Heterogeneous Hydrocarbon Oxidation - American Chemical Society


Heterogeneous Hydrocarbon Oxidation - American Chemical Societyhttps://pubs.acs.org/doi/pdf/10.1021/bk-1996-0638.ch032au...

0 downloads 72 Views 709KB Size

Chapter 32

Photocatalytic Destruction of Automobile Exhaust Emissions 1

Downloaded by UNIV MASSACHUSETTS AMHERST on October 8, 2012 | http://pubs.acs.org Publication Date: August 13, 1996 | doi: 10.1021/bk-1996-0638.ch032

P. D. Kaviratna and C. H. F. Peden

Pacific Northwest National Laboratory, P.O. Box 999, Richland, WA 99352

Hydrocarbons, carbon monoxide, and nitrogen oxides contained in automobile exhaust emissions are among the major atmospheric air pollutants. During the first few minutes of a cold start of the engine, the emission levels of unburned hydrocarbon and CO pollutants are very high due to the inefficiency of the cold engine and the poor activity of the catalyst at lower temperatures. Therefore, it is necessary to provide an alternative approach to deal with this specific problem in order to meet near-term regulatory requirements. Our approach has been to use known photocatalytic reactions obtainable on semiconducting powders such as titanium dioxide. In this paper, we describe our recent studies aimed at the photocatalytic oxidation of unburned hydrocarbons in automobile exhaust emissions. Our results demonstrate the effective destruction of propylene into water and carbon dioxide. The conversion was found to be a strong function of the propylene flow rate. The reaction rate was studied as a function of time, humidity and temperature. The effect of the power of the UV source on conversion is also discussed.

A number of technologies exist for the removal or destruction of dilute levels of organic contaminants in air in enclosed living or working environments, off-gas treating of water treatment facilities, and emission control in process and manufacturing plants. Since most of these common contaminants are oxidizable, a chemical oxidation process is in principle a viable alternative. However, because of the diluted levels of organics, combined with large volumes of air to be treated, it is necessary that the process be extremely fast and energy efficient. 1

Corresponding author

0097-6156/96/0638-0428$15.00/0 © 1996 American Chemical Society In Heterogeneous Hydrocarbon Oxidation; Warren, B., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

Downloaded by UNIV MASSACHUSETTS AMHERST on October 8, 2012 | http://pubs.acs.org Publication Date: August 13, 1996 | doi: 10.1021/bk-1996-0638.ch032

32. KAVIRATNA & PEDEN

Photocatalytic Destruction of Automobile Exhaust 429

Moreover, it would be desirable if the oxygen in the air could be used as the oxidant, so that would not be necessary to employ additional chemicals. Novel processes based on photocatalyzed reactions with semiconducting powders such as titanium dioxide show great promise for rapid efficient destruction of environmental pollutants that have proven difficult or expensive to treat using established remediation methods (1-3). For example, the ability of ultraviolet illuminated titanium dioxide to affect the partial or complete oxidation of gaseous paraffins, olefins and alcohols is well documented (4-6). Most efforts to apply photoassisted reactions to the destruction of environmental contaminants have focused on the purification of aqueous solutions in slurry reactors using relatively intense levels of ultraviolet light (7-15). However, the photodegradation of gaseous organic pollutants has gained attention in recent years (15-22). Ground level ozone, formed by photochemical reactions of nitrogen oxides (NO ) and hydrocarbons in the air, is one of the major air pollutants of concern. Automobiles contribute nearly one-third of the total hydrocarbon emissions. Therefore, reducing hydrocarbon emissions from automobiles has become a center of attention by the government regulators. The problem is particularly severe in heavily populated urban areas. Modern automobiles sold in the US are equipped with computer-controlled three-way catalytic converters to reduce the emissions of N O as well as CO and hydrocarbons. However, both hydrocarbon and carbon monoxide emission during the first few minutes of a cold start of the engine is still much higher than the permitted emission levels of these pollutants. This is a direct result of the inefficiency of the cold engine as well as the poor activity of the catalyst at lower temperatures. Therefore, it is necessary to either improve the activity of the catalyst in automobile catalytic converters or develop new catalysts superior to those currently being used. This work is aimed at assessing the potential of a photocatalytic process for the reduction of unburned hydrocarbons, carbon monoxide and oxides of nitrogen in automobile exhaust emissions (23). At this stage, the emphasis is given to the particular problem of unburned or partially burned hydrocarbon emissions in automobile exhaust during cold-start. An attempt is made to identify some of the factors that control the photocatalytic destruction of these pollutants. In particular, we report here a determination of the reaction rate as a function of gas flow (space velocity), the stability of the rate over time, the effect of impurity water-vapor concentration, and the dependence of the reaction rate on UV flux. x

x

Experimental As shown in Figure 1 the photocatalytic reactor system used for these studies is composed of a mass flow controller (Tylan General), a UV lamp (Aquafine SP-2), a catalytic reactor, and a sampling valve. The catalytic reactor is made out of titania coated quartz capillary tubes assembled in a 3/8" quartz tube and connected to a gas flow system made out of 1/8" stainless steel tubing. The capillaries are coated both inside and outside with a titania film by a sol-gel process to ensure the maximum contact area of the catalyst. After calcination in air, Raman spectroscopy confirmed that the films were anatase Ti02 (24). The test gas contains 300 ppm propylene ( C 3 H 6 ) in dry, carbon dioxide free air. This particular hydrocarbon and its concentration is considered to be prototypical for an automobile exhaust composition. Gas samples were analyzed by using a Hewlett Packard 5890 Series II gas chromatograph equipped with a TCD detector, and Porapak Q and molecular sieve packed columns.

In Heterogeneous Hydrocarbon Oxidation; Warren, B., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

HETEROGENEOUS HYDROCARBON OXIDATION

430

Propylene/Air

Downloaded by UNIV MASSACHUSETTS AMHERST on October 8, 2012 | http://pubs.acs.org Publication Date: August 13, 1996 | doi: 10.1021/bk-1996-0638.ch032

Quartz reactor T i 0 -anatase internally and externally coated capillaries 2

UV lamp

Septum for G C sampling Vent Figure 1. A schematic illustration of the photocatalytic reactor system.

Flow Rate (cc/min)



2





1

4







1

6







1







1

*



'

1

'

8 10 12 Retention Time (min)





1



1

1

14

16

Figure 2. Gas chromatographic traces of the products generated in the photochemical destruction of propylene in the presence of a UV light source.

In Heterogeneous Hydrocarbon Oxidation; Warren, B., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

32. KAVIRATNA & PEDEN

Photocatalytic Destruction of Automobile Exhaust 431

Downloaded by UNIV MASSACHUSETTS AMHERST on October 8, 2012 | http://pubs.acs.org Publication Date: August 13, 1996 | doi: 10.1021/bk-1996-0638.ch032

Results and Discussion Traces from the gas chromatograph (GC) at different propylene/air flow rates are shown in Figure 2. The GC peak around 4 minutes retention time is due to the presence of carbon dioxide, the peak around 10 minutes indicates the presence of water vapor, and the peak around 13 minutes is due to the reactant, propylene. Each GC trace shown in Figure 2 is obtained after an equilibration time of one hour at each flow rate. A blank experiment is shown in the bottom GC trace, and was obtained with a flow rate of 10 cc/min with the UV source turned off. In the absence of UV illumination, no reaction is observed. However, a GC peak corresponding to the product of reaction, carbon dioxide, is present upon irradiation. Concurrently a decrease in the GC peak corresponding to propylene is observed. Importantly, note the absence of any other reaction product from incomplete combustion, such as carbon monoxide or partially oxidized hydrocarbon species. Thus, the overall reaction proceeds according to equation 1: C H 3

6

+ 9/2 0

2

—> 3 C 0 + 3 H 0 2

2

(1)

The appearance of carbon dioxide and the decrease in propylene clearly depend on the flow rate of the propylene/air mixture as illustrated in Figure 3. At the lowest flow rates studied here, greater than 50% of propylene is converted to carbon dioxide and water. However, at higher flow rates, the propylene conversion rate drops dramatically due to the decrease in residence time. Based on the volume of the reactor containing catalyst, we estimate that the highest flow rates studied here (-900 cc/min) correspond to a space velocity of approximately 6800 hr*. We found the photocatalytic propylene oxidation activity to be remarkably stable over time as shown in Figure 4. The data plotted in this figure were obtained with a flow rate of 36 cc/min. During the entire reaction time the propylene conversion rate was steady demonstrating the photostability of the catalyst. Experiments carried out for still longer times (> 24 hours, not shown in Figure 4) showed no change in the reaction rate. In fact, the same set of titania coated capillary tubes have been used for over a year under a variety of reaction conditions without any loss in activity. Another important consideration is the effect of water vapor on the photocatalytic reactivity of titania because of the potential for relatively high concentrations of water in automobile exhaust, particularly in high humidity climates. It has also been reported in the literature (79) that while the presence of water in the reactant gas mixture is critical for the gas-phase photocatalytic destruction of chlorohydrocarbons using titania, high water-vapor concentrations lead to catalyst deactivation. Figure 5 depicts the effect of water vapor concentration on photocatalytic propylene conversion rates. In the figure, the diamond symbols represent the conversion of propylene when the reaction is carried out with the as-received propylene gas mixture. The open squares indicate the activity of titania if this gas mixture is first passed through an isopropanol/dry ice trap designed to remove impurity water. The open circles are for reactions carried out after saturating the propylene gas mixture with water vapor by bubbling the gas through a room temperature water reservoir. It is clear from these data that the presence of water vapor is not critical for the photocatalytic destruction of hydrocarbons, nor does water poison the catalyst. Finally, the effect of the UV light source power on the photocatalytic activity was studied. There are two competing issues of concern here. First, sufficient source power is needed to deliver UV photons to all of the catalyst in the reactor. However, a practical design will ultimately be constrained by the power

In Heterogeneous Hydrocarbon Oxidation; Warren, B., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

Downloaded by UNIV MASSACHUSETTS AMHERST on October 8, 2012 | http://pubs.acs.org Publication Date: August 13, 1996 | doi: 10.1021/bk-1996-0638.ch032

432

HETEROGENEOUS HYDROCARBON OXIDATION

200

400

1000

600

Flow Rate (cc/min) Figure 3. Flow rate dependence of propylene conversion.

40 Flow Rate = 36 cc/min

c jo 30 *u S >

O u

n

n Q

o °

c o

20 e a

£

10

-iJL

J-A.

•JLJ-

J-L.

3 4 Time (Hours) Figure 4. Time dependence of propylene conversion at a flow rate of 36 cc/min.

In Heterogeneous Hydrocarbon Oxidation; Warren, B., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

32. KAVIRATNA & PEDEN

Photocatalytic Destruction of Automobile Exhaust 433

deliverable by the battery. Our current reactor design uses very low (< 100 watts) power and can be scaled up well within the current available power. Of more immediate concern is an accurate assessment of reaction rates (space velocities) that will not be obtained if some of the catalyst is not irradiated. The data plotted in Figure 6 shows a strong dependence of propylene conversion rates on the UV flux at least at low power. However, at higher power levels, the conversion seems to level off with further increases in power. This result may indicate that the reaction rates (space velocities) we measured at the highest power levels used here represent the intrinsic activity of the present catalyst in our reactor. Downloaded by UNIV MASSACHUSETTS AMHERST on October 8, 2012 | http://pubs.acs.org Publication Date: August 13, 1996 | doi: 10.1021/bk-1996-0638.ch032

Conclusions In the present studies, we have demonstrated that a prototypical automobile exhaust hydrocarbon, propylene, can be completely and efficiently combusted to carbon dioxide and water by a photochemical process over a titania catalyst. Importantly, we have also shown that the catalyst does not deactivate over time, and that the propylene conversion rates are insensitive to atmospheric water vapor even at relatively high concentrations. We compared our measured reaction rates (space velocities) to those typically desired in present 3-way catalytic converters. The current catalyst system needs to handle a total gas flow rate of at least 4.4 x 10^ cc/min. Because of the need to minimize both the amount of precious metal used in the three-way catalyst as well as the weight of the converter, space velocities on the order of 20,000-50,000 hr"l with destruction efficiencies of 80-90% or better are desirable for the current systems. To date, we have demonstrated reaction rates that are considerably below these levels. For example, we find that for space velocities of about 1000 hr"l, we obtain propylene conversion rates of 10-20%. However, there are a number of (thus far) unexplored areas of improvement that could dramatically improve these numbers. For example, neither the catalyst geometry (e.g., surface area) or the catalyst composition have been optimized. Recent reports have demonstrated that small additions of a second oxide material to titania leads to greatly enhanced photocatalytic oxidation rates (12-14). Additionally, small amounts of precious metal on titania can lead to significant increases in electron transfer rates and, thus, photoreactivity (14). We are currently exploring these and other potential improvements to the photocatalytic activity of our system. It is important to keep in mind, however, that the "targets" described above for the 3-way catalysts (> 80% destruction at space velocities > 20,000 hr*) are likely to be much higher than necessary for the particular system described here. First, as discussed above, these targets are dictated to some extent by the need to minimize the precious metal content in the catalyst while the titania photocatalyst may not require any precious metal. Thus, a larger reactor may be possible to overcome lower activities. Perhaps even more important to consider is the fact that this approach is designed to deal with the specific problem of cold-start hydrocarbon emissions. The current EPA Federal Test Procedure (FTP) makes an integrated measure of emissions over a lengthy (several minutes) engine cycle from cold-start through several accelerations, decelerations and cruising modes. A large fraction of the hydrocarbon emissions occur during the cold-start period and any reduction during this period can result in significantly more flexibility in the operating envelope (e.g., air/fuel ratio) for the remainder of the FTP test.

In Heterogeneous Hydrocarbon Oxidation; Warren, B., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

434

HETEROGENEOUS HYDROCARBON OXIDATION

0 Bubbled through water 1 | Bubbled through isopropanol/dry ice O 300 ppm propylene/air mixture as received

Downloaded by UNIV MASSACHUSETTS AMHERST on October 8, 2012 | http://pubs.acs.org Publication Date: August 13, 1996 | doi: 10.1021/bk-1996-0638.ch032

0

o

oo oo o

@B

D8

-L

@

d

J-



111

3 4 5 Time (Hours) Figure 5. Effect of humidity on propylene conversion in the photocatalytic destruction process.

20

30 Flux (mW)

40

60

Figure 6. Propylene conversion as a function of the UV lamp source power.

In Heterogeneous Hydrocarbon Oxidation; Warren, B., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

32. KAVIRATNA & PEDEN

Photocatalytic Destruction of Automobile Exhaust 435

Pacific Northwest National Laboratory is a multiprogram national laboratory operated for the U.S. Department of Energy by Battelle Memorial Institute under contract DE -AC06-76RLO 1830.

Downloaded by UNIV MASSACHUSETTS AMHERST on October 8, 2012 | http://pubs.acs.org Publication Date: August 13, 1996 | doi: 10.1021/bk-1996-0638.ch032

Acknowledgments This work was supported by Laboratory Technology Transfer Funds provided by the U. S. Department of Energy's (DOE) Office of Energy Research. One of us (PDK) wishes to thank the Associated Western Universities, Inc., Northwest Division (AWU-NW) for a post-doctoral fellowship. AWU-NW is supported by the grant DE-FG06-89ER-75522 with the U. S. DOE. Literature Cited (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24)

Frank, S.; Bard, A. J. Phys. Chem. 1977, 81, 1484. Ollis, D. Environ. Sci. Tech. 1983,17,628. Gratzel, C.; Jirousek, M.; Gratzel, M. J. Mol. Catal. 1987, 39, 347. McLintock, I.; Ritchie, M. Trans. Faraday Soc. 1965, 61, 1007. Pichat, P.; Hermann, J.; Courbon, H.; Disdier, J.; Mozzanega, M. Can. J. Chem. Eng. 1982, 60, 27. Blake, N.; Griffin, G. J. Phys. Chem. 1988, 92, 5697. Pruden, Α.; Ollis, D. J. Catal. 1983, 60, 404. Ollis, D.; Hsiao, C.; Budiman, L.; Lee, C. J. Catal. 1984, 88, 89. Matthews, R. J. Catal. 1988, 113, 549. Sabate, J.; Anderson, Μ. Α.; Kikkawa, H.; Edwards, M.; Hill, C. G., Jr. J. Catal. 1991, 127, 167. Brezova, V.; Stasko, A. J. Catal. 1994, 147, 156. Papp, J.; Soled, S.; Dwight, K.; Wold, A. Chem. Mater. 1994, 6, 496. Cui, H.; Dwight, K.; Soled, S.; Wold, A. J. Solid State Chem 1995, 115, 187. Hoffmann, M. R.; Martin, S. T.; Choi, W.; Bahnemann, D. W. Chem. Rev. 1995, 95, 69. Augugliaro, V.; Loddo, V.; Palmisano, L.; Schiavello, M. J. Catal. 1995, 153, 32. Formenti, M.; Juillet, F.; Meriaudeau, P.; Teichner, S. J. Chem. Tech. 1971, 1, 680. Dibble, L. Α.; Raupp, G. B. Catal. Lett. 1990, 4, 345. Dibble, L. Α.; Raupp, G. B. Environ. Sci. Technol. 1992, 26, 492. Raupp, G. B.; Junio, C. T. Appl. Surf. Sci. 1993, 72, 321. Peral, J.; Ollis, D. F. J. Catal. 1992, 136, 554. Sampath, S.; Uchida, H.; Yoneyama, H. J. Catal. 1994, 149, 189. Tada, H.; Honda, H. J. Electrochem. Soc. 1995, 142, 3438. Janata, J.; McVay, G.L., Peden, C. H. F.; Exarhos, G. J. patent pending. Exarhos, G. J. unpublished results.

In Heterogeneous Hydrocarbon Oxidation; Warren, B., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1996.