Replacement Page 1
Process and Catalyst for the Oxidation of SO2 to SO3
BACKGROUND OF THE INVENTION
The present invention relates to a process and catalyst for the catalytic oxidation of sulphur dioxide to sulphur tri- oxide. More particularly, the invention relates to reticu¬ lated ceramic foam catalysts having a high surface to vol¬ ume ratio containing a catalytically active phase for use in the oxidation process.
DESCRIPTION OP RELATED ART
Commercial heterogeneous catalysts generally consist of po¬ rous inorganic solids optionally in combination with one or more metals. These catalyst systems need to be shaped in a rational way in order to fulfil the requirements regarding e.g. catalytic performance, mechanical strength and pres¬ sure drop. Fixed bed catalysts can traditionally be in the form of pellets, extrudates or monoliths as known by those skilled in the art.
In the last decade, however, foam catalysts have received considerable attention due to the combination of several attractive features (M. V. Twigg and J. Richardson, Stud. Surf. Sci. Catal . Prep, of Catalysts VI. 91 (1995) 345) . Ceramic foams are characterised by a continuous, highly po¬ rous structure constituted of interconnecting cells. This megaporosity, in which the pore size typically varies from 0.04 to 1.5 mm, gives rise to a highly tortuous flow pat- tern in which turbulence is significantly enhanced. This leads to forced convective flow and better convective heat
transfer. Due to the megaporosity, the pressure drop is relatively low enabling a high space velocity. Finally, the reticulated cellular structure provides a high surface area to volume ratio which simulates very small pellet diameters and gives rise to low diffusion resistance. Additionally, a large surface area can be applied at the foam by wash coat¬ ing it with a layer of a high surface area oxide.
In general, the optimal catalyst shape is dependant on the detailed process conditions. For a process that is not dif¬ fusion limited the catalytic activity depends on the sur¬ face area of active material. For a process which is diffu¬ sion limited the activity is increased by increasing the geometric surface area per unit volume. While decreasing the size of pieces constituting a catalyst bed has the ef¬ fect of increasing the surface area per unit volume, it has the adverse effect of increasing the pressure drop result¬ ing from the flow of reactants. Usually, it is desirable to minimise pressure drop. Furthermore, catalytic reactions generally involve the absorption or evolution of heat and the geometrical shape affects the transfer of heat to or from the reactants to a significant extent.
In relation to a fixed bed composed of pellets, ceramic foam possesses clear advantages with respect to a rela¬ tively low pressure drop and improved heat and mass trans¬ fer characteristics . Catalysts in the form of monoliths possess to a considerable extent the same advantages as ce¬ ramic foams, but they lack the highly tortuous flow pattern and the improved convective heat transfer. On the contrary, monoliths show even lower pressure drops enabling extremely high space velocities.
The pressure drop by gas flow through a fixed bed is deter¬ mined by the void fraction and the equivalent particle size [R.B. Bird, W.E. Stewart and E.N. Lightfoot, Transport Phe- nomena, John Wiley & Sons, 1960] . The equivalent particle size (hydraulic diameter) is calculated from the geometric surface area per reactor volume. High void fraction or large equivalent particle size results in lower pressure drop than low void fraction or small equivalent particle size at given gas flow rate. For foams the equivalent par¬ ticle size relates to the pore diameter and the skeleton geometry. Small pores corresponds to high surface area per unit volume and consequently to a small equivalent particle size [J.T. Richardson, Y. Peng and D. Remue, Applied Ca- talysis A: General volume 204 (2000), 19-32] .
The relations between the equivalent particle diameter area per unit bed volume and void fraction is described by R. B. Bird, W.E. Stewart and E.N. Lightfoot in Transport Phenom- ena, John Wiley & Sons, 1960, and is given by:
a = av* (1-e) and Dp = 6/av, where
a is the specific particle area per reactor volume (m2/m3) av is the specific particle area per particle volume (m2/πι3) ε is the void fraction and Dp is the equivalent particle diameter (m) .
Typical values for area per unit volume for ceramic foams are given by Richardson et al. [J.T. Richardson, Y. Peng
and D. Remue, Applied Catalysis A: General volume 204 (2000) , 19-32] .
The prior art reveals many examples of catalytic processes in which the specific advantages of foam based catalysts as outlined above are demonstrated. These processes involve production of synthesis gas by catalytic partial oxidation
(WO patent application 01/60515, US patent No. 5,658,497,
WO 96/16737) steam reforming (US patent No. 6,409,976, EP 0260826) and autothermal reforming (WO 0076651) . Other ex¬ amples involve the oxidation of ammonia (US patent No. 5,336,656), selective catalytic reduction (US patent Nos . 6,040,266, US 5,422,085), oxidative dehydrogenation (US patent No. 6,072,097), CO oxidation (US patent No. 5,112,787) and catalytic combustion (US patent No. 6,077,600) . However, oxidation of sulphur dioxide to sul¬ phur trioxide, which is the basis for commercial sulphuric acid manufacturing, has apparently not been published using foam based catalysts.
The production of sulphuric acid is commercially achieved by alkali promoted vanadium based catalysts as is known to those skilled in the art. In addition, Pt based catalysts are well known and have higher activity compared to the va- nadiurα based catalysts, though they are not applied commer¬ cially due to their much higher price. US patent No. 5,264,200 contains an extensive description of the prior art and reveals that both vanadium and Pt based catalysts have been described on particulate and monolithic sub- strates. In addition, US patent No. 5,.264,200 discloses monolithic catalysts both with respect to vanadium and Pt based systems. In detail, the monoliths are preferably sil-
ica extruded in nominally 100 to 300 cpsi with square cells. Furthermore, a wash coat is applied involving SiO2 and preferably ZrO2 as a promoter compound. Finally, the Pt catalysts are obtained using conventional ion-exchange techniques. These catalysts are applied in an adiabatic process comprising four beds with SO3 absorption after the second and fourth bed. More specifically, the first, second and third bed contain monolithic Pt based catalysts while the fourth bed contains a conventional Cs-V catalyst.
SUMMARY OF THE INVENTION
It has now surprisingly been found that the use of ceramic foam substrates, wash coated with a high surface area oxide and impregnated with platinum for oxidation of sulphur di¬ oxide to sulphur trioxide, utilises the platinum in a much more efficient way. This enables the use of less platinum catalyst compared to the above mentioned monoliths in the first and third beds of a reactor and also allows the use of Pt catalyst in the fourth bed of such a reactor.
It is therefore an object of the invention to provide a process for the catalytic oxidation of SO2 to SO3 utilising a Pt catalyst based upon reticulated ceramic foam that is wash coated with a high surface area oxide. This could for instance be TiO2.
Reticulated ceramic foam is available in many forms from a number of commercial suppliers. The ceramic foam substrates may be in the form of a monolithic structure or as pellets. By varying the pellet size distribution it is possible to vary the void fraction in the reactor bed. By proper choice
of pore size in the foam skeleton and foam pellet size it is thus possible to adjust the bed properties to the spe¬ cific requirements such as pressure drop at a given gas flow rate and amount of catalyst per unit reactor volume. Also, catalyst in the form of foam pellets is easier to load into an existing reactor than a monolithic structure.
DETAILED DESCRIPTION OP THE INVENTION
The invention relates to a platinum catalyst based upon re¬ ticulated ceramic foam wash coated with a high surface area oxide for use in the catalytic oxidation of SO2 to SO3. Foam material is preferred as it is stable at the operating conditions and has good mechanical strength and a well- defined porosity. The foam can be made of different materi¬ als such as oxides, carbides or nitrides, preferably oxides of aluminum, titanium, zirconium or mixtures thereof. The pore density is in the range of 10 to 80 pores per inch (PPI) , preferably in the range of 10 to 30 PPI. The foam substrates are sintered at high temperature to a low sur¬ face area and have a skeleton porosity in the range of 0% to 50%, preferably in the range of 10% to 40%.
The foams can for example be manufactured by a method simi- lar to that described by Schwartzwalder et al. in US 3,090,094. They can also be purchased from commercial sup¬ pliers such as Drach Umwelttechnik GmbH (Diez, Germany) , Ceramiques Techniques et Industrielles s.a. (Salindres, France) or Selee Corporation (Hendersonville, North Caro- lina, USA) . The foam substrates may be used in the form of monoliths or smaller pellets for fixed bed reactors.
The void fraction in the reactor bed is in range of 0.40 to 0.95, preferably in the range of 0.5 to 0.9. The total void fraction is composed of the open volume inside the foam to¬ gether with the open volume between the pellets. A prefer- able range for the total void fraction is 0.45 to 0.95. The choice of combination of pore density and void fraction is determined by the required conversion for a given bed to¬ gether with the maximum allowable pressure drop.
A platinum load of the monoliths of approximately 54.4 g.ft"3 is used in a preferred embodiment of this invention.
An embodiment of the invention is the process where the re¬ ticulated foam is wash-coated with a high surface area ox¬ ide comprising of one or more metal oxides such as titania (Tiθ2) , zirconia (ZrO2) or silica (SiO2) . After application of the high surface area wash coat, the catalytically ac¬ tive material platinum is introduced. The platinum can be introduced either by conventionally impregnation techniques or by chemical vapour deposition.
An embodiment of the invention is the process which uses a macroporous monolithic foam based catalyst. The foam is wash coated with a high-surface area oxide, thereby ensur¬ ing a high dispersion of the catalytically active phase, platinum.
A further embodiment of the invention comprises the process where the total void fraction of the catalyst is composed of the open volume inside the foam together with the open volume between the pellets is in the range of 0.45 to 0.95. Catalyst in the form of foam pellets and having the above
total void fraction is suitable for placement in a fixed bed reactor.
The catalyst, thus obtained, is suitable for catalytic oxi- dation of SO2 to SO3 and can be used in all the reactor beds .
Example 1
Three catalysts were made. Catalyst No. 1 was made from ce- ramie foam (20 PPI, zirconia-alumina) obtained from Cerami- ques Techniques et Industrielles s.a. (Salindres, France) was wash coated with an aqueous suspension of TiO2 made by suspending 8Og TiO2 powder in a mixture of 7Og TiO2 sol, 3Og water and dispersing agents. After drying at room tem- perature, the sample was calcined in air at 6000C. The wash coated reticulated foam was used as cylinders, 10 mm in di¬ ameter and 20 mm in length, impregnated with Pt using an aqueous solution of [Pt (NH3) 4] (NO3) 2, dried and calcined at 6000C.
Catalyst No. 2 was made by impregnation of a 9 mm Daisy shaped ring of TiO2 (surface area 70 m2/g with the same Pt precursor) .
Catalyst No. 3 was made according to Monsanto' s method de¬ scribed in US patent No. 5,175,136.
The catalyst compositions and platinum loads are shown in Table 1.
Table 1
These catalysts were loaded in the first, third and fourth beds of a reactor and the activity, expressed in terms of the conversion and rate constant, for SO2 oxidation to SO3 at different conditions was measured. The oxidation process utilising catalyst No. 1 illustrates the process of the in¬ vention. Oxidation processes utilising catalysts Nos. 2 and 3 are comparative. All conversions were measured at 4000C and are shown in Table 2.
Table 2
Dividing the rate constant with the platinum load (rate/Pt load) gives a number for the utilization of the platinum. Clearly the use of foam utilizes the platinum far better than a monolithic catalyst in third and fourth bed in ac¬ cordance with the invention. For first bed the utilization is at the same level as a monolithic catalyst both being below a pellet catalyst.
Example 2
The pressure drop over a pellet bed with a length of 0.33 m and a diameter of 0.0855 m was measured as a function of different air flow rates at room temperature. The void fraction was calculated by weighing and measuring the amount of dry sand filling the interstices between the pel¬ lets constituting the bed.
The pellets used for this example were extrudates of a com¬ mercial SO2 oxidation catalyst of the type VK-69, 9 mm daisy, obtained from Haldor Topsøe A/S, and crushed and sieved pellets of 10 PPI TiO2 foam obtained from Ceramiques Techniques et Industrielles s.a., Salindres, France. The specific area for foam particles was estimated from the geometric area of the foam and the measured void fraction. A specific area of 3000 m2/m3 and a void of 0.85 were used for the un-crushed foam. The specific area was measured for the commercial SO2 oxidation catalyst. The values are given in Table 3.
Table 3
The pressure drop as a function of linear velocity for dif¬ ferent pellet types is shown in Fig. 1. The figure shows a
graph of the pressure drop as a function of linear air ve¬ locity through a fixed bed of different particles.
It is seen that by altering the foam pellet size distribu- tion it is possible to vary the bed void fraction. Thus, with the same initial size of the reticulated foam pellets it is possible to obtain lower as well as higher pressure drops compared to a commercial SO2 oxidation catalyst.
This is in accordance with the theory of flow through a fixed bed [R.B. Bird, W.E. Stewart and E.N. Lightfoot, Transport Phenomena, John Wiley & Sons, 1960] . Lowering the void results in higher pressure drop. On the other hand more active material per unit volume is placed in the cata- lyst bed. Using foam pellets of different size it is thus possible to optimize the acceptable pressure drop at a given flow rate towards the desire to have a certain cata¬ lytic activity per unit volume.
Furthermore, it is seen from Table 3 that a much lower pressure drop is obtained for large foam particles than for the extrudates even though the specific area per reactor volume is much larger. For film diffusion limited reac¬ tions, as is the case here, this is a great improvement to prior art.