WO2008145486A1

WO2008145486A1 - Ceramic filter element having a small deep-bed filtration contribution

Info

Publication number: WO2008145486A1
Application number: PCT/EP2008/055543
Authority: WO
Inventors: Frank Miller; Christian Schiller; Thomas Willam
Original assignee: Robert Bosch Gmbh
Priority date: 2007-05-29
Filing date: 2008-05-06
Publication date: 2008-12-04
Also published as: DE102007024871A1

Abstract

The invention relates to a process for producing a ceramic filter element (114) for decreasing pollutants in an offgas stream, in particular a process for producing a diesel particle filter. The filter element (114) has a filter wall (130) through which the offgas stream flows. In the process, a ceramic starting material is shaped to produce a shaped body and subjected to a sintering process. Here, the starting material and/or at least one process parameter of the process are/is selected so that the following relationship applies to the pore size distribution: (I). Here, D50 is the cumulated pore size below which 50% of the pores come. D10 is the cumulated pore size below which 10% of the pores come. L is the wall thickness of the filter wall (130) through which the offgas stream flows.

Description

description

title

Ceramic filter element with low depth filtration

State of the art

The invention is based on known ceramic filter elements, as they are used in particular in the field of pollutant reduction of exhaust gases. Such ceramic filter elements are used for example in the context of diesel particulate filters (DPFs), but can also be used in other exhaust aftertreatment devices, such as in catalysts or other pollutant reducing elements.

The ceramic filter elements can be used in different structures, which depend on the nature of the exhaust gas purification. In particular, the filter elements can be designed as functional elements which have a honeycomb structure. For use in diesel particulate filters, for example, the functional element may have mutually closed channels whose walls have a certain porosity. The gas stream entering through inlet channels is then forced to pass through the pores of the channel walls.

As a rule, the filter elements have a ceramic carrier material which has a coherent porous structure through which the exhaust gas flow to be cleaned can pass. The ceramic carrier material used is usually ceramic materials such as silicon carbide, cordierite, aluminum titanate, sintered metals or other ceramics. Cordierite and / or aluminum titanate, in particular, have proven to be the most favorably priced materials of the cited ceramic materials, which also have suitable thermal properties for use in filter elements.

The filter effect of the filter elements is based essentially on the fact that a filter wall of the filter elements is flowed through by the exhaust gas flow, wherein particles in the ceramic

Material of the filter wall are deposited. Accordingly, these particles, for example diesel particles, build a filter layer in or on the filter wall (the so-called filter). cake) on. This filter cake is itself an effective filter for separating the particles of which it consists.

During the deposition process, the thickness of the filter layer increases continuously. However, this in turn leads to impairments of the gas flow, which has an increase in the back pressure result.

In addition to a deposition on the surface of the filter walls, however, at least in the initial range of the operating range of the filter and / or after a filter regeneration and a deposition of the particles in the interior of the porous structures of the filter walls. This deposition is also referred to as depth filtration.

However, a disadvantage of this depth filtration, which precedes surface filtration, is that it is associated with a comparatively strong initial differential pressure increase. This means that the pressure drop Δp, which occurs at the filter element, rises sharply at the beginning of the operation of the filter element and that the differential pressure is generally higher for filter elements with a high depth filtration fraction than for filter elements with a lower depth filtration fraction. However, the differential pressure or its increase with the introduced soot quantity or particle quantity represents a decisive quality criterion for such filter elements. For example, after reaching a certain differential pressure regeneration of the filter element is required to undisturbed operation of an internal combustion engine, arranged in the exhaust line of the filter element is to ensure. The sooner the maximum differential pressure is reached, the earlier the said regeneration is required.

Disclosure of the invention

Accordingly, a ceramic filter element for pollutant reduction in an exhaust gas flow is proposed, which may in particular comprise a diesel particulate filter, and a manufacturing method for producing a ceramic filter element, in particular a ceramic filter element according to the invention. The proposed filter element and the proposed method largely avoid the aforementioned disadvantages of known filter elements or methods.

In particular sintered ceramics can be used in the ceramic filter elements, which comprise silicate ceramics, in particular aluminosilicate ceramics. The proposed method is particularly suitable for cordierite ceramics. In this case, the material properties of the proposed filter element or the process parameters are selected such that the filter element, in particular a cordierite honeycomb filter, has improved functional properties in terms of counterpressure and filtration compared to standard filters. The back pressure of the filter can be permanently reduced over the entire functional range and over the entire service life.

As a result, with the filtration efficiency remaining the same or even improved, the time span between two regeneration phases of the filter element can be significantly increased. This reduces overall exhaust emissions, which plays a significant role in the context of the ongoing climate debate. The reduced frequency of the regeneration phases can also reduce the energy consumption, since any regeneration, which is usually thermally initiated, is accompanied by an increased energy requirement.

Another advantage of the inventive design of the method or the filter element is that due to the reduced frequency of the required regeneration processes and the entire life of the filter element can be significantly increased. Each regeneration process, which usually occurs at temperatures of 1000 ⁰ C or more, represents a significant thermo-mechanical load on the filter element, which in the long run can lead to fractures or other defects and thus limits the life of the filter element. A reduction in the frequency of these regenerations may thus prolong the life span.

The proposed method is a method in which a ceramic starting material is shaped into a shaped body and subjected to a sintering process. Several shaping processes and / or sintering processes can also be provided.

In this case, the starting material of the production method and / or at least one process parameter of the method is selected such that the following relationship applies to the pore size distribution in the finished ceramic of the filter element, in particular in the region of a filter wall of the filter element through which the exhaust gas flows.

D50 - D10. ., _{/ τ r} . _n ... <9 mil / Lrmill (1).

D50

D50 denotes the cumulative pore size below which 50 percent of all pores fall, ie an integral from 0 to the pore size D50 via a (normalized) frequency distribution H (D) of the pore sizes, which covers 50 percent of all pores: (-D50

J ₀ H (D) dD = 0.5 (2).

Accordingly, the quantity D10 is defined as the cumulative pore size below which 10 percent of the pores fall. L denotes the wall thickness of the filter wall through which the exhaust gas stream flows. If filters with varying filter wall thickness are used, then, at least approximately, the average filter wall thickness can be used for the size L.

The mentioned relation was determined by determining the depth filtration fraction to be minimized for many different filter walls with different pore size distributions. In this case, a maximum value of 20 mbar was set for this depth filtration portion, which is defined by an initial, non-linear increase in the differential pressure, which is still tolerable for most filter elements used in practice. Almost all of the investigated filter elements which fulfill the abovementioned condition have depth filtration fractions of 20 mbar or less and thus fall within the stated tolerable range.

The pore size distribution mentioned is preferably a pore size distribution measured by means of mercury porosimetry (MIP). The measurement of this pore size distribution by means of the said mercury porosimetry method represents a standard method in the production of ceramics, so that the fulfillment of the stated condition can be carried out by default with means already available today. The filter elements which fulfill the stated condition can then be selected according to a trial-and-error method, or the pore size distributions can be adjusted in a targeted manner (see below) so that the stated condition is satisfied.

The filtration efficiency of the filter elements increases with increasing thickness L of the filter walls (wall thickness). Thus, the wall thickness L, as done in the formula described above, must be included in the pore size distribution. Here, as is common practice in filter manufacture, the wall thickness L is given in the unit mil. One mil represents 1: 1000 inches, ie 25.4 μm. The condition (1) described above can thus be represented in SI units as follows:

^{D50 ~} D10 = 228.6 μm / L [μm]. (3)

D50 It has proven to be particularly advantageous if a pore size distribution is used or adjusted, in which the described pore size D50 is a maximum of 20 microns.

On the one hand, as described above, it is of course possible to leave the fulfillment of the described condition of the pore size distribution to reduce the depth filtration to chance, and to select only those filter elements which fulfill the said condition.

However, it is preferred if the described method is carried out in such a way that the pore size distribution is specifically set in such a way that the stated condition is met. For such a targeted adjustment of a pore size distribution, there are several possibilities. These different possibilities can also be used in combination.

A first possibility is to use a starting material for the ceramic filter element which contains particles. These particles should be designed such that they are replaced in the manufacturing process of the ceramic, in particular in the sintering process, by pores of appropriate size. Accordingly, analogously to the above relation 1, the following relation can be established for these particles or their particle size distribution:

^{P50 ~ P 10} <9 mil / L [mil] (4)

P50

Here P50 represents the cumulative particle size under which 50 percent of the particles fall. The size PlO is defined analogously.

In this proposed method variant, particulate formers are thus used in a targeted manner, which are replaced in the finished ceramic by corresponding pores. For example, these pore formers may comprise organic pore formers which are burned at a sintering step. The released CO2 or other combustion products can, with a suitable choice of process parameters, for example in a correspondingly slow sintering process, escape, so that corresponding pores remain.

As an alternative to the use of organic pore-forming agents, other types of pore-forming agents are also known from the prior art. A correspondingly known further possibility which can be used in the context of the present invention, the use of salts as pore formers, which are for example also thermally decomposed or which can be washed out after sintering by using appropriate solvents.

A further, alternatively or additionally employable method for deliberately effecting a pore size distribution according to the above relation is to use reaction sintered ceramics. For example, the ceramic filter element may comprise at least one sintered ceramic produced by reaction sintering, wherein the sintering process comprises a reaction sintering step in which a first starting material is melted and reacts with at least one second starting material. In this reaction, voids are formed which substantially conform to the shape of the first raw material since the surrounding matrix substantially accommodates the first raw material. An important, preferred example of such a reaction sintered ceramic is again, for example, a silicate ceramic, in particular an aluminosilicate ceramic, preferably a cordierite ceramic. In this case, the first starting material may in particular comprise talc (ideal formula 3 MgO: 4SiO ₂ : 1H ₂ O) or SiO ₂ , which may be clad with other starting materials, in particular quartz-free clays or kaolins (ideal formula 1 Al ₂ O ₃ : 2 SiO ₂ 2 H ₂ O) to form cordierite (ideal formula 2 MgO: 2 Al ₂ O ₃ : 5 SiO ₂ ).

In this reaction sintered production of cordierite, the pore structure is essentially determined by the talcum particles or SiO ₂ particles, so that the pore size distribution is determined by appropriate preparation of the talcum starting powder, for example a corresponding milling and / or granulation of these particles can be targeted.

Brief description of the drawings

Embodiments of the invention are illustrated in the drawings and explained in more detail in the following description.

Show it:

Figure 1 shows a longitudinal section through a known ceramic filter element;

FIG. 2 shows a pore structure of a filter wall of a known filter element in a schematic representation; FIG. 3 shows the differential pressure of a filter element as a function of the amount of soot introduced;

FIG. 4 shows a cumulative pore volume as a function of the pore diameter in a schematic diagram to clarify the determination of the values D10, D50 and

D90;

FIG. 5 shows a correlation of the depth filtration and the pore size distribution; and

FIG. 6 shows a schematic profile of the pressure increase and the filter efficiency as a function of the loading for different filter elements.

Embodiments of the invention

FIG. 1 shows a filter element in the form of a honeycomb body 114, which can be used in an exhaust gas aftertreatment device. In particular, this exhaust aftertreatment device may be a filter in which soot particles are removed from an exhaust gas stream of an internal combustion engine (diesel particle filter, DPF). The filter is arranged in an exhaust line of the internal combustion engine. The honeycomb body 114 is configured cylindrical in this embodiment, with an axis of symmetry 116. The honeycomb body 114 can then be used for example in a rotationally symmetrical filter housing in the exhaust system.

The honeycomb body 114 is configured as a monolithic ceramic functional element, which can be produced for example by means of an extrusion process. For example, for this purpose, the above-described ceramic materials may be used, for example magnesium aluminum silicate, preferably cordierite.

The honeycomb body 114 is traversed in use in a diesel particulate filter in the direction of arrows 118 from the exhaust gas. The exhaust gas enters the honeycomb body 114 via an entry surface 120 and leaves it via an exit surface 122.

Parallel to the axis of symmetry 116 of the honeycomb body 114 extend a plurality of inlet channels 124, in alternation with outlet channels 126. The inlet channels 124 are closed at the outlet surface 122. In the embodiment shown here, sealing plugs 128 are provided for this purpose. Accordingly, the outlet channels 126 are closed at the entrance surface 120. However, instead of the sealing plug 128, it is also possible for the inlet channels 124 to taper towards the outlet surface 122 until they touch the walls of an inlet channel 124 and the inlet channel 124 is closed in this way. In this case, the inlet channel 124 in the direction parallel to the axis of symmetry 116 has a triangular cross-section.

Accordingly, the outlet channels 126 are open at the outlet surface 122 and closed in the region of the inlet surface 120. The flow path of the unpurified waste gas thus leads into one of the inlet channels 124 and from there through a filter wall 130 into one of the outlet channels 126. This is illustrated by the arrows 132 in FIG. 1 by way of example.

FIG. 2 shows the ceramic substrate of the filter wall 130 in a schematic detail. The filter wall has a (mean) wall thickness L, which is usually given in mil.

It can be seen that the filter substrate in addition to a plurality of Mirkorissen or cracks 112 a plurality of micro channels 134, micropores 136 and pores 138 has. Transitions between large pores 138 are designated by the term "microchannels" 134. Figure 2 also shows a representation of the flow path 132 through the filter wall 130, which is not to scale and which illustrates that the flow path 132 extends substantially through the pores 138. In the following, "micropores" 136 are bulges of larger pores 138. With the above-described mercury porosimetry method, essentially the pores 138 and the micropores 136 are detected. As a rule, however, those pores 138 and micropores 136 which are arranged in the interior of the filter, ie pores 138 and micropores 136, which have no connection to the filter surface and / or to other pores (blind pores) are generally not detected.

As described above, the filtering action upon passage of the exhaust stream 118 through the filter wall 130 is comprised of surface filtration and depth filtration. Strong depth filtration (i.e., incorporation of diesel soot into filter walls 130) results in poor filtration efficiencies and high back pressure. The lower the proportion of depth filtration, i. H. an early transition to surface filtration, the lower the back pressure of the filter under soot loading.

In Figure 3, loading curves of a conventional filter (reference numeral 310) and a filter according to the invention (reference numeral 312) are shown schematically in comparison. listed here is the differential pressure Δp in mbar against the amount of soot introduced in g / m ² filter surface. Here, only the geometric filter surface is taken into account, ie the filter surface without pores, which is available for the filtration.

Characteristic of the surface filtration (soot deposition on the top of the filter wall 130) is a linear increase of the loading curve shown in FIG. This linear increase is indicated in each case by the rising dashed lines 314 in FIG.

In small loading areas, however, d. H. for small values of M, there is no linear increase in differential pressure with soot loading, but a non-linear increase due to depth filtration. Metrologically, the depth filtration of a filter substrate can be effected by the straight-line extrapolation of the surface filtration straight line 314 shown. The intersections of these straight lines 314, which are denoted by b, are compared with the actual value at M = 0. The difference, designated by reference numeral 316 in FIG. 3, results in depth filtration. This depth filtration is thus the difference between the pressure at surface filtration and the initial pressure of the fresh filter without visualization:

dDp (depth filtration) = b (pressure at surface filtration) - Dp (MR = O) (initial pressure at

0 g / m ² Berußung). (5).

As described above, FIG. 3 contains pressure curve curves for two different filter materials. The curve 310 describes a standard filter material that corresponds to the prior art, whereas the curve 312 denotes a pressure curve for a filter material having optimized, so minimized depth filtration and thus a minimized back pressure under soot loading. The depth filtration accordingly also has the unit mbar.

Decisive for the characterization of a filter are therefore on the one hand, the filtration efficiency, which is given in percent and represents a measure of the retention capacity of soot particles of the filter. Furthermore, the counter-pressure shown in Figure 3, so the differential pressure at a current soot load, decisive, which is indicated in mbar and the resistance of a filter material (under soot loading) in the exhaust system features. The filtration efficiency generally increases with increasing thickness of the filter walls (wall thickness L). Thus, the wall thickness must be included in the pore size distribution.

Another decisive parameter for characterizing a filter wall 130 is the pore size distribution. In the present case, this pore size distribution is preferably measured by means of mercury porosimetry (MIP). From the cumulative pore volume cumulative curve, a characteristic distribution value D10, D50 ("average" pore diameter) and D90 can be determined, as shown in Figure 4. In this case, the cumulative pore volume Vkum is given in percent on the y-axis, whereas on the x For example, a gaussian frequency distribution of the pore diameters may be present, integration over this frequency distribution then yielding the curve rising steeply in the vicinity of D50 to the value 100% shown in FIG.

The value D50 is obtained, for example, by forming the intersection of the curve Vkum with a 50% straight line and determining the corresponding associated value for the pore size D. Accordingly, the values D10 and D90 are determined.

By difference formation of D50 and D10, with subsequent normalization to D50 [d. H. (D50-D10) / D50] gives a pure ratio which characterizes the width of the pore size distribution. This ratio is a measure of the width of the pore fine fraction in a filter substrate.

By means of the measurement method for the size (D50 - D10) / D50 illustrated with reference to FIG. 4, various types of filter substrates were examined in accordance with the plot in FIG. 3, and the respective depth filtrations 316 were determined in each case. The results of these screening tests are plotted in FIG. In this case, the depth filtration 316 in mbar, which was determined according to FIG. 3, is plotted on the y-axis. On the x-axis the characteristic value [(D50 - D10) / D50] ^• L is plotted in the unit mil.

In this case, shaded deposited an area shown in which a depth filtration is present, which is below 20 mbar. The following conditions apply to the shaded area:

1. [(D50 - D10) / D50] L <9 mil 2. ddp <20 mbar.

In addition, it has been found in the investigations that for the hatched area the condition D50 <20 microns, i. all points that lie in this area in FIG. 5 have a maximum D50 of 20 micrometers.

From a pore fine ratio [(D50 - D10) / D50] ^• L <9 mil and an average pore diameter D50 <20 μm, a significant minimization of the depth filtration dDp 316 can be observed. These conditions apply to a depth filtration dDp <20 mbar.

Adhering to these limits gives a filter which, as Figure 3 shows, is significantly improved in terms of back pressure and filter efficiency and has significant advantages over standard filters. In particular, the curve 312 of the filter according to the invention runs parallel, but clearly below the curve 310 of the standard filter. The initial nonlinear region, which designates the depth filtration 316, is also significantly reduced in the filter according to the invention compared to the prior art. Overall, thus, the entire pressure drop across the filter element can be reduced, at the same time the filtration efficiency is improved.

This improvement in filtration efficiency is shown schematically in FIG. In each case, the course of the filtration efficiency (curves 610, 612) and the pressure rise (curves 614, 616) as a function of the soot load M (soot amount in grams per area in m ² ) is plotted schematically. Curves 610 and 614 show traces of a conventional filter material (dots outside the hatched area in FIG. 5), whereas curves 612 and 616 show traces for a filter material according to the invention (dots within the hatched area in FIG. 5). The filtration efficiency or filter efficiency η is in each case the number of particles filtered out of an air flow per total number of particles penetrating the filter. The information is given here in percent.

Curves 614 and 616 thus correspond to curves 310 and 312 in FIG. 3, respectively. A slight difference arises in the initial region of curve 614, which, in contrast to curve 310, shows a slight S-turn so that curves 614 and 616 cut at low load. This effect can basically also occur in FIG. 3, but was neglected there. This does not change at the depth filtration 316 determination, and at high loadings, the curve 614 extends above the curve 616. As can also be seen from FIG. 6, a minimal or minimized depth filling region 316 has an improved initial (initial) filter efficiency η. This is shown in a comparison of curves 610 and 612. For improved filter material (ie, points within the hatched area in Figure 5), the filter efficiency as a function of the soot load of the filter increases more rapidly to its final value of approximately (nearly) 100% then to be constant (curve 612). By contrast, with a poorer filter material (points outside the hatched area in FIG. 5), the filtration efficiency η only approaches the final value later (curve 610).

Claims

claims

1. A method for producing a ceramic filter element (114) for reducing pollutants in an exhaust gas stream, in particular a diesel particulate filter, wherein the filter element (114) has a filter wall (130) through which the exhaust gas stream flows, wherein in the method a ceramic starting material is formed into a shaped body and a sintering process, wherein the starting material and / or at least one process parameter of the method are selected such that the following relationship applies to the pore size distribution:

D50 - D10

^■ <9 mil / L [mil],

D50

where D50 is the cumulative pore size below which 50% of the pores fall, where D10 is the cumulative pore size below which 10% of the pores fall, and L is the wall thickness of the filter wall (130) through which the exhaust stream flows.

2. The method according to the preceding claim, wherein the starting material and / or at least one process parameter of the method are selected such that in the filter element (114) adjusts a pore size distribution with a pore size D50, which is a maximum of 20 microns.

3. The method according to any one of the two preceding claims, wherein a starting material is used with a first component having particles having a particle size distribution, wherein the space occupied by the particles of the first component in the production of the filter element (114), in particular at the

Sintering process is replaced by pores of appropriate size.

4. The method according to the preceding claim, wherein for the particle size distribution, the relation

P50 - P10

■ <9 mil / L [mil]

P50

where P50 is the cumulative particle size below which 50% of the particles fall, where PlO is the cumulative particle size below which 10% of the particles fall.

5. The method according to one of the two preceding claims, wherein the first component comprises an organic pore-forming agent.

6. The method according to any one of the three preceding claims, wherein the first component comprises soluble salt as a pore-forming agent.

7. The method according to claim 4, wherein the ceramic filter element comprises at least one sintered ceramic produced by reaction sintering, wherein the sintering process comprises a reaction sintering step in which a first starting material melts and reacts with at least one second starting material to form pores which substantially correspond to the shape of the first starting material.

8. Method according to the preceding claim, wherein the sintered ceramic comprises a silicate ceramic, in particular an aluminosilicate ceramic, in particular a cordierite ceramic

Ceramics.

9. The method according to one of the two preceding claims, wherein the first starting material comprises at least one of the following substances: talc; SiC> second

10. Ceramic filter element (114) for pollutant reduction in an exhaust gas flow, wherein the filter element (114) has a flow through the exhaust flow filter wall (130) having a wall thickness, wherein the filter wall (130) is porous, wherein the following relation applies to the pore size distribution :

D50 - D10

■ <9 mil / L [mil],

D50

11. Ceramic filter element (114) according to the preceding claim, wherein the pore size D50 is at most 20 micrometers.