WO2021152043A1

WO2021152043A1 - Polycrystalline ceramic with improved properties

Info

Publication number: WO2021152043A1
Application number: PCT/EP2021/052032
Authority: WO
Inventors: Jürgen Rödel; Lukas PORZ; Xufei FANG; Arne Jan KLOMP
Original assignee: Technische Universität Darmstadt
Priority date: 2020-01-31
Filing date: 2021-01-28
Publication date: 2021-08-05
Also published as: DE102020102527A1

Abstract

The present invention relates to a polycrystalline ceramic having improved properties, in particular having improved crack opening fracture toughness (also referred to as intrinsic fracture toughness in metal science) and preferably ductility. The invention further relates to a method for producing the ceramic.

Description

Polycrystalline ceramics with improved properties

The present invention relates to a polycrystalline ceramic with improved properties, in particular with improved crack tip fracture toughness (in metallurgy also referred to as intrinsic fracture toughness) and preferably plastic deformability. The invention also relates to a method for producing the ceramic.

Polycrystalline ceramics are known from the prior art. Ceramic materials have many advantages over metals, e.g. corrosion resistance, higher melting point or functional properties. However, all technical ceramics are brittle at room temperature, which severely limits the areas of application. Metals are tough and ductile and fail “good-natured”, while ceramics fail due to catastrophic breakage. This means that, on the one hand, stress peaks are not reduced, which means that cracks can easily develop and grow in the ceramics, and on the other hand, the materials are almost impossible to deform plastically. This prevents their use in numerous areas, severely restricts the processing methods and thus the exploitation of the potential of ceramic materials.

Against this background, the object of the present invention is to overcome the disadvantages of the prior art and to provide correspondingly improved polycrystalline Kera mics.

The object is achieved by the subject matter of the claims.

The object is achieved in particular by a polycrystalline ceramic, a) the ceramic with nanoindentation with a tip radius of 50-200 nm at least 30 out of 50 impressions with a minimum distance from each other of 20 pm at a maximum load of 5 mN below preferably 1 .5 GPa, more preferably less than 1 GPa, even more preferably less than 0.5 GPa, exhibits non-elastic behavior and / or the ceramic exhibits a transition from elastic to plastic behavior with at most 20 out of 50 impressions. "Pop-in") above 1 GPa, and b) wherein the ceramic has a crack tip fracture toughness of at least 3.5 MPa m ^1/2 , more preferably at least 4 MPa m ^1/2 , even more preferably at least 5 MPa m ^{1 / 2} when implemented in any ceramic with the exception of ceramics with a Pero- Has vskitkristallstruktur. The ceramic preferably has a crack tip fracture toughness of 3.5 MPa m ^1/2 to 40 MPa m ^1/2 , preferably from 5 MPa m ^1/2 to 20 MPa m ^1/2 . For ceramics with a perovskite crystal structure, a crack tip fracture toughness ^{of at least 2 MPa m 1/2 is} preferred, more preferably of at least 2.5 MPa m ^1/2, still further preferably of at least 3.5 MPa m ^1/2 . For ceramics with a perovskite crystal structure, a crack tip fracture toughness of 2 MPa m ¹ ' ² to 35 MPa m ^{1/2 is} preferred, more preferably from 3 MPa m ^1/2 to 20 MPa m ^1/2 Clearly distinguishable from the total fracture toughness (commonly referred to as fracture toughness), which also includes mechanisms far behind the crack tip such as crack bridging. Only with a high crack tip fracture toughness can short cracks of, for example, 1 μm or 10 μm in length be effectively kept open, whereby a massively increased strength and operational reliability can be achieved.

The increased ductility from a) preferably leads to reduced crack formation, so that the ceramic with a Vickers impression of at least 10 g, preferably of at least 15 g, further preferably of at least 25 g, even more preferably of at least 50 g, even more preferred of at least 100 g, less than an average of 1.5 cracks. Optionally, the Vickers impression carries a maximum of 150 g, preferably a maximum of 100 g, preferably a maximum of 60 g, preferably a maximum of 30 g, preferably a maximum of 20 g. Optionally, the ceramic has more than 0.01, preferably more than 0.05, cracks. As described in particular in b), the operational reliability of the ceramic is preferably controlled by controlling short cracks. In this case, the crack length is particularly preferably reduced to zero.

The ceramic preferably has at least 1 * 10 ⁹ dislocation sources per cm ³ but not more than 10 ¹⁸ cm ^-3 and / or at least one dislocation source per grain (in contrast to one-dimensional dislocations, which ^{are measured in cm -2} , zero-dimensional displacements are measurement sources in cm ^-3 ).

Sources of dislocation include, for example, vacancy clusters, dislocation rings (slip rings) which are left behind by dislocation movement, grain boundaries (under certain conditions), vacancy complexes which are left behind by dislocation movement and kinematic sources ("Theory of Dislocations" 3 ^rd edition Peter M. Anderson, John P. Hirth and Jens Lothe, Cambridge University Press).

Dislocations are one-dimensional crystal defects which are responsible for the plastic deformability of most metals and which also allow very local plastic deformation there. This results in a reduction in voltage peaks at stress concentrators like notches or pores. They also form at the tip of a crack by nucleating there and allowing ductile deformation through their movement, which leads to a massively increased fracture toughness. This region at the top of the crack is known as the "plastic zone". According to the relevant opinion in the specialist literature, (Ritchie, RO The conflicts between strength and toughness. Nat Mater 10, 817-822, doi: 10.1038 / Nmat3115 (2011) and Evans, AG, Perspective on the Development of High-Toughness Ceramics. Journal of the American Ceramic Society, 1990. 73 (2): p. 187-206.), This mechanism for ceramics was never considered for interlocking.

However, there are many ceramics in which dislocation movement is possible at room temperature and thus the central basic requirement for a plastic zone, that dislocations are mobile, is given. Their single crystals have often been successfully deformed at room temperature (Brunner, D., et al., Surprising Results of a Study on the Plasticity in Strontium Titanate. Journal of the American Ceramic Society, 2001. 84 (5): p. 1161-1163. Amodeo, J., et al., Dislocations and Plastic Deformation in MgO Crystals: A Review. Crystals, 2018. 8 (6). Gilman, JJ and WG Johnston, Dislocations in Lithium Fluoride Crystals. Solid State Physics-Advances in Research and Applications, 1962. 13: pp. 147-222). Despite the deformability of the single crystals, polycrystalline ceramics cannot be plastically deformed as is the case with metals. To explain this, the Taylor criterion must be explained: It states that five independent sliding systems are necessary for any deformation of a polycrystal. In 1963 it was discussed that many ceramics did not have enough independent slip systems (Groves, G.W. and A. Kelly, Independent slip systems in crystals. Philosophical Magazine, 1963. 8 (89): p. 877-887.). If there are not enough independent sliding systems, a material fails after minimal plastic deformation, e.g. below 0.5% (Scott, W.D. and J.A. Pask,

Deformation and Fracture of Polycrystalline Lithium Fluoride. Journal of the American Ceramic Society, 1963. 46 (6): p. 284-294.). However, there are also ceramics that fail completely without plastic deformation. From an application point of view, the difference is negligibly small and has not been dealt with further in research. This difference, even if it is minimal, cannot be described physically correctly by the usual mechanisms.

It has now been found that, in addition to the mobility of dislocations and the number of independent slip systems, the dislocation source density also play a role for the ductility and crack-tip fracture toughness. In contrast to metals, the dislocation density in ceramics is generally very low and is usually below 1 * 10 ⁶ cm ^-2 (see Figure 1). Even by multiplication of dislocations in the event of deformation, which is only possible in single crystals, the dislocation density can be reduced to a maximum value in the range of 1 * 10 ⁹ cmr ² (for strontium titanate e.g. 1, 8 * 10 ⁹ cmr ² , see Figure 2). If there are sources or other crystal defects, new dislocations can easily be generated from them - also in ceramics. However, these sources are extremely rare in ceramics. Their absence at the crack tip, the area around the crack tip and near or at the grain boundaries thus prevents the formation of an effective plastic zone in the polycrystalline. One object of the present invention is therefore to provide polycrystalline ceramics with improved mechanical properties, in particular with improved crack tip fracture toughness.

For millennia, ceramics have been manufactured using the sintering process, where there are many variations. We are now proposing to substantially change the sintering manufacturing process, to replace it completely or to supplement it with new types of pre- or post-treatment.

The mechanisms known for metals cannot easily be transferred to ceramics for the following reasons.

1) The dislocation density in the known ceramics is several orders of magnitude lower than in metals, which means that the dominant mechanisms are fundamentally different.

2) In the case of metals, the "Frank-Read source" and the mechanisms derived from it are the most important sources of dislocation. Often only the Frank-Read source is taught in the relevant lectures as a displacement source, which is sufficient to largely understand the behavior of metals. However, this plays almost no role in ceramics. In contrast, a number of other mechanisms are important in ceramics.

3) In most metals it is possible to create dislocations directly at the crack tip. This is very helpful for the creation of a “plastic zone”, which substantially increases the fracture toughness. However, this is a key problem in ceramics. In turn, the generation of dislocations at the crack tip as a starting point for a plastic zone in metals is completely natural.

4) In metals, the number of dislocations is often increased by deformation in order to harden the metals, with cold rolling being a known method for this, with dislocations in ceramics, as opposed to metals, being said to lead to softening. Apart from other reasons, the deformation is already due to the Taylor criterion in ceramics not possible. All ceramics according to the state of the art fail brittle when attempted to deform. The dislocation density can therefore not be increased by deformation in polycrystals. ) Polycrystalline ceramics that do not contain any dislocations fail without any significant plastic expansion. Polycrystalline ceramics, which contain (few) misalignments, fail after less than 0.5% plastic elongation. From an engineering point of view, this difference appears completely negligible and experimental examples of this are very rare. Furthermore, the Taylor criterion states that most ceramics do not have enough sliding systems for any deformation, which means that they must fail after less than 0.5% deformation. So far, this suggests that dislocations in ceramics are not worth discussing further. The transfer of deformability from one grain to the next is not only limited by the Taylor criterion, but also by the possibility of nucleating dislocations in the neighboring grain, even if the observed difference in mechanical properties is initially very small. ) In addition to the irrelevance for direct technical application mentioned in point 5), important mechanisms are extremely difficult to observe experimentally in ceramics. Since they occur in the same area in which the Taylor criterion calls for a failure of the ceramic, delimitation is difficult. Therefore, the level of knowledge about which mechanism is transferable and which is not is patchy to this day. ) In ceramic single crystals, dislocations can be introduced by mechanical deformation based on metals. A maximum dislocation density in the range of 1 * 10 ⁹ cm ^-2 is established. An increase in the crack tip fracture toughness is, however, not noticeably observed even in areas of maximum dislocation density and the crystals often shatter during the deformation attempt. No case is known in which deformation has led to a significant increase in the crack tip fracture toughness. ) The aforementioned points affect ductility and fracture toughness in an intertwined manner. For a correct understanding, it is necessary to do several of the aforementioned thought steps at the same time and to understand their effects on one another. If one of the points is initially left out or discussed separately, the result presented here will not be reached. This includes the knowledge a) that dislocations are mobile in many ceramics, b) that the dislocation density is extremely low, c) the dislocation source density is extremely low, d) that the multiplication of dislocations is very simple, but the nucleation of dislocations does not take place easily, e) the dislocation density achievable by dislocation multiplication has a clear upper limit f) dislocation densities beyond this natural upper limit are required g) the knowledge nis that, in addition to the Taylor criterion, the possibility of nucleating new dislocations in the neighboring grain also determines the transfer of deformation from one grain to the next, although this initially makes an almost negligible difference in macroscopic deformation tests, and, h) that the nucleation of dislocations at the crack tip in metals is simple, but a central problem in ceramics, whereby rather unusual dislocation sources play the central role.

In ceramic single crystals, the few mobile dislocations in each crystal volume can be moved and multiplied, which means that the entire crystal can be easily deformed. However, very small volumes in individual grains are responsible for the mechanical properties of a polycrystal, which must be able to perform a deformation independently, e.g. at a grain boundary or at a crack tip. There are enough sources of dislocation in metals and new dislocations can be nucleated at a crack tip. In ceramics, however, due to the small number, statistically speaking, dislocations are rarely present in the relevant volume and a nucleation of dislocations under relevant conditions is not possible. Therefore, the critical volumes, e.g. at a grain boundary or a crack tip and the area around the crack tip, cannot deform, which means that the behavior of ceramics is brittle.

The solution to the problem by the present invention is to increase the displacement source density. With a particularly high density of dislocation sources, contrary to all expectations, a plastic zone can be made possible.

The dislocation density in polycrystalline ceramics cannot be increased by deformation as in metals, since deformation is not possible according to the prior art. Even in ceramic single crystals, which are often deformable, there is an upper limit dislocation density caused by the multiplication mechanisms, which is not sufficient for a plastic zone. Deformation must be possible in any volume at low stresses. The possibility of this can be tested using nanoindentation (see Figure 3 and Figure 4). A substantially increased dislocation source density can provide the necessary dislocation density in the relevant volumes.

In metals it is possible to nucleate dislocations at a crack tip. The nucleation he follows with metals in the case of stress in the desired volume without further action. this leads to a blunted crack tip and thus to an increased crack tip fracture toughness. The high crack tip fracture toughness ensures that the stresses are well distributed and dislocation movement can take place in a very large area around the crack tip. The larger the area, the lower the dislocation density at which the plastic zone can strengthen itself. In contrast, nucleation of dislocations at the crack tip is a central problem with ceramics at room temperature. Furthermore, dislocations and sources of displacement in ceramics according to the state of the art are also not sufficiently present in the vicinity of the crack tip and on grain boundaries and their surroundings.

The crack tip fracture toughness in a ceramic of the prior art is extremely low. This means that the area in which the stress is large enough for dislocation movement to take place at all is extremely small. The size of the area is presumably inversely quadratic proportional to the crack tip fracture toughness. In this area, suitable dislocations must occur in this area, which is much smaller in the case of ceramics than in metals. This means that an even greater dislocation density or dislocation source density in the environment is necessary in order to obtain a self-reinforcing mechanism of the plastic zone. In addition, grain boundaries without nucleation in neighboring grains would restrict the deformation to individual grains and thus significantly reduce the size of the plastic zone. The restriction of the deformability to one grain also excludes the plastic deformability of a polycrystal. Therefore, a nucleation in the desired volume (e.g. the grain boundary or its proximity and the area around the crack tip) must be brought about.

The solution to the problem by the present invention consists in increasing the dislocation source density, which on the one hand enables dislocations to be provided in any relevant volume and is also the key to achieving the local dislocation density. In this way, contrary to all expectations, local deformation in any volume and thus also a plastic zone is possible. On the one hand, dislocation sources enable nucleation at or near the crack tip. On the other hand, a particularly high dislocation source density can locally increase the dislocation density by more than a hundred times the value that can be achieved by multiplication. It is also possible to activate dislocation sources under special conditions (e.g. additional hydrostatic pressures of e.g. over 500 MPa or temperatures of e.g. over 800 ° C) already during production.

Novel synthesis routes are necessary for this, as sintering, which is the established manufacturing process for ceramics, very effectively reduces the dislocation density and the dislocation source density to a minimum. The basic feasibility of increasing the offset source density was recently demonstrated (Jin, K., et al., Quantifying early stage Irradiation damage from nanoindentation pop-in tests. Scripta Materialia, 2018. 157: p. 49-53). So far it was however, it is not known that the dislocation source density is an important parameter to improve the above properties of polycrystalline ceramics. If enough independent sliding systems are mobile, the introduction of sufficient dislocation sources results in complete plastic deformability. A local deformation and thus the formation of a plastic zone and consequently an increase in the crack tip fracture toughness is also possible with less than 5 independent sliding systems. In Groves, GW and A. Kelly, Independent slip Systems in crystals. Philosophical Magazine, 1963. 8 (89): p. 877-887 some metals are named which do not have enough independent sliding systems. Nevertheless, these metals are much tougher than all ceramics. A significantly higher ductility is also observed.

If the grain size, which is preferably determined in accordance with DIN EN 623-3, is particularly small, disadvantageous competing deformation mechanisms can occur particularly easily. Furthermore, the distance that a dislocation can cover and thus the effectiveness of the dislocations and dislocation sources for the desired properties is particularly low in the case of small grains. A grain size of preferably 100 nm to 250 μm, more preferably from 200 nm to 100 μm, is therefore advantageous.

In order to increase the dislocation source density (in particular in the entire volume), a number of methods are possible according to the invention. Since, for example, the melting point of the claimed materials extends over a range of over 2000 ° C (e.g. of lithium fluoride: 848 ° C, e.g. of magnesium oxide: 2852 ° C), the process parameters vary significantly between the material classes. In particular, these methods are preferably selected from the group consisting of:

- Dislocation creep at high temperature, in particular between 50% and 90% of the melting temperature in Kelvin with mechanical loads that lead to plastic deformation of more than 0.5% within a period of less than 10 days,

- Plastic deformation at high temperature, in particular between 30% and 98% of the melting temperature in Kelvin at mechanical loads that are at least the yield point of the ceramic at that temperature, with plastic deformation between 0.5% and 80%, preferably between 0, 5% and 25%, more preferably between 0.5% and 20% and elongation rates in the range from 10 ⁵ s ^_1 to 10 ³ s ^_1 ,

- Reduction and thereby condensation of oxygen vacancies and formation of dislocation loops in an atmosphere with an oxygen partial pressure in the preferred range from 10 ⁸ bar to 10 ³ ° bar, more preferably from 10 ¹⁵ bar to 10 ²⁷ bar, - Irradiation preferably with massive radiation particles such as ions and neutrons and formation of radiation damage that develops into dislocations and dislocation sources,

- micro-precipitates (see Figure 11),

- Electric fields during the sintering process, including flash sintering or spark plasma sintering with currents in the preferred range of, for example, at least 0.1 Acnr ² and at most an average of 2 MAcnr ² ,

- aerosol deposition (a process of room temperature deposition of grains with plastic deformation),

- Irradiation in the grid with femtosecond laser pulses with a focus point in the material volume. Pulse energies of, for example, 5 to 10 microjoules, especially about 7.5 microjoules, are preferred. Multiple pulses of e.g. 10 to 1000 pulses are preferred,

- Temperature cycles with cooling rates of more than 25 Kelvin per minute, and

- combinations of two or more of them.

It is emphasized that the invention can be applied to many classes of materials, e.g. perovskites, diborides of transition metals, ceramics with a sodium chloride crystal structure, ceramics with a zinc blende crystal structure and ceramics with a wurtzite crystal structure. Some semiconductors can also be regarded as ceramics. Specifically mentioned are e.g. the ceramics of the titanates and niobates of the perovskite structure, e.g. BaTiCh, SrTiCh, KNbCh, then diborides such as Zrß2, T1B2, HfB2, materials of the sodium chloride structure such as LiF, MgO and ceramics of the zinc blende or wurtzite structure such as ZnS and ZnO.

The transferability to other material classes is given in particular if in that material class dislocation movement is possible at the relevant temperature, e.g. room temperature. This can be shown, for example, by the deformation of single crystals or by indentation. All ceramics and ceramic classes in which the mobility of dislocations is given are according to the invention. The person skilled in the art can easily determine whether a certain class fulfills this requirement without inventive intervention, for example by deforming single crystals or by indentation. This can be checked individually for each ceramic or is already known from various literature sources. An example of this transferability is given below. In the following study, indentation in CaF2 at room temperature was investigated (Shikimaka, O. & Grabco, D., Deformation created by Berkovich and Vickers indenters and its influence on surface morphology of indentations for Li F and CaF ₂ single crystals. J Phys D Appi Phys 41, doi: Artn 074012 10.1088 / 0022-3727 / 41/7/074012 (2008)). There it is shown by etch pits in Figure 1 that many dislocations have formed around the indentation. Therefore it is shown that dislocations in CaF2 are sufficiently mobile at room temperature. This means that it can also be implemented with CaF2. Furthermore, it can be implemented for all ceramics with the same crystal structure.

Ceramics are preferably selected from the group which is characterized by possible displacement mobility at room temperature.

The following material classes, defined by their crystal structure, are particularly preferred for the invention, with particularly preferred specific materials being given for each crystal structure:

Ceramics with a perovskite crystal structure (e.g. BaTiCh, SrTiCh, KNbCh), ceramics with a sodium chloride crystal structure (e.g. NaCI, LiF, MgO), ceramics with a zinc blende crystal structure (e.g. ZnS), ceramics with a wurtzite crystal structure (ZB ZnO) and / or ceramics with a calcium fluoride crystal structure (ZB CaF2).

Further examples of preferred ceramics are e.g. ZrB2 and HfB2.

As an alternative or in addition, the invention can be applied to any polycrystalline ceramic, and therefore this ceramic is used, which is characterized by the natural property of displacement mobility at room temperature.

The invention is described in more detail below.

The invention preferably relates to a polycrystalline ceramic a) with a dislocation source density of at least one dislocation source per grain, a combination of dislocation source density and dislocation source quality is characterized by the ability to be preferably 1.5 GPa, more preferably 1, in any volume in the polycrystal GPa, more preferably of 0.5 GPa displacements (with nanoindentation, the ceramic preferably has at least 30 out of 50 indentations with a minimum distance from one another on a surface that has been prepared from the volume and properly polished with a tip radius of 50-200 nm of 20 pm at a maximum load of 5 mN below preferably 1.5 GPa, more preferably below 1 GPa, even more preferably below 0.5 GPa, and / or the ceramic exhibits non-elastic behavior at a maximum of 20 out of 50 impressions a transition from elastic to plastic behavior (English: "pop-in") above 1 GPa.), and b) with a crack tip fracture toughness of at least 3.5 MPa m ^1/2 , more preferably min at least 4 MPa m ^1/2 , even more preferably at least 5 MPa m ^1/2 when implemented in any ceramic with the exception of ceramics with a perovskite crystal structure. For ceramics with a perovskite crystal structure, a crack tip ^{fracture toughness of 2 MPa m 1/2 is} preferred, more preferably of at least 2.5 MPa m ^1/2, even more preferably of 3.5 MPa m ^1/2 . The crack tip fracture toughness only describes the toughness at the crack tip and can be clearly distinguished from the total fracture toughness (commonly referred to as fracture toughness), which also includes mechanisms far behind the crack tip such as crack bridging. Only with a high crack tip fracture toughness can short cracks of, for example, 1 μm or 10 μm in length be effectively stopped, whereby a massively increased strength and operational reliability can be achieved.

The values in a) with the unit GPa are preferably maximum stress values in the polycrystalline.

Sources of dislocation are in particular those that are activated by the stress field of a crack instead of a homogeneous stress field. Are dislocation sources for the basic book "Theory of dislocations" (3 ^rd edition Peter M. Anderson, John P. Hirth and Jens Lothe, Cambridge University Press 2017), for example, vacancy clusters, seal rings, grain boundary sources and kinematic sources. The dislocation sources are preferably selected from the group consisting of vacancy clusters, slip rings, grain boundary sources, kinematic sources, mobile dislocations and combinations thereof. The dislocation sources are of particular importance for the desired properties, but quantification is difficult since a dislocation source can have different shapes and can generate different numbers of dislocations.

The lower limit of the dislocation source density that is present in a material can be estimated. In the event of a load, the dislocation source must have generated a dislocation. With the exception of small-angle grain boundaries, which is rare in polycrystals, grain boundaries are not permeable to dislocations in the case of relevant stresses. Thus, in every grain in which at least one mobile dislocation can be observed after the load case, at least one dislocation source must have been present. However, one dislocation source can result in exactly one or 100 dislocations, for example, whereby this depends on several factors. A dislocation source density in the range of 1 * 10 ¹² cm ^-3 (this corresponds to one source per pm ³ ) appears to be a desirable dislocation source density. If 100 dislocations with a length of 1 pm are nucleated per dislocation source, then this results in a dislocation density of 1 * 10 ¹⁰ rrr ² in the case of loading. Furthermore, a displacement source density in the preferred range of 1 * 10 ⁹ cm ^-3 to 1 * 10 ¹⁵ cm ^{-3 is} desirable. The displacement source density is preferably in a range from 10 ⁹ cm ^-3 to 10 ¹⁸ cm ^-3 , more preferably from 10 ¹ ° cm ^-3 to 10 ¹⁷ cm ^-3 , more preferably from 10 ¹² cm ^-3 to 10 ¹⁷ cm ^-3 , more preferably from 10 ¹³ cm ^-3 to 10 ¹⁶ cm ³ , more preferably from 10 ¹⁴ cm ³ to 10 ¹⁶ cm ³ . A dislocation source density of at least 10 ⁹ cm ³ , more preferably at least 10 ¹ ° cm ³ , more preferably at least 10 ¹¹ cm ³ , more preferably at least 10 ¹² cm ³ , more preferably at least 10 ¹³ cm ³ , is particularly advantageous for achieving the properties described above.

After appropriate exposure, dislocation sources according to the invention must have generated dislocations; these dislocations must be detected with: transmission electron microscopy, etching pits, rocking curve measurements, dark field X-ray microscopy, electron channeling contrast imaging, electron backscattering diffraction or other methods.

The dislocation source density (per cm ³ ) can be estimated as follows. Sources of dislocation generate local dislocations in the relevant load case. A relevant load case is, for example, the break, whereby the relevant volume is in the vicinity of the crack tip. If the crack advances, dislocation sources according to the invention produce dislocations in the vicinity of the crack tip. The line-shaped crack tip creates a new surface as it progresses, which later becomes the fracture surface. By definition, every point on the fracture surface was in the immediate vicinity of the crack tip at a certain point in time. Dislocations can therefore be found under the fracture surface, for example at a depth of 200 nm, 1 pm, 2 pm, 5 pm or 25 pm. The resulting dislocations on the fracture surface can with established methods can be demonstrated. To prove this, the ceramic is broken, for example in accordance with ASTM C 1161 (Standard Test Method for Advanced Ceramics at Ambient Temperature). Dislocation sources according to the invention have generated dislocations during fracture, which can be detected on the fracture surface. Evidence of the dislocation can be provided by various imaging methods. To do this, a sample must be prepared from the fracture surface. For this purpose, one or more pieces can be prepared from the fracture surface. This also includes the option of cutting out a slice perpendicular to the fracture surface. For observation, the cut surface, but not the fracture surface, can be properly processed, for example polished.

Preferred methods of detecting dislocations are as follows.

• "Electron channeling contrast imaging" is a scanning electron microscope-based imaging method which can directly visualize dislocations under a properly polished surface.

• Transmission electron microscopy. For this purpose, a tiny sample is dissected from the area of the fracture surface. The sample then has a thickness of a few 100 nm in order to be permeable to the electron beam. Dislocations are clearly visible in the corresponding images as lines in the dark field contrast. In the examples two images are given, one with and one without offsets. (Figure 1 with almost no dislocations, as in the prior art and Figure 12 with many dislocations, according to the invention)

Other methods of visualizing dislocations are also possible.

In the case of polycrystalline ceramics with a dislocation source density of at least one dislocation source per grain, the dislocation density after breakage is at a depth of 200 nm,

1 pm and 2 pm, preferably also in 5 pm, more preferably also in 20 pm below the surface, more than one dislocation per grain; The dislocation source density per grain was at least 1.Furthermore, with the above-mentioned calculation example, assuming that a dislocation source nucleates 100 dislocations with a length of 1 pm, a dislocation density of 10 ¹⁰ cm ^{-2 can be} approximately a dislocation ^{source density of 10 12} cm ^-3 , whereby the two assumptions mentioned must be checked on a case-by-case basis.

In all cases there can be areas in which no dislocations have occurred. This is due, for example, to the Taylor criterion. Therefore, evidence of the absence or presence of dislocations is to be carried out at preferably at least ten randomly selected locations on the fracture surface. The choice of positions is to be documented and illustrated. the Dislocation density is determined at distances of 200 nm, 1 pm, 2 pm, 5 pm, 10 pm, 20 pm and 50 pm to the fracture surface. Images of a size of at least 1 megapixel, preferably of at least 4 megapixels, with a pixel size of less than 10 nm, preferably of a pixel size of less than 3 nm, preferably of a pixel size of less than 1 nm, are preferred. Overview images, each with a five times larger field of view, are to be enclosed and the selected point for the measurement of the dislocation density must be marked to prove that the selection of images is representative. In the overview images, the grains must be recognizable as a whole. In the case of polycrystalline ceramics with a dislocation source density of at least one dislocation source per grain, the dislocation density after the break at a depth of 200 nm, 1 pm and 2 pm, preferably also at 5 pm, more preferably also at 20 pm below the surface, is more than one dislocation per grain; So the dislocation source density per grain was at least 1. If a grain only partially protrudes into the area of less than 2 μm, that grain must also be considered. 60% of the grains in the aforementioned depths should preferably have dislocations. The phenomenon of a “dislocation free zone” at a crack tip should be mentioned, according to which dislocations directly at a crack tip can often hardly be detected because they do not remain there, but only come to a stop a short distance away from the crack tip. Extensive temperature treatment above 200 ° C, for example, should be avoided between the break and the imaging analysis.

A combination of dislocation source density and dislocation source quality, which describes the ability to generate dislocations in any volume in the polycrystal, preferably 1.5 GPa, more preferably 1 GPa, more preferably 0.5 GPa, can be quantified more clearly than the dislocation source density alone. A ceramic according to the invention is preferably distinguished by this ability. This can be measured by means of nanoindentation on any surface that has been prepared and properly polished from the volume. With a tip radius of 50-200 nm, at least 30 out of 50 impressions with a minimum distance from one another of 20 μm at a maximum load of 5 mN are preferably below 1.5 GPa, more preferably below 1 GPa, even more preferably below of 0.5 GPa to demonstrate non-elastic behavior and / or to demonstrate a transition from elastic to plastic behavior (English: "pop-in") above 1 GPa with a maximum of 20 out of 50 impressions.

The values in the first sentence of the previous paragraph with the unit GPa are preferably maximum stress values in the polycrystal.

The crack tip fracture toughness (Kio) of the polycrystalline ceramics of the present invention is preferably in a range from 3.5 MPa m ^1/2 to 40 MPa m ^1/2 , more preferably from 4 MPa m ^1/2 to 25 MPa m ^1/2 , more preferably from 4.5 MPa m ^1/2 to 17.5 MPa m ^1/2 except of ceramics with a perovskite crystal structure. For ceramics with a perovskite crystal structure, the crack tip fracture toughness (Kio) of the polycrystalline ceramics of the present invention is preferably in a range from 2 MPa m ^1/2 to 30 MPa m ^1/2 , more preferably from 2.5 MPa m ^1/2 to 25 MPa m ^1/2 even more preferably from 3.5 MPa m ^1/2 to 20 MPa m ^1/2 . The crack tip fracture toughness is preferably determined by determining the crack opening profile (“crack opening displacement” (COD)). For this purpose, detailed scanning electron microscopy images are made and compared. The crack opening (the distance between the two crack flanks at an applied load of K _| > 0.8 ^* K _{| C} , preferably K _| > 0.9K _{| C} , preferably KFKI _c ) at periodic intervals of 100 nm, preferably measured at periodic intervals of 50 nm from the crack tip (see Figure 5). The measurement takes place over a length, measured from the crack tip, of at least 1 pm, preferably of 2 pm, more preferably of 5 pm, further preferably of 10 pm and even more preferably of 50 pm. In the case of a crack caused by Vickers indentation, only the first third of the crack length needs to be taken into account. Vickersein pressures are to be produced according to ISO 14705, for example. Before indentation, the sample must be coated with carbon in order to have a brittle, conductive surface for electron microscopy.

All tough ceramics developed to date required mechanisms that lead to an R-curve (Ritchie, RO The conflicts between strength and toughness. Nat Mater 10, 817-822, doi: 10.1038 / Nmat3115 (2011) and Evans, AG, Perspective on the Development of High-Toughness Ceramics. Journal of the American Ceramic Society, 1990. 73 (2): pp. 187-206.). This describes a crack length-dependent fracture toughness, which has the disadvantage that only the presence of long cracks leads to a tough ceramic. In the ceramics of the present invention, however, the tip of the crack is preferably made tougher, which is efficient for short cracks and enables high strengths. In order to distinguish this effect from other mechanisms, the so-called crack tip fracture toughness is recorded quantitatively by measuring the crack opening near the crack tip (preferably at a distance of 0.1 - 50 pm from the crack tip with a resolution of better than 10 nm) and using the Modulus of elasticity of the stress intensity factor is determined with

(u = half the crack opening, E ‘is the modulus of elasticity in the corresponding stress state, Kio is the crack tip fracture toughness, p is the circle number 3.1415 ..., and x is the distance to the crack tip).

This method is to be determined on an equilibrium crack, approximated by a radial crack from an indentation or a stressed crack in a compact sample (Fünfschilling, S. et al. Determination of the crack-tip toughness in silicon nitride ceramics. J Mater Sei 44, 335- 338 (2009); Seidel, J. & Rödel, J. Measurement of crack tip toughness in alumina as a function of grain size. J Am Ceram Soc 80, 433-438 (1997); Haubensak, F. & Argon, AS A new method offracture toughness determination in brittle ceramics by open-crack shape analysis. J Mater Sei 32, 1473-1477 (1997).) The implementation was demonstrated in Figures 6 and 7. Alternatively, the crack opening is determined with an atomic force microscope (F. Meschke, P. Alves-Ricardo, GA Schneider, and N. Claussen, “Failure Behavior of Alimina and Alumina / Silicon Carbide Nanocomposite with Natural and Artificial Fiaws,” J. Mater. Res., 12, 3307-15 (1997); Z. Burghard, A. Zimmermann, J. Rödel, F. Aldinger, and BR Lawn, “Crack Opening Profiles of Indentation Cracks in Normal and Anomalous Glasses,” Acta Mater., 52/2, 293-7 (2004).)

The polycrystalline ceramic of the present invention preferably also has improved ductility, in particular as long as there are sufficient independent mobile sliding systems in the crystal structure. Without a sufficient number of dislocation sources, as proposed according to the invention, it is not possible to observe complete plasticity in ceramics, even if five or more sliding systems are mobile at the same time, or texturing has been achieved so that fewer than five sliding systems are needed. Therefore, it has not been possible to invent fully ductile ceramics in the past. It is only through this invention that the prerequisites for a completely ductile ceramic are created. Therefore, ceramics which, due to dislocation movement at room temperature, have a preferred ductility of greater than 0.5% are also according to the invention. The ductility is preferably determined as plastic elongation in a tensile or compression test using a stress-strain curve. The ductility should preferably be determined in accordance with DIN 50106 "Testing of metallic materials - pressure test at room temperature". The ceramics according to the invention preferably have a ductility of more than 0.5%. A ductility of 2% to 20% is more preferred, more preferably 5 to 75%.

A grain texture is preferably present which allows ductility in all directions, since the texturing means that fewer than five independent sliding systems are required for plastic deformation.

A grain texture is preferably present which allows ductility in the relevant directions of the load, since less than five independent sliding systems are required for the plastic deformation due to the texturing.

Preferably, the polycrystalline ceramic of the present invention also exhibits improved resistance to cracking in the case of a Vickers impression. Ideally, the ceramic according to the invention shows no crack formation in the case of a Vickers impression, in particular if the load applied is greater than or equal to 5 g and / or less than or equal to 100 g. Alternatively or additionally, the ceramic shows cracking, in particular on average, less than 4 cracks, preferably less than 3 cracks, preferably less than 2 cracks, preferably less than 1.5 cracks, with a Vickers impression on a surface of the ceramic , per Vickers impression, with the applied load preferably greater than or equal to 10 g, preferably greater than or equal to 15 g, preferably greater than or equal to 25 g, preferably greater than or equal to 50 g, preferably greater than or equal to 100 g, preferably greater than or equal to 150 g, and / or less or equal to 170 g, preferably less or equal to 150 g, preferably less or equal to 100 g, preferably less or equal to 80 g, preferably less or equal to 60 g, preferably less or equal to 40 g, preferably less or equal 30 g, preferably less than or equal to 30 g, preferably less or equal to 20 g, preferably less or equal to 15 g, preferably less or the same h 10 g, wears.

For example, with a Vickers impression on a surface of the ceramic, the ceramic shows cracking of less than 4 cracks, preferably less than 3 cracks, preferably less than 2 cracks, preferably less than 1.5 cracks, per Vickers impression at a load of 25 g.

For example, with a Vickers impression on a surface of the ceramic, the ceramic shows cracking of less than 4 cracks, preferably less than 3 cracks, preferably less than 2 cracks, preferably less than 1.5 cracks, per Vickers impression at a load of 100 g.

In one embodiment, with a Vickers impression on a surface of the ceramic, the ceramic shows cracking averaging at least 0.1 cracks. This can be, for example, an average value over 100 Vickers impressions.

In any case, the surface of the ceramic can be a, in particular properly prepared, surface with a roughness of less than 0.1 μm without previously existing cracks in order to enable a Vickers impression.

The number of cracks can easily be determined manually and / or automatically using the ceramic. For example, images such as those in Figures 8 and 9, which will be discussed in detail later, can be used and evaluated for this purpose.

If the load is steadily increased during the test by a Vickers impression, the ceramic according to the invention can at some point also have cracks, which, however, are significantly shorter than the cracks in a ceramic according to the prior art. In one embodiment, the crack length of the cracks at 25 g load is between 5 pm and 14 pm, preferably between 5 pm and 12 pm, in particular between 6 pm and 10 pm.

The material SrTiCh is particularly preferred for short crack lengths.

A ceramic according to the invention which has the improved resistance to cracking is documented in FIGS. 8b, 9d and 9f. Ceramics from the state of the art do not meet this criterion, as shown in Figures 8a, 9c and 9e. In this regard, reference is made to the corresponding explanations below for the further discussion of the figures.

Preferably, the polycrystalline ceramic of the present invention also has improved thermal shock resistance. In a brittle ceramic, this is linked to the strength, the coefficient of thermal expansion and the coefficient of elasticity. However, the local plasticity limits the stresses with the dislocations. The thermal shock resistance is preferably in a range of> 500 ° C., more preferably in a range from 600 ° C. to 1600 ° C., but preferably not more than the temperature difference between the melting point of the ceramic and room temperature. The thermal shock resistance is preferable fol gender extent determined by the volume of ceramic to an elevated temperature of t _ma x (for example) 600 ° C is placed and is then quenched in boiling water (100 ° C). The limit temperature at which this does not result in a reduction in strength is the thermal shock resistance (Tmax - 100 ° C). The strength is determined by a four-point bending test according to DIN EN 843-1.

The present invention also relates to a method for producing a polycrystalline ceramic according to the invention, comprising at least one, preferably exactly one of the following steps:

The conscious avoidance of sintering. Ceramic powders are usually ground as small as possible in order to be baked together by the diffusion-controlled process of sintering, with the grains usually also growing significantly. By preventing grain growth by more than a factor of 3 compared to the starting powder size and by avoiding holding times of fractions of minutes to hours at temperatures in the respective sintering range, e.g. 70-90% of the melting point, it is avoided that the dislocation density massively reduced.

Dislocation creep at high temperature, in particular between 50% and 90% of the melting temperature in Kelvin with mechanical loads that lead to plastic deformation of more than 0.5% within a time of less than 10 days. In perovskite BaTiCh, stresses between 5 and 300 MPa at temperatures between 1000 and 1350 ° C are useful. The deformation time depends on temperature, pressure and the desired degree of deformation. A degree of deformation of greater than 0.5% is preferred.

Plastic deformation at high temperature, in particular between 30% and 98% of the melting temperature in Kelvin at mechanical loads that are at least the yield point of the ceramic at that temperature, with plastic deformation between 0.5% and 80%, preferably between 0 , 75% and 25%, more preferably between 1% and 20%. Sufficient independent sliding systems are preferably thermally activated to ensure ductility. This requires a temperature at which a sufficient number of sliding systems are thermally activated to be able to overcome the Peierls barrier. At the same time, the temperature must be chosen so that sliding systems do not become too immobile due to climbing dissociation of partial dislocations. This temperature range is different for each ceramic material class and also varies within a class. There is often an exponential or quadratic relationship with the temperature, especially the process temperature. The deformation should preferably be carried out in air, since an inert gas atmosphere or vacuum through Cottrell atmospheres of oxygen vacancies around the dislocations increases the required stresses and can also have other effects. For the Pero vskit SrTiOs, grain sizes greater than 2 μm and a temperature of 1050 ° C.-1350 ° C. are preferred. The required stress depends extremely on the temperature and the strain rate and should be between 1 MPa and 1500 MPa. Can strain rates in egg nem extremely wide range of preferably 10 ⁸ s ^_1 to 10 ⁵ s ^_1, preferably from 10 ⁷ s ^_1 to 10 ² s ^1, more preferably from 10 ⁶ s ^_1 to 10 ¹ s ^_1, still more preferably from 10 ⁵ s ^_1 to 10 ° s ¹ . The time over which the tension is held is often proportional to the degree of deformation. Cooling during the deformation increases the dislocation density further. A temperature between 30% and 80% of the melting temperature in Kelvin is preferred for mechanical loads that are at least the yield point of the ceramic at that temperature, with plastic deformation between 0.5% and 80%, more preferably between 0.75 % and 25%, more preferably between 1% and 20%.

Reduction and thereby condensation of oxygen vacancies and formation of dislocation loops. In order to allow the exchange of oxygen vacancies with the environment, elevated temperatures are necessary, the exact level of which depends on the material in question. Temperatures at which dislocations do not occur are preferred Change climbing-dissociated configuration, but oxygen ion expansion and diffusion is possible. For the perovskite SrTiCh this is below 850 ° C, but above 300 ° C. With the help of light, the installation and thus the necessary temperature can be massively changed.

Irradiation and the formation of radiation damage that develop into dislocations; pore formation should preferably be avoided here. A combination with temperature treatment can improve the result. The radiation dose, which depends on the specific material, is the central parameter. A fluence of 1 * 10 ¹² nr ² to 1 * 10 ¹⁵ nr ² of nickel ions with 10 MeV is an example. However, the effectiveness of the irradiation can often only emerge in a temperature treatment step. In addition, the effectiveness depends on the combination of ion, energy and fluence used and is different for each material.

Solution heat treatment and aging with possible quenching as an intermediate step. From a supersaturated solid solution, produced by solution annealing at preferably 50% to 95% of the melting temperature, homogeneous micro-precipitates, similar to Guinier-Preston zones in the precipitation hardening of aluminum alloys, are introduced by aging at preferably 30% to 75% of the melting temperature . The elimination of a dissolved component is important, which is carried out by annealing in the solution and a further aging step. It should be emphasized that, in contrast to metals, these lead to softening and not to hardening. It depends on the degree of supersaturation, the aging temperature and the aging time. The temperature-dependent limited solubility of the dissolved component is a basic requirement.

Introduction of micro-precipitates with a different crystal structure or thermal expansion coefficient. Due to the mismatch, these locally create tensions which favor dislocation nucleation. This can be done, for example, by adding a powder to a second, insoluble ceramic. The grain and powder size is preferably in the range from 3 nm to 3 μm, more preferably from 5 nm to 750 nm.

Sintering with simultaneous application of high electrical currents in the preferred range of 0.1 Acnr ² to 2 MAcnr ² . As a result, the material is heated differently from region to region (higher at the sinter necks) and experiences a high, internal, but locally very limited temperature gradient, which translates directly into local chip- transfer gradients. In addition, the high current ensures the formation and movement of charged defects, which cause additional disorder in the crystallites, which can serve as nucleation points.

- Aerosol separation.

Irradiation in the grid with femtosecond laser pulses with a focus point in the material volume. Pulse energies of e.g. 7.5 microjoules are preferred. Multiple pulses per raster point of e.g. 10 to 1000 pulses are preferred.

Special temperature cycles with a cooling rate of more than 25 Kelvin per minute.

The melting of ceramics into the liquid phase with casting as a possible process step (similar to metals).

Preferably, the method comprises at least the step of dislocation creep at high temperature and / or the step of plastic deformation at high temperature.

It has already been noted elsewhere, in particular in connection with the manufacturing process, that "flash sintering" is particularly preferred according to the invention in order to possibly increase the dislocation source density (in particular in the entire volume). A detailed description of an example of a manufacturing method of a polycrystalline ceramic according to the invention is disclosed below:

A particularly preferred production method according to the present invention, in particular in order to produce the desired dislocations in a meaningful density, is “flash sintering” (sometimes also referred to as “ultra-fast fi ring”). The method has in particular:

Providing a ceramic green body, in particular producing the green body from ceramic powder, preferably comprising undoped SrTi0 ₃ , by pressing and / or with the help of binders and (optionally cold isostatic) compression of the powder,

Contacting the green body with electrodes by means of which an electric field can be applied, in particular formed, within the green body, and / or

Formation of the electric field by means of the electrodes within the green body, preferably (i) at an ambient temperature of the green body of between 500 ° C and 1500 ° C, in particular between 900 ° C and 1300 ° C, in particular around 1100 ° C, and / or (ii) for between 1 second and 15 minutes, preferably between 10 seconds and 10 minutes, preferably between 10 seconds and 5 minutes, in particular about 2 Minutes, with an electric current flowing between the electrodes through the green body at least intermittently as a result of the electric field and the green body is thereby heated at least in certain areas.

The procedure could be designed practically as follows, for example:

A ceramic green body is pressed from ceramic powder, for example with a particle size of between 100 nm and 500 nm, preferably between 200 nm and 400 nm, in particular 300 nm, and (optionally cold isostatic), for example at between 300 MPa and 800 MPa, preferably compressed at between 400 MPa and 600 MPa, preferably at 500 MPa. This green body is provided with electrodes which apply an electrical field, for example of between 10 V / mm and 30 V / mm, preferably of between 10 V / mm and 20 V / mm, in particular of about 15 V / mm, the Current flow, in particular through the green body, preferably to a current density of at least regionally between 30 mA / cm ² and 100 mA / cm ² , preferably between 40 mA / cm ² and 70 mA / cm ² , in particular of about 60 mA / cm ² , is limited. The ambient temperature of the green body is set. For example, with an oven in which the green body is placed. The furnace with the green body is heated, preferably at 5 Kelvin per minute, to between 900 ° C. and 1300 ° C., preferably to between 1000 ° C. and 1200 ° C., in particular about 1100 ° C., between 1 second and 15 minutes, preferably held at the temperature for between 10 seconds and 10 minutes, preferably between 10 seconds and 5 minutes, in particular about 2 minutes, and then cooled again, in particular to the starting temperature. The electric field is applied during and / or after the furnace is heated up and in particular before it cools down. When the furnace has reached a sufficient temperature, in particular a temperature that depends on the green body, the green body becomes electrically conductive, causing electrical current to flow. The electrical resistance in the green body then heats the green body, in particular very quickly, for example within between 1 second and 100 seconds, preferably between 10 seconds and 80 seconds, preferably between 15 seconds and 30 seconds, far - in particular between 10% and 100%, preferably between 15% and 80% or between 20% and 100%, of the oven temperature - above the oven temperature. The temperature of the sample preferably reaches a value close to the sintering temperature. However, the sample temperature preferably does not have to be measured and / or controlled directly. After the current intensity has reached the set limit, the current is maintained for a further 2 seconds to 30 minutes, preferably between 10 seconds and 10 minutes, preferably between 10 seconds and 5 minutes, in particular about 2 minutes, and then switched off, after which the sample up again quickly the oven temperature cools down. Instead of a furnace, the ambient temperature can also be set and / or controlled in another suitable manner.

The sample properties are preferably influenced by the temperature profile inside the sample. The electric field and the current density can preferably be adapted to the respective sintering temperature of the material.

The parameters can vary between the individual materials so that the settings can be adjusted depending on the materials used.

Description of the images

Figure 1 shows a transmission electron microscope image of a conventionally sintered SrTiCh ceramic. Only very isolated dislocations can be found. An example of the rare dislocations is marked with the number 1 in the picture on the right.

Figure 2 shows a scanning electron microscope image of a deformed strontium titanate single crystal with etched pits which visualize the dislocations. In strips, areas with a quasi constant maximum dislocation density, in particular too common due to mechanical deformation, can be seen separately from areas without dislocations.

Figure 3 shows an exemplary load curve of a nanoindentation test with a tip radius of 80 nm. Marked with number 2, the ceramic according to the state of the art shows a clearly recognizable transition from elastic to plastic behavior. This feature cannot be seen in the ceramic according to the invention, with a deviation from the elastic behavior already being seen below 0.01 mN.

Figure 4 shows the statistical evaluation of the measurement shown in Figure 3. It is carried out as an example for two different tip radii. The areas with maximum dislocation density, as shown in Figure 2, show only a very slight change in behavior compared to the completely dislocation-free ceramic. The area with a gray background and marked with the number 3 is preferred for a measurement on a ceramic according to the invention, a measured value for which no pop-in could be detected automatically falls within this value range.

Figure 5 shows an example of how a crack from a Vickers impression is selected and visualized at a higher magnification. With number 5 in Fig. 5c) the crack opening is marked in two different places. The distance from the crack tip is marked at intervals of 100 nm. The crack tip is marked with number 4. Figure 6 shows scanning electron microscopy images of cracks at Vickers indentations with 25 g load a) reference crystal b) and b) ceramic according to the invention with high dislocation density. The tip of the crack is marked with an arrow and number 4 and a small line. Number 6 and the black line mark a section of the edge of the Vickers impression. The double arrow marked with number 7 indicates the [110] crystal direction. The cracks in Figure b) and c) spread in different crystal directions.

Figure 7 shows a quantitative evaluation of scanning electron microscopy images of cracks on Vickers indentations with 25 g load. For two ceramics according to the invention of the perovskite crystal structure (SrTiCh) and for a reference crystal, cracks were evaluated in such a way that the crack opening was determined as a function of the distance from the crack tip. These values were plotted against one another (COD value in nm (y-axis) versus distance from the crack tip in pm (x-axis)). The black squares show the value pairs determined for the reference crystal. The values of the two ceramics according to the invention are shown as white circles and white triangles. For comparison, the theoretical curve has a stress intensity factor of 0.6 MPaVrn (solid line) and 2.9 MPaVrn (dashed line shows). It can be seen that the reference crystal has approximately a crack tip fracture toughness of 0.6 MPaVrn and the ceramic according to the invention has approximately a crack tip fracture toughness of approximately 2.9 MPaVrn. The course of the crack opening values is compared with the theoretical course, which results from the formula already mentioned. The crack tip fracture toughness should be selected in the theoretical course so that the course reflects the experimental data as best as possible. This value is then the experimentally determined crack tip fracture toughness u = K _I0 / E '* V (8x / 7r) (u = half crack opening, E' is the modulus of elasticity in the corresponding stress state, Kio the crack tip fracture toughness, p is the circle number 3.1415 ... , and x is the distance to the crack tip).

Figure 8 shows Vickers impressions with 10 g each in a) and b) and 25 g in c) and d). The ceramic in Figure b) and d) was treated as described in the example. This considerably reduces the formation of cracks. In comparison, the dislocation density of the reference crystal is low and in the range of 10 ⁵ cm ^-2 .

Figure 9 shows Vickers impressions with 200 g each in a) and b) and 10 g in c), d), e) and f).

The ceramic in figure b), d) and f) was treated as described in the example. This considerably reduces the formation of cracks. The ceramic in Figure a), c) and e) is state-of-the-art. The ceramics in figures a), b), c) and d) have a grain size of approx. 9 μm. The ceramics in Figures e) and f) have a grain size of approx. 2 μm. Figure 10 shows dark field X-ray microscopy images in which all dislocations are visible, e.g. marked with number 9. It can be seen that no dislocations previously existed in the crystal. The dislocation bands are clearly recognizable (e.g. marked with number 10) and signs of various multiplication mechanisms can be recognized, for example cross-slip, which is marked with number 11. Image artifacts are marked with number 8 as an example.

Figure 11 shows optical microscopy images of rutile (T1O2) crystals. The surface was etched so that dislocations through etch pits are visible. It can be seen that in the vicinity of “pollution particles” (marked with number 12) there are many dislocations (marked with number 13). This shows that foreign particles can be used as sources of dislocation far away from surfaces.

Figure 12 shows transmission electron microscopy images of an SrTiOs ceramic that was deformed at high temperature. The deformation took place at 1150 ° C with a strain rate of 2.5 * 10 ⁵ s ¹ . A degree of deformation of approx. 4% was achieved, the maximum loads being between 20 and 200 MPa compressive stress. It can be clearly seen that a very high dislocation density can exist in ceramics.

Figure 13 shows results of a nanoindentation using the example of the material Magnesi umoxid (MgO) according to the invention. The representation in Figure 13 is basically very similar to that used in Figure 4 and shows that the reduction in the pop-in voltage, which was shown in Figure 4 for the material SrTiOs, can be directly transferred to MgO.

In this respect, Figure 13 shows the statistical evaluation of the measurement on magnesium oxide shown in Figure 3. The area with a gray background and marked with the number 3 is preferred for a measurement on a ceramic according to the invention, a measured value for which no pop-in could be determined automatically falls within this value range. A significant reduction in the “pop-in” voltage from approx. 10 GPa to less than 1.5 GPa was found, whereby all values for the ceramic according to the invention fall into the range marked with number 3.

Figure 13 shows the statistical evaluation of a measurement of a load curve of a nanoindentation test with a tip radius of 80 nm in relation to the material MgO. This shows that the results and conclusions presented in this application, as well as the teaching of the invention, can also be applied equally to a large number of other ceramics.

Figure 14 shows results in connection with the crack tip toughness using the example of the material magnesium oxide (MgO). The evaluation of the crack tip toughness shown in Figures 5, 6 and 7 for SrTiCh is carried out identically for magnesium oxide, the results being presented in a) of Figure 14. In b) and c) of Figure 14, the crack formation according to a Vickers impression with 10 g load in a ceramic according to the prior art (in b) and a ceramic according to the invention (in c) is compared. This shows the direct transferability of the results from SrTiCh in Figure 8 to the material MgO.

Figure 14 shows results and illustrations related to the crack tip toughness in MgO.

For this purpose, cracks were evaluated for MgO ceramic according to the invention after polishing in such a way that the crack opening was determined as a function of the distance from the crack tip. These values were plotted against one another (COD value in nm (y-axis) versus distance from the crack tip in pm (x-axis)). The values of the two ceramics are shown as black squares with measurement error bars. For comparison, the theoretical curve has a stress intensity factor of 3.2 MPaVrn (solid black line in Figure 14 a)). It can be seen that the ceramic according to the invention has approximately a crack tip fracture toughness of approx. 3.2 MPaVrn after polishing. The course of the crack opening values is compared with the theoretical course, which results from the formula already mentioned. The crack tip fracture toughness in the theoretical course is selected in such a way that the course reflects the experimental data in the best possible way. This value is then the experimentally determined crack tip fracture toughness u = K _I0 / E '* (8x / n) (u = half crack opening, E' is the modulus of elasticity in the corresponding stress state, Kio the crack tip fracture toughness, p is the circle number 3.1415. .., and x is the distance to the crack tip).

Furthermore, Figure 14 shows half of a Vickers indentation before and after polishing in order to emphasize the contrast. Before polishing, that is, according to the prior art, clear cracks can be seen, as shown in FIG. 14b. Please refer to the table below for the length of the crack. In contrast, the ceramic according to the invention has no cracks in FIG. 14c. The representation in Figure 14a is basically very similar to that used in Figure 7. The representations in Figures 14b and 14c are basically very similar to those used in Figures 8a and 8b or Figures 8c and 8d.

Figure 15 shows an illustration for determining the length of cracks (“crack length”) that can arise with a Vickers impression in a ceramic, such as a ceramic according to the invention.

The crack length is therefore preferably (in particular also in general) determined by placing horizontally straight lines H1 and H2 and vertically straight lines V1 and V2 on the most distant crack tip as an extension of the indentation diagonals. The distance D1 and D2 between two opposing lines is determined as the crack length. For each impression there are two values for the crack length and the crack length is the mean of these two values. An average value over several, for example 10, impressions is preferably determined as the respective crack length. In other words, from the multiple, for example 10, values for the distance D1 and the multiple, for example 10, values for the distance D2, an average value is calculated as a single crack length. This determination of the crack length is particularly suitable for single crystals and polycrystals of strontium titanate and / or single crystals of magnesium oxide. In particular, in each case for ceramics according to the prior art and according to the invention.

Examples

The invention consists of two functional steps. First, a high dislocation source density must provide displacements in the corresponding volume before or during the load case, so that local deformation becomes possible. During production, dislocation sources can be activated which can only be used due to special boundary conditions (e.g. extreme pressures or temperature) during production and then become inactive again (see example 5). Second, the dislocations provided must lead to the desired improvement in properties (increase in crack tip fracture toughness and ductility). A high density of dislocation sources can be introduced and used in many different ways, one of which we will exemplify in the following. The sources of displacement are preferably only activated when there is a load. This is preferably done at the crack tip and its surroundings as well as at grain boundaries and their surroundings.

In particular, it was recognized in the present case that this is possible inside a ceramic and thus far away from a surface. Examples are presented below. For this purpose, examples are first shown in which the parameters of the present invention lead to the desired property improvements. It is also shown that the introduction of dislocation sources in the entire volume of a polycrystalline ceramic is possible.

Single crystal

For an experimental proof of the importance and effectiveness of the dislocation source density or the dislocation density made available in the relevant volume in order to improve the properties, a single crystal made of SrTiCh of the perovskite crystal structure is used. SrTiCh (strontium titanate) is one of the materials which, as described in accordance with the invention, have the basic requirement of dislocation mobility at room temperature. In order to ensure that the results presented are based solely on the claimed parameters and that no other parameters are changed at the same time, a single crystal is selected first.

When a very small tip comes into contact with a ceramic surface, extremely high tensions of often more than 15 GPa are generated locally. These tensions are sufficient for homogeneous dislocation nucleation. Homogeneous nucleation below a pointed contact point is indeed the most inefficient source of dislocation and is limited to the surface, but this dislocation source is particularly suitable for demonstration purposes, as chemical changes are also excluded and thus perfect comparability is guaranteed. If the dislocation source is moved over the surface, it nucleates new dislocations at each contact point. This can be done very efficiently with SiC sandpaper, for example. This introduces a dislocation density of approx. 10 ¹¹ cm ^-2 , which can be assigned to an approximately displacement source density of 10 ¹³ cm ^-3 and thus approximately 10 contact points per mht ² in this case. The dislocation density in the crystal, which was produced by means of the Verneuil method, is typically 10 ⁵ cm ^-2 . In the reference sample, approx. 50 μm are removed from the surface by increasingly gentle polishing. This can be achieved, for example, with a polishing in four steps of 40 minutes each at 150 rpm with diamond paste of 6 pm, 3 pm, 1 pm and% pm each. Finally, a very gentle polishing step with OPS solution is carried out in a vibration polishing machine overnight.

On the other hand, the dislocations introduced on the surface with SiC paper can be retained in their particularly high density due to the dislocation source density if the polishing steps are carried out for only 30 seconds instead of 40 minutes and the vibration polishing step is limited to one hour. The polishing quality and roughness of the two polishing processes are almost identical, apart from the occasional scratches that can easily be avoided.

The required deformability at any point was made possible by the treatment. In Figure 4 it was shown that, according to the state of the art, deformation is only observed above 10 GPa stress. On the treated surface, the threshold voltage for plastic deformation for all indentations was well below 1 GPa and even so low that it could no longer be reliably quantified with nanoindentation (gray area marked with number 3 in Fig. 4). However, the deviation from the elastic behavior shows that the behavior is plastic. For all the nanoindentations carried out, a deviation of the load curve from the elastic behavior could be seen from the beginning, whereby all values are located close to zero on the X-axis in Figure 4 and therefore fall into the number 3 marked value in Figure 4.

Figure 6 shows that the treatment significantly shortens the cracks. The crack tip fracture toughness was determined on these cracks with the aid of the crack opening profile, which is presented in Figure 7. An increase by a factor of five from 0.6 MPam ^/ to 2.9 MPam ^/ , which was previously believed to be impossible, is visible.

Polycrystalline ceramics

The same surface treatment was also carried out on two polycrystalline SrTi0 ₃ ceramics of the perovskite crystal structure with a grain size of approx. 2 pm and approx. 9 pm. Figure 8 shows exemplary Vickers impressions on the single crystal, whereas Figure 9 shows exemplary Vickers impressions for the polycrystal. The influence on the length of the crack and its origin is directly comparable.

Further key points are demonstrated below.

1) Dislocations or sources of dislocation are almost non-existent in ceramics after sintering. A plastic zone is therefore not possible in ceramics that are manufactured according to the state of the art (see Figure 1).

For this purpose, a polycrystalline ceramic was made from 99.99% pure SrTi0 ₃ . It was sintered for compression at 1425 ° C for one hour and then at 1350 ° C for 15 hours in an oxygen atmosphere. This ceramic was obtained using Ultra High Voltage Electron Microscopy examined for dislocations in a transmission electron microscope. Only three dislocations were found in approx. 130 examined grains. Exemplary recordings are shown in Figure 1.

2) There are a large number of dislocation nucleation and dislocation multiplication mechanisms which are relevant for ceramics, but were not previously known with regard to ceramics. The knowledge often comes from a time when technical ceramics were not yet used industrially on a larger scale. Technical ceramics currently use different interlocking mechanisms, which means that the fields of knowledge are not linked. The room for maneuver is therefore significantly greater than would be assessed by a ceramic expert.

3) The multiplication of dislocations was demonstrated in a previously completely dislocation-free volume. This shows that the nucleation, i.e. the dislocation source density, is a central parameter, because the prerequisites for mobility and multiplication can easily be observed. At the same time, there is a certain maximum dislocation density, which cannot be overcome only by multiplication. For this purpose, recordings were made using synchrotron-based X-ray dark field microscopy (see Figure 10). In addition, the dislocations generated by the multiplication were visualized with etching pits, whereby a constantly high but insufficient dislocation density can be seen (see Figure 2).

4) In titanium dioxide it has been proven that microscopic precipitates can create dislocations in their surroundings at any depth inside the material (see Figure 11). In the past, such “impurities” were avoided as much as possible. We propose that these are deliberately used in a size of 10-2000 nm as a dislocation source density according to the invention. This can be done by adding foreign phases.

5) Deformation of ceramics according to the state of the art and thus increasing the dislocation density at room temperature is, based on the knowledge presented here, not possible for three reasons: A) the Taylor criterion, B) the lack of mobility of the dislocations of some sliding systems at room temperature (five mobile sliding systems are necessary to meet the Taylor criterion), and C) the absence of dislocation sources. By drastically increasing the temperature to, for example, 1150 ° C, these three problems are solved at the same time, but only until the temperature drops again. The temperature range in SrTi0 ₃ enables the dislocation movement on sliding systems, which are not mobile at room temperature. This overcomes the Taylor criterion and thus the area in which deformation is enabled is expanded. At the same time, thermal activation can cause dislocations at the grain boundaries. Nucleation is possible at many points on the grain boundary and thus the temperature allows an extremely high density of temporary dislocation sources and thus the achievement of a dislocation density which is not limited by the multiplication. After cooling, the grain boundaries can no longer function as dislocation sources, but they have already created a high dislocation density.

A high dislocation density was introduced into the above-mentioned ceramic by plastic deformation at 1150 ° C. and a stress of 20 MPa. Local dislocation densities of approx. 5 * 10 ¹⁰ cm ^-2 could be detected in the deformed ceramic by counting the number of dislocations in a certain observed area. This value is in the range according to the invention. In particular, this was possible 2 mm from the surface of a polycrystal and thus implemented inside a polycrystal. Exemplary transmission electron microscopy images are shown in Figure 12.

Crack length

The following table shows average crack lengths, determined from 10 impressions, in micrometers (pm) for ceramics according to the prior art.

The following table shows average crack lengths, determined from 10 impressions, in micrometers (pm) for ceramics according to the invention.

There is optionally the possibility of making the delimitation either via crack lengths or via a threshold load at which noticeable crack formation occurs.

In the case of low loads, in particular of less than 25 g and / or when no or hardly any soot formation occurs, the crack length is preferably determined as follows. If no crack can be seen in a diagonal direction, e.g. in the direction bounded by V1, the line V1 demonstrated in Figure 15 is placed in the center of the impression. If a crack can be seen in the opposite direction and is marked by V2, the distance D2 from the center of the indentation to V2 is measured accordingly. If no cracks can be seen in both directions V1 and V2, the distance D2 is zero. This procedure applies analogously to D1, which is limited by H1 and H2.

In particular, if there are few or no cracks at all, the average crack length determined by the method described above with reference to Figure 15 can be smaller than the diagonal of the indentation. In the previous table, this is indicated by the expression “No or only a few cracks”.

Reference list

1 Dislocation in the dark field contrast of a transmission electron microscope

2 Transition from elastic to plastic behavior (English: "pop-in")

3 Value range according to the invention for the probability of the "pop-in"

4 crack tip

5 crack opening

6 Contour of a Vickers impression

7 [110] crystal direction

8 artifacts

9 dislocations

10 dislocation band

11 Cross slide

12 particles

13 Etch pits created by dislocations D1 distance

D2 Distance H1 Horizontal line H2 Horizontal line V1 Vertical line

V2 vertical line

Claims

Expectations

1. Polycrystalline ceramic, a) where the ceramic with nanoindentation with a tip radius of 50-200 nm at least 30 out of 50 impressions with a minimum distance from each other of 20 pm at a maximum load of 5 mN below 1.5 GPa non-elastic behavior and / or the ceramic shows a transition from elastic to plastic behavior (English: "pop-in") above 1 GPa with a maximum of 20 out of 50 impressions, and b) where the ceramic has a crack tip fracture toughness of at least 2 MPa m ^{1 / 2} has.

2. Polycrystalline ceramic according to claim 1, wherein the ceramic has a ductility of more than 0.5%.

3. Polycrystalline ceramic according to at least one of the preceding claims, wherein the ceramic has a thermal shock resistance in a range from 600 ° C to 1600 ° C.

4. Polycrystalline ceramic according to at least one of the preceding claims, wherein the ceramics are selected from the group consisting of perovskites, diborides of transition metals, ceramics with sodium chloride crystal structure, ceramics with zinc blen decrystal structure and ceramics with wurtzite crystal structure and preferably ceramics with calcium fluoride crystal structure.

5. Polycrystalline ceramic, in particular according to claim 4, wherein the ceramic has Zrß2 or HfB2 and / or the ceramics of the perovskites BaTiOs, SrTiOs and / or KNbOs aufwei sen, the ceramics with sodium chloride crystal structure NaCl, LiF and / or MgO, the ceramics with Have zinc blende crystal structure ZnS, the ceramics with Wurtzitkris tall structure ZnO and / or the ceramics with calcium fluoride crystal structure have CaF2.

6. Polycrystalline ceramic according to at least one of the preceding claims, wherein the ceramic has at least 1 * 10 ⁹ dislocation sources per cm ³ and / or at least one Verset tion source per grain.

7. Polycrystalline ceramic according to at least one of the preceding claims, wherein the dislocation sources are selected from the group consisting of, vacancy clusters, Slip rings, grain boundary sources, kinematic sources, and combinations of two or more thereof.

8. Polycrystalline ceramic according to at least one of the preceding claims with a grain size in a range from 100 nm to 250 μm.

9. Polycrystalline ceramic according to at least one of the preceding claims with a crack tip fracture toughness in a range from 2 MPa m ^1/2 to 40 MPa m ^1/2 .

10. Polycrystalline ceramic according to at least one of the preceding claims, wherein the ceramic in a Vickers impression on a surface of the ceramic crack formation, in particular on average, of less than 4 cracks, preferably less than 3 cracks, preferably less than 2 cracks, preferably of less than 1.5 cracks per Vickers impression, the applied load preferably being greater than or equal to 10 g, preferably greater than or equal to 15 g, preferably greater than or equal to 25 g, preferably greater than or equal to 50 g, preferably greater than or equal to 100 g, preferably greater than or equal to 150 g, and / or less or equal to 170 g, preferably less or equal to 150 g, preferably less or equal to 100 g, preferably less or equal to 80 g, preferably less or equal to 60 g, preferably less or equal to 40 g, preferably less than or equal to 30 g, preferably less or equal to 30 g, preferably less than or equal to 20 g, vorzugswe ise is less than or equal to 15 g, preferably less than or equal to 10 g.

11. Polycrystalline ceramic according to claim 10, wherein the ceramic has a Vickers impression on a surface of the ceramic crack formation of less than 4 cracks, preferably of less than 3 cracks, preferably of less than 2 cracks, preferably of less than 1, 5 cracks , per Vickers impression at a load of 25 g.

12. Polycrystalline ceramic according to at least one of the preceding claims, the crack length of the cracks at 25 g load between 5 pm and 14 pm, preferably between 5 pm and 12 pm, in particular between 6 pm and 10 pm.

13. A method for producing a polycrystalline ceramic according to at least one of the preceding claims, comprising at least one of the following steps:

Dislocation creep at high temperature, in particular between 50% and 90% of the melting temperature in Kelvin in the case of mechanical loads that lead to plastic deformation of more than 0.5% within a time of less than 10 days, Plastic deformation at high temperature, particularly preferably between 25% and 99% of the melting temperature in Kelvin, preferably from 35% to 90%, with mechanical loads that are at least the yield point of the ceramic at the temperature, with a plastic deformation between 0, 5% and 80%, preferably between 0.5% and 25%, preferably between 0.5% and 20%,

Reduction and thereby condensation of oxygen vacancies and formation of dislocation loops

Irradiation and formation of radiation damage such as vacancies, vacancy clusters or atoms on interstitial spaces,

Introduction of micro-excretions,

- Use of electric fields, e.g. at the sintering temperature, especially when high currents occur, e.g. preferably at least 0.1 Acnr ² and at most on average 2 MAcnr ² ,

- aerosol separation,

Pouring from the liquid phase,

Irradiation in the grid with femtosecond laser pulses with a focus point in the material volume.

Temperature cycles with a cooling rate of more than 25 Kelvin per minute.

14. The method according to claim 13, comprising at least one of the following steps:

Dislocation creep at high temperature,

Plastic deformation at high temperature.

15. A method for producing a polycrystalline ceramic according to at least one of claims 1 to 12, comprising at least one of the steps of claim 13 or 14 and / or at least one of the following steps:

Providing a ceramic green body, in particular producing the green body from ceramic powder, preferably comprising undoped SrTi0 ₃ , by pressing and / or with the aid of binders and, preferably cold isostatic, compression of the powder, Contacting the green body with electrodes by means of which an electric field can be applied, in particular formed, within the green body,

- Formation of the electric field by means of the electrodes within the green body, preferably (i) at an ambient temperature of the green body of between 500 ° C and 1500 ° C, in particular between 900 ° C and 1300 ° C, in particular around 1100 ° C, and / or (ii) for between 1 second and 15 minutes, preferably between 10 seconds and 10 minutes, preferably between 10 seconds and 5 minutes, in particular about 2 minutes, with an electric current between the electrodes through the at least intermittent due to the electric field The green body flows and the green body is thereby heated at least in some areas.