxygen could also induce the desorption of the adsorbed propylene because atomic oxygen can be easily fixed over a surface and survive for a longer time.26 Another unique phenomenon is the change occurring in the catalyst temperature when the plasma is switched on. The heat originating from the exothermic reaction by active catalytic reaction is one reason.8 The other reason is the ohmic (conduction) and/or dielectric heating of a catalyst placed in a time-varying alternating high voltage. This implies that the electrical characteristics of a catalyst play an important role in generating stable plasmas with enhanced synergistic effects, and future studies are required to evaluate catalyst-related plasma characteristics that depend upon the dielectric or conductive nature of a catalyst. In conclusion, we developed a hybrid reactor that presents an effective combination of plasma and catalyst with an enhanced synergistic effect of low-temperature catalytic activity by plasma using a widely used honeycomb-structured monolith catalyst. The dielectric and conductive nature of the catalyst enabled the plasma to be effectively produced over the catalyst by the application of a moderate voltage with low consuming power. Dependent upon the operation temperature, the catalyst affected the plasma conditions. The generated plasma exhibited typical current patterns of DBDs. The synergistic mechanism between the catalyst and plasma influencing total HC variations is driven by the generated reactive species that affects both the volumetric reaction (oxidation) by plasma and surface reaction (adsorption/desorption) by the catalyst. In this hybrid reaction, the applied voltage and ambient temperature determined the balance of two reactions: volumetric and surface reactions. The plasma−catalyst synergistic reaction was more effective at low temperatures. The hybrid reaction reduced the temperature required to achieve the same level of DRE when compared to the temperature of the reaction under the influence of the catalyst alone. On the basis of this work, we verified that the combined plasma−catalyst technology is feasible to control exhaust emissions from next-generation LTC-based diesel engines. Future research should focus on the scale-up study of the developed reactor concept at a commercial level for controlling actual automobile exhaust gas.
Figure 5. Variations of THC with the change in the temperature from 140 to 170 °C.
Figure 6. Variation of DREs of HCs with the temperature.
a synergistic effect over ambient temperatures, reducing the light-off temperature. The prompt decrease of THC was more pronounced under the low-temperature conditions. At high temperatures, the plasma−catalyst synergistic reaction was weak because the catalyst was sufficiently active without any additional external assistance from the plasma. By selecting an appropriate type of catalyst for a specific exhaust gas composition, the light-off temperature is expected to be decreased further. By simple structure, the plasma−catalyst reactor developed is capable of being practically used in automobile exhaust treatment. In addition, we found that the reactor exhibited a plasma−catalyst synergistic effect by HC variation. Because the reactor that we developed showed typical DBD characteristics that generated electrons at low temperatures with energies less than 9 eV that can enable oxygen dissociation,20 propylene did not decompose directly; oxidation of propylene with the dissociated oxygen may have occurred with higher probability. When external voltage was applied, the plasma dissociated the oxygen with the accelerated electrons and generated atomic oxygen. The density of atomic oxygen dissociated from oxygen molecules typically decreases while generating ozone. However, at high temperatures, the reaction coefficient of ozone generation decreases and synthesized ozone decomposes into O and O2; more atomic oxygen remains, which is a major oxidant species.23 Therefore, it is probable that the flowing gas first came in contact with the plasma, resulting in the oxidation caused by the atomic oxygen generated by the plasma. Subsequently, the catalytic reaction occurred after the plasmaassisted reaction. It is noteworthy that electron-induced reactions occur within a few nanoseconds, which is small when compared to the time scales of chemical reactions or flow-induced reactions that take place over microseconds. This
■
AUTHOR INFORMATION
Corresponding Author
*Telephone: +82-42-8687435. Fax: +82-42-8687284. E-mail:
[email protected]. Notes
The authors declare no competing financial interest. 11361
dx.doi.org/10.1021/es402477a | Environ. Sci. Technol. 2013, 47, 11358−11362
Environmental Science & Technology
■
Article
(19) Lin, H.; Guan, B.; Cheng, Q.; Huang, Z. An investigation on the principal paths to plasma oxidation of propylene and NO. Energy Fuels 2010, 24 (10), 5418−5425. (20) Eliasson, B.; Kogelschatz, U. Nonequilibrium volume plasma chemical processing. IEEE Trans. Plasma Sci. 1991, 19 (6), 1063− 1077. (21) Kang, W. S.; Park, J. M.; Kim, Y.; Hong, S. H. Numerical study on influences of barrier arrangements on dielectric barrier discharge characteristics. IEEE Trans. Plasma Sci. 2003, 31 (4), 504−510. (22) Harling, A. M.; Kim, H.-H.; Futamura, S.; Whitehead, J. C. Temperature dependence of plasma−catalysis using a nonthermal, atmospheric pressure packed bed; the destruction of benzene and toluene. J. Phys. Chem. C 2007, 111 (13), 5090−5095. (23) Peyrous, R.; Pignolet, P.; Held, B. Kinetic simulation of gaseous species created by an electrical discharge in dry or humid oxygen. J. Phys. D: Appl. Phys. 1989, 22 (11), 1658−1667. (24) Kim, H.-H.; Otaga, A. Nonthermal plasma activates catalyst: From current understanding and future prospects. Eur. Phys. J. Appl. Phys. 2011, 55 (1), 13806. (25) Lee, D. H.; Kim, K.-T.; Song, Y.-H.; Kang, W. S.; Jo, S. Mapping plasma chemistry in hydrocarbon fuel processing processes. Plasma Chem. Plasma Proc. 2013, 33 (1), 249−269. (26) Kim, H.-H.; Ogata, A.; Schiorlin, M.; Marotta, E.; Paradisi, C. Oxygen Isotope (18O2) evidence on the role of oxygen in the plasmadriven catalysis of VOC oxidation. Catal. Lett. 2011, 141 (2), 277− 282.
ACKNOWLEDGMENTS This work was financially supported by the Korea Research Council for Industrial Science and Technology (ISTK).
■
REFERENCES
(1) Kim, H.-H. Nonthermal plasma processing for air-pollution control: A historical review, current issues, and future prospects. Plasma Process. Polym. 2004, 1 (2), 91−110. (2) Amouroux, J.; Cavadias, S.; Doubla, A. Carbon dioxide reduction by non-equilibrium electrocatalysis plasma reactor. IOP Conf. Ser.: Mater. Sci. Eng. 2011, 19 (1), 012005. (3) Kraus, M.; Eliasson, B.; Kogelschatz, U.; Wokaun, A. CO2 reforming of methane by the combination of dielectric-barrier discharges and catalysis. Phys. Chem. Chem. Phys. 2001, 3 (3), 294− 300. (4) Hammer, Th.; Kappes, Th.; Baldauf, M. Plasma catalytic hybrid processes: Gas discharge initiation and plasma activation of catalytic processes. Catal. Today 2004, 89 (1−2), 5−14. (5) Song, Y.-H.; Kim, S.-J.; Choi, K.-I.; Yamamoto, T. Effects of adsorption and temperature on a nonthermal plasma process for removing VOCs. J. Electrostat. 2002, 55 (2), 189−201. (6) Chen, H. L.; Lee, H. M.; Chen, S. H.; Chang, M. B.; Yu, S. J.; Li, S. N. Removal of volatile organic compounds by single-stage and twostage plasma catalysis systems: A review of the performance enhancement mechanisms, current status, and suitable applications. Environ. Sci. Technol. 2009, 43 (7), 2216−2227. (7) Son, G. S.; Yun, S. W.; Song, J. W.; Lee, K. Y. Application of photocatalyst-plasma-honeycomb system for cold start emission of gasoline vehicles. SAE [Tech. Pap.] 2002, DOI: 10.4271/2002-012706. (8) Kirkpatrick, M. J.; Odic, E.; Leininger, J.-P.; Blanchard, G.; Rousseau, S.; Glipa, X. Plasma assisted heterogeneous catalytic oxidation of carbon monoxide and unburned hydrocarbons: Laboratory-scale investigations. Appl. Catal., B 2011, 106 (1), 160− 166. (9) Rico, V. J.; Huseo, J. L.; Cotrino, J.; Gonzales-Elipe, A. R. Evaluation of different dielectric barrier discharge configurations as an alternative technology for green C1 chemistry in the carbon dioxide reforming of methane and the direct decomposition of methanol. J. Phys. Chem. A 2010, 114 (11), 4009−4016. (10) Pham, M. H.; Goujard, V.; Tatibouet, J. M.; Batiot-Dupeyrat, C. Activation of methane and carbon dioxide in a dielectric-barrier discharge-plasma reactor to produce hydrocarbonsInfluences of La2O3/γ-Al2O3 catalyst. Catal. Today 2011, 171 (1), 67−71. (11) Manley, D. K.; McIlroy, A.; Taatjes, C. A. Research needs for future internal combustion engines. Phys. Today 2008, 61 (11), 47−52. (12) Flowers, D. L.; Aceves, S. M.; Martinez-Frias, J.; Dibble, R. W. Prediction of carbon monoxide and hydrocarbon emissions in isooctane HCCI engine combustion using multizone simulations. Proc. Combust. Inst. 2002, 29 (1), 687−694. (13) Juttu, S.; Thipse, S.; Marathe, N. V.; Gajendra Babu, M. K. Homogeneous charge compression ignition (HCCI): A new concept for near zero NOx and particulate matter from diesel engine combustion. SAE [Tech. Pap.] 2007, DOI: 10.4271/2007-26-020. (14) Hensel, K.; Katsura, S.; Mizuno, A. DC microdischarges inside porous ceramics. IEEE Trans. Plasma Sci. 2005, 33 (2), 574−575. (15) Hensel, K. Microdischarges in ceramic foams and honeycombs. Eur. Phys. J. D 2009, 54 (2), 141−148. (16) Manley, T. C. The electric characteristics of the ozonator discharge. J. Electrochem. Soc. 1943, 84 (1), 83−96. (17) Filimonova, E. A; Kim, Y.; Hong, S. H.; Song, Y.-H. Multiparametric investigation on NOx removal from simulated diesel exhaust with hydrocarbons by pulsed corona discharge. J. Phys. D: Appl. Phys. 2002, 35 (21), 2795−2807. (18) Dorai, R.; Kushner, M. J. Consequences of unburned hydrocarbons on microstreamer dynamics and chemistry during plasma remediation of NOx using dielectric barrier discharges. J. Phys. D: Appl. Phys. 2003, 36 (9), 1075−1083. 11362
dx.doi.org/10.1021/es402477a | Environ. Sci. Technol. 2013, 47, 11358−11362
