As used herein, the phrase and similar phrases to"without external addition of any gas phase additive"refers to the absence of any non-exhaust input.

As used herein, the term"pore size"is the network ring opening size in the absence of a charge compensating cation or a"bare"material.

As used herein, the term"free aperture"is the size of an opening through the pore after a cation has been inserted into the pore of the bare material.

BACKGROUND OF THE INVENTION Combustion exhaust theoretically should be simply carbon dioxide and water vapor. However, because most fuels are complex compounds, and because the source of oxidant is the atmosphere with about 21 vol. % nitrogen, combustion exhaust generally contains unburned hydrocarbons and one or more versions of nitrogen oxide (NOx) converted from the nitrogen. Even a methane or hydrogen fuel source can have unburned fuel and produce nitrogen oxide.

There has been much effort expended to obtain both the greatest possible energy from a fuel and achieve a clean exhaust free of hydrocarbons and nitrogen oxide. One approach toward achieving clean exhaust has been plasma assisted catalytic reduction operated downstream of the combustion chamber.

An example of this approach is U. S. patent 5, 711, 147 to Vogtlin et al. wherein a non-thermal plasma gas is combined with y-AI203 catalyst and a

stream of added hydrocarbon to achieve an 80% reduction of NOX. A disadvantage of this method is the added hydrocarbon defeating the goal of maximizing energy extraction from the fuel.

The U. S. patent 5, 715, 677 to Wallman et al. achieves an unspecified reduction of nitrogen oxide (NOx) from diesel exhaust by collecting particulate and NOx onto an adsorbent bed with bead size from 5-10 mm then exposing the adsorbent bed to plasma. Materials for the adsorbent bed included metal oxides, dolomite, zeolites of ZSM-5, Cu-ZSM-5, and H-mordenite, and perovskites. A disadvantage of this method is the semi-batch or non-continuous mode of operation requiring separate adsorption and plasma exposure steps.

The U. S. patent 5, 746, 984 to Hoard is directed to treating cold-start emissions by adsorbing the cold start emissions on a storage device then desorbing them as the exhaust temperature increases followed by destroying the desorbed cold start emissions in a non-thermal plasma. A honeycomb monolith is prepared to receive a coating of an adsorbing material as the storage device.

Adsorbing materials are recited as oxides of copper, barium and lanthanum, and CuZSM-5 zeolite. Hoard reports that operation of the plasma reactor consumes no more than about 2% of engine energy consumption.

However, it has been shown in Analysis of Plasma-Catalysis for Diesel NOX Remediation, J Hoard, ML Balmer, Diesel Engine Emissions Reduction Workshop Proceedings, 1998 that in the case of Cu-ZSM-5 placed either in the plasma or downstream of (after) the plasma, converts the gas phase species back to NOx, an undesirable product. Thus, Hoard's report of 2% consumption is not for complete conversion of NOx to N2.

The U. S. patent 5, 458, 748 to Breault et al. show NOx conversion as a function of applied voltage for the plasma, with conversion up to about 98%.

However, no catalyst other than plasma is used and the products of conversion include HN03 which is corrosive unless removed. In addition it has been shown by BM Pentrante, WJ Pitz, MC Hsaio, BT Merritt, and GE Vogtlin, in the Proceedings for the 1997 Diesel Engine Emissions Reduction Workshop that in the absence of a catalyst, the plasma chemistry primarily converts NO to N02

with some NOx disappearing to unmeasured products NOx species converted in the plasma are converted back to NO.

U. S. patent 5, 609, 736 to Yamamoto describes a vessel formed of opposing conductive plates and opposing dielectric plates between the opposing conductive plates. The volume of the vessel is filled with beads having a catalytic coating thereon. Catalytic coatings specified are oxidation catalysts of noble metals and metallic oxides. Removal of carbon tetrachloride is reported.

Nitrogen oxide conversion is mentioned but not quantitatively reported.

UK patent application GB 2, 274, 412 A describes a chamber filled with pellets held between screens that act as electrodes and showing removal of toluene from a gas stream. The pellets are Pb or Ba niobate, titanate or zirconate. Nitrogen oxide removal is not quantified.

Japanese Kokai JP 6-269635 describes a plasma exhaust gas processing apparatus and absorbent apparatus connected in order. No catalyst is used.

The conversion product N02 is absorbed in the absorbent apparatus. A disadvantage of this technique is the need to regenerate the absorbent apparatus.

The German Offenlegungsschrift DE 195 10 804 A1 discusses combining an exhaust gas stream with a reduction agent that is made into a plasma and contacted with a catalyst. The reduction agent is a nitrogen compound, for example nitrogen, ammodia, hydrazine or cyanuric acid or is aliphatic or olefin hydrocarbon or hydrogen. The catalyst may be zeolites doped with elements of the platinum group, copper group, or iron group, oxides doped with elements of the platinum-, aluminum-, titanium-or lanthanide groups or their mixtures, mixed oxides of wolfram, chrome, or vanadium. NOx reduction is not quantified. A disadvantage of this process is the addition of a reduction agent.

The paper PLASMA-ASSISTED HETEROGENOUS CATALYSIS FOR NOX REDUCTION IN LEAN-BURN ENGINE EXHAUST, BM Penetrante, MC Hsiao, BT Merritt, GE Vogtlin, Proceedings of the 1997 Diesel Engine Emissions Reduction Workshop discusses the advantage of using plasma in combination with"a new class of catalysts that are potentially more durable, more active,

more selective and more sulfur tolerant". Because the inventors of 5, 711, 147 (discussed above) are included as authors of this paper, and the paper was published after the patent application was filed, it is hypothesized that the"new class of catalysts"includes gamma alumina mentioned in the patent and possibly other catalyst materials mentioned in the literature. FIG. 4 (b) of the paper shows maximum conversion of NOx to N2 of 77% at about 25 J/L energy density in real diesel exhaust. It is assumed that the added hydrocarbon mentioned in the patent is also used to obtain the data in the paper although an added hydrocarbon is not specifically mentioned in the paper.

The abstract of the paper CATALYST ASSISTE RECTION BY USING NON-THERMAL PLASMA ON NITRIC OXIDES REMOVAL, K Shimizu, T Oda, 1997, identifies Na-ZSM-5 as an effective catalyst in combination with non- thermal plasma for NOx removal. FIG. 4 of the paper shows a maximum removal rate of less than 55% for input power up to 30 W.

Japanese patent application Hei 6-106025, K Isogai et al. discusses a plasma reactor with catalyst material for decomposition of NOx. Catalyst material is metal oxide, and zeolites of H-ZSM-5, H-Y, H-Mordenite, Na-ZSM-5, and Cu-ZSM-5. Decomposition rate is not shown. The ineffectiveness of Cu- ZSM-5 because of reconversion to NO was mentioned above.

Japanese patent application Hei 6-15143 K Isogai discusses a plasma reactor with a photocatalytic dielectric material. The photocatalytic dielectric material may be Ti02, ZnO, SrTiO2, and ZnS. Nitrogen oxide destruction is not quantified.

As may be seen from the review of prior art literature provided herein, the energy efficiency of NOx conversion is often not reported and when it is reported, it is difficult to compare because of different bases of quantification. The destruction rate of NOX decreases logarithmically with energy density as shown in Tonkyn (Vehicle Exhaust Treatment Using Electrical Discharge methods," Tonkyn, Barlow, Balmer, Orlando, Hoard, and Goulette, SAE Paper 971716, May 1997), specifically

[NO] = [N0] o*e' [NOx] [NOx] f + ([NOx] O- [NOx] f) * e~Eß where p (or beta parameter) is the first order decay parameter in Joules/standard liter, [NOX] o is the initial NOx concentration and [NOx] f is the final NOx concentration. The beta parameter is therefore a quantitative measure of energy consumption of the conversion of NOx to N2. In general, a low beta value indicates that the energy efficiency is high, or alternatively, that the amount of energy expended to achieve the maximum NOx destruction is relatively low.

The overall NOx destruction efficiency is affected by both the type of plasma reactor and the type of catalyst. Other measures of conversion efficiency include percent of total fuel consumption, and percent destruction for a given power (Watts).

Although much attention and work has been done to clean up combustion gas exhaust streams, there is still a need in the art for an apparatus and method that is capable of converting NOx to N2 in a real or actual exhaust stream without any gas phase additive and with greater energy efficiency.

SUMMARY OF THE INVENTION The present invention is based upon the discovery that microporous materials having a free aperture greater than 3 angstroms in combination with a plasma are capable of catalyzing conversion of NOx to N2 in a real or actual exhaust stream without any gas phase additive.

It is an object of the present invention to provide a method of NOx conversion to N2 using a plasma in the presence of a microporous catalyst.

It is a particular advantage of the present invention that the optimum performance is achieved at temperatures and gas compositions typically found in the exhaust of light duty vehicles.

The subject matter of the present invention is particularly pointed out and distinctly claimed in the concluding portion of this specification. However, both

the organization and method of operation, together with further advantages and objects thereof, may best be understood by reference to the following description taken in connection with accompanying drawings wherein like reference characters refer to like elements.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a cross section of a single stage apparatus of the present invention.

FIG. 2A is a cross section of a two-stage apparatus of the present invention.

FIG. 2B is a cutaway of a tube array reactor encased in a metallic housing.

FIG. 2C is a cutaway of an alternate tube array reactor.

FIG. 3 is a single stage NOx conversion graph for Z-Y-Na.

FIG. 4 is a single stage NOX conversion graph for Z-Y-K.

FIG. 5 is a single stage NOX conversion graph for Z-Y-Cs.

FIG. 6 is a single stage NOX conversion graph for Z-Y-Ga.

FIG. 7 is a single stage NOX conversion graph for Z-Y-Co.

FIG. 8 is a single stage NOX conversion graph for Z-Y-Ag.

FIG. 9 is a single stage NOX conversion graph for Z-Y-Ni.

FIG. 10 is a single stage NOx conversion graph for Z-X-Na.

FIG. 11 is a two-stage NOX conversion graph for Z-Y-Na.

FIG. 12 is a double single stage NOx conversion graph for Z-Y-Na.

FIG. 13 is a single stage NOX conversion graph for LTA-K.

FIG. 14 is a single stage NOx conversion graph for LTA-Na.

FIG. 15 is a single stage NOx conversion graph for LTA-Ca.

FIG. 16 is a single stage NOx conversion graph for ETS-10.

DESCRIPTION OF THE PREFERRED EMBODIMENT (S) The present invention shown in FIG. 1 includes a method and an apparatus for catalytic plasma removal of at least one nitrogen oxide from an exhaust gas stream 100 containing the at least one nitrogen oxide by converting the at least one nitrogen oxide to a nitrogen gas (N2). The method and apparatus rely upon a vessel 102 with an inlet 104 and an outlet 106, the vessel 102 having a microporous catalyst 108 therein together with electrodes 110 for creating a plasma. The exhaust gas stream 100 is passed into the inlet 104 through the vessel 102 and contacted with the microporous catalyst 108 and the plasma thereby converting the at least one nitrogen oxide to a nitrogen gas, forming a nitrogen oxide reduced exhaust gas stream 112, and flowing the nitrogen oxide reduced exhaust stream 112 through the outlet 106. If the dielectric constant of the microporous catalyst 108 is not sufficient to prevent arcing between the electrodes 110, a dielectric barrier 114 may be added. If the microporous catalyst 108 is not self supporting, a support frit 116 may be used.

The microporous catalyst 108 may be supported in a binder including but not limited to attapulgite clay, alumina and combinations thereof. It is important that the binder permit gas (hydrocarbon and NOx) access to the microporous catalyst 108.

The vessel 102 may have separate chambers as shown in FIG. 2A, a plasma chamber 200 and a catalyst chamber 202.

Alternatively, two vessels 102 having microporous catalyst 108 therein may be placed in series (not shown) as a double single stage configuration, or a double two-stage configuration.

The plasma chamber may comprise a packed bed reactor or a tube array reactor. Packed bed reactors suffer from electrical current flow across the particles of the packed bed and higher pressure drop through the packed bed which contribute to higher energy requirements than are needed with tube array reactors. The tube array is the preferred reactor for the plasma chamber due to its lower energy requirement and higher efficiency design resulting from factors

such as a higher transfer of electrical energy into the plasma and the openness of the tube array design which facilitates a lower pressure drop across the tube array. A tube array reactor 250 encased in a metallic housing 252 is shown in Fig. 2B. Dielectric tubes 254 are metalized on the inside 256 and electrified by fused contacts 258 to a high voltage power source (not shown). In this tube array design, the exhaust gas stream enters the chamber inlet 260 and flows in a transverse direction across the dielectric tubes before exiting the chamber outlet 262. In an alternate configuration shown in Fig. 2C, taken from U. S. Patent 5, 458, 748, the exhaust gas stream enters the chamber inlet 260 and flows through the tubes 254 in a parallel direction with the tubes 254.

A preferred microporous catalyst is a zeolite in the class of metallic Faujasites, Linde Type A (LTA) and combinations thereof. The microporous catalyst may have all pores as microporous (pore size of 15 angstroms or less) or may have mixed micropores (15 angstroms or less) and mesopores (greater than 15 angstroms).

The metallic Faujasite must be crystalline and has a structure selected from the group of zeolite X, zeolite Y and combinations thereof. Microporous material composition includes but is not limited to aluminosilicate and titanosilicate. Other network site cations include iron, gallium, germanium and combinations thereof. The metal in the exchange site of metallic Faujasite may be alkali, including but not limited to Mg, Li, Na, K, Rb, Cs, alkali earth, including but not limited to Ca, Sr, Ba, transition metals including but not limited to Sc, Y, Ti, Zr, Hf, V, Nb, Ta, Cr, Mo, W, Mn, Tc, Re, Fe, Ru, Os, Co, Rh, Os, Co, Rh, Ir, Ni, Pd, Pt, Cu, Ag, Au, rare earths and combinations thereof. Rare earths include but are not limited to Ce and Sm. The amount of metal atoms in the metallic Faujasite may range from about 5% to about 100% of available sites for metal atoms. The metallic Faujasite has a micropore size of at least 6 angstrom units. Generally, zeolite Y has a pore size of about 7 angstroms and zeolite X has a pore size of about 10 angstroms. Metallic Faujasite in or after plasma converts more than 55% of NOx to N2 for certain cations at a temperature of about 180 °C. The optimum temperature is from about 150°C to about 250°C.

This is an advantage because diesel exhaust from a light duty vehicle is typically near 150°C to 200°C.

Free aperture size greater than 3. 5 angstrom units permits flow of unburned hydrocarbon into the pores of the metallic Faujasite during conversion with at least one benefit of preventing the formation of nitric acid. Free aperture size greater than 6 angstrom units is preferred.

The catalyst structure may be a perforated or porous monlith, beads, powder or combination thereof.

Zeolites of Linde Type A (LTA) are characterized by pore sizes on the order of 4 angstroms. Small cations, including but not limited to Li, Na, Ca and combinations thereof permit NOx conversion to N2. Large cations, for example K, fill the pore preventing entrance of unburned hydrocarbons and substantially reduce conversion to the point of inoperability.

Other microporous materials with structures different from Faujasite or Type A including titanosilicate (ETS-10) templated mesoporous materials with micropores have been found useful for NOX conversion to N2.

Experimental Apparatus and Method for Examples Simulated exhaust gas streams were obtained from gas tanks with mass flow controllers. The exhaust gas test conditions were 100 °C to 300 °C in a lean mix of 02 (2-8 vol%), C02 (< 7 vol%), H20 (2-10 vol%), CO (<1 vol%), hydrocarbon (25-3100 ppm (as propylene or nonene, etc.)), and NO (10-500 ppm) at a space velocity of 2, 000-45, 000 hr'. Two catalyst configurations were used, either a single stage of a packed bed of catalyst material in the plasma region or two-stages of a plasma region followed by a packed bed of the catalyst material (FIG. 2).

NO and NOx were monitored with a chemiluminescent NOX detector with differences attributed to N02. End products of C02 and N2 were monitored with a mass spectrometer and gas chromatograph.

Example 1 Catalysts known in the prior art were used to measure NOx conversion with plasma. The amount of C02 was less than 4 vol%. The catalyst was in the form of a bead or pellet with particles held with a binder. Table E1-1 shows the catalysts and conversion results.

Table E1-1 Prior Art Catalyst Performance Catalyst Stages Temp (°C) Beta %NOx H-ZSM-5'10 Na-ZSM-5 1 200 52 Cu-ZSM-5"1 150 42 Cu-ZSM-5 1 200 41 Cu-ZSM-5 1 180 15 Cu-ZSM-5 1 250 36 Na-ZSM-5 1 200 52 H-Z-Y 2 200/300 6.5 0-67B H-Z-Y 1 200 --c H-Mordenite 1 180 33 20 AI203 1 180 20 Auto Beads 1 180 21 (PGM, Al203, Ce02) ANOX conversion excludes thermal NOX reduction for Temp > 200 C B Conversion high at onset is close to 67%, however rapid degradation of activity observed after several hours. In general, protonated zeolites adsorbed large amounts of hydrocarbon then failed by deactivation or conduction. c Sample adsorbs large amounts of hydrocarbon and possibly NOX, the reactor (plasma) fails due to conduction.

Example 2 Metallic Faujasites according to the present invention were used for NOX conversion. The Z-X materials were obtained from Aldrich Chemical Milwaukee, Wl. Zeolite Y (Z-X materials) with a Na as the exchange cation was obtained from Zeolyst International Valley Forge, PA. To obtain zeolite Y with other cations, an ion exchange or solution impregnation technique was used on the Zeolyst material. Pellets were made from powders using either a gamma alumina binder or an attapulgite clay binder and 500°C heat treatment.

The Z-Y with sodium was tested on a monolith honeycomb. The others were in the form of a bead or pellet secured with a binder. The single stage, conditions typically were : Temp. 180 °C, space velocity 12, 000 ho (4 L/min), Os (8 vol%), C02 (7 vol%), H2O (7 vol%), CO (0. 0 vol%), hydrocarbon (C3H6 525 ppm, C3H8 75 ppm), and NO (250 ppm). In addition, there was argon (9, 000 ppm) and hydrogen (130 ppm). For 2-stage experiments, the conditions were similar with water content reduced to 2 vol%, NO reduced to 150 ppm, and hydrocarbon levels reduced to C3H6 400 ppm.

Results are shown in Table E2-1 and FIG.'s 3-12. The Z-Y with sodium was tested on a monolith honeycomb as well as in pellet and bead form. The others were in the form of a bead or pellet secured with a binder.

FIG.'s 3-10 are for single stage operation with various cations. Single stage operation achieves 36-88% conversion with beta parameters ranging from 20-78 J/L. For Ag zeolite Y, which converts 73% of NOx at 50 J/L, or 3. 3 Watts, or an equivalent of approximately 5% of the vehicle fuel would be used to reduce NOx to N2.

For two-stage operation using Z-Y-Na, 29-81 % conversions are obtained for the beta parameter ranging from about 2 to 13 (fuel equivalent range from about 0. 7% to about 4%). In FIG. 9 the beta parameter is 5 for a catalyst in the two stage configuration that achieves 81 % NOx destruction at 20J/L, or 0. 35 Watts, or 0. 002 J/gram NOx, or an equivalent of approximately 2% of the vehicle fuel would be used to reduce NOx to N2 nitrogen (a Ford Taurus was used to calculate fuel equivalent). For comparison, current catalytic converter systems need to operate at stoichiometric air/fuel ratios which results in a reduction of fuel efficiencies of about 6%.

Table E2-1 NOx conversion with Faujasite

Material t-ree Si : Al %H2O Cation Temp %NOx ß Reactor Stages Aperture (°C) Reduced NO/NOx Type" (Å) Z-Y 7. 4 5 8 Na 100 36-/46 PB Z-Y 7. 4 5 8 Na 180 67-77 51/57 PB 1 Z-Y 7. 4 5 8 Na 300 40-/37 PB. 1 Z-Y 7. 4 5 8 K 180 75 35/43 PB 1 Z-Y 7.4 5 8 Cs 180 80 32/40B PB 1 Z-Y 7. 4 5 8 Ca 180 55 24/47 PB Z-Y 7. 4 5 8 Ga 180 50 31/36 PB Z-Y 7.4 5 8 Co 180 76 24/20D PB 1 Z-Y 7. 4 5 8 Mn 180 18-/48 PB 1 Z-Y 7.4 5 8 Ag 180 88 23/34E PB 1 Z-Y 7. 4 5 8 Ni 180 74 17/22 PB 1 Z-Y 7. 4 5 8 Fe 180 37 67/40 PB Z-Y 7. 4 5 8 Ce 200 43-/78 PB Z-X 9-10 5. 6 8 Na 180 36 33/20 PB Z-Y 7. 4 5 2 Na 180 65 4. 5 TAR 2 Z-Y 7.4 5 2 Na 180 81 5/5 TAR 2 Z-Y 7. 4 5 2 K 150 67 10/9 TAR 2 Z-Y 7. 4 5 2 K 225 50 6/10 TAR 2 Z-Y 7. 4 5 2 Ce 180 33-/5. 5 TAR 2 Z-X 9-10 5. 6 2 Na 180/164 ! 70 13/9. 6 TAR 2 Z-X 9-10 5. 6 2 Na 180/224 ! 63 3. 6/3. 6 TAR 2 Z-X 9-10 5. 6 2 Na 180/303 ! 29 2/1. 6 TAR 2 Z-Y 9-10 5. 6 8 Na 180 36 6/6 TAR 2X2 -No Data ! Reactor temperature different from the catalyst temperature A FIG. 3 B FIG. 5 c FIG. 6 D FIG. 7 E FIG. 8 F FIG. 9 G FIG. 10 H PB is packed bed and TAR is tube array reactor.

For a double single stage operation with Z-Y-Na with a conversion of 36%, FIG. 12 shows a beta parameter of about 6 which are about the same as a two-stage configuration.

Example 3 An experiment was conducted to demonstrate the effect of free aperture on the present invention. The catalyst was zeolite LTA with a Si : AI ratio of 1 : 1 in pellet form. For the single stage, conditions were Temp. 180 °C, space velocity 12, 000 hr' (4 L/min), 02 (8 vol%), C02 (7 vol%), H2O (7 vol%), CO (0. 0 vol%), hydrocarbon (C3H6 525 ppm, C3Hô 75 ppm), and NO (250 ppm). In addition, there was argon (9, 000 ppm) and hydrogen (130 ppm). Results are shown in Table E3-1.

Table E3-1 NOx conversion with Various Free Aperture Cation Free Temp % H20 Stages % Nox Aperture (°C) Reduce NOx (A) d K 3 180 8 1 4 10' Na 4 180 8 1 4 NA Ca 5 180 2 1 50 K3226227 Na 4 221 2 2 23- Ca 5 219 2 2 57 35 A FIG. 11 B FIG. 12 c FIG. 13 Results are further shown in FIG.'s 13-15 for single stage conversion showing both NO and NOx conversion. FIG. 13 is for zeolite LTA with potassium (K), FIG. 14 is for zeolite LTA with sodium (Na), and FIG. 15 is for zeolite LTA with calcium (Ca). For conversion rates from 4-57%, the beta parameter ranges from 10 to 30 respectively.

Example 4 An experiment was conducted to demonstrate titanosilicate as a catalyst for NOX conversion to N2. Catalyst beads were synthesized from the titanosilicate ETS-10 powder (obtained from Engelhard corporation, Iselin, NJ) using an attalpulgite clay or alumina binder. The free aperture of this catalyst

was oval shaped with opening of 4. 9Ax7. 6A. The titanosilicate catalyst in a single-stage configuration with 8% water, at 180 °C achieved 47% NOx conversion with a beta parameter of 39 (FIG. 16).

CLOSURE While a preferred embodiment of the present invention has been shown and described, it will be apparent to those skilled in the art that many changes and modifications may be made without departing from the invention in its broader aspects. The appended claims are therefore intended to cover all such changes and modifications as fall within the true spirit and scope of the invention.