Patent Grant
8178068
This application is the U.S. National Phase application of PCT International Application No. PCT/GB2004/001783, filed Apr. 27, 2004, and claims priority of British Patent Application No. 0309747.4, filed Apr. 29, 2003, and British Patent Application No. 0315643.7, filed Jul. 4, 2003.
The present invention concerns an improved catalyst charge design, suitable for an ammonia oxidation process. More particularly, it concerns a two- or multi-stage catalyst charge, which is believed to offer technical and economic advantages.
The oxidation of ammonia using air to form nitric oxide for nitric acid production, for nitrate fertilisers or explosives, has been established since the early 20th Century (the Ostwald process), and is in widespread use across the world. The established process uses a platinum or platinum alloy gauze as catalyst for the reaction of ammonia with the oxygen in air. The Andrussow process for the production of hydrogen cyanide uses essentially similar technology, and should be considered as being within the scope of the present invention. The Pt gauze is usually supplied as a “pack” of several gauzes mounted one above the other. Successive gauzes may vary in filament thickness, alloy composition, and in other ways.
During operation, platinum is lost physically or by volatilisation from the gauzes, to a greater or lesser extent depending upon the process conditions, e.g. pressure. Whilst the platinum may collect in cooler parts of the plant, recovery of the platinum can prove difficult. It is known to include one or more “catchment” or “getter” gauzes, which are palladium (Pd)-based, which are mounted closely downstream of the catalyst gauzes. The catchment gauzes collect a high proportion of the Pt lost, although a proportion of the Pd is itself simultaneously displaced from the catchment gauze. Instead of having a separate Pd catchment gauze, it is possible to incorporate Pd filaments in a second or third gauze. It has been suggested that in such a case, the Pt recovered continues to be available to act as a catalyst. Recent research work by the Applicants indicates strongly, however, that Pd gauzes may increase N2 and/or N2O levels compared to a Pt gauze, up to several times. One mechanism may be:
It is not presently known whether N2 and N2O generation is the result of incomplete oxidation of residual ammonia or by reaction of residual ammonia with NOx, but indications are that residual ammonia is undesirable. Also undesirable, because of the effect on overall yield of NO and NO2, is either loss of NO, for example by the above reaction, or conversion of ammonia or NOx to N2.
There have been few advances in the Pt gauze-based technology over the last 20 years, apart from the development and introduction of knitted gauzes, in place of woven gauzes, by the present Applicants.
The substantial capital cost of Pt gauzes has led to some exploration of base metal catalysts for ammonia oxidation. These catalysts are generally based upon cobalt compounds such as Co oxides or perovskites. However, although the best Co perovskite catalysts offer some technical advantages over Pt-based gauze catalysts in addition to the lower capital cost, it appears that sufficient disadvantages remain, primarily a significant sensitivity to poisoning from atmospheric sulphur compounds, so that there has been only very limited commercial scale applications, in the manner of trials, as far as is presently in the public domain.
There have been a few proposals to combine platinum gauze with a Co-containing material. For example, Chemical Abstracts 114:188430 describes the combination of 1-2 Pt grids with a downstream Co3O4 catalyst. Such a catalyst is stated to show maximum selectivity to NO formation at 350° C., which is very much lower than the operating temperature of a conventional Pt gauze catalyst, and hence lower than the normal operating temperature of ammonia oxidation plants. Low temperatures are necessary for Co3O4 catalysts because of a phase change they are known to undergo above 850° C. resulting in the formation of CoO which has lower selectivity for NO in the ammonia oxidation process. Furthermore, no reference was made to the resulting N2O levels. WO 99/64352 describes a platinum gauze catalyst followed by a Co-containing catalyst, where a reduction in side reactions; such as nitrous oxide formation, is claimed. However the combined catalyst system only demonstrated at best N2O levels of 700 ppm in laboratory apparatus. This patent application was abandoned, and we are not aware of any commercialisation of the technology. Indeed, platinum or platinum alloy gauzes remain the only commercially available and technically acceptable technology in the marketplace.
The gauze manufacturers design gauze packs individually for each plant, or for each burner on a plant. It is state of the art practice that pack design aims for maximum conversion possible under the plant operating conditions, since there is a gradual fall-off of conversion efficiency as the platinum gauze deteriorates in use. Eventually, when conversion drops to unacceptable levels, the plant campaign is terminated and the catalyst is replaced. It would therefore be counter-intuitive to design a catalyst pack or catalyst charge where the conversion over a first platinum group metal catalyst element is deliberately chosen to be less than complete. Since incomplete conversion results in residual ammonia in the gas stream, the opportunity to make N2O by side reactions is likely to be increased. Indeed, loss of yield is established in the art as being very important; ammonia slip can cause the side reaction yielding N2, which clearly causes loss of yield, and ammonia slip also raises the possibility of forming explosive ammonium nitrate downstream of the catalyst.
There remains a need for improved choices for the ammonia oxidation plant operator, and for the catalyst charge designer, not least because of the capital cost of a platinum or platinum-alloy gauze catalyst, combined with the losses of valuable platinum from the catalyst in use. We consider that the objectives of the present invention, one or more of which we believe are satisfied in the preferred embodiments and when operating under preferred conditions, may be stated as:—
Hereinafter, sometimes the abbreviation “PGM” is used for “platinum group metals”.
Accordingly, the present invention provides a process for the oxidation of ammonia, including the Andrussow process, comprising passing a feed gas comprising a source of ammonia (eg, ammonia itself or offgas from a urea plant) together with a source of oxygen over a catalyst, and operating at temperatures of from 700 to 1000° C., characterised in that the feed gas is passed over a first stage catalyst for ammonia oxidation comprising a high surface area platinum group metal catalyst to yield a first stage product comprising nitrogen oxides, oxygen and unreacted ammonia and a source of oxygen and which first stage product is used in combination with a second stage catalyst for ammonia oxidation, and in that ammonia oxidation is not completed during passage of the feed gas over the first stage catalyst, but is completed during passage of the first stage product over the second stage catalyst, to yield a second stage product. Under preferred operating conditions, there is insignificant ammonia slip from the second stage catalyst combined with low N2O make, in the second stage product.
It will, of course, be understood that the first and second stage catalysts are not identical.
It is envisaged that the process of the invention can, under ideal conditions, offer increased campaign life. Other potential advantages of the present invention include reduced operating costs due to reduced precious metal costs, and improved environmental performance arising from reduced overall emissions of process by-products.
The invention in a second aspect provides a method of designing a catalyst charge for the oxidation of ammonia, including the Andrussow process, for a particular feedstock comprising ammonia and a source of oxygen, comprising incorporating a first, upstream, catalyst for ammonia oxidation comprising a high surface area platinum group metal catalyst in sufficient quantity and in adequate physical arrangement to oxidise an incomplete quantity, from 20% up to 99%, of the ammonia in the feedstock, to generate a designed first stage product containing nitrogen oxides, residual ammonia and a source of oxygen, and incorporating a second ammonia oxidation catalyst to be installed downstream of the first catalyst, which second catalyst is present in excess of the quantity and physical arrangement to oxidise the residual ammonia in the designed first stage product. In a preferred embodiment, the catalyst charge is designed to ensure that a second stage catalyst is used with significant N2O destruction activity, and the temperature within the second stage is in excess of that at which 50% N2O is destroyed, preferably greater than 80% of N2O is destroyed, more preferably, greater than 90% is destroyed.
The invention also provides a catalyst charge for ammonia oxidation, designed in accordance with the second aspect of the invention. In a further aspect, the invention provides the use of the catalyst charge in a high yield extended campaign ammonia oxidation process. High yield and extended campaign are each relative to a conventional, state-of-the-art, Pt or Pt alloy gauze catalyst pack.
Catalyst charges according to the invention desirably are direct replacements for the convention gauze packs. Depending upon the burner size and handling requirements, they may be constructed as a single charge unit, or in sections, e.g. quadrants, hinged or linked quadrants, or the like. Desirably, the catalyst charges, whether single charge units or divided in some way, comprise all the components in an integral manner.
It should be understood that in defining first and second stage catalysts, the invention includes variations, such as the inclusion of one or more other layers before, interposed between or after the essential two catalyst layers, even where such other layer(s) has a catalytic effect.
An advantage of the present invention is that conventional plant operating parameters, such as pressure, pre-heat and throughput (“load”) do not need to be altered in order to obtain at least some of the advantages of the present invention. Desirably, the Gas Hourly Space Velocity “GHSV”) for the total first and second stage catalysts is of the order of 100,000 hr−1, suitably from 50,000 to 200,000 hr−1.
A suitable design of first stage catalyst provides for from 20 to 99% conversion under normal operating conditions, preferably 60 to 99% conversion, more preferably >75%, particularly 50 to 70%, although in some embodiments it may be desirable to convert 85 to 95% of the ammonia over the first stage. The invention desirably provides a second stage product comprising less than 1600 ppm N2O, preferably less than 600 ppm, more preferably less than 500 ppm, for example in the range 100 to 200 ppm, N2O. In the case where either the second stage catalyst is particularly active for N2O destruction, or where there is a separate, downstream N2O destruction catalyst, we have discovered that many of the gas components in a process of ammonia oxidation actually inhibit N2O destruction. Therefore, it is important to carry out representative tests before selecting such catalysts.
In all aspects of the invention, the first stage catalyst is a high surface area platinum group metal ammonia oxidation catalyst, preferably Pt or a Pt alloy such as Pt5Rh or Pt10Rh, or other conventional Pt alloy. The actual alloy chosen is not critical to the present invention, and it may comprise other alloying components, for example Co or Ir. The Pt or Pt alloy may be in filamentary form, such as drawn filaments or wires, conveniently as a knitted or a woven gauze, or may be in ribbon form and/or as a non-woven layer. Successive gauzes or layers may comprise different alloys or metals, mixtures of different alloys or metals and/or different size filaments. Such gauzes or layers may be readily manufactured in conventional manner. The first stage catalyst may be or comprise a “pad” of knitted gauze, prepared by three-dimensional knitting to give enhanced thickness, with loops of filaments extending through the depth of the pad.
However, the present invention includes alternative high surface area, low pressure drop catalysts, where the catalytic Pt or Pt alloy is supported on a base metal or ceramic support, which may be filamentary, in bead or shaped form, in the form of a metal or ceramic foam or in monolith form such as a metal or ceramic honeycomb, grid, static mixer or the like. The Pt or Pt alloy catalyst may be itself supported on oxidic particles, or may be deposited onto an oxidic coating on the metal or ceramic support. Such oxidic particles or supports are well established in catalyst technology. Such supported catalysts may offer advantages in campaign length compared to gauze catalysts, where the structural integrity of the gauze may be quite badly affected by loss of Pt from the filaments.
A deposition route of possible interest for Pt and Pt alloys onto supports such as ceramics or metals comprises combustion flame spraying or vacuum or air plasma spraying, which yields a high surface area, porous deposited layer. Other, conventional, techniques for depositing a precious metal salt or complex onto a support and firing or reducing to yield a metal may of course be used. The metal spray technique may be used to deposit a high surface area Pt layer onto a Pt or Pt alloy gauze; such a layer may further improve catalyst light-off for process start-up.
The invention permits a reduction in the total mass of Pt or Pt alloy used for a given design duty. Such “thrifting” of precious metal may have significant economic value to the plant operator. In particular, where the second stage ammonia oxidation catalyst is a bed of particulate lanthanum-cerium cobaltate catalyst as described in WO 98/28073, we have found that it is possible to replace >20%, preferably >30% and most preferably >40% by weight of platinum alloy gauzes used before inclusion of the cobaltate catalyst in the catalyst assembly without significant loss in NO efficiency whilst reducing nitrous oxide levels to below 50% of that obtained before inclusion. However, certain preliminary studies also suggest that particular advantages may result from the use of Pt or Pt alloy in an amount up to 90% of the weight of a traditional gauze only catalyst pack, with a relatively small quantity of second stage catalyst.
The second stage catalyst may be selected from a wide variety of catalysts. In one aspect, the second stage catalyst may be a platinum group metal catalyst, which may be supported, such as a Rh- and/or Ir-based catalyst, or may be or comprise a base metal or base metal oxide, especially where the base metal is a transition metal or a rare earth metal, and may, for example, comprise one or more of iron, nickel, copper, cobalt, manganese, or be or comprise silver, or supported platinum, palladium or ruthenium. The second catalyst may be a mixture of one or more base metals with one or more precious metals. For example, initial trials indicate that a Rh catalyst dispersed and supported on a shaped catalyst support, can be effective to complete the oxidation of residual ammonia, whilst at the same time reducing unwanted side reactions. Although such a Rh catalyst may comprise other components or promoters, such as other platinum group metal catalysts, initial indications are that such a second stage catalyst is preferably free of catalytic or promoting components, and consists essentially of Rh on its support.
Catalysts suitable to be considered for the second stage catalyst include Cobalt-containing and non-Cobalt-containing ammonia oxidation catalysts and mixtures of these. Included within these are Au, supported PGM catalysts, La2O3, CO3O4 optionally with minor quantities of Li2O, certain spinels such as CoAl2O4, certain substituted ABO3 materials, certain perovskites such as LaCoO3, including LaCoO3 in which partial substitution (e.g. up to 20 mole %) of the A-site has been made by e.g. Sr or Ce, or partial substitution (e.g. up to 50 mole %) of the B-site has been made by e.g. Cu), La2CoO4, CO3O4 supported on alumina, thoria, ceria, zinc oxide or calcium oxide, CO3O4 or Bi2O3 promoted by rare earth elements or thorium and optionally containing one or more of oxides of Mn, Fe, Mg, Cr or Nb, CoOx with Pt on a support.
The second stage catalyst may be a Co-containing ammonia oxidation catalyst, especially a mixed Co and one or more rare earth oxides, most preferably as described in WO 98/28073.
In the case of a supported catalyst, this may be prepared in conventional manner, generally incorporating impregnation and firing steps. Such a supported catalyst may carry catalyst dispersed throughout the support, or may be loaded on a surface skin or in or around pores. However, it is not believed that catalyst deposition methods are critical to the present invention. We wish to mention the possibility of depositing metals, especially platinum group metals, by metal spraying using a combustion flame spraying process, or using vacuum or air plasma spraying. Such methods may deposit high surface area (porous) metal layers onto ceramic or even metal catalyst supports. Similar technology may be used to deposit base metal oxides, particularly when spraying is in the presence of oxygen or air.
Suitable catalyst loadings may be established by experiment for the particular conditions, but for platinum group metals (of which Rh is the metal of choice) are generally from 0.1 to 10% by weight, more usually from 0.3 to 5% by wt, most suitably from 0.5 to 1.5 by wt and for base metals are generally from 1 to 20 wt %, more usually from 5 to 15 wt %. Base metal catalysts may comprise minor amounts of promoters.
The catalyst supports may be any refractory metal or ceramic support offering adequate surface area combined with low pressure drop. Suitable metal supports include heat resistant stainless steels, which may incorporate passivating components or components capable of improving catalyst adhesion; the alloy known as “Fecralloy” may be considered. Suitable ceramics include those based on alumina, alumina-silicate, cordierite, zirconia, zircon-mullite or the like. Alumina appears to be the preferred support for a Rh second stage catalyst. Although metal supports usually do not have significant intrinsic porosity unless in the form of a metal foam, the surface area, and the ability of the support to carry high surface area dispersed catalyst may be improved by a washcoat coating. Suitable washcoats and coating technology are available to those skilled in the art, and may comprise one or more of alumina, zirconia, ceria, alumina-silica, lanthana-alumina and mixtures thereof. It is known from automotive emission control catalysts that the washcoat can contribute not just to catalyst surface area (including stabilising the washcoat at high temperatures), but also to more effective use of the catalyst. Therefore, the washcoat may comprise other components, which are generally one or more metal oxides or mixed metal oxides, especially from the transition metals or rare earth metals.
Shaped supports are conveniently shaped ceramic supports pellets or extrudates, such as cylinders or rings, multi-holed cylinders or multi-lobe supports, or other high surface area, low pressure drop, shaped support. The shaped support may be a high surface area monolith such as a flow-through honeycomb or a foam, or even a ceramic woven or non-woven fabric. A presently preferred support is well known as “Raschig rings”, which are desirably highly porous, in suitable sizes, for example 5 to 40 mm diameter. The supports may be impregnated or coated with the chosen catalyst composition.
The second stage catalyst may be used alone or in admixture with non-catalysed materials, or with materials carrying other catalysts. It may be supported (deposited) on a monolithic catalyst support or if particulate, supported on a metal grid, or between metal grids.
There may be an intermediate layer positioned between the first and second stage catalysts. This may be inert, serving as a support for one or both of the catalysts, or may have a function. Functional intermediate layers include those which absorb gas components, or those which convert gas components into products which are inert or relatively innocuous as regards the first stage product or the second stage product, or into product NO or products which are converted into NO, or into products which do not cause undesirable side reactions in the second stage reaction. For example, if the second stage catalyst is sensitive to, e.g. poisoned by, atmospheric components such as sulphur, or other materials found in the gas feedstock (Fe, Se) the intermediate layer may act as an absorber or getter. In this regard, we include herein the teaching of our co-pending United Kingdom patent Application No. 0309747.4.
Other components, such as pre-first stage or post-second stage layers may be incorporated into the catalyst charge or pack, as desired. Such a layers may, as in the case of an intermediate layer, be inert or functional. Functional layers may improve the campaign lifetime or improve the quality of the overall reaction in some aspect. In addition to physical layers in the catalyst charge of the invention, the invention may comprise additional steps, perhaps remote from the catalyst charge itself, which improve efficiency or other aspects of performance. For example, sulphur removal is seen as advantageous for LaCoOxide-type second stage catalysts, and an example of a suitable sulphur removal guard bed comprises lanthana or a similar oxide.
Whilst the first and second stage catalysts are not identical, it is possible to conceive of one or other being supported on a support such as a monolith to which state-of-the-art automotive catalyst coating technology is applied to give a “striped” catalyst, in which bands of different catalytic components are applied to the same monolith. This could, of course, also be used to deposit first and second stage catalysts onto a single monolith, and such a construction should be considered as falling within the present invention.
The present inventors have considered a number of embodiments of catalyst charges etc, all of which should be regarded as being according to the invention, and which are listed below for the assistance of readers. It will be understood that certain of the embodiments may overlap, or may be a sub-set of another, and also that the invention should not be limited to this listing.
Embodiment A. Pt or Pt alloy first stage catalyst plus second stage ammonia oxidation catalyst (which may additionally inhibit N2O formation or catalytically convert N2O to relatively innocuous by-products), especially where: (i) the first stage catalyst is a supported Pt or Pt alloy catalyst and the second stage catalyst is Co-based and/or the process of the invention incorporates designed and specified ammonia slip past the first stage catalyst, or (ii) a non-Co-based second stage catalyst in combination with an additional layer or incorporating an N2O destruction catalyst. Embodiment A(i) may desirably include a first stage catalyst designed to operate to react 90 to 99% of the ammonia in the feed gases, in combination with a second stage catalyst that oxidises the remainder of the ammonia, whilst making low levels of N2O, desirably less than 500 ppm, more desirably less than 200 ppm, of N2O.
Embodiment B. Non-gauze PGM first stage catalyst, for example, where the first stage catalyst is a supported Pt or Pt alloy catalyst, and where the second stage catalyst has high activity for N2O destruction.
Embodiment C. Pt or Pt alloy first stage catalyst and a supported platinum group metal second stage catalyst.
Embodiment D. Pt or Pt alloy first stage catalyst, an intermediate layer or additional component in the downstream part of the first stage catalyst or the upstream part of the second stage catalyst, especially where the intermediate layer or additional component acts as a guard bed to remove or reduce the effects of sulphur components in combination with the second stage catalyst. The second stage catalyst may suitably be a Co-based perovskite catalyst.
Embodiment E. First and second stage catalysts, optionally with intermediate layers or beds, designed and operated such that the ammonia slip from the first stage is in the range from 20 to 99 wt %, preferably 50 to 70 wt %, and especially where the initial design ammonia slip is negligible and the N2O content of the second stage product is less than 1600 ppm, preferably less than 500 ppm.
Embodiment F. Pt or Pt alloy first stage catalyst and non-Co-based second stage catalyst, in combination with an N2O destruction catalyst as a bed/layer downstream of the second stage catalyst or incorporated within the second stage catalyst. For example, the second stage catalyst is a supported platinum group metal catalyst and the N2O destruction catalyst is a Co-based catalyst. This embodiment, essentially a three-layer system, may prove particularly advantageous for high yield systems and processes.
Embodiment G. First and second stage catalysts in combination with a Pt catchment component, suitably a Pd-based catchment gauze, for example mounted upstream or downstream of the second stage catalyst. We have found surprisingly that while Pd-catchment gauzes may contribute to N2O production, in the present invention where ammonia slips from the first stage catalyst to the second stage catalyst that the positioning of a Pd-gauze below, i.e. downstream of the second stage catalyst may offer reduced N2O make compared with the placement of the Pd-catchment between the first and second stage catalyst. It is believed that any Pd-containing gauze will have a Pt catchment effect if placed downstream of a Pt catalyst. However, preferably the catchment gauze contains palladium>10% wt, more preferably >40% wt, especially >70% wt. The Pd-based catchment may comprise minor amounts (≦10% wt) one or alloying elements, such as Ni, Au, Co etc, and may also contain platinum. A typical catchment gauze is 5% Ni:95% Pd. Preferably the Pd-catchment gauze is formulated to reduce N2O levels and comprises <5% wt rhodium. Preferred catchment gauzes comprise palladium, platinum and rhodium. Such gauzes may comprise 8-25% wt, preferably 10-20% wt platinum. Examples of suitable catchment gauze materials include >92% wt palladium, 24% wt rhodium and the remainder platinum, or alternatively 82-83% wt palladium, 2.5-3.5% wt rhodium and the remainder platinum.
Embodiment H. First stage catalyst and a second stage catalyst which is platinum group metal coated or an admixture of second stage catalyst with platinum group metal supported catalyst.
Embodiment I. First stage catalyst and second stage catalyst, with an intermediate bed, especially a sulphur-protecting guard bed, and with an additional bed or layer downstream of the second stage catalyst of a specific N2O destruction catalyst.
Embodiment J. A catalyst charge with a guard bed upstream of the first stage catalyst. The guard bed may especially remove poisons, whether poisons for Pt or poisons for the second stage catalyst. Thus the guard bed may remove materials such as Fe, Si, Cl, Pb, As and S, P etc.
Embodiment K. A first stage catalyst and non-Co-based second stage catalyst, in combination with two downstream beds, especially an N2O destruction catalyst and a catchment bed. For example, the first stage catalyst, a platinum group metal-based second stage catalyst, a Co-based N2O destruction catalyst and a Pd gauze catchment gauze.
In general, tests performed in reaching the present invention show that high conversions of N2O are obtained over about 850° C. (lower with the best second stage catalysts). This indicates that an active first stage catalyst is preferably chosen, and the catalyst is designed and positioned to produce a temperature in the second stage above the light-off temperature for N2O destruction, suitably of at least 700° C., preferably at least 850° C. Accordingly, it appears desirable to oxidise a high proportion (i.e. >50% by volume) of ammonia in the first stage, to give sufficient exotherm for efficient N2O destruction in the second stage.
The invention will now be described by reference to working bench-scale examples, small pilot scale examples and a full-scale plant trial example.
A single 95%Pt5%Rh knitted gauze was used to catalyse a feed stream of 10% NH3 in air to produce a first stage product comprising NOx, N2O, steam, nitrogen, and unreacted NH3 and oxygen. The first stage product was passed directly into a second reactor containing each of the second stage catalysts A, B and C. The equipment was such that the temperature over the second stage catalyst could be varied in a range from 450 and 800° C., and the composition of the exit stream could be analysed for N2O, with the results shown in accompanying
A large number of tests were performed in a high throughput parallel reactor on a variety of metal or metal oxide catalysts carried on a variety of commercially available and proprietary oxidic supports, including mixed oxide supports, and made in conventional manner. Testing for N2O decomposition led to the following conclusions:
The highest activity N20 destruction catalyst from Example 2, 0.5% Rh on alumina, was compared with a LaCeCo oxide catalyst prepared according to WO 98/28073, in a representative gas comprising NOx, oxygen and water vapour, at various temperatures, both “fresh” and after hydrothermal aging at 1000° C. in air, to represent a used catalyst. The fresh Rh catalyst showed the best performance from about 500° C., reaching 100% conversion at about 800° C., however, even the aged LaCeCo oxide catalyst showed greater than 90% conversion at 900° C.
The preferred supported catalysts from the preceding Examples were tested in the laboratory for ammonia oxidation, measuring both conversion to NOx and conversion to N2O. The results are shown in the table below:
From the results below, as regards ammonia oxidation alone (ignoring N2O destruction activity) the WO 98/28073 LaCeCoOxide demonstrates good ammonia oxidation performance with low N2O make, better than the supported platinum group metal catalysts tested. The activity of the platinum group metals is the reverse of that for N2O decomposition, that is:
Pt>Pd>Rh
Pd produces high levels of N2O, because of equal selectivity to NOx and N2O.
A catalyst charge according to the invention, consisting of 6 “Prolok” 95%Pt5%Rh commercial knitted catalyst gauze, available from Johnson Matthey, as first stage catalyst mounted upstream of a 50 mm bed of LaCeCoOxide 3 mm cylindrical pellets according to WO 98/28073 as second stage catalyst (“Invention”), was compared to a bed of identical 6 “Prolok” gauzes (“Comparison”). The tests were run for 30 days in a tubular reactor of 40 mm id, at conditions simulating a commercial ammonia oxidation plant: 200° C. preheat; 10.5% NH3 in air at 4.5 bar absolute pressure, and at feedstock flow rates of 7.7 m3/hr.
Infra-red and UV-vis spectrometry were used to measure [NO], [NO2], and [N2O] concentration in the dried product gas. Infra-red spectrometry was used to measure [NH3] concentration in the gaseous feed. The efficiency of the catalyst was calculated on the basis of the nitrogen balance using the following formula: Efficiency to NOx={[NO]+[NO2]}/[NH3] with a correction for molar volume changes in the reaction.
The following conclusions could be deduced from studying the results:
* Typical SO2 levels on a plant may be 5 ppb, so this test gives accelerated effects.
A further pair of 30 day runs were carried out under the same conditions as example 5, except ammonia concentration was 10%+/−0.5%, but with the invention charge consisting of 3 Prolok gauzes and a 25 mm depth bed of the same second stage catalyst
Conclusions were:
A further pair of runs were carried out using the same Comparison catalyst on the same equipment using 4 sheets of the Prolok gauze combined with 50 mm bed depth of the same second stage catalyst. Test conditions modelled a different commercial plant: pressure was 10 bar absolute; flow rate was 10 m3/hr; ammonia concentration was 10% and pre-heat temperature was 237° C.
Conclusions were:
A further pair of runs were carried out over 24 days under the same conditions as example 5, except the Comparison consisted of 1 sheet of Prolok gauze over 5 mm of pelleted alumina and the invention example consisted of 1 sheet of Prolok gauze on 50 mm of the LaCeCoOxide pellets. These configurations demonstrate a considerable reduction in the PGM catalyst loading and a higher level of ammonia slip to the second stage catalyst.
Conclusions were
A further pair of runs were carried out according to the method of Example 8, both according to the present invention, wherein 2 sheets of a Pd-catchment comprising a 5% wt Ni—Pd alloy were placed either between the single Prolok sheet and the 50 mm of LaCeCoO3 pellets (arrangement 1) or after (i.e. downstream of) the LaCeCoO3 pellets (arrangement 2).
The conclusions were
Ammonia was oxidised with air in a tubular laboratory reactor of 28-mm internal diameter. The reaction conditions were as follows;
The pressure and exit temperatures varied and are given in the table below.
A mass spectrometer was used to analyze the composition of NO, NO2 and N2O in the product gas, with Ar used as an internal standard. Infra red spectrometry was used to measure [NH3] concentration in the gaseous feed. The efficiency of the catalyst was calculated on the basis of the nitrogen balance using the following formula: Efficiency to NOx={[NO]+[NO2]}/[NH3] with a correction for molar volume changes in the reaction using Ar as an internal standard.
The catalyst arrangements placed in the reactor comprised either:
The tests were performed over 15-21 hours. The results obtained over 3-5 hour periods under steady state conditions are set out in the following table:
In comparison, using the same equipment under similar conditions, 4-plies of platinum-rhodium gauze make 2345 ppm N2O (at an efficiency of 90.0%). Furthermore, 50 mm of pelleted CuAl2O4 (a nitrous oxide decomposition catalyst) placed under 1 ply platinum-rhodium gauze gives an efficiency of only 58%, with a N2O make of 278 ppm. Thus the catalyst assemblies of the present invention are able to provide the combination of high efficiency and low N2O make where the amount of platinum has been markedly reduced resulting in ammonia slip between the first and second stage catalysts.
A full-scale plant trial was performed using a catalyst assembly comprising a first stage catalyst consisting of 6 gauzes of Pt:Rh 92:8 (available commercially as Nitrolok 800 and Nitrolok 760 from Johnson Matthey PLC), 4 gauzes of Pt:Rh 95:5 (available commercially as Prolok 750 from Johnson Matthey PLC) and 1 gauze of Pt: Rh:Pd 90:5:5 (available commercially as Nitrolok 820 from Johnson Matthey PLC), having a total weight of 24.25 kg, and a second stage catalyst consisting of a 100 mm deep bed of pelleted La1-xCexCoO3 prepared according to WO 98/28073 in the form of 4-holed pellets, weighing 480 kg.
The plant operating parameters were as follows:
The results were as follows;
Mean N2O in flue gas, according to the invention 154 ppm, compared to 463 ppm for a single stage Pt gauze-only process. Thus the present invention has reduced the N2O made by >50%. The single-stage Pt gauze-only process requires an extra 15 kg of platinum catalyst indicating that the present invention offers considerable savings in platinum to the catalyst assembly. The actual trial duration was 112 days compared to design of 90 days (average efficiency at 112 days was 91%). Thus the present invention offered a >20% increase in campaign length. Inspection of the gauzes after the trial indicated that the ammonia was slipping from the first stage catalyst onto the cobalt catalyst. By calculation of the weight loss from the platinum catalysts it was established that 29.3% of the overall ammonia oxidation happened over the La1-xCexCoO3 catalyst.
| Number | Date | Country | Kind |
|---|---|---|---|
| 0309747.4 | Apr 2003 | GB | national |
| 0315643.7 | Jul 2003 | GB | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/GB2004/001783 | 4/27/2004 | WO | 00 | 10/2/2006 |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2004/096702 | 11/11/2004 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 4774069 | Handley | Sep 1988 | A |
| 5296435 | Kitaguchi et al. | Mar 1994 | A |
| 5985230 | Vlaming et al. | Nov 1999 | A |
| 6451278 | Zolotarsky et al. | Sep 2002 | B1 |
| 6743404 | Schumacher et al. | Jun 2004 | B1 |
| 20030124046 | Gorywoda et al. | Jul 2003 | A1 |
| 20030133849 | Schumacher et al. | Jul 2003 | A1 |
| 20040105803 | Schwefer et al. | Jun 2004 | A1 |
| 20040179986 | Burckhardt et al. | Sep 2004 | A1 |
| Number | Date | Country |
|---|---|---|
| 0 275 681 | Jul 1988 | EP |
| 0 611 041 | Aug 1994 | EP |
| 0 835 686 | Apr 1998 | EP |
| 0 835 686 | Apr 1998 | EP |
| 1 347 491 | Feb 1974 | GB |
| 05058607 | Mar 1993 | JP |
| 2 186 724 | Aug 2002 | RU |
| 1636332 | Mar 1991 | SU |
| WO-9828073 | Jul 1998 | WO |
| WO-9955621 | Nov 1999 | WO |
| WO-9964352 | Dec 1999 | WO |
| WO-0013789 | Mar 2000 | WO |
| WO-0151415 | Jul 2001 | WO |
| WO-0187771 | Nov 2001 | WO |
| WO-03011448 | Feb 2003 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20070031314 A1 | Feb 2007 | US |
