CA1036794A - Removal of nitrogen oxides from exhaust gases and catalyst therefor - Google Patents

Abstract

REMOVAL OF NITROGEN OXIDES FROM EXHAUST GASES AND CATALYST
THEREFOR

Abstract of the Disclosure A catalyst is described which is effective as a contact catalyst for reducing nitrogen oxides found in exhaust gas streams and is resistant to poisoning even in the presence of SO2. The catalyst comprises an alumina carrier on which is carried copper together with (a) at least one member selected from alkaline earth metals, alkali metals and transition elements and (b) a small amount of precious metal, especially rhodium or ruthenium.

Description

~0;~67~4 Nitrogen oxides dis arged from various burners, chemical plants, and cars represent a serious source of air pollutants. There is, therefore a great need for an effective method of removing these oxides from exhaust gas streams.
`Basically, the methods for removal of NOx may be divided into two methods, one for suppressing the production of NOx and the other for making NOx harmless after it is produced~ As for the former method, two-step combustion processes, lo~-oxygen combustion processes and exhaust gas circulation processes are ~0 being studied. The latter method includes a process for conver-sion into nitrogen through a catalytic reaction and a second process for absorption removal by using an absorptive liquid.
These processes, however, have their own merits and demerits.
No process has yet been established that is industrially satis-factory. Seven different forms of nitrogen oxides are known, but the principal air pollutants are NO and N02. These two forms are collectively termed NOx. It is said that the NOx in an ordinary exhaust gas contains 90 - 95% or more NO, the balance being N02. Therefore, a method for removing NOx must be primarily capable of removing at least NO.
One method of removal of NOx, as is known from U.S.
Patent No. 3,454,355 and other literature (e.g. Bartok et al, "System Study of Nitrogen Oxide Control Methods for Stationary Sources" Final Report-Vol. II, Esso Research and Engineering Company Government Research Laboratory, November 20, 1969), comprises using as a catalyst a metal, e.g. platinum, supported on an alumina carrier and allowing NOx to react with carbon monoxide, hydrogen or methane to reduce the NOx into nitrogen thereby making it harmless. Such methods produce no harmful by-products and present few problems in putting them into~
practical use.

10367~4 However, we have found that when tested with actual exhaust gaseq these catalysts were not effective in removing NOx.
It appears that the catalyst becomes poisoned by the oxygen, moisture and sulfur oxides in the exhaùst gases. This is illuQtrated in the following comparative Example A.
In the attached drawingsj Figures 1 - 5 represent a series of graphs showing amounts of N0 removed with different catalysts and under differing conditions.
Experiment Example A
In this experiment, a typical catalyst found in conventional literature, that is, a material comprising copper carried on y-alumina was used. Such catalyst is pre-pared by immersing commercially available y-alumina sieved to 8 - 14 mesh in an aqueous solution of copper nitrate for a fixed period of time, separating the aqueous solution, drying at 100-120 C and baking at 540 C. The amount of Cu carried in this catalyst was 5 wt%.
The reaction was conducted in a quartz pipe with an inner d$ameter of 30 mm which was ins~alled within an annular furnace, it being so arranged that the reaction temperature could be preset at a predetermined level. The reductlon of NOx (in the experiment, N0 being used) by C0 and H2 was carried out by filling the reaction pipe with said catalyst, followed by treatment with hydrogen gas at 540 for 1 hour. The reaction gas was prepared by mixing bombed mixed gases C0 + N2, H2 + N2'

2 + N2- N0 ~ N2 and S02 + N2 of predetermined controlled concentrations in accordance with the experimental conditions and was then admitted into the reaction pipe.
In addition, moisture was added by detonating the 2 + N2 mixed gas in water in a bubbler installed in a thermostatic water tank controlled to a predetermined temperature before the flows of said various mixed gases meet each other.
The amount of N0 removed was determined by gas analyzers at the inlet and outlet ports of the reaction pipe, an N02 meter produced by Mast Company being used for the analysis of N0. The N0 in the sample gas was oxidized into N02 by an oxidizin~ device and then admitted into the analyzer. S02, C0 and 2 were continuously nnalyzed using an infrared type analyzed (produced by Horiba Saisaku-Sho) an infrared type analyzer (produced by ~ugi Denkl Seizo) and a magnetic type analyzer (produced by Slmazu Seisaku-Sho), respectively.
Further, H2 was analyzed by a gas chromatograph (produced by Simazu Seisaku-Sho). In addition, the moisture concentration was determlned by calculation.
The condltions and results of the experiment are shown ln Tables 1 and la.

Table 1 Reactlon gas composition Reaction Space Rates of reaction (the balance being N2) temperature velociOy of N0 and S02 (%) (Dry gas basis) (at 20 C

N0 2 2 C0 - ( C) (hr~1) N0 ¦ S0 ppm % ppm % % 4 300 0 2500 0.75 0 540 10 98 100 300 1.0 1500 2.75 10 540 104 280 0 0 2.57 0 540 104100 __ _ 280 1.1 0 2.57 l10 540 104 - -~ ~~

NOTE: H 0 concentration is indicated on a wet gas basis.

- 3 -10~6794 Table la ~eaction gas composition Reaction ¦Space velocity j Rates of (the balance being N2) temperature (converted for reaction of (Dry gas basis) 20C) NO

¦ NO I H2 I 2 I H2 ¦ ( C) ¦ (hr ) ¦ (X) ppm¦ % ~ % ¦ %
500 1.0 O I O 450 10 100 I . I I .
0.4 l 10 L 450 104 O

NOTE: H20 concentration is indicated on a wet gas basis.

10 ~ Table 1 corresponds to the case wherè CO~ is used as a reducing agent, while Table la indicates the results when H2 is used as a reducing agent. In each case, when there is neither 2 nor H20 in the reaction gas, substantially complete removal of NO and S02 is achieved. However, with 2 and H20 admitted into the reaction gas, this catalyst loses its activity. This may also be said of the case of the single reduction of NO.
Apart from S02, actual exhaust gases from boilers, chemical plants and cars contain 2 and H20 without exception.
Further, it is almost impossible to remove these gases, particularly H20. Therefore, it may be concluded that this catalyst cannot be used in actual exhaust gases. In addition, further tests were also carried out using this catalyst with the reducing agent replaced by a hydrocarbon such as CH4, but similarly the catalyst lost its activity under the influence of 2 and H20-In our co-pending application Serial No. 181,840, filed September 25, 1973 we have described a catalyst comprising an alumina carrier, on which is carried copper together with at least one further member selected from alkaline earth metals, 10367~4 alkali metals and transition elements. That catalyst was e~fective in the presence of oxygen and moisture as illustrated by the following Experiment Example B.
Experiment Example B
The equipment used in Experiment Example A was used in this experiment, and the experimental method was also approximately the same~ In this experiment, however, C02 was also added to a reaction gas in order to make the latter resemble the actual exhaust gas more closely. In addition, a gas chrom-atograph was used for the analysis of C02 and CH4. In this case,an example is shown in which C0, H2 and CH4 are used~ as reducing agents, but there was observed some difference in the reaction temperature at which the effective reduction of N0 was achieved. That is, in the case of C0 and H2, temperatures of about 300 C or above were required, whereas temperatures of 590 C or above in the case of CH4. Thus, the reduction experiments were conducted at 450C for C0 and H2 and a~t 600C for CH4.
Further, the prepared catalyst was used in its state established when it had been baked at 540C. In the previous experiment, H2 reduction treatment was carried out, but there was no substantial difference in the result, except that an activation period of about 1 hour at maximum was taken before a fixed high activity was obtained. The experimental conditions are as follows:
(1) C0 reduction experiment conditions Reaction gas composition;
500ppmN0 + 4%2 + 1%C0 + 13%C02 + 13%H20 + N2 Reaction temperature; 450C
Space velocity (converted for 20C); 104 hr 1 (2) H2 reduction experiment conditions Reaction gas composition;

500ppmN0 + 4%2 + 1.1%H2 + 15%C02 + 10%H20 + N2 Reaction temperature; 450 C

~a3679~
Space velocity (converted for 20 C); 104 hr 1 (3) CH4 reduction experiment conditions Reaction gas composition;
500 ppmNO + 4%2 + 2.3%CH4 + 13%C02 + 10%H20 + N2 Reaction temperature; 600C
Space velocity (converted for 20C); 104hr 1 The experimental results sre shown in Table 2. In addition, the percen~age value preceeding each metal element indicates the analy~ed value.
Table 2 Rate of reduction-wise Catalyst removal of NO (%) ~ ~ CH4 5.3%Cu-4.4%Mg-A1203100L 100 ~:
5.0%Cu-5.0%Ba-A1203100 100 100 5.0%Cu-0.2%K-A1203 89 80 5.1XCu-5.4XCr-A1203100 100 85 8.lXCu-5.1%Mn-A1203100 100 90 5.8%Cu-6.8%Fe-A1203 100 5.6XCu-6.8%MO-A120389 85 85 5.1%Cu-0.2%Pr-A120340 40 10 As shown in Table 2, these are decidely superior catalysts. However, when 500-lOOOppm S02 was added to said reaction gas, all of these catalysts were gradually poisoned by S02, so ehat the NO reduction activity was decreased.
When the concentration of the individual reducing agent was extremely increased, however, the rate of reduction of NO was achieved to some extent t30-80%), but S02 was reduced sub-103679~
stantially completely into H2S and COS. Therefore, when it is desired to use these catalysts, except for the speclal case of the exhaust gas not containing S02, it is necessary to remove S2 in advance in order to prevent the production of H2S
and COS.
~ n ideal NOx reduction catalyse would have to be such that it is immune from the influence of S02 contained in reaction gases and that it has a suitable activity not causing the conversion of S02 itself into H2S or COS-The de8ree to which the various catalysts described above are influenced by S02 differs wieh the kind of additive~
metals, that is, metals other than Cu, and also with the concentration. Of these catalysts, those which are relatively immune from the influence of S02 were Cu-Mg-A1203, Cu-Mn-A1203, etc. Further, in the case of the Cu-Mg-A1203 catalyst, relatively good results were obtained when the Mg concentration was about 3%. Thus, the results of a chec~ on the S02-poisoned conditions of the 5%Cu-3%Mg-A120 catalyst are shown in Experiment Example C.
ExPeriment Example C
The experimental equipment and method were the same as those described in Experiment Example B. In order to observe the influence of S02, pure S02 collected in advance was intermittently in~ected at a fixed rate by a syringe into the reaction gas immediately before the reaction pipe. In addltion, the rate of in~ection was ad~usted so that the S02 concentra~ion in the reaction gas was about lOOOppm. The experimental conditions were:
Catalyst; 5Z Cu - 3% Hg - A1203 Reaction gas composition: 500ppm N0 + 0-4% 2 + 1.1~ H2 + 15% C02 + 13% H20 + N2 Reaction temperature; 450 C

Space velocity; 104hr 1 (converted for 20 C) The results of the experiment are shown in Fig. 1, in which the number of in~ection of S02 is plotted on the horl~ontal axis and the reduction of N0 on the vertical axis. While Fig. 1 shows the results when H2 is used as a reducing agent, there was observed almost no difference in the degree of poisoning by S02 when C0 or CH4 was used. That is; it has been ascertained that the S02 poisoning has relation to S02 and catalysts but has nothing to do with the kind of reducing gases.
On the other hand, as is generally known a catalyst carrying thereon a precious metal such as platinum, ruthenium, rhodium and palladium is effective for use as a reduction catalyst and is utilized in various contact reactions. Thus, we have made a check on the N0 reducing characteristics of such catalysts. The results are shown in Experiment Example D.
Experiment Example D
The experimental equipment and method were the same as those previously described, but in this expèriment, 2 was not contained in the reaction gas. The reaction gas 20 composition is as follows and the experimental results are shown in Table 3.
Reaction gas composition;
300ppmNO + 1700ppm S02 + 0.6%C0 + 10%H20 + N2 Reaction temperature; 540 C
Space velocity (converted for 20C); 5000 hr 1 Table 3 Catalyst IRate of r~ action (%) I N0 S02 , 10.5%Ru-Al2o3 j 40 88 ¦
¦0.5%Rh-Al2o3 17 83 .5%Pd-A1203 10 84 .5%Pt-A1203 16 73 The rate of removal of N0, though low, is retained.
Further, the rate of reaction of S02 is about 70 - 90% but the amount of H2S detected on the outlet side was 1410ppm for Ru-A1203, 1420ppm for Rh-A1203, 1400ppm for Pd-A1203, and 1200ppm for Pt-A1203~ Further, slight amounts of COS were detec-ted. Similarly, when H2 was used as a reducing agent, S02 was converted into H2S. NOx reduction experiments were conducted under a condition in which S02 has been removed from the reaction gas. However, in the case of C0 reduction and H2 reduction, NH3 was formed in the reaction gas and the making harmless of NOx could not be achieved. In C0 reduction, it is believed that the reason for the production of NH3 is that the reaction between C0 and H20 produces H2, which reacts with N0 to produce NH3. As described above, if a catalyst prepared by carrying a precious metal alone on a y-alumina carrier is used, conversion of S02 or conversion of N0 into NH3 takes place and under coe~istence with S02 it has not sufficient NO-reduction activity. Thus, it may be concluded that it is not an industrially satisfactory catalyst.
Summary of the Invention According to one aspect of the present invention there is provided a method of removing nitrogen oxides from nitrogen oxide-containing exhaust gases, wherein the nitrogen oxides in said exhaust gases are reacted with a reducing gas in the presence of a reducing catalyst comprising an alumina carrier on which is carried copper in an amount of less than lOX by weight together with at least one member in an amount of less than 10% by weight selected from alkaline earth metals, alkali metals and transition elements, characterized in that the catalyst also contains about 0.005 to 0.06% by weight of a precious metal selected from rhodium and ruthenium.
According to another aspect of the invention there provided a process for treating an internal combustion engine exhaust gas to decrease its content of nitrogen oxides which comprises passing the exhaust gas under reducing conditions over a catalyst comprising an alumina support on which is carried a metal from the group iron, cobalt and nickel in an amount of less than 10% by weight together with copper in an amount of less than 10% by weight and a platinum group metal selected from rhodium and ruthenlum in an amout~t of 0~005 to 0.06~ by weight.
Certain preferred embodlments of the inventive catalyst are illustrated by the following examples:
Example 1 In this ex~ample, 5%Cu - 3%Mg - A1203 was used as the catalyst base. Thus, catalysts prepared by immersing 5%Cu--3%Mg - A1203 (~ - 14 mesh, crushed product) in aqueous solutions of salts of Ru, Rh, Pd and Pt, respectively, were used. The Ru, Rh, Pd and Pt contents were as follows.
0.051%Ru - Cu - Mg - A1203 0.051%Rh - Cu - Mg - A1203 0.053%Pd - Cu - Mg - A1203 0.052%Pt - Cu - Mg - A1203 The experimental equipment and method were substantially the same as in the preceding experiment examples, and the influence of S0 was determined by intermittently injecting the same. C0 was used as a reducing agent. The conditions are shown below and the experimental results are shown in Fig. 2.
The results in Figure 2 were obtained with the following con-ditions:
Reaction gas composition: 660ppm N0 + 4%2 + 1%C0 +
13%C02 + 10%H20 + N2 Reaction temperature: 450 C

Space velocity: 10 hr (converted for 20C) ~ - 10-_ _ . , As seen in Fig. 2, even if S02 was injected, Rh - Cu - Mg - A1203 and Ru - Cu - Mg - A1203 did not exhibit any change in their NO-removing capacities, and NO was completely removed. In - lOa -contrast, Pt - Cu - Mg - A1203 and Pd - Cu - Mg - A1203 were strongly influenced by S0. Indeed, they were even more adversely influenced than was the Cu - M8 - A1203 not carrying such precious metal. The fact that very good results are obtained when Rh and Ru are added and that the results are not good in the case of Pt and Pd, is contrary to our expectations and shows that mere addition of the precious metals in slight amounts is not effective to avoid the poisoning by S02.
Next, an example in which the rate of reduction-removal of N0 at different temperatures while a reaction gas is contin-uously flowing, by using H2 and CH4 as reducing agents, isshown in Example 2.
Example 2 In this example, the case of the 0.051% Rh - Cu - Mg -A1203 is shown, but the conditions for reaction are as follows:
(i) H2 reduction experiment Reaction gas composition; .-SSOppmN0 + 600ppmS02 + 1.1%H2 + 4%2 + 15%C02+ 10%Y20 + the balance H2 Space velocity (converted for 20C); 10 hr 1 (ii) CH4 reduction experiment Reaction gas composition;
SOOppmN0 + 600ppmS02 ~ 2.5%CH4 + 4%2 + 13%C02 + 10%H20 +the balance N2 Space velocity; 104hr 1 Graphs corresponding to the above conditions are shown in Figs. 3 and 4. As seen in these Figures, even under the presence of S02, this catalyst was not influenced by S02 at all, and extremely effective removal of NOx was possible provided that the temperature was above 300 C for reduction with H2 and above 600C for reduction with CH4. These reaction temperatures are as superior as ~ e results obtalned wlth the Cu - Mg - A1203 catalyst, using a gas not containing S02.
These experiments were continuously conducted, each expending about 50 hours, during which the catalysts themselves displayed no sign of deterioration. Further,` S02 on the outlet side was analyzed, and there was observed a decrease due to adsorption by the catalyst itself in the initial stage of the experiment, but about 2 hours after the start of passage of the gas, ie coinclded with the inlet concentration. That is, as previously described and as seen in the precious metal-A1203 ~atalyst, there was observed almost no conversion into H2S.
Further, there was observed no NH3 caused by excessive reduction of N0. These properties fully satisfy the requirements for an industrial catalyst previously described.
Example 3 In this example, the amount of a precious metal to be added is excessively decreased. The catalyst used was 0.005%Rh-Cu-Mg-A1203, and the reducing agent used was H2.
~he conditions for experiment are as follow~, and thc result~ are sho~n in ~ig.5.
Reaction gas compo~itions;
700ppmN0 ~ 700ppmS02 ~ 1.2~H2 ~ -4~2 + 11%C~
+ 1O~6H2 0 ~ N2 Rcaction tempcraturc; 450C
Space velocity (converted for 20 C); 104hr~1 As seen in Fig. 5, the amount of N0 removed becomes constant 40 hours after the start. There is observed some `
difference as compared with the fact that in a similar experiment using a 0.051XRh - Cu - Mg - A1203 catalyst the removal of N0 was maintained at a value of 100%. During that time, ~ 036794 o~ course, ncithcr H2S nor ~I3 was dctected in the gas on the reaction pipe outlet side.
Detail~d results have been described with reference to Experiment Examples and examples in which sli~ht amounts o~ Rh and Ru are carried on Cu-M~-Al203typc catalysts ha~e been described. ~s for othor catalysts in which a slight amount o~ Rh or ~u is carried on other catalyst type, that is, the ca~alyst type which carries Cu a~d the alkali metals, nlkaline earth me~als or t~ansition metals, such as Cu-K-Al2 03~ Cu-Ba-A1203~
Cu-Cr-Al203, Cu-Mn-Al20~ and C~l-Mo-Al203, there were obtained substantially the same xesults as in the case of the Cu-Mg-A1203 type, though there was some difference in the removal of NO~ Also the catalyst obtained by a precious metal while using as the base the catalyst having no precious metal carried thereon, that is, Cu-Me-A1203, which is less liable to be influenced by S02, tended to a higher rate of removal of NOx. . .
As for the method of producing catalysts, the catalysts used in the experiment examples described herein were prepared by the so-called immersion process in which ~ -~lumina i~ immersed in an aqueous solution of the salt of e motal to cerry it thereon. however~ the method of !

jf . .
^13-~ 036794 producin~ cntaly5ts i~ not limitcd to t~o immcr~ion proces~. In~lecd, ~ood rcsults were obtained by usin~ a c~talyst havinG Cu and Rh or ~u imprc~nation-carricd on Me-~ O~ (Mc: one or more mctals selected from the group COnQisting of the ~lkali metals, alkaline earth met~ls and transition met~ls) produced by the so-called co-preci~itation process or kncading process. Further, a cat~lyst prepared by impre~n~tion carrying Rh or Ru on Cu-Me-Al~O~,produced by the kneading pxocess or co-pr~cipi-ta~ion process, was also ef~ective. There~ore, thc term '`carry`' aæ used in claim i9 not intendcd to ~imit the p~ocess to the i~mersion process.
As ~or the reducing gas, examples have been described in which H2, CH4 and Co were used, and only in the case of CH~ it was necessary to raise the reaction temperature, but there was no substantial difference in the required temperature between CO and H2.
In the experiments, CH4 was selected as a typical example of hydrocarbon, but it is known to use hydro-carbons as reducing agents. ~or example, according to a report by J.W. Ault, R.~. Ayen, et al. (A.I. Ch. E.
Journal Vol.17,No.2, page 26~- 271, 1971), CH4 belongs to a group having the lo~est reducing power in the hydro- !
carbons. ~here~ore, whcn a catalyst according to t~e present invcntion is used, it is readily seen that it is possible to replace CH4 by other hydrocarbon~ while achieving sufficient removal of N0x. Further, the use of said gaseS, that is, H2, hydrocaxbons C0 and other single gases, of course, makes possible the -14- !
. . ~

removal or NOx in an industrially satisractory m~nner.
llowever, it cnn be easily surmiscd that industrial gases havin~ these gases as the principal in~redients, for example, city gas may be used.
~ h~ ~eatures of the catalys~s of the present invention, that is, Rh-Cu-Me-Al2 ~ and ~u-Cu-~le~ 03 catalysts (Me;
one or more metals selected ~rom the ~roup consisting of the al~aline earth me~als, alXali metals and tran~ition m~als) may be sun~ariz~d as follows.
~a) ~hey are not poisoned by 2 H20 and S02 contained in actual ~ases, providing a stabili~ed NOx removal rate. ~he activity is very high.
(b) S02 passes through the present catalysts as it isj and S~2 is never converted into other substances~
; ~ (c) Reduction of ~Ox stops upon formation of N2 -and there is no NH3 by-produced.
, ; ~d) As reducing agents, hydrocarbons such as H? and , C~4 and, C0 gas may be used and industrial gases ` i having these 6ases as the principal in~redients may also be used.
(e) ~he production of the catàlysts is easy, and in use there is no need for preparatory treatment such as H2 reduction treatment.
; As described above, the catalysts invented herein have many features, and it is believed that they are also industrially superior as they can be used for reducing or making harmless the nitrogen oxides contained in exhaust ' gases discllurged ~rom steam-power plants, ~arious furnaces, ¦ chemical plants, etc.

` -15- -~

Claims (7)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A method of removing nitrogen oxides from nitrogen oxide-containing exhaust gases, wherein the nitrogen oxides in said exhaust gases are reacted with a reducing gas in the presence of a reducing catalyst comprising an alumina carrier on which is carried copper in an amount of less than 10% by weight together with at least one member in an amount of less than 10% by weight selected from alkaline earth metals, alkali metals and transition elements, characterized in that the catalyst also contains about 0.005 to 0.06% by weight of a precious metal selected from rhodium and ruthenium.

2. A method according to claim 1, wherein the copper is present in an amount of up to 5% by weight.

3. A method according to claim 2, wherein the alkaline earth metal, alkali metal or transition element is present in an amount of up to 5% by weight.

4. A method according to claim 3 wherein the catalyst is selected from Ru - Cu - Mg - Al2O3 and Rh - Cu - Mg - Al2O3.

5. A method according to claim 1 wherein the reducing gas is selected from CO, H2 and CH4.

6. A method according to claim 5 wherein the exhaust gas also contains SO2.

7. A process for treating an internal combustion engine exhaust gas to decrease its content of nitrogen oxides which comprises passing the exhaust gas under reducing conditions over a catalyst comprising an alumina support on which is carried a metal from the group iron, cobalt and nickel in an amount of less than 10% by weight together with copper in an amount of less than 10% by weight and a platinum group metal selected from rhodium and ruthenium in an amount of 0.005 to 0.06% by weight.