Material for a Thermoelectric Element and Method for Producing a Material for a Thermoelectric Element

Abstract

A material for a thermoelectric element and a method for producing a material for a thermoelectric element are disclosed. In an embodiment the thermoelectric element includes a material comprising calcium manganese oxide that is partially doped with Fe atoms in positions of Mn atoms.

Claims

1-15. (canceled)

16. A thermoelectric element comprising: a material comprising calcium manganese oxide that is partially doped with Fe atoms in positions of Mn atoms.

17. The thermoelectric element according to claim 16, wherein a doping with Fe atoms provides a content of z20% at the positions of Mn atoms.

18. The thermoelectric element according to claim 16, wherein the material is further doped with an element that provides electrons for electrical conductivity in positions of Ca.sup.2+ atoms.

19. The thermoelectric element according to claim 18, wherein the element is selected from the group consisting of rare earth metals, Sb.sup.3+, and Bi.sup.3+.

20. The thermoelectric element according to claim 18, wherein a doping with the element provides a content of 0<y0.5 at the positions of Ca atoms.

21. The thermoelectric element according to claim 16, wherein the material is further doped with a divalent element in positions of Ca.sup.2+ atoms.

22. The thermoelectric element according to claim 21, wherein the divalent element is selected from a group consisting of Mg.sup.2+, Sr.sup.2+, Ba.sup.2+, Zn.sup.2+, Pb.sup.2+, Cd.sup.2+ and Hg.sup.2+.

23. The thermoelectric element according to claim 21, wherein a doping with the divalent element provides a content of 0<x0.5 at the positions of Ca atoms.

24. The thermoelectric element according to claim 23, wherein the doping with the divalent element provides a content of x0.05.

25. The thermoelectric element according to claim 16, wherein the material is represented by the general formula Ca.sub.1-x-yISO.sub.xDON.sub.yMn.sub.1-zFe.sub.zO.sub.n, wherein ISO denotes a divalent element that can replace Ca.sup.2+ in a crystal lattice, wherein DON denotes an element that can replace Ca.sup.2+ in the crystal lattice and provides electrons for electrical conductivity, and wherein 0x0.5; 0<y0.5; 0.0001z<0.2; n2.

26. The thermoelectric element according to claim 16, further comprising a second material based on the composition (Ca.sub.3-xNa.sub.x)Co.sub.4O.sub.9-, wherein 0.1x2.9 and 0<2.

27. A method for producing a material for a thermoelectric element, the method comprising: firing a material, wherein, for a maximum temperature T.sub.max, T.sub.maxT.sub.S75 C. is true, wherein T.sub.S denotes a melting temperature of the material, and wherein a maintenance time of at least 30 minutes is observed on cooling at a preset temperature.

28. The method according to claim 27, wherein the temperature during the maintenance time is in a range of 700 C. to 800 C.

29. The method according to claim 27, wherein the maximum temperature is greater than or equal to T.sub.S75 C. for at least 10 hours.

30. The method according to claim 27, wherein a cooling rate of less than or equal to 1 C./min is used in cooling.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In the following, the subject matter described here is explained in further detail based on working examples depicted schematically and not to scale.

[0047] The figures show the following:

[0048] FIG. 1 is a diffractogram of a material for a thermoelectric element,

[0049] FIG. 2 is a diagram of electrical conductivity as a function of maximum firing temperature for two materials,

[0050] FIG. 3 is a micrograph of a material,

[0051] FIG. 4 is a diagram of electrical conductivity as a function of temperature for a material,

[0052] FIG. 5 is a diagram of the Seebeck coefficient as a function of temperature for the material of FIG. 4,

[0053] FIG. 6 is a diagram of thermal conductivity as a function of temperature for the material of FIG. 4,

[0054] FIG. 7 is a diagram of figure of merit as a function of temperature for the material of FIG. 4,

[0055] FIG. 8 is a diagram of thermal conductivity as a function of temperature for two further materials,

[0056] FIG. 9 is a diffractogram of two materials,

[0057] FIG. 10 is a diagram of sintering density as a function of Fe content in a material,

[0058] FIG. 11 is a diagram of the Seebeck coefficient as a function of Fe content in the material of FIG. 10,

[0059] FIG. 12 is a diagram of sintering density as a function of Fe content in two materials,

[0060] FIG. 13 is a diagram of the Seebeck coefficient as a function of Fe content in the two materials of FIG. 12, and

[0061] FIG. 14 is a working example of a thermoelectric generator having a plurality of thermoelectric elements.

DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS

Method for Producing the Material

Example: Preparation of Ca.SUB.0.85.Sr.SUB.0.10.Dy.SUB.0.05.Mn.SUB.0.975.Fe.SUB.0.025.O.SUB.3

[0062] A method is first described for producing a material for a thermoelectric element.

[0063] For example, the method is used to produce a material of the composition Ca.sub.0.85Sr.sub.0.10Dy.sub.0.05Mn.sub.0.975Fe.sub.0.025O.sub.3. However, the method is not limited to this material, but is also suitable for producing other materials for thermoelectric elements.

[0064] For example, the material, preferably a complex metal oxide, can be produced by means of the so-called mixed oxide process. However, it is also possible to use other production methods, such as wet-chemical routes or mechanical alloying.

[0065] Stoichiometric amounts of CaCO.sub.3, SrCO.sub.3, Mn.sub.3O.sub.4, Fe.sub.2O.sub.3, and Dy.sub.2O.sub.3 are weighed in and wet-ground (using deionized water). A microfine grain size is achieved using suitable fine milling technology such as a planetary mill or an agitator bead mill. The grain size distribution is preferably d(0.5)<1 m and d(0.9)<1.5 m. This makes it possible to achieve sufficient reactivity in the subsequent calcining process. The milled suspension is dried and sifted.

[0066] Calcination, in which a solid-state reaction takes place to form a complex metal oxide, is carried out, for example, at 1100 C. in an air atmosphere for several hours. In this reaction, a largely single-phase material is preferably obtained. Small amounts of unreacted raw materials from second phases can further react in subsequent firing to form a complex metal oxide.

[0067] FIG. 1 shows an x-ray diffractogram (XRD) of the working example. The measured radiation intensities I are plotted against the angle of the radiation source, sample, and detector (2 angle). A comparison with the values reported in the literature for CaMnO.sub.3 shows that incorporation of Fe atoms has taken place without any substantial change in the structure of the ABO.sub.3 unit cell.

[0068] In order to provide good sinterability for firing of the components, it is advantageous to repeat the step of micronization. For this purpose, the powder is again mixed with deionized water and finely milled. One should preferably aim for a grain size distribution having roughly the following properties: d(0.5)=0.5 m and d(0.9)1 m. In the next step, a pressable powder or granulate is produced from the milled suspension. This can be carried out directly by spray-drying of a suspension mixed with a binder, orin the case of small amounts, for example,by drying the suspension and then manually adding a binder component.

[0069] Shaping of the component is now carried out. Components are preferably molded by means of dry pressing. For the production of conversion modules, for example, rod-shaped or cylindrical components are required. Before subsequent firing of the components, pre-decarbonization is advantageous (using thermal releasing agents). It has been found that firing of the components is of great importance in configuring the thermoelectric properties of the material described.

[0070] The measurements of sintering density were carried out on a cylindrical component having a diameter of 11 mm and a height of 5.5 mm. The measurements of electrical conductivity and thermopower were carried out on a cylindrical component having a diameter of 10 mm and a height of 1 mm. The measurements of thermal conductivity were carried out on a cylindrical component having a diameter of 11 mm and a height of 1 mm.

Optimization of the Firing Process

Example: Ca.SUB.0.95.Dy.SUB.0.05.MnO.SUB.3 .and Ca.SUB.0.95.Gd.SUB.0.05.MnO.SUB.3

[0071] The optimized firing process developed is described below by way of example for the materials Ca.sub.0.95Dy.sub.0.05MnO.sub.3 and Ca.sub.0.95Gd.sub.0.05O.sub.3. The method is not limited to these materials, but was successfully used in producing all of the tested formulations of a complex metal oxide.

[0072] A particularly high maximum firing temperature is used in the method. However, the maximum firing temperature should be below the melting temperature, as the component could otherwise melt and be destroyed. The firing temperature is preferably just below the melting temperature of the material used.

[0073] For example, the maximum firing temperature T.sub.max is 100 C. below the melting temperature T.sub.S or above it, i.e., T.sub.maxT.sub.S100 C. In an embodiment, T.sub.maxT.sub.S75 C. is true, e.g., T.sub.maxT.sub.S50 C. The firing temperature is preferably at least 10 C. below the melting temperature, i.e., T.sub.maxT.sub.S10 C. For example, the firing temperature is in the range of 10 C. to 50 C. below the melting temperature. For the materials tested here, for example, the melting temperature is approx. 1400 C.

[0074] In the method, a very long maintenance time at the maximum temperature is preferred. In particular, the maintenance time is at least 10 h. For example, the maintenance time is at least 15 h.

[0075] Sintering is preferably carried out in an atmosphere having sufficient oxygen. For example, sintering is carried out in an air atmosphere or an oxygen-enriched atmosphere.

[0076] Moreover, the method is characterized by a slow cooling rate. In particular, a cooling rate of less than or equal to 1 C./min is used in cooling from 1000 C. to 600 C.

[0077] Furthermore, in cooling from 1000 C. to 600 C., an additional maintenance time of at least one hour is preferably used.

[0078] The slow cooling rate and additional maintenance time allow the most complete conversion from Mn.sup.3+ to Mn.sup.4+, so that the compound obtained is as stoichiometric as possible and has particularly favorable thermoelectric properties. For this purpose, cooling below a specified temperature is required. On the other hand, the rate of diffusion of the required oxygen in the ceramic decreases with falling temperature. There is therefore an optimum temperature for the maintenance time. In sintering in air and at atmospheric pressure, this temperature is in the range of 700 C. to 800 C., e.g., 750 C. Oxygen uptake is accompanied by phase transitions, which can easily cause the brittle ceramic to crack. A slow cooling rate in the range of the phase transition and below makes it possible to produce a ceramic having few or no cracks.

[0079] It has been found that by means of this method, it is possible to find a process window within which favorable grain growth with advantageous properties can be achieved without melting of the ceramic. Moreover, it has been found that a material manufactured in this manner is highly resistant to air and oxygen. In particular, the material remains stable in air up to high temperatures (800 C.).

[0080] The following table shows the electrical conductivity and density of the fired ceramic for the two formulations at various maximum firing temperatures.

TABLE-US-00001 Max. firing Electrical Density of temperature conductivity ceramic Formulation ( C.) (S/cm) (g/ml) Ca.sub.0.95Dy.sub.0.05MnO.sub.3 1150 148 4.27 1250 304 4.66 1350 428 4.66 Ca.sub.0.95Gd.sub.0.05MnO.sub.3 1150 123 4.07 1250 285 4.62 1350 416 4.62

[0081] As can be seen from the table, at a maximum firing temperature of T.sub.max=1150 C., the electrical conductivity of the two formulations is below 150 S/cm. At this firing temperature, the density of the ceramic is <4.3 g/ml for the two formulations. When the maximum firing temperature is increased to T.sub.max=1250 C., electrical conductivity increases sharply. The sintering density also increases. On a further increase in the maximum firing temperature to T.sub.max=1350 C., the electrical conductivity of the two formulations increases to a value of >400 S/cm. The density of the ceramic is >4.6 g/ml.

[0082] FIG. 2 shows a graphical representation of electrical conductivity G as a function of maximum firing temperature T.sub.max for the two formulations. The electrical conductivity shows virtually linear dependency on maximum firing temperature.

[0083] FIG. 3 shows the microgram obtained in sintering as an example for one of the working examples.

[0084] By means of the method used, taking a primary grain size of 0.5 m as a starting point, a stable and dense ceramic composed of grains measuring 10 m in diameter can be produced. The growth of the grains was therefore greater than one order of magnitude. The favorable electrical conductivity can be attributed to the large grain diameter, as in this case only minor dispersion of the charge carriers takes place at the grain boundaries.

[0085] In the following, various materials and components containing the materials are characterized. All of the materials or components were produced by the above-described method. In particular, the components of a complex metal oxide can be determined by comparing their properties.

Example: Ca.SUB.0.97.La.SUB.0.03.MnO.SUB.3

[0086] As a first example, a ceramic based on calcium manganese oxide (calcium manganate) is tested in which Ca.sup.2+ has been partially replaced by a suitable atom with a valence of 3+, corresponding to donor doping in the A position. The ceramic is represented by the formula Ca.sub.0.97La.sub.0.03MnO.sub.3. Sintering was carried out at a maximum temperature of 1320 C.

[0087] The following properties in particular are relevant for thermoelectric conversion. Characterization was conducted at room temperature.

TABLE-US-00002 Sintering density = 4.61 g/cm.sup.3 Electrical conductivity = 258 S/cm Thermopower = 125 V/K Power factor ( .Math. .sup.2) PF = 4.06 .Math. 10.sup.4 W/(mK.sup.2) Thermal conductivity = 3.89 W/(mK) Figure of merit ZT = 0.033

[0088] For thermoelectric conversion, the dependency of the properties on the surrounding temperature is of particular interest. The ends of a thermoelectric component are at different temperature levels. The amount of energy converted increases with increasing temperature difference, provided that the figure of merit does not decrease disproportionately with temperature.

[0089] FIG. 4 shows the temperature dependency of electrical conductivity for the Ca.sub.0.97La.sub.0.03MnO.sub.3 ceramic. The measurements were carried out in two components. The components were produced under the same conditions. The virtually identical measurement results demonstrate the favorable reproducibility of component production and of the measurement method.

[0090] Electrical conductivity decreases with increasing temperature. This reduction in conductivity with temperature is also referred to as metallic behavior.

[0091] FIG. 5 shows the temperature dependency of the Seebeck coefficient for the two components. In this case, an increase in the absolute value with increasing temperature can be observed.

[0092] FIG. 6 shows the temperature dependency of thermal conductivity K for one of the components. Thermal conductivity was measured by means of a laser flash method. Thermal conductivity decreases with increasing temperature.

[0093] Based on these measurements, the figure of merit ZT can be derived by means of equation (1).

[0094] FIG. 7 shows the course of the figure of merit ZT, measured in the two components of the Ca.sub.0.97La.sub.0.03MnO.sub.3 ceramic. The figure of merit reflects the efficiency of thermoelectric conversion.

Example: Ca.SUB.0.9.Sr.SUB.0.05.Yb.SUB.0.05.MnO.SUB.3

[0095] As a further example, a ceramic based on calcium manganate was tested in which donor doping with Yb.sup.3+ was carried out instead of donor doping with La.sup.3+. The doping content was also increased from 3% to 5%. In this case, an increase in the number of charge carriers and thus improved electrical conductivity is to be expected. However, the number of charge carriers also affects the result (See, e.g., Heike's formula). At a donor content of y>50%, the conduction mechanism usually changes to hole conduction, so the donor content should be less than 50%.

[0096] In addition, 5% of the Ca.sup.2+ atoms were replaced by specific heavier Sr.sup.2+ atoms. With an unchanged unit cell of the perovskite structure, this should make it possible to increase the density of the material and reduce thermal conductivity.

[0097] The material is therefore represented by the formula Ca.sub.0.9Sr.sub.0.05Yb.sub.0.05MnO.sub.3. The above-described method was again used for production.

[0098] Characterization of the component at room temperature was again conducted:

TABLE-US-00003 Sintering density = 4.70 g/cm.sup.3 Electrical conductivity = 399 S/cm Thermopower (Seebeck coefficient) = 101 V/K Power factor PF = 4.05 .Math. 10.sup.4 W/(mK.sup.2) Thermal conductivity = 3.08 W/(mK) Figure of merit ZT = 0.040

[0099] It can be derived from these values that the improved electrical conductivity is compensated for by the reduced thermopower, so that the power factor remains approximately the same. An increase in sintering density of approx. 2% and a decrease in thermal conductivity of approx. 20% can be observed, which therefore also improves the figure of merit ZT by approx. 20%.

[0100] Overall, it was found that material modifications that specifically lead to denser structures with reduced thermal conductivity constitute an interesting alternative to material changes that only alter the electronic properties of the oxide ceramic.

Example: Ca.SUB.0.85.Sr.SUB.0.10.Dy.SUB.0.05.MnO.SUB.3

[0101] As a further example, a ceramic based on calcium manganate was tested in which even more Ca.sup.2+ atoms (10%) were specifically replaced by heavier Sr.sup.2+ atoms. The donor doping content was kept at 5%, but doping was carried out in this case with Dy.sup.3+.

[0102] The material is therefore represented by the formula Ca.sub.0.85Sr.sub.0.10Dy.sub.0.05MnO.sub.3. The above-described method was again used for production.

[0103] At room temperature, the following properties are seen compared to the preceding examples:

TABLE-US-00004 Thermal Comparison Sintering density conductivity example Material (g/ml) (W/mK.sup.2) 1 Ca.sub.0.97La.sub.0.03MnO.sub.3 4.61 3.89 2 Ca.sub.0.90Sr.sub.0.05MnO.sub.3 4.70 3.08 3 Ca.sub.0.85Sr.sub.0.10Dy.sub.0.05MnO.sub.3 4.74 2.88

[0104] The Ca.sub.0.85Sr.sub.0.10Dy.sub.0.05MnO.sub.3 and Ca.sub.0.9Sr.sub.0.05Yb.sub.0.05MnO.sub.3 ceramics thus show increased sintering density and reduced thermal conductivity.

[0105] FIG. 8 shows the dependency of temperature on thermal conductivity for the materials

[0106] Ca.sub.0.85Sr.sub.0.10Dy.sub.0.05MnO.sub.3 and Ca.sub.0.9Sr.sub.0.05Yb.sub.0.05MnO.sub.3. It can be seen that there is reduced thermal conductivity in the entire range of 300 to 1000 Kelvins.

[0107] The three examples show that the efficiency of thermoelectric conversion can be improved by means of structures having increased density and reduced thermal conductivity.

[0108] It would be expected that this effect could be further increased by further or complete replacement of Ca.sup.2+ atoms by specific heavier Sr.sup.2+ atoms. However, it was found that at a content of over 20% Sr.sup.2+ atoms, there is an increasing change in the unit cell of the perovskite, which thus alters the electronic properties (conductivity, thermopower) in an unfavorable manner. The changed structure of the unit cell can be seen, for example, on the x-ray diffractogram (XRD).

[0109] It has been found that a further increase in efficiency can be achieved by incorporating suitable atoms that are even heavier than Sr.sup.2+. For example, Ba.sup.2+ and Pb.sup.2+ are suitable for this purpose.

Example: Ca.SUB.0.85.Sr.SUB.0.10.X.SUB.0.05.Mn.SUB.1-z.Fe.SUB.z.O.SUB.3.(XDy, Bi)

[0110] As a working example of a material based on CaMnO.sub.3 doped with Fe atoms that replace Mn atoms, a material is characterized below that is represented by the formula Ca.sub.0.85Sr.sub.0.10X.sub.0.05Mn.sub.1-zFe.sub.zO.sub.3, where X is equal to Dy or Bi. A portion of the Mn atoms in the B positions is therefore replaced by Fe atoms. The vast majority (>80%) of the B positions are occupied by Mn atoms. For this reason, the crystal structure and stability of the manganate compound that are beneficial for thermoelectric conversion are largely retained.

[0111] FIG. 9 shows a comparison of the x-ray diffractograms for the compounds

Ca.sub.0.85Sr.sub.0.10Bi.sub.0.05MnO.sub.3 and Ca.sub.0.85Sr.sub.0.10Bi.sub.0.05Mn.sub.0.90Fe.sub.0.10O.sub.3.

[0112] A virtually identical reflex pattern can be seen, although 10% of the Mn atoms in the B position are replaced by Fe atoms. This means that incorporation of the Fe atoms took place without any substantial change in the structure of the ABO.sub.3 unit cell.

[0113] In the following, the effect of the content of incorporated Fe atoms is investigated in further detail. In particular, the content z of Fe atoms in the material of formula Ca.sub.0.85Sr.sub.0.10Dy.sub.0.05Mn.sub.1-zFe.sub.zO.sub.3 is varied.

[0114] FIG. 10 shows the dependency of sintering density on the content z of Fe atoms in this material. Fe contents of z=o %, 0.5%, 1%, 2.5%, 5% and 10% were tested. The compensation curve was roughly estimated.

[0115] It can be seen from FIG. 10 that with an added amount of up to 5% Fe, density is higher than the value for the Fe-free compound. At 10% or above, density again decreases sharply. Because of the increased density at up to 5% Fe, and because the Fe atoms in the lattice are to be seen as defects for phonons, it can be concluded that the thermal conductivity in this range is also below the value for the Fe-free compound.

[0116] FIG. 11 shows the dependency of the Seebeck coefficient on the Fe content z of this material. Measurements were conducted at room temperature. Fe contents of z=0%, 0.5%, 1%, 2.5%, 5% and 10% were again tested. The compensation curve was roughly estimated.

[0117] Up to approx. 10% Fe content, thermopower is negative (the material is of the n-type). Up to 5%, the absolute value of the Seebeck coefficient increases. With addition of slightly more than 5% Fe, thermopower again decreases sharply.

[0118] The parameters of thermoelectric conversion can therefore be optimized by means of the measurement values from FIGS. 10 and 11. It has been found that a material having an Fe content in the range of 0.0001 to 0.2 shows advantageous properties. At an Fe content of z>0.2, electronic conductivity is extremely low.

Example: Ca.SUB.1-x-0.05.Sr.SUB.x.Dy.SUB.0.05.Mn.SUB.1-z.Fe.SUB.z.O.SUB.3

[0119] As a further working example, a material is characterized in which the Sr content is increased from 10% to 20% compared to the preceding working example. In particular, a material of the formula Ca.sub.1-x-0.05Sr.sub.xDy.sub.0.05Mn.sub.1-zFe.sub.zO.sub.3 is characterized. Again, a variation of the content z of Fe atoms is tested.

[0120] FIG. 12 shows the dependency of sintering density on Fe content z for Sr content of x=10% and x=20%.

[0121] The incorporation of a greater number of heavier Sr atoms increases the density of the ceramic produced and reduces its thermal conductivity. However, it has been found that with an Sr content of x>50%, the properties strongly resemble the more unfavorable properties of SrMnO.sub.3.

[0122] Even with an Sr content of 20%, an additional positive effect on sintering density is seen with Fe addition of up to 5%.

[0123] FIG. 13 shows the dependency of thermopower a on Fe content z for an Sr content of x=10% and x=20%.

[0124] The resulting course is similar to that shown in the working example of FIG. 11. Up to a content of approx. 10% of added Fe, the thermopower is negative (the material is of the n-type). Up to approx. 5% Fe content, the absolute value of thermopower increases in an advantageous manner.

[0125] FIG. 14 shows a working example of a thermoelectric element 1, in particular a thermoelectric generator.

[0126] The generator has a so-called H structure. The generator is configured as a module having a plurality of materials 2, 3 of different types. The materials 2, 3 form the legs of the generator. The first material 2 is of the n-type, and as described above, is based on calcium manganese oxide. The second material 3 is of the p-type. The two materials 2, 3 preferably have comparable figures of merit. In this case, particularly favorable energy conversion can be achieved overall.

[0127] For example, a sodium cobaltate based on the general formula (Ca.sub.3-xNa.sub.x)Co.sub.4O.sub.9-, where 0.1x2.9 and 0<2, and particularly where 0.3x2.7 and 0<1, is used for the second material 3.

[0128] The legs comprising the materials 2, 3 are thermally parallel and electrically connected in series. Contacts 6 composed, e.g., of an Ag paste are provided for electrical connection purposes.

[0129] The generator has two electrical connections 4, 5. Thermal contact elements 7, 8 are also present that simultaneously form electrical insulators. Examples of compounds used for this purpose include Al.sub.2O.sub.3, AlN and/or Si.sub.3N.sub.4. For example, the materials 2, 3 are sintered together with the electrical contacts 6 and the thermal contact elements 7, 8.

[0130] When there is a temperature difference between the two contact elements 7, 8, a voltage referred to as thermopower is generated between the electrical connections 4, 5.

[0131] In an alternative embodiment, a thermoelectric element, in particular a thermoelectric generator, has only two legs composed of different materials 2, 3.