Method of producing a body comprising porous alpha silicon carbide and the body produced by the method

Abstract

The present invention relates to a method of producing porous alpha-SiC containing shaped body and porous alpha-SiC containing shaped body produced by that method. The porous alpha-SiC containing shaped body shows a characteristic microstructure providing a high degree of mechanical stability.

Claims

1. A method of producing a porous silicon carbide product, said method comprising: preparing a mixture comprising silicon, carbon, at least one silicide forming and carbide forming agent and at least one alloy forming agent; wherein said at least one silicide forming and carbide forming agent comprises a group 13 element, or a group 12 element, or a group 2 element; wherein said at least one alloy forming agent comprises Cu; producing a monolith structure by processing said mixture; applying heating to said monolithic structure, wherein said applying heating comprises performing pyrolysis of said produced monolithic structure in a controlled atmosphere at a temperature between 700 and 1000 C. for a period between 1 and 24 h; wherein said applying heating further comprises: applying to said produced monolithic structure at least three temperature treatments wherein the temperature is varied between room temperature and 2100 C., said room temperature being about 20 to 26 C. with an average of 23 C., wherein said at least three temperature treatments comprise: a first temperature treatment between 850 and 1500 C. for a period between 1 and 24 h in inert atmosphere or vacuum; a second temperature treatment between 1500 C. and 2100 C. and back to room temperature for a period between 1 and 24 h in a controlled atmosphere; and a third temperature treatment between room temperature up to 1150 C. and back to room temperature in an oxidizing atmosphere; and wherein said at least one alloy forming agent is, after performing said pyrolysis, in a concentration between 0.1 and 0.9 at. %.

2. The method according to claim 1, wherein said group 13 element is Al.

3. The method according to claim 1, wherein said group 12 element is Zn.

4. The method according to claim 1, wherein said group 2 element is Mg.

5. The method according to claim 1, wherein said at least one alloy forming agent is an alloy forming agent supporting oxidation of silicon carbide.

6. The method according to claim 1, further comprising: exposing said heat treated porous silicon carbide product to acidic environment.

7. The method according to claim 1, further comprising: depositing a catalyst onto said temperature treated monolithic structure.

8. The method according to claim 1, wherein said processing is by extrusion or by application of pressure to said mixture.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) The silicon carbide filters and the method according to some aspects of the invention will now be described in more detail with regard to the accompanying figures. The figures show one way of implementing the present invention and is not to be construed as being limiting to other possible embodiments falling within the scope of the attached claim set.

(2) FIG. 1 shows the visual comparison between a DPF after the oxidation step when the starting mixture does not comprise an alloy forming agent (1) and when the mixture comprises Cu (2).

(3) FIG. 2 is a graph showing the temperature and atmosphere profile for the process for producing porous alpha silicon carbide filters. For example, SIC filters with 58% accessible porosity with an average pore size of 18 m. The line reflects the different steps.

(4) FIG. 3A-B are SEM images showing the microstructure of the monolith before oxidation, (FIG. 3A) when alloy forming agent is not present in the starting mixture and when Cu is present in trace amounts in the starting mixture (FIG. 3B). The microstructure is evidently different: both shapes and size of the crystals are different and the degree of interconnectedness is much more pronounced in the Cu containing monolith.

(5) FIG. 4A-B are high magnification SEM images of the monolith containing trace amounts of Cu before (FIG. 4A) and after (FIG. 4B) oxidation. It can be noticed that after the oxidation step a thick surface layer is produced filling the minor cavities in the microstructure and thus increasing the overall mechanically stability of the filter.

(6) FIG. 5A-F show SEM images in different magnifications comparing the microstructure of the oxidized monoliths produced with trace amounts of Cu added to the starting mixture (FIGS. 5B, 5D and 5F) and without trace amounts of an alloy forming agent added to the starting mixture (FIGS. 5A, 5C and 5E). Clearly, the Cu reduces the ternary carbide formation and results in a more interconnected microstructure of smaller less euhedral crystals. After oxidation the surface of the crystals are covered by a thick silica layer.

(7) FIG. 6 is a graphical representation showing wall strength values, determined according to ASTM 1674-08, for monoliths produced from a starting mixture comprising Cu and monoliths where Cu was not present.

(8) FIG. 7 shows the pore size distributions before and after oxidation for a monolith produced from a starting mixture comprising Cu.

(9) FIG. 8A-C are graphical representations showing the specific weight before oxidation and after each oxidation step, i.e. FIG. 8A, the differential increase after each step, i.e. FIG. 8B, and the first two oxidation steps accumulated in comparison to FIG. 8B, i.e. FIG. 8C, respectively.

(10) FIG. 9A is a graphical representation showing the measured porosities of the different samples after each oxidation step.

(11) FIG. 9B is a graphical representation showing the total pore volumes corresponding to the porosities measured in FIG. 9A.

(12) FIG. 10 is a graphical representation of the measured particle numbers in the exhaust mass flow over a European Transient Test Cycle with and without the filters.

(13) FIG. 11 is a workflow of the method according to one aspect of the invention.

DETAILED DESCRIPTION OF EMBODIMENTS

Example 1

(14) An extrusion batch starting mixture was created according to the following formulation:

(15) TABLE-US-00001 Weight Material % Electrode graphite 0-40 m .sup.17% Liquid starch 8.35% Metallic Silicon 20-75 m .sup.44% Aluminium flakes 4.8% Metallic copper 0-50 m 1.15% Methyl cellelose ether 3.8% Water 20.4% Polyoxyethylene 0.5%

(16) The electrode graphite was a commercially available Elektrodengrafit from Richard Anton; the liquid starch is the product CS76 from Zschimmer & Schartz; the metallic silicon was the Silgrain HQ from Elkem in a size of 20-75 m; the aluminum flakes are the Exoandit BE1160 from Grimm Metallpulver; the copper powder was also from Grimm Metallpulver and has the commercial name Rogal Kupfer 50 and the Methyl Cellulose ether was the Methocel K15M from Dow. The Polyoxyethylene was also from Zschimmer & Schwartz and has the commercial name Zusoplast G72.

(17) First, the dry components were mixed and then the wet components added. The whole mixture was ready after a mixing time of 3 hours in a Z-blade mixer. This starting mixture was filled into a piston extruder and extruded at a pressure of 70 bar into a honeycomb shape with a cell density of 200 cpsi and a wall thickness of 16 mil/400 m. They were cut to a length of 220 mm.

(18) These honeycomb monoliths were dried in continuous airflow of 0.9 m/s for 5 hours.

(19) After drying the monoliths went through a step-heat-treatment comprising a pyrolysis and three further heating steps (I-II-III) as described below and illustrated graphically in FIG. 2.

(20) The Pyrolysis occurs under 16.5 L/min flowing nitrogen gas (N.sub.2) of 99.8% purity at 850 for 1 h in a 240L tube furnace;

(21) In the subsequest reaction step the monoliths were placed in a graphite crucible in high temperature furnace (maximum; 2500 C.) of 70L; in the reaction step at 1450 C. for 2 h, and subsequently at 1600 C. for another 1 h under argon (Ar) (99%) with a flow rate of 4L/min, Si and C react to form 3C-SiC.

(22) Before the recrystallization step the atmosphere is renewed without cooling. In the recrystallization step the monoliths were heated from the 1600 C. to 1960 C. and held there for 2 h and the 3C-SiC is converted to alpha-SiC and opening of the pores in the microstructure takes place through recrystallization. The monoliths are subsequently cooled, still under argon, by turning off the furnace

(23) Following the recrystallization step, oxidation of the monoliths in air to 1100 C. for 4 h in an electrical furnace during the oxidation step. Cooling was achieved by turning off the furnace.

(24) Monoliths were taken out for characterization before and after this last oxidation step.

(25) The samples were characterized with X-Ray Diffractrometry (XRD), Scanning Electron Microscope (SEM), argon ion sputtering combined with Auger Electron Spectroscopy (AES), Energy Dispersive Spectroscopy (EDS) and Mercury Intrusion Porosimetry (MIP). In addition, the mechanical strength was tested on test bars with 77 cells and a length of 120 mm according to ASTM 1674-08.

(26) Furthermore, honeycomb monoliths were taken and cut to 203 mm length. The channels on each side were closed alternatively on each side with a cement based on the composition of 64.5 weight-% SiC, 1 weight-% Methyl cellulose ether, 23.5 weight-% silica sol with 45% solid content and 11 weight-% of water in a way that each open inlet channel is closed on the outlet and vice versa.

(27) These 49 mm49 mm and 203 mm long filter segments were tested in respect to back pressure at volume flow rates of 1000 L/min at 23 C. (cold flow test).

Example 2

(28) An extrusion batch starting mixture was created according to the following formulation:

(29) TABLE-US-00002 Weight Material % Electrode graphite 0-40 m 15.5% Liquid starch 7.25% Metallic Silicon 20-75 m 40.2% Aluminum flakes 4.4% Metallic copper 0-50 m 1.05% Methyl cellulose ether 5.6% Water 19.3% Polyamide powder 3.0% Polypropylene powder 3.2% Polyoxyethylene 0.5%

(30) The mixing sequence was the same as for example 1. However, in example 2 the polymer powders polyamide and polypropylene have been added to the dry components as pore forming agents.

(31) The polyamide powder was the commercial product Vestamelt 730 P1 from Evonik; the polypropylene powder is the commercial available PropyolTex 270S with a mean particle diameter of 15-25 m from Micro Powders Inc.

(32) The same type of plugged filter segments have been prepared to check the back pressure and also the same type of 77 channels test bars for the flexural strength according to ASTM 1674-08.

Comparative Examples 1b and 2b

(33) In addition to the examples 1 and 2 comparative examples 1b and 2b were conducted. The honeycomb monoliths were prepared exactly according to the formulation given in the two examples, 1 and 2, except for the fact that in both cases the copper was left out.

Evaluation of the Examples 1, 1b, 2 and 2b

(34) By comparing the monoliths, with trace amounts of alloy forming agent (examples 1 and 2) and without (examples 1b and 2b), a clear difference in crystalmorphology, microstructure, mechanical stability, crystal by-product formation and the creation of in situ silica layer can be noticed.

(35) The monoliths produced through the method described in examples 1 and 2 and represented in FIG. 2, according to one aspect of the invention, has some clear distinguishing features due to the presence of an alloy forming agent, such as Cu, in trace amounts, such as 0.5 at % (after pyrolysis 1 wt % before) in the starting mixture.

(36) FIG. 1 shows the visual comparison between a DPF-monoliths after the oxidation step when the starting mixture does not comprise an alloy forming agent (referred as 1) and when the mixture comprises Cu (referred as 2), (examples 1 and 2).

(37) Between the oxidized monoliths produced without Cu, FIG. 1-1, and the one produced with trace amounts of Cu, FIG. 1-2, there is a visible difference. The monolith produced from a starting mixture that does not contain Cu is brittle, bluish-black in colour and sparkling whereas the monolith produced from a starting mixture that contains Cu is neither of these.

(38) The heat treated monoliths clearly show individual anhedral interconnected 4H-SiC crystals, comparison of the unoxidized microstructure making up the monoliths without, FIG. 3A, and with Cu in trace amounts, FIG. 3B.

(39) When an alloy forming agent is not present in the starting mixture (examples 1b and 2b) a by-product in the form of large ternary carbide crystals are intergrown with the SiC microstructure that can be identified by human eye as a sparking effect on the filter or monolith surface. This effect is not present when Cu is added to the starting mixture (examples 1 and 2). Thus, a significant difference introduced by the addition of Cu is the suppression or reduction of ternary carbide crystal formation.

(40) For monoliths produced without Cu (examples 1b and 2b) contain large aluminum containing ternary carbide crystals predominantly in the composition Al.sub.4SiC.sub.4 and Al.sub.4Si.sub.2C.sub.5 (or the isostructural equivalent Al.sub.5NC.sub.3 and Al.sub.6N.sub.2C.sub.3) and the solid solutions that can lie between these compositions. These crystals are translucent yellow in color and with a platy hexagonal shape. These yellow translucent crystals of a distinct hexagonal and platy shape may be deposited on the furnace walls and the surface of the monoliths upon condensation from vapour and may protrude the monolith surface. Generally, these crystals consist of a mixture of Al.sub.4SiC.sub.4 and Al.sub.4Si.sub.2C.sub.5 and appear as intergrown with the 4H-SiC microstructure, FIG. 5A. When adding Cu in trace amounts to the starting mixture (examples 1 and 2) these by-product crystals are smaller and not disrupting the microstructure, FIG. 5B.

(41) When adding trace amounts of Cu to the staring mixture (examples 1 and 2), the 4H-SiC microstructure has a narrow grain size distribution (GDS) of small grains characterized by normal grain growth, FIGS. 5C and 5D.

(42) From this, it can also be observed that the grains size is smaller and less platy when adding trace amounts of Cu to the starting mixture. It appears that the viscosity of the AlSiCu alloy increases with increasing Cu content. Higher viscosity in turn corresponds to a lower diffusivity in the Cu containing liquid. Since euhedral shapes are a function of the diffusivity, the low diffusivity when adding Cu to the AlSi alloy will result in less euhedral crystal shapes as also observed. In addition the AlCuSi alloy is more likely to be present in this viscous liquid phase since the addition of Cu to either Al or Si or a combination thereof lowers the melting point of the alloy, and the presence of Cu introduces an eutectic to the system.

(43) The above described effects contributes to a more mechanically stable microstructure and thus a more mechanically stable monolith, even before oxidation and this effect is even more pronounced after oxidation as shown in FIG. 4. As the monolith is oxidized in air a thick surface layer of silica is formed on the microstructure increasing the mechanical stability even further. The oxidized monoliths with (examples 1 and 2) and without trace amounts of Cu in the starting mixture (examples 1b and 2b) are shown in FIGS. 5F and 5E respectively. For the monolith with trace amounts of Cu in the starting mixture the silica layer fills the minor cavities in the microstructure stabilizing the overall structure and thus increasing mechanical stability.

(44) The more rough surface morphology results from the silica formation which is an advantage when the addition of a catalyst is needed, as the catalyst adhesion to the surface is improved when the surface has a rough surface morphology.

(45) The silica layer fills deep cavities providing a further advantage when the addition of a catalyst is needed. The filtered fluid does not get into direct contact with catalyst located in deep cavities in the microstructure, thus filling these with silica prevents catalyst from being deposited in these cavities and thus reducing the amount of catalyst needed to cover the microstructure.

(46) The specific weight of the honey combs prepared according to the described examples are listed together with the back pressure at a flow rate of 10001/min in table 1.

(47) TABLE-US-00003 TABLE 1 weight before weight after weight back pressure oxidation, oxidation, increase, at 1000 L/min, g/L g/L % mbar Example 1 506 539 6.52% 20 Example 1b 509 515 1.18% 20.8 Example 2 441 469 6.35% 19.7 Example 2b 441 443 0.45% 19.9

(48) One can clearly notice that the weight increase is significantly higher for the examples 1 and 2 prepared by adding copper to the initial mixture.

(49) The wall strength, determined according to ASTM 1674-08 is shown in the diagram in FIG. 6.

(50) The results for the pore analysis with mercury intrusion porosimetry are listed in table 2.

(51) TABLE-US-00004 TABLE 2 Accessible modal pore Porosity, diameter, % m Example 1, before oxidation 62.5 19.9 Example 1, after oxidation 57.5 19.8 Example 1b, before oxidation 63 20.5 Example 1b, after oxidation 61 19.6 Example 2, before oxidation 67.1 23.1 Example 2, after oxidation 62.35 22.55 Example 2b, before oxidation 65.8 23.1 Example 2b, after oxidation 65.5 22.65

(52) The increase of the mechanical strengthfor the examples with the lower porosity is not very high, but for the examples with more than 65% porosity it is significant. This means, that for the porosity level of 62% or lower, the mechanical strength is mainly a result of the interconnectedness of the 4H-SiC microstructure itself and for higher porosity level the mechanical strength can be improved significantly by the SiO.sub.2 layer. The SiO.sub.2 layer is thicker in the examples prepared by adding copper to the starting mixture. This is confirmed by the significant increase of the specific weight and by the significant decrease of the porosity after oxidation. The formation of the SiO.sub.2 layer leads to the reduction of the width of the pore size distribution as the silica fills up very small pores. Thus, the formation of the SiO.sub.2 layer narrows the distribution of the pore size. This can be seen in table 3, in which the D10, D50 and D90 values are listed before and after oxidation for the examples 1 and 2.

(53) TABLE-US-00005 TABLE 3 D10, m D50, m D90, m Example 1, before 13.8 19.14 24.28 oxidation Example 1, after 14.65 19.1 22.1 oxidation Example 2, before 17.62 22.7 27.25 oxidation Example 2, after 17.13 21.57 24.85 oxidation

(54) The pore size distributions are plotted for the example 2 before and after oxidation in FIG. 7.

(55) In particular, the presence of an alloy forming agent comprising Cu in the starting mixture results in a porous alpha-silicon carbide product characterized by a modal pore diameter between 19 and 24 m, by a silicon oxide protection layer having a thickness between 50 and 100 nm, having an increase in weight after the oxidation step between 6 and 7%, a significant increase in mechanical strength when the porosity is higher than 65% following the oxidation step and the absence or limited presence of large ternary carbide crystals.

(56) The impact of the copper on the oxidation of the SIC and the resulting build up of a SiO.sub.2 layer has been investigated by oxidation trials. Samples of honey combs prepared according to example 2 and 2b have been taken after temperature step III, i.e. after the siliconization and recrystallization step. These samples have been oxidized in different steps and after each step the change in specific weight, porosity and pore volume has been investigated. The following oxidation steps have been performed: 1. 4 hours at 950 C.; 2. 4 hours at 1100 C.; 3. 4 hours at 1100 C.

(57) The specific weight before oxidation and after each oxidation step are shown in FIG. 8A, the differential increase after each step is shown in FIG. 8B and in FIG. 8C the first two oxidation steps are accumulated in comparison to FIG. 8B. One can clearly see, at the end of the step at 950 C., a significant weight increase is produced for the copper-containing example 2 when compared to the example 2b without copper. Without copper a significant weight increase can be observed at the second oxidation step at 1100 C. At this oxidation step, the copper containing sample shows a higher weight increase. These results clearly show, that the copper strongly promotes the oxidation of the SIC. The differential increase of the specific weight between the different oxidation steps as shown in FIG. 8C shows that this accelerated oxidation also decreases with subsequent oxidation.

(58) The measured porosities of the different samples after each oxidation step are shown in FIG. 9A. The corresponding total pore volumes are shows in FIG. 9B. These results correspond very well with the results shown in FIGS. 8 A-C. The total accessible porosity and the corresponding total pore volume is reduced by each oxidation step.

Example 3

(59) An extrusion batch mixture was created according to recipe of example 1.

(60) TABLE-US-00006 Electrode graphite 0-40 m .sup.17% Liquid starch 8.35% Metallic Silicon 20-75 m .sup.44% Aluminum flakes 4.8% Metallic copper 0-50 m 1.15% Methyl cellelose ether 3.8% Water 20.4% Polyoxyethylene 0.5%

(61) This mix was filled into a piston extruder and extruded at a pressure of 70 bar at the extrusion die into a honey comb shape with a cell density of 200 cpsi and a wall thickness of 12 mil/305 m. They have been cut to a length of 220 mm. The drying and complete heat treatment procedure to create a final oxidized SiC honey comb segment was exactly the same as described in example 1.

(62) After the complete heat treatment steps samples have been taken for MIP.

(63) The honey comb segments have been taken and cut to 203 mm length. The channels on each side have been closed alternatively on each side with a cement based on the composition of 64.5 weight-% SiC, 1 weight-% Methyl cellulose ether, 23.5 weight-% silica sol with 45% solid content and 11 weight-% of water in a way, that each open inlet channel is closed on the outlet and vice versa. These 49 mm49 mm and 203 mm long filter segments have been used to assemble filter blocks of 33 segments. To glue the segments to each other the commercial cement IsoFrax DPF cement from the company UniFrax was used. This Filter block was then cut to a round shape with a Diameter of 143.5 mm. The outer skin was sealed using the same cement as for the assembling. After drying the whole filter block was hardened at 700 C. for 2 hours.

(64) The resulting filter has had a diameter of 144 mm and a length of 203 mm.

Comparative Examples 3b

(65) In addition to the examples 3 a comparative example 3b has been worked out. The honey comb samples have been prepared exactly according to the recipes given in example 3, except the fact, that the copper was left out.

(66) The resulting segments have been assembled and cut out in the same way as in example 3 to build up a filter with a diameter of 144 mm and a length of 203 mm.

Evaluation of the Examples

(67) The specific weight of the honey combs prepared according to the described examples are listed together with the values from the MIP analysis and the weights of the resulting assembled filters in table 4. Also listed in table 4 is the back pressure of the filters in a cold flow test (20 C., 1013 mbar) at an air flow of 800 m.sup.3/h. The surprising effect is, that the back pressure of both filters are at 800 m.sup.3/h on a similar level although the pore diameter and porosity of example 3 with the Copper as the additive is significantly lower.

(68) TABLE-US-00007 TABLE 4 back Spec. pressure weight of Weight of Mean po- of the segments assembled pore ros- filter at after oxid., filter, diameter, ity, 800 m.sup.3/ g/l g m % h, mbar Example 3 472 1845 12 52 62 Example 3b 456 1776 17 60 67

(69) The filters have been tested on an engine bench test.

(70) The test setup is described below:

(71) Engine: VM Motori R425 with 2.5 displacement.

(72) T250 engine dynamometer

(73) Condensation Particle Counter from TSI

(74) The back pressure as tested at different load points. This was done with the fresh filters as well as with the filters at a soot load level of 5 g/l. In table 5 the results are listed for the two examples.

(75) TABLE-US-00008 TABLE 5 Back pressure at different load points for the filters according to examples 3 and 3b fresh and at a soot load level of 5 g/l. Filter Example 3, fresh Filter Example 3, at 5 g/l soot load exhaust back exhaust back T mass flow pressure T mass flow pressure C. kg/h kPa C. kg/h kPa 459.9 233.8 4.0 470.6 214.7 16.0 418.4 210.9 3.4 432.4 195.8 14.0 346.9 184.6 2.6 357.9 175.3 11.4 204.7 153.1 1.6 212.8 150.4 8.0 117.3 60.9 0.3 121.9 60.9 2.7 Filter Example 3b, fresh Filter Example 3b, at 5 g/l soot load exhaust back exhaust back T mass flow pressure T mass flow pressure C. kg/h kPa C. kg/h kPa 452.5 234.6 3.7 524.8 220.1 23.5 411.2 212.2 3.1 462.9 196.8 20.2 342.5 185.8 2.3 375.3 174.7 16.7 202.1 153.8 1.4 220.4 148.9 11.9 117.0 61.8 0.3 121.3 61.6 4.1

(76) Again, both filters show fresh (without or with very low soot load) a similar back pressure but after a loading to 5 g/l the filter of example 3 shows a much lower increase under soot load. This is an effect of the different structure caused by the copper additive: the smaller pores lead to a reduced deep bed filtration and therefore to a faster build up of a soot layer on the inlet surface of the filter channels. The reduced deep bed filtration causes the significant lower back pressure under soot load in comparison to the structure without copper additive of example 3b

(77) This is a clear advantage of the structure created with Copper as the additive. After the back pressure test of the fresh filters the filtration efficiency was tested running an European Transient Test cycle (ETC). This test was done before the soot load up to 5 g/l. The measured particle numbers in the exhaust mass flow with and without the filters is shown in the diagram in FIG. 10. One can clearly see that the filter according to Example 3 shows a better filtration right from the beginning. The accumulated particle numbers over the whole test cycle led to the following filtration performance:

(78) Filter Example 3: 99.9% filtration efficiency

(79) Filter Example 3b: 98.3% filtration efficiency

(80) This means that a Diesel particle filter according to Example 3 and therefore according to this invention shows improved filtration performance and soot load characteristic in comparison to a corresponding filter without the copper as the additive.

(81) FIG. 11 is a workflow of the method according to one aspect of the invention. The method (10) of producing a porous silicon carbide product comprises the steps: (S1) preparing a mixture comprising silicon, carbon, at least one silicide forming and carbide forming agent and at least one alloy forming agent; (S2) producing a monolith structure by processing the mixture; (S3) applying heating to the monolithic structure.

(82) The method of producing a porous silicon carbide product may further comprise the step (S4), i.e. exposing the heat treated porous silicon carbide product to acidic environment.

(83) This step (S4) as the effect of removing, at least partially, undesired oxides. For example, when an alloy forming agent comprising Cu is used in the stating mixture, CuO may be formed.

(84) In general, CuO may have a desired catalytic activity. CuO in combination with CeO.sub.2 can reduce NO.sub.x to N.sub.2 and water and generate ammonia in a high oxygen content gas flow. This effect is advantageous if a SCR catalyst is deposited on the porous SIC according to this invention. On the other hand CuO can also produce unwanted by-products, like dioxines, when the porous silicon carbide product is used as DPF with a special type of fuel like Biodiesel. In this case it is advantageous to remove at least a part of the CuO from the surface of the porous SiC.

(85) Removal of at least part of the CuO from the surface of the porous SIC may have the benefit of reducing the oxidation performance after the initial oxidation at 1100 C. (heat treatment III). This could be necessary if the porous SIC is used as a DPF in applications with severe regeneration conditions which produce temperature peaks inside the DPF of 1100 C. or higher. In this case it would be optimal to reduce the amount of CuO on the SIC surface to a minimum.

(86) Removal of at least part of the CuO from the SIC surface is not straightforward. According to the invention removal of the CuO may be achieved in acidic media, such as mineral or organic acids with and without additional oxidizing agents. For example, removal of CuO was performed on the sample prepared according to example 2. Following the final oxidation step at 1100 C., the sample was placed into nitric acid at a concentration of 65% at a temperature of 65 C. for 4 hours. After this the sample was washed with demineralized water and dried. This treatment resulted in a weight decrease of 0.7%. The sample was subsequently oxidized again at 1100 C. for 5 hours. This oxidation led to a weight increase of only 0.5%, i.e. the oxidation performance was reduced to a minimum.

(87) Other examples of acidic media may be hot (about 80 Celsius) aqueous mixture of HCl(aq) and cupric chloride CuCl.sub.2(aq).

(88) Although the present invention has been described in connection with the specified embodiments, it should not be construed as being in any way limited to the presented examples. The scope of the present invention is set out by the accompanying claim set. In the context of the claims, the terms comprising or comprises do not exclude other possible elements or steps. In addition, the mentioning of references such as a or an etc. should not be construed as excluding a plurality. The use of reference signs in the claims with respect to elements indicated in the figures shall also not be construed as limiting the scope of the invention. Furthermore, individual features mentioned in different claims, may possibly be advantageously combined, and the mentioning of these features in different claims does not exclude that a combination of features is not possible and advantageous.