NTC composition, thermistor, multilayer thermistor, multilayer thin-film thermistor, use of a formula for determining material properties and for material optimisation, and method for material property optimisation of ntcs in ceramics

Abstract

The invention relates to a NTC composition based on manganese cobalt oxide spinel corresponding to the molecular formula Mn3-x-y-zCoxAlyCuzO4, where 0.80 ≤ x ≤ 1.40, 0.50 ≤ y ≤ 1.10 and 0.15 ≤ z ≤ 0.35, or the amounts of an addition of Al and Cu are determined with the aid of formulas B(T) = Bmax × tanhyp(3 T / T0) and T0 = (h × ν0) / (2 k arcsinh γ), wherein the amounts are adjusted such that a value Bmax is 3400 to 4000 K or / and a lower limit temperature T0 ≤ 600 Kelvin is reached.

Description

[0001] Description

[0002] NTC composition, thermistor, multilayer thermistor, multilayer thin-film thermistor, use of a formula for determining material properties and material optimization, and methods for material property optimization of NTCs in ceramics

[0003] The present application relates to an NTC composition, a thermistor comprising this, a multilayer thermistor comprising this, and a multilayer thin-film thermistor comprising this. Furthermore, the application relates to the use of a formula for determining material properties, the use of the formula for material property optimization, and a method for material property optimization.

[0004] This application claims priority from German patent application No. 10 2023 134 260 . 6 , filed on December 7, 2023, the contents of which are incorporated herein by reference in their entirety.

[0005] NTC ceramics are used in thermistors. Such thermistors are used for a wide variety of applications, for example in electronics in general, but also in combustion-engine vehicles and electric vehicles. For some applications, the smallest possible components can be advantageous. Some applications require the widest possible temperature sensing over a large range. In turn, some applications require temperature sensing at low temperatures. In view of the above requirements, an improved NTC composition and a thermistor using this composition should be provided.

[0006] The object is at least partially achieved by an NTC composition according to claim 1 or claim 2. Further preferred embodiments are specified in the subclaims. Other subject matter is claimed in subordinate claims or described below.

[0007] According to one embodiment, an NTC composition with a manganese-cobalt oxide spinel-based main component is described. This composition has aluminum with the element symbol Al and copper with the element symbol Cu as additives. The corresponding composition can, for example, be represented by the molecular formula Mn3-xy-zCo x Al y CUz04. The composition may contain oxygen gaps, but preferably has no oxygen gaps.

[0008] In addition to the main component, which makes up the majority of the NTC composition, the NTC composition can also contain a portion of a second phase as a so-called precipitate. Precipitations can form during sintering, particularly if sintering is carried out at an excessively high sintering temperature. The ceramic is preferably selected in terms of its starting materials as if the entire NTC composition would ultimately consist of the main component as the main phase. During production, however, a precipitation phase or phases can form alongside the main component, whereby portions of the ceramic still meet the target composition and then represent the main component. The precipitate preferably makes up a small portion of the composition. The main component preferably has a phase portion of over 90%, i.e. if the main component is represented by Mna-xy-zCOxAlyCUzCh, this spinel component has a phase portion of over 90%.Even more preferably, it represents a proportion of over 95%. With careful sintering, it can also be produced as a largely single-phase material, e.g., with a phase proportion of over 99%.

[0009] According to one embodiment, the proportions of the addition of Al and Cu can be determined using the formulas B ( T ) = Bmax x tanhyp ( 3 T / To ) and To = (hx vo ) / ( 2 k arcsinh y) can be determined. Here, B ( T ) is the temperature-dependent B value of the ceramic, T is the temperature, B max is the temperature-independent B value of the dominating small polaron hopping conduction, To is the lower limit temperature of the dominating small polaron hopping conduction (SPH conduction for short), h is the Planck constant, vo is the characteristic oscillation frequency of the crystal lattice, k is the Boltzmann constant, and Y is the coupling constant. The B value is a basically known quantity in the NTC ceramic field, which represents a characteristic measure of the composition of a thermistor made with it, or can be used as a measure of the measurement sensitivity of a thermistor made with it. Small polaron hopping conduction is the generally accepted dominating conduction mechanism in NTC thermistor ceramics. In addition to SPH conduction, a certain temperature-dependent proportion of the charge carriers is also transported by band semiconductor. Both mechanisms operate in parallel, resulting in a certain deviation from linear behavior.

[0010] In particular, the use of the formulas allows

[0011] Determination of the otherwise difficult to access material parameters Bmax, To and y can be determined easily, accurately and reproducibly.

[0012] For example, material optimization can be carried out in a short time.

[0013] According to one embodiment, the proportions of ceramic components, or in particular the proportions of Al and Cu, are adjusted using the above-mentioned formulas so that a value B max 3400 to 4000 K. According to a preferred embodiment, proportions are adjusted so that a value B max 3500 to 3800 K. According to a particularly preferred embodiment, proportions are adjusted so that a value B max 3500 to 3600 K.

[0014] According to one embodiment, proportions of ceramic components, or in particular the proportions of Al and Cu, are adjusted using the above-mentioned formulas so that a value To of at most 600 K. According to a preferred embodiment, proportions are adjusted so that a value To of at most 550 K. According to a particularly preferred embodiment, proportions are adjusted so that a value To of at most 500 K.

[0015] A manganese-cobalt oxide spinel with Al and Cu as additives, which has a value B maxfrom 3400 to 4000 K and / or a value To of max. 600 K, can be suitable for use at temperatures below 0 ° C. For such a thermistor, for example, a B value can more easily remain above a value of 3000 K at temperatures below 0 ° C. Furthermore, the linear range of the B value can be more extensive and, for example, extend to lower temperatures. Such a thermistor can therefore be particularly well suited to extending the technically useful measuring range into temperature ranges below 0 ° C.

[0016] A manganese-cobalt oxide spinel with Al and Cu as additives, for example, can be used effectively in a temperature range where typical thermistors based on a manganese-iron-nickel oxide spinel with Cu as an additive, for example, can no longer be used effectively. Such manganese-iron-nickel oxide spinels, such as Mna-xy-zFexNiyCUzCh with 0.80 < x < 0.85 and 0.60 < y < 0.65 and 0.001 < z < 0.10, exhibit poor linearity and sensitivity of the measurement signal below 0°C and therefore cannot be used effectively in this measuring range.

[0017] According to one embodiment, the B value of the manganese-cobalt-oxide-spinel-based NTC composition can be adjusted via the Cu content. The addition of Al makes it possible to increase the specific resistance of the composition. For example, the addition of Cu can increase conductivity. The inventors have thus recognized that the composition can be optimized through the targeted interplay of these components, whose proportions are adjusted, for example, using the formulas mentioned above. In this way, these two components can contribute preferred properties and mutually balance out less preferred properties. Consequently, as an alternative and independent of the features mentioned above or below, a generalized NTC composition with a main component based on manganese-cobalt-oxide-spinel with an addition of aluminum and copper is also disclosed. This can have further features of other embodiments.According to one embodiment, the ceramic is adjusted to a resistivity of 200 Ω-cm or more. Preferably, the resistivity is 250 Ω-cm or more, and even more preferably, the resistivity is 400 Ω-cm or more. A value of 800 Ω-cm or higher is even more preferred. Higher resistivities allow thermistors to be made smaller or, in multilayer thermistors, to reduce layer thicknesses. In particular, a ceramic with the above-mentioned B values ​​and the resistivities described here is preferred.

[0018] Furthermore, a ceramic NTC composition with a main component having the molecular formula Mna-xy-zCOxAlyCUzCh is described as a further embodiment. In this embodiment, 0.8 < x < 1.4, 0.5 < y < 1.1, and 0.15 < z < 0.35 apply. According to a variant of this embodiment, the more precise parameter ranges 0.80 < x < 1.40, 0.50 < y < 1.10, and 0.15 < z < 0.35 can be maintained for the formula.

[0019] Here and elsewhere in the description or claims, the term "<" is understood to mean "less than or equal to." It therefore includes both the alternatives "less than" and "equal to," both of which are therefore to be considered disclosed.

[0020] According to one embodiment, 0.90 < x < 1.30, 0.55 < y <

[0021] 1.00 and 0.15 < z < 0.35.

[0022] The aforementioned features may be combined with one another and with the features mentioned below. In particular, a composition prepared using the above-mentioned formulas may exhibit the properties set forth here or below.

[0023] The inventors of the present invention have recognized that such a composition can achieve improved performance of the NTC composition or of a thermistor formed therefrom. For example, the B value can exhibit better linearity, i.e., can be linearly approximated over a different or wider range without significant error. Furthermore, an excessive drop in the resistance of such a ceramic composition can be prevented. Furthermore, a thermistor can be provided that enables improved temperature measurement at low temperatures. In particular, the sensitivity at low temperatures can be improved.

[0024] According to a preferred embodiment, 0.15 < z < 0.30 can apply for z. Even more preferably, 0.15 < z < 0.29 can apply, and even more preferably 0.15 < z < 0.28 can apply. According to an embodiment with certain properties, 0.15 < z < 0.27 can apply. According to alternative embodiments, 0.19 < z < 0.28 or 0.195 < z < 0.275 can apply for z. According to further embodiments, 0.21 < z < 0.35 can apply for z. Values ​​of z above 0.21 can be particularly preferred here. For example, 0.219 < z < 0.35 can apply.

[0025] For z < 0.35, it was found that this can prevent an excessive drop in the specific resistance of the composition. For values ​​of z < 0.28, or sometimes even slightly lower, it was found that such an NTC composition exhibits a preferentially high specific resistance.

[0026] In particular, by using a composition for which x and y, for example, correspond to the values ​​mentioned above and z satisfies one of the ranges mentioned here, it is easier to obtain an NTC composition that has a high B value at low temperatures and, at the same time, has a high specific resistance. The specified lower limits for z have proven preferable, as they help to adjust the B value. They can also have beneficial effects on the B ma x-value can be achieved.

[0027] Preferred ranges for x are 0.90 < x < 1.30, such as 0.95 < x < 1.25. Even more preferred ranges are, for example, 1.00 < x < 1.21 or 1.00 < x < 1.10. Particularly preferred material properties have been identified for these value ranges.

[0028] For y, the range 0.55 < y < 0.95 was considered preferred. Even more preferred is a range of 0.60 < y < 0.95 or 0.60 < y < 0.90 or 0.70 < y < 0.95 or 0.75 < y < 0.95 or 0.76 < y < 0.95 or 0.75 < y < 0.85. Alternatively, the range 0.55 < y < 0.94 can also be regarded as preferred. Particularly preferred material properties have also been identified for this. Also preferably, the upper or lower limit of these preferred ranges can be linked to the respective other limit from the broadest range mentioned above. For example, the range 0.60 < y < 1.10 can be preferred. Accordingly, in this example, not only the range 0.60 < y < 0.95 but also the range 0.95 < y < 1.10 can be considered preferred in addition to or instead of this.

[0029] The same applies to the other value ranges of x and z, as discussed above. Here, too, the upper and lower limits of these preferred ranges can be linked to the other limit from the broadest range mentioned above.

[0030] In particular, all material compositions in which two or, even better, three of x, y and z have a preferred value or a very preferred value were recognized as preferred.

[0031] According to a further embodiment, an NTC composition with a manganese-cobalt oxide spinel-based main component is described, to which aluminum and copper are added, and in which the aluminum to copper ratio is > 2.4, i.e., y:z > 2.4. With such an Al to Cu ratio, preferred properties with respect to the B value and the resistivity can be achieved. Even more preferred in this context are embodiments with y:z > 2.9 or even y:z > 3.4.

[0032] If an NTC composition with a main component having the molecular formula Mn3-xy-zCOxAl2Cu2Ch is considered as an embodiment of the NTC composition based on manganese-cobalt oxide spinel, then in the case of the previous embodiment, y:z > 2.4 applies. x can assume the values ​​described above, y can assume the values ​​described above, and z can be determined according to the ratio of y:z. Alternatively, z can also assume the values ​​described above, and y can be determined from the ratio of y:z. According to a preferred embodiment, the NTC composition with the molecular formula Mn3-xy-zCo x AlyCu z 04 both the above-described values ​​for x, y, and z and the ratio y:z > 2.4. The advantages can be particularly pronounced in this case. According to one embodiment, the composition is free of nickel, except for unavoidable impurities.

[0033] Furthermore, a thermistor is disclosed. This thermistor has the above-mentioned NTC composition in an NTC ceramic. This means that the ceramic NTC composition constitutes at least a portion or a partial volume of the ceramic of a thermistor.

[0034] According to one embodiment, the thermistor can be a monolithic NTC thermistor or a multilayer thermistor. A monolithic NTC thermistor can be either a component sintered from a single piece or a component formed and sintered from individual layers, which, however, unlike a multilayer thermistor, does not have internal electrodes.

[0035] According to a further embodiment, the thermistor can be a monolayer NTC thin-film thermistor or a multilayer thin-film thermistor. A monolayer NTC thin-film thermistor can be either a component made from a single thin-film layer or a component constructed from several individual thin-film layers stacked one above the other, even in successive manufacturing steps. However, unlike a multilayer thin-film thermistor, the component does not have internal electrodes.

[0036] According to a further embodiment, the thermistor can be a thermistor in a so-called T-design. A T-design can be understood as what is understood as such in the field. In particular, the T-design can be a structure in which, starting from two outer electrodes, at least two inner electrodes extend from the respective outer electrode into the interior of the thermistor ceramic. Preferably, inner electrodes of different polarity lie opposite one another. With regard to a stacking direction, these can be arranged at the same height. A free electrode layer which is not connected to any polarity is arranged between the pair of electrodes of the same polarity. For example, this can be arranged centrally between the outer electrodes.For example, a thermistor in T-design can particularly preferably be realized with a thermistor material with a specific resistance value at 25 ° of more than 1300 Ω-cm.

[0037] According to a further embodiment, the thermistor can be designed in a so-called tip design. This can also be referred to as a gap design. The tip design can be designed such that, starting from two outer electrodes, two inner electrodes protrude into the ceramic thermistor material for each outer electrode and thus for each polarity. The electrodes of different polarity, which protrude from the respective outer electrodes into the thermistor material, are preferably arranged at the same height. An equal height can in particular relate to a stacking direction of the material. In such a structure, the current can flow from one inner electrode tip to the opposite electrode tip of different polarity. Ceramic materials with a specific resistance value of more than 500 Ω-cm can be particularly suitable for a tip design.

[0038] The ceramic composition allows for small designs for the thermistors mentioned, as it enables a sufficiently high specific resistance. In particular, this allows for miniaturization of the components. For example, according to one embodiment, a multilayer thermistor with a volume between 0.075 and 10 mm 3 According to a preferred embodiment, a volume of such a multilayer thermistor can be 0.3 to 5.5 mm 3 amount to .

[0039] According to one embodiment, a multilayer thermistor contains at least one ceramic layer containing or consisting of the NTC composition. A similar embodiment can also be embodied as a multilayer thin-film thermistor, wherein the multilayer thin-film thermistor contains a ceramic thin-film layer containing or consisting of the NTC composition.

[0040] According to one embodiment, a ceramic layer may have a thickness of 10 to 100 pm, and preferably of 20 to 50 pm. According to one embodiment, a thin-film layer may have a thickness of 0.001 to 10 pm, and preferably of 0.01 to 1 pm.

[0041] In particular, a ceramic composition with the above-mentioned values ​​for z , such as 0 . 20 < z < 0 . 30, has particularly suitable properties to ensure a sufficiently high resistance for corresponding layer thicknesses.

[0042] In particular, such a composition can be used for thermistors with nominal resistances at 25°C of 1 to 20 kΩ. In particular, a nominal resistance at 25°C can be 10 kΩ. The thermistors, and in particular the multilayer thermistors, can be designed as SMD components (surface-mounted device components). The thin-film thermistors, regardless of whether they are designed in a multilayer design or as a monolayer thin-film thermistor, can also be applied to or in substrates or be suitable for embedding (field of application: SESUB technology) or be intended as a component for a MEMS component.

[0043] Furthermore, the use of the formula B(T) = B ma x x tanhyp(3 T / To) is described for determining the material properties of an NTC spinel ceramic. Here, B(T), Bmax, T, and To correspond to the quantities described above.

[0044] Until now, NTC spinel ceramics, such as nickel-manganese spinels or nickel-cobalt spinels, have often been described exclusively using the linear approximation B(T1, T2) = ln(R2 / R1) / (1 / T2 - 1 / T1). For example, T1 = 25°C and T2 = 100°C, where R1 and R2 are resistances at T1 and T2, respectively. Therefore, only the ideal RT characteristic of a thermistor ceramic was considered. However, this simplified view does not do justice to such ceramics, especially not at lower temperatures. In the linear range, the generally accepted conduction mechanism of "small polaron hopping" (SPH for short) dominates. This is particularly prevalent at higher temperatures, as sufficient thermal energy is available to excite the crystal oscillations. At room temperature and especially below, this is no longer entirely the case. In addition to SPH conduction, a smaller proportion of the charge carriers are also transported by band semiconductors. Both mechanisms operate in parallel.This results in a deviation from the linear approximation. If the B (TI, T2) value is calculated in this temperature range, the lower the TI selected, the lower the value. If a close-meshed measurement of TI and T2 is chosen, a constant decrease in the B(T1, T2) value becomes apparent at lower temperatures.

[0045] To date, mechanisms and parameters have been recorded only very imprecisely. The approaches published by Casado et al. (J. Phys. : Condens. Matter 6 (1994) 4685 - 4698) offer a physical description but only an inadequate basis for recording material properties in technically relevant systems. Thus, according to Casado et al., the formulas To = (hv o ) / (2k arcsinh y) and B max = (yhv o ) / (2k). Here, To is the lower limit temperature of the dominant SPH line, B max the constant B value of the dominant SPH line, v othe characteristic vibration frequency of the crystal lattice, h the Planck constant, k the Boltzmann constant, y the coupling constant of the vibrational energy with the electrical energy.

[0046] Formula B max = (yhv o ) / (2k) contributes a lot to the understanding of the deviation from linearity at low temperatures, but the parameters To, B max and y are difficult to determine with this method. Until now, a graphical method has generally been used, in which ln(R) is plotted against (1 / T) over the largest possible temperature range. In the range of smaller (1 / T) values, a straight line is fitted as an asymptote. This allows the values ​​T o (coincidence of the RT values ​​with the straight line) and B max(slope of the asymptote) can be estimated. However, the placement of the straight lines and their drawing is partly arbitrary and at least subject to considerable uncertainty. The inventive approach using the formula B(T) = Bmax x tanhyp(3 T / To) offers an improved, simple, and accurate method. When analyzing the B(T1, T2) curve, the inventors realized that it is very well represented by the hyperbolic tangent function. Furthermore, the hyperbolic tangent function exhibits the physically correct behavior: As the temperature increases, the function value asymptotically approaches a constant final value. This final value corresponds to B. max .

[0047] Since B(T) changes asymptotically (T co) to B maxAccording to one embodiment, it can be defined which approximation is considered sufficient. For practical application, To can be defined such that it is considered to be reached when 99.5% of the final value for B (i.e., B max ) can be achieved. The result is: B(To) = tanhyp ( 3 ) -Bm a x = 0.995-B max

[0048] Preferred are the support points for B(T2,T1) , to which an adjustment according to the formula B(T) = B max x tanhyp (3 T / To) is chosen. Thus, according to one embodiment, a sampling point spacing with T2-T1 < IOK is selected over a measuring range of -50°C to 200°C. However, T2-T1 can also be chosen more widely. According to one embodiment, T2-T1 can be < 25K, while still achieving a good fit. This is achieved according to a

[0049] This design is particularly good if at least 9 support points are available in the measuring range -50°C to 150°C.

[0050] According to one embodiment, the coupling constant can be obtained via To = (hx vo) / (2 k arcsinh y). Using the described procedure, material properties can be determined quickly and easily. In particular, it facilitates the evaluation of large data sets. Furthermore, the procedure can be applied automatically.

[0051] According to one embodiment, the use of the formula B(T) = Bmax x tanhyp(3 T / To) for material property optimization of NTC spinel ceramics is described.

[0052] Since the procedure described above enables an improved recording of material properties, material development including material property optimization can be carried out in a targeted manner.

[0053] Furthermore, a method using the formula Bmax x tanhyp(3 T / To) is described. This method can have the features described above.

[0054] According to one embodiment, the method is a method for optimizing the material properties of NTC spinel ceramics, wherein a fit is applied to a B-value curve of a starting material composition using the formula B(T) = Bmax x tanhyp(3 T / To). Together with the formula To = (hx \>o) / (2 k arcsinh y), the starting material properties of the starting material composition are determined. Based on these starting material properties, additives are selected that specifically influence the B-value at a specific temperature, the To value, and / or the specific resistance. The statements made above regarding the influences of the additives can apply accordingly according to exemplary embodiments. However, the use of the procedure is not limited to the compositions mentioned. For example, in the case of copper and aluminum, these properties can be adjusted by adding them.

[0055] In this way, material optimization can be carried out starting from a starting composition, whereby the advantages mentioned above apply.

[0056] According to one embodiment, for example, the value B max can be adjusted by the amount of copper. By adding copper, B max be lowered.

[0057] According to one embodiment, the resistivity can be adjusted by adjusting the amount of aluminum. The resistivity can be increased by adding aluminum.

[0058] According to one embodiment, the coupling factor can be adjusted by adjusting the aluminum content. The coupling factor can be reduced by adding aluminum.

[0059] According to one embodiment, To can be influenced by the copper content. For example, an increasing copper content can help lower To.

[0060] According to one embodiment, the process can be carried out as a multi-stage process. In this case, the composition already optimized by the process undergoes the process again by being subjected to the process sequence again as the starting material composition. In other words, the ceramic composition already optimized by additives is subjected to the described process sequence again at least once as the starting material composition.

[0061] Furthermore, a method for producing a multilayer thermistor is provided. According to this method, starting materials for a ceramic composition with the molecular formula Mn3-xy-zCOxAl2Cu2Ch are first provided. These starting materials can be, for example, oxides, carbonates, or hydroxides or similar materials of the formula Mn3-xy-zCo x AlyCu z04 mentioned metal atoms. The starting materials are weighed in such a way that 0.8 < x < 1.4, 0.5 < y < 1.1 and 0.15 < z < 0.35 is obtained for the ceramic composition. After weighing, the starting materials are ground and then calcined. The grinding can be wet grinding, for example. This makes it possible to achieve an average grain size between 0.5 and 1.5 pm, and preferably between 0.7 and 1.0 pm. Calcining can take place at 800 °C to 1000 °C, e.g. at 825 °C to 925 °C. The calcinate obtained in this way is then ground, which can be referred to as re-grinding. The targeted grain sizes can correspond to those of the first grinding step. Green films are then produced using the ground calcinate. The calcinate can be mixed with additives such as binders, wetting agents, dispersants, or similar. This mixture can be applied to a carrier film. The green films are then stacked.Stacking takes place together with starting materials for internal electrode layers. For example, the starting material for internal electrodes can be a metal paste and this can be printed onto several green films before stacking. One or more green thermistor components are cut out of the resulting stack. This or these are then debindered. The debindered green thermistor components are sintered. Sintering is preferably carried out at 1000 °C to 1200 °C. At this temperature, as few secondary phases as possible and as many main phases as possible are formed in the resulting ceramic. Decomposition can be reduced accordingly. A range of 1050 °C to 1150 °C is preferred in this regard. External electrodes are also applied, each of which contacts internal electrodes.

[0062] A monolithic thermistor can also be obtained in a similar manner. For this purpose, the steps for forming internal electrodes are omitted.

[0063] Furthermore, the ceramic obtained by the process can have the above-mentioned properties.

[0064] In the following, the invention is described with reference to exemplary embodiments and figures. Schematic representations of components are not to scale. Parts of these may be shown distorted with regard to their size, length or length ratio compared to other components. Accordingly, no sizes or ratios can be derived from the schematic drawings. Similar or similarly acting components are provided with the same reference numerals. This does not imply that all features of these components are always identical or that the invention is limited to the specifically described features of the exemplary embodiments.

[0065] Figure 1 schematically shows a first embodiment of an NTC thermistor in cross-section. Figure 2 shows a cross-sectional microscopy image of a second embodiment of an NTC thermistor.

[0066] Figure 3 shows the curve B(T) for a first thermistor ceramic and a fitting curve.

[0067] Figure 4 shows the curve B(T) for a second thermistor ceramic and a fitting curve.

[0068] Figure 5 shows the curve B(T) for a third thermistor ceramic and a fitting curve.

[0069] Figure 6 shows the curve B(T) for a fourth thermistor ceramic and a fitting curve.

[0070] Figure 7 shows a cross-sectional microscopy image of a third embodiment of an NTC thermistor.

[0071] Figure 8 shows the comparison of the curves of B(T) for the first and fourth thermistor ceramics.

[0072] Figure 9 shows a schematic cross-section of the third embodiment of the NTC thermistor from Figure 7.

[0073] Figure 10 shows a fourth embodiment of an NTC thermistor in schematic cross section.

[0074] Figure 11 shows a graph of the influence of the copper content on the value B ma x.

[0075] Figure 12 shows a graph of the influence of the aluminum content on the specific resistance at 25 ° . Figure 13 shows a graph of the influence of the

[0076] Aluminum content on the coupling factor y.

[0077] Figure 14 shows a graph showing the relationship between the coupling factor y and the temperature To .

[0078] Figures 1 and 2 each show two exemplary embodiments of a multilayer thermistor 1. Figure 1 shows a schematic cross-sectional view, and Figure 2 shows a cross-sectional microscopy image taken with a scanning electron microscope.

[0079] Both exemplary embodiments of the multilayer thermistor 1 have a plurality of ceramic layers 2 which are alternately stacked with internal electrode layers 3. The ceramic layers 2 consist of a ceramic material, as described below with reference to Figures 4 to 6. The internal electrode layers 3 consist of a palladium-silver alloy. The ceramic layers 2 and the internal electrode layers 3 can be obtained using conventional multilayer technology. They can thus be formed from green films and by applying metallization. One or more green films can be laminated for a ceramic layer 2. In the stacking or layer direction, end regions 5 are located at the top and bottom of the layer stack. These also comprise a ceramic material, which can be the same material as the ceramic layers 2.Even though no dimensions or proportions can be deduced from the schematic Figure 1, it can be seen that, depending on the embodiment, any termination areas 5 can be selected with different thicknesses. Furthermore, both multilayer thermistors 1 have external electrodes 4, which are obtainable using conventional metallization processes.

[0080] The multilayer thermistors 1 according to Figure 1 or Figure 2 have dimensions of length x depth x height = (1.0 ± 0.1) mm x (0.5 ± 0.05) mm x 0.6 mm or of length x depth x height = (1.6 ± 0.15) mm x (0.8 ± 0.15) mm x 0.9 mm. The ceramic layers in Figure 2 have a thickness of 30 pm.

[0081] Figure 3 shows the curve for B(T1, T2) for a ceramic composition with sample number 2917. This is a manganese-iron-nickel-copper spinel that was sintered at 1070°C. This is an NTC ceramic mentioned at the beginning, which is common in the prior art, but which does not meet the desired requirements, particularly in the temperature range below 0°C. Sufficient data on the relevant parameters that describe an NTC thermistor with regard to its properties are known in the prior art for this ceramic to also provide a basis for comparison for the improved determination method for these parameters according to the invention.

[0082] The composition, measured linearly with B(T1, T2) = ln(R2 / Rl) / (1 / T2 - 1 / T1), has a B(25°C, 100°C) value of 3454 K and a specific resistance of 1700 Ω-cm. This allowed the fabrication of a component with a nominal resistance of 10 kΩ. Numerous sampling points (T2-T1 < 10 K) were recorded.

[0083] The measured data were based on the function B(T) = B ma x x tanhyp(3 T / To) was applied. Here, B(T) is the B value of the ceramic, T is the temperature, and To is the lower limit temperature of the dominant small polaron hopping conduction (SPH conduction). Furthermore, the following approximation was used: To is achieved when B(To) = tanhyp (3) -Bmax = 0.995-B max , i.e. when 99.5% of the final value has been reached.

[0084] B was able to max = 3870 K and To = 680 K (407°C).

[0085] Furthermore, using To = (hx \>o) / (2 k arcsinh y), the coupling constant y was determined, which is 3.10. Here, h is the Planck constant, vo is the characteristic vibration frequency of the crystal lattice, and k is the Boltzmann constant.

[0086] Thus, relevant material parameters can be obtained. The composition shows unsatisfactory linearity and unsatisfactory sensitivity at T < 0°C. B(T) decreases below 3000 K in the analyzed range. The analysis shows which parameters have potential for improvement: To achieve a lower T o To obtain this, both the coupling factor y should be increased towards 4 and B max be reduced, e.g., to approximately 3400 to 4000 K or, preferably, to 3500 - 3600 Kelvin. This should make it possible to achieve a B(T1, T2) value of > 3000 Kelvin at subzero temperatures and improve the linearity of the RT characteristic curve.

[0087] This approach can replace a previously used less reliable approach. Previously, material properties were determined based on the results of fundamental research by Casado et al. (J. Phys. : Condens. Matter 6 (1994) 4685 - 4698) using To = (hv o ) / (2k arcsinh y) and B max = (yhv o ) / (2k). Here, To is the lower limit temperature of the dominant SPH line, v o the characteristic vibration frequency of the crystal lattice, h the Planck constant, k the Boltzmann constant, y the coupling constant of the vibrational energy with the electrical energy. B max = (yhv o ) / (2k) contributes a lot to the understanding of the deviation from linearity at low temperatures, but the parameters T o , B maxand y are difficult to determine from this. A graphical method is used in the literature, where ln(R) is plotted against (1 / T) in a (largest possible) temperature range. In the range of smaller (1 / T) values, a straight line is fitted as an asymptote. This allows the values ​​T o (coincidence of the RT values ​​with the straight line) and B max (slope of the asymptote) can be estimated. But ultimately, the drawing of the straight line is arbitrary and subject to great uncertainty.

[0088] For this reason, the improved and more accurate method presented above was developed, based on the analysis of the B(T1, T2) curve. It was found that the B-value curve (Figure 3) is very well represented by the hyperbolic tangent function. Furthermore, the hyperbolic tangent function exhibits the physically correct behavior: with increasing temperature, the function value asymptotically approaches a constant final value.

[0089] Figure 4 shows the B(T1, T2) curve for a ceramic composition according to the invention and an approximation using the method described above. The composition follows the empirical formula Mna-xy-zCOxAlyCUzCh, where x = 1.18, y = 0.60, and z = 0.28, i.e., it contains a main phase with this composition. The composition was sintered at 1150°C. The composition is assigned sample number 15. This composition shows improved values ​​compared to sample 2917, which corresponds to the conventional state of the art. The specific resistance is 280 Ω-cm, the B(25, 100) value is 3441 K, B max is 3549 K, To is 467 K and y is 3.7. Thus, compared to sample 2917, the coupling factor y has been increased towards 4, and B maxto a value of 3500-3600 K. This made it possible to achieve a B(T1, T2) value of > 3000 Kelvin at subzero temperatures and to improve the linearity of the RT characteristic, as shown in Figure 4. An excessive drop in the resistivity was also prevented. However, this could still be improved.

[0090] Figure 5 shows the B(T1, T2) curve for a preferred ceramic composition according to the invention and an approximation using the method described above. The composition follows the empirical formula Mna-x- y — z C o x Al y Cu z 04 , where x = 1.21, y = 0.90, and z = 0.22. The composition was sintered at 1100°C. The composition is assigned sample number 31.

[0091] This composition shows at least partially improved values ​​compared to sample 2917 and also to sample 15. The specific resistance is 1049 Ω-cm, the B (25, 100) value is 3581 K, B max is 3774 K, To is 543 K, and y is 3.5. Although the coupling factor is slightly reduced, the resistivity is significantly improved compared to sample 15. This allows a component with a nominal resistance of 10 kΩ to be easily achieved.

[0092] Figure 6 shows the B(T1, T2) curve for a particularly preferred ceramic composition according to the invention and an approximation using the method described above. The composition follows the empirical formula Mna-xy-zCOxAlyCUzCh, where x = 1.02, y = 0.80, and z = 0.27. The composition was sintered at 1100°C. The composition is assigned sample number 49.

[0093] This composition shows at least partially improved properties compared to the previously described samples. The specific resistance is 832 Ω-cm, the B (25, 100) value is 3446 K, B ma x is 3542 K, To is 457 K, and y is 3.8. Thus, the sample exhibits a very good balance of material properties.

[0094] The production process is also explained using sample 49.

[0095] First, a powder is prepared. The weight and ingredients used for an 80 kg powder batch can be found in Table 1.

[0096] Table 1

[0097] The components were then mixed with water and ground in a stirred ball mill until a target grain size d(50%) of 0.7 to 1.0 pm was obtained. After obtaining the target grain size, the suspension was dried and sieved. The powder was then calcined. For this purpose, the powder was filled into cordierite capsules and reacted in a stationary furnace. The heating rate was 5 K / min, the top temperature was 875°C, the holding time was 8 hours, and the cooling rate was -5 K / min. An XRD measurement showed a conversion rate of > 90%.

[0098] The reacted powder was then reground. The calcinate was first sieved dry, then mixed with water and reground until a target particle size d(50%) of 0.7 to 1.0 pm was obtained. After the target particle size was achieved, the suspension was dried and sieved again.

[0099] Green films were then produced. The powder is mixed with organic solvents and admixed with additives commonly used in this technology (binders, wetting agents, dispersants, etc.). After adjusting the appropriate viscosity and degassing the suspension, a ceramic film is produced on a carrier film using a film drawing machine. This film can be further processed using multilayer technology.

[0100] In the specific example, a film with a nominal thickness ("green thickness") of 28.5 pm was produced.

[0101] The sintered body is then produced. This is first explained using the illustration in Figure 7, which shows an SEM image of a sintered component in longitudinal section. In this exemplary embodiment, the so-called "tip design" was used. Two inner electrode layers 3 (electrode tips) protrude from each outer electrode 4 into the component, but without the inner electrode layers 3 overlapping. The current therefore flows primarily from the electrode tips to the opposite electrode tips of the counter electrode. This minimizes the cross-section of the current flow. The further apart the electrode tips are from each other, the higher the resistance.

[0102] First, internal electrodes were created by printing the ceramic foil. For this purpose, an AgPd paste (60% Ag, 40% Pd) was applied to the designated areas using screen printing.

[0103] The printed and unprinted films are then stacked on top of each other in a defined sequence.

[0104] Stacking sequence based on the embodiment shown in Figure 7:

[0105] 14 unprinted foils (form one of the final areas 5)

[0106] 1 printed foil (internal electrode tips of a level 1)

[0107] 5 unprinted foils (space to a level 2)

[0108] 1 printed foil (internal electrode tips of level 2)

[0109] 13 unprinted foils (which form an additional cover layer or finishing area)

[0110] The stacked films are then pressed (laminated) to ensure stable adhesion of the individual layers to one another.

[0111] The stack was then cut into components ("green parts"). The dimensions of the unsintered components (green parts) were 1.8 x 0.9 mm (L x W). The components were then debound. The binders required for the green processes were burned out before sintering. For this purpose, the components were heated slowly (< 1 K / min) to 450 °C under air purge (holding time 6 hours).

[0112] The components are then sintered. The components were sintered in an air atmosphere under the following sintering conditions:

[0113] Heating rate = 5 K / min

[0114] Top temperature = 1100 ° C (1080 ° C - 1140 ° C possible)

[0115] Holding time = 2 hours

[0116] Cooling rate = -5 K / min

[0117] For the evaluation measurements, the two outer electrodes are applied. For this purpose, the caps are coated with silver paste and baked. The formation of the metallization caps is clearly visible in the SEM image in Figure 7.

[0118] Furthermore, the components of the exemplary embodiment were passivated on the surface (glass coating) and the caps were nickel-plated and tin-plated by electroplating in order to be easy to solder.

[0119] Figure 8 also shows a comparison of the B(T) curves for samples 49 and 2917. Sample 49 exhibits significantly improved properties with respect to the B value at low temperatures and linearity. Sample 49 thus has significantly improved properties compared to the NTC ceramics commonly used in the state of the art, corresponding to sample 2917, and can therefore extend the technically feasible measuring range for NTC thermistors into the temperature range well below 0 °C.

[0120] Furthermore, Figure 9 shows a schematic cross-section of the third exemplary embodiment of an NTC thermistor. The properties mentioned above generally apply. The NTC thermistor is designed in a so-called "tip design". This can also be referred to as a gap design. In this case, two inner electrode layers 3 protrude from each outer electrode 4 into the component as electrode tips. These do not overlap. In such a structure, the current flows from one electrode tip to the opposite counter electrode.

[0121] Figure 10 shows a fourth exemplary embodiment of an NTC thermistor. In principle, the structure of this fourth exemplary embodiment can be similar to that of the third exemplary embodiment. However, while otherwise identical in size, the internal electrode layers 3 tend to protrude further into the internal volume or into the NTC ceramic. Furthermore, a free internal electrode 6 is provided, which has overlapping regions with the internal electrode layers 3.

[0122] In Tables 2 to 4 below, the results of a series of tests are listed and explained in connection with the values ​​from the tables and in connection with Figures 11, 12, 13 and 14.

[0123] In Table 2 below, the compositions of the components manganese (value : 3-xy- z ), cobalt (value : x ), aluminum (value ; y) , and copper (value : z ) resulting from the recipes are shown as indices of the

[0124] The composition formula is Mny-xy-zCOxAl yCUzCh. The ratio y:z is also given. Test reference number #49 corresponds to the exemplary embodiment described in Figure 6 (referred to there as sample 49). Please note that the values ​​for x, y, and z are rounded differently than in the tables.

[0125] Furthermore, the structure of the respective examples can be seen from Table 3. In particular, the Design column indicates whether it is a tip design, as shown in Figure 9 (term "GAP"), or a T-design. The total number of internal electrodes is always 4 for the tip design. For the T-design, this is always 5, with one of the internal electrodes being a free internal electrode, as shown in Figure 10.

[0126] The so-called gap widths are also specified. This always refers to the smallest distance between an outer electrode and the opposite outer electrode. The gap width here therefore refers to the distance between the outer electrodes and not the distance between the inner electrodes. For example, Figure 7 shows that the metallization of the outer electrodes 4 extends over the edges of the ceramic body. This is achieved by dipping into a paste by means of which the metallization is produced. The distance between the sections of the opposite outer electrodes that extend over the edges determines how much current can flow over the component surface. This current flow is added to the current inside the component. By varying the gap width, the resistance of the component can be readjusted even after sintering if necessary, since the immersion depth in the paste for producing the outer electrodes can be varied.This allows adaptation without altering the internal structure. The examples shown were sintered at a maximum sintering temperature of 1100 °C and with a holding time of 2 hours.

[0127] The properties determined for the test series are given in Table 4. The values ​​determined in Table 4 were adapted accordingly, as shown for example for test reference number #49 above, using the formulas B(T) = B ma x x tanhyp(3 T / To) and To = (hx vo) / (2 k arcsinh y) are determined, as explained at the beginning.

[0128] Table 2

[0129] Table 3

[0130] Table 4

[0131] The tables show clear trends for the effects of the various copper and aluminum additives. The corresponding graphs in Figures 11, 12, 13, and 14 reflect some of these trends.

[0132] First of all, it can be seen from the tables that x lies between 1.020 and 1.206, y lies between 0.559 and 0.937 and z lies between 0.195 and 0.274.

[0133] Figure 11 shows that the addition of copper is an important lever for reducing B ma x. This is also evident in the test reference examples T19, T20, and T21, which have the same design and the same gap dimensions and also have the same x- and y-values. The z-value of these examples was varied. It is clear that B ma x can be reduced by the copper content.

[0134] Comparing examples T19, T20, and T21, it is also clear that To can be influenced by the copper content (z-value). Thus, To can be reduced as the copper content increases.

[0135] As can be seen from Figure 12, the addition of aluminum results in an exponential increase in resistivity at 25 °C. This is also reflected in the comparison of the test results for T20, T22, and T23, which have the same design and the same gap width and in which x and z remain constant. The y value is varied here. As can be seen from Table 4, there is a clear trend of increasing resistivity with increased aluminum content.

[0136] As can be seen in combination in Figures 13 and 14, the aluminum content is also an important lever for adjusting the coupling factor and thus for reducing To. As the aluminum content increases, the coupling factor y decreases, which in turn increases To, as shown in Figure 14. The inventors explain this by saying that a high coupling factor, which is a measure of the conversion of lattice vibrations in electrical charge transport, is necessary to enable the onset of small polaron hopping conduction at low temperatures To. Again, this trend can be confirmed by comparing the test results for T20, T22, and T23. As the value y increases, y decreases and To increases. The scatter in Figure 13 is attributed, among other things, to the fact that different designs are being compared here.If the examples T20, T22 and T23, in which only the aluminum value is varied, are compared, the trend becomes even clearer.

[0137] Reference symbol list

[0138] 1 multilayer thermistor

[0139] 2 Ceramic layer 3 Internal electrode layer

[0140] 4 Outer electrode

[0141] 5 Final area

[0142] 6 free electrode

Claims

Patent claims 1. NTC composition comprising a main component based on manganese-cobalt oxide spinel, wherein the proportions of an addition of Al and Cu are determined using the formulas B(T) = Bmax x tanhyp(3 T / To) and To = (hx vo) / (2 k arcsinh y), where B(T) is the B-value of the ceramic, T is the temperature, To is the lower limit temperature of the dominant small polaron hopping conduction, h is Planck constant, vo is the characteristic vibration frequency of the crystal lattice, k is the Boltzmann constant, and Y is the coupling constant, Bmax is the temperature-independent B value of the dominant small polaron hopping conduction, where the proportions are adjusted so that a value B ma x is 3400 to 4000 K and / or a lower limit temperature To d of 600 Kelvin is reached.

2. NTC composition having a main component with the molecular formula Mn 3 -xy-zCo x AlyCUz04 , where 0.8 < x < 1.4, 0.5 < y < 1.1 and 0.15 < z < 0.

35.

3. NTC composition according to claim 2, wherein 0.15 < z < 0.30 or preferably 0.15 < z < 0.

28.

4. NTC composition according to claim 2 or 3, wherein 0.21 < z < 0.

35.

5. NTC composition according to one of claims 2 to 4, wherein y:z > 2.

4.

6. NTC composition according to one of claims 2 to 5, wherein 0.90 < x < 1.30 or preferably 0.95 < x < 1.

25.

7. NTC composition according to one of claims 2 to 6, wherein 0.55 < y < 0.95 or preferably 0.60 < y < 0.

95.

8. NTC composition according to one of claims 2 to 7, wherein 0.6 < y < 1.

1.

9. NTC composition according to any one of claims 1 to 8, wherein the main component constitutes at least 90% of the phases of the NTC composition.

10. A thermistor comprising the NTC composition according to any one of claims 1 to 9.

11. Multilayer thermistor comprising a ceramic layer comprising the NTC composition according to one of the claims 1 to 10.

12. Multilayer thermistor according to claim 11, wherein it has a volume of 0.075 to 10 mm 3 has.

13. A multilayer thermistor according to claim 11 or 12, wherein the ceramic layer has a thickness of 10 to 100 pm and the thermistor is constructed from one or more layers.

14. Multilayer thermistor according to one of claims 11 to 13, which has a nominal resistance at 25 °C of 1 to 20 kΩ.

15. Multilayer thin film thermistor comprising a Ceramic layer comprising the NTC composition according to one of claims 1 to 9.

16. A method for producing a multilayer thermistor, wherein Starting materials for a ceramic composition with the molecular formula Mna-xy-zCOxAlyCUzCh are provided, wherein the starting materials are weighed in such a way that 0.8 < x < 1.4, 0.5 < y < 1.1 and 0.15 < z < 0.35 is obtained for the ceramic composition, the starting materials are ground, the powder obtained after grinding the starting materials is calcined, the calcinate obtained is ground, green sheets are produced with the ground calcinate, the green sheets are stacked together with starting materials for internal electrodes, green thermistor components are cut from the stack, the green thermistor components are debindered, the debindered green thermistor components are sintered to sintered thermistor components at 1000 to 1200 °C, and external electrodes are applied to the sintered thermistor components.

17. The method according to claim 16, wherein the starting material for internal electrodes is a metal paste and this is printed on some green sheets before stacking.

18. The method according to claim 16 or 17, wherein the sintering is carried out at 1050°C to 1150°C.

19. Use of the formula B(To) = B ma x x tanhyp (3 T / To) , where B(T) is the B-value of the ceramic, Bmax is the temperature-independent B-value of the dominant small polaron hopping conduction, T is the temperature, To is the lower limit temperature of the dominant small polaron hopping conduction, for determining the material properties of an NTC spinel ceramic.

20. Using the formula B(T) = B ma x x tanhyp (3 T / To) , where B(T) is the B-value of the ceramic, Bmax is the temperature-independent B-value of the dominant small polaron hopping conduction, T is the temperature, To is the lower limit temperature of the dominating small polaron hopping conduction, for material property optimization of NTC spinel ceramics.

21. Use according to claim 19 or 20, wherein To is defined as being reached when B(T) = tanhyp ( 3 ) -E x = 0.995-B ma x is valid .

22. Method for material property optimization of NTC spinel ceramics, wherein a curve of the B value of a starting material composition is applied using the formula B(T) = Bmax x tanhyp (3 T / To) an adjustment is applied, where B(T) is the B-value of the ceramic, Bmax is the temperature-independent B-value of the dominant small polaron hopping conduction, T is the temperature, To the lower limit temperature of the dominant small polaron Hopping conduction is , and with this together with the formula To = (hx \> o ) / ( 2 k arcsinh y) , where h is Planck's constant, vo is the characteristic vibration frequency of the crystal lattice, k is the Boltzmann constant, and Y is the coupling constant, a determination of the raw material properties of the raw material composition is obtained, and based on these raw material properties, additives are selected which specifically influence the B value at a certain temperature, the To value and / or the specific resistance.

23. Process according to claim 22, wherein the process is carried out as a multi-stage process, wherein the ceramic composition already optimized by additives is subjected as starting material composition again to the described process sequence at least once.

Images