Alloy, sintered article, thermoelectric module and method for the production of a sintered article

Abstract

An alloy is provided that consists essentially of
(Ti.sub.xTa.sub.yV.sub.zA.sub.cNb.sub.1-x-y-z-c)(Fe.sub.1-dMn.sub.d).sub.a(Sb.sub.1-eSn.sub.e).sub.b,
wherein 0.06≤x≤0.24, 0.01≤y≤0.06, 0/08≤z≤0.4, 0.9≤(a, b)≤1.1, 0≤c≤0.05, 0≤d≤0.05 and 0≤e≤0.1 and A is one or more of the elements in the group consisting of Zr, Hf, Sc, Y, La, and up to 5 atom % impurities.

Claims

1. An alloy consisting essentially of
(Ti.sub.xTa.sub.yV.sub.zA.sub.cNb.sub.1-x-y-z-c)(Fe.sub.1-dMn.sub.d).sub.a(Sb.sub.1-eSn.sub.e).sub.b wherein 0.06≤x≤0.24, 0.01≤y≤0.045, 0.075≤z≤0.3, 0.9≤(a, b)≤1.1, 0≤c≤0.05, 0≤d≤0.05, 0≤e≤0.1, wherein A is one or more of the elements in the group consisting of Zr, Hf, Sc, Y, La, and up to 5 atom % impurities, and wherein the alloy has a thermoelectric figure of merit ZT of ZT≥0.8, where T=500° C.

2. An alloy according to claim 1, wherein the alloy has a positive Seebeck coefficient.

3. An alloy according to claim 1, wherein c=0, d=0 and e=0.

4. A sintered article comprising an alloy according to claim 1.

5. A sintered article according to claim 4, wherein the sintered article has an average grain size of greater than 1.25 μm.

6. A sintered article according to claim 4, wherein the sintered article has a density D, D being ≥90% of the theoretic density D.sub.i.

7. A thermoelectric module comprising at least one thermoelectric element made of an alloy according to claim 1 and having at least one thermoelectric element made of an N-type alloy.

8. A thermoelectric module comprising at least one thermoelectric element comprising a sintered article according to claim 4 and having at least one thermoelectric element made of an N-type thermoelectric alloy.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows a schematic view of a thermoelectric module.

(2) FIG. 2 shows a scanning electron microscope image of the microstructure from Example 1.

(3) FIGS. 3a-3c show scanning electron microscope images of the microstructure of the compounds from Example 2 with the molecular formula Ti.sub.0.2Ta.sub.yV.sub.0.24Nb.sub.0.56-yFeSb.

(4) FIG. 4 shows an enlarged section of a scanning electron microscope image of the compound Ti.sub.0.2Ta.sub.0.03V.sub.0.24Nb.sub.0.53 from Example 2.

(5) FIG. 5 shows a graph of ZT values of the samples from Example 2 at 500° C.

(6) FIGS. 6a-6e show thermoelectric properties of the materials from Example 3 at 500° C.

(7) FIGS. 7a-7e show thermoelectric properties of the Ti.sub.0.2Ta.sub.yV.sub.0.18Nb.sub.0.62-yFeSb compounds from Example 4.

(8) FIGS. 8a-8e show thermoelectric properties of the Ti.sub.0.16Ta.sub.yV.sub.0.25Nb.sub.0.59-yFeSb compounds from Example 5.

(9) FIGS. 9a-9e show thermoelectric properties of the Ti.sub.0.16Ta.sub.yV.sub.0.13Nb.sub.0.71-yFeSb compounds from Example 6.

(10) FIGS. 10a-10e show thermoelectric properties of the Ti.sub.0.12Ta.sub.yV.sub.0.13Nb.sub.0.75-yFeSb compounds from Example 7.

DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS

(11) FIG. 1 shows a schematic view of a thermoelectric module 10. The thermoelectric module 10 has thermoelectric elements 1 and 2 arranged in pairs, also referred to as legs, that are connected to one another by electrically conductive contact layers in the form of electrodes 3 and 4. The thermoelectric elements 1 and 2 each have a first face 13 and a second face 14 opposite the first face 13. In this arrangement the first electrode 3 is arranged partially on the first face 13 of the thermoelectric elements 1 and 2 and the second electrode 4 is arranged partially on the second face 14 of the thermoelectric elements 1 and 2.

(12) By way of example, an n-type thermoelectric material that has a negative Seebeck coefficient is used for the first leg of a pair of elements, e.g. element 1, and a p-type thermoelectric material that has a positive Seebeck coefficient is used for the second leg, e.g. element 2.

(13) A first side 11 of the thermoelectric module 10 is connected to a heat source 5 and an opposite second side 12 of the thermoelectric module 10 is connected to a heat sink 6. During operation of the thermoelectric module 10 the first side 11 thus forms a hot side and the opposite second side 12 forms a cold side of the thermoelectric module 10.

(14) In the embodiment shown, the legs of a pair of elements, i.e. thermoelectric elements 1 and 2, are electrically connected in series. Owing to the opposite or complementary Seebeck effect, the current in the n-type leg, i.e. in thermoelectric element 1, flows from the cold to the hold side and the current in the p-type leg, i.e. in thermoelectric element 2, flows from the hot to the side back to the cold side. The external connections of thermoelectric module 10 can thus both be located on the cold side. The direction of the current flow is illustrated schematically in FIG. 1 by means of arrows.

(15) As the current generated by the single pair of elements and the voltage are typically relatively low, in a thermoelectric module a plurality of thermoelectric elements 1 and 2 are preferably connected to one another, FIG. 1 showing only two pairs with thermoelectric elements 1 and 2 for reasons of clarity. It is possible to provide a current-voltage characteristic suitable for a given application by means of combinations of parallel and series connections, FIG. 1 showing a series connection. In this arrangement an electrical consumer 9 in the form of a resistor is illustrated schematically in FIG. 1.

(16) In the thermoelectric module 10 operated as a thermoelectric generator a temperature gradient is produced by the leg by connecting the first side 11 of the thermoelectric module 10 to the heat source 5 and the opposite second side 12 to the heat sink 6. In the embodiment shown the thermoelectric elements 1 and 2 and the contact layers in the form of electrodes 3 and 4 are electrically insulated by means of insulating layers 7 and 8 in relation to the heat source 5 and the heat sink 6 to avoid short circuits. The insulating layers 7 have good heat conductivity to enable effective heat conduction from the heat source 5 and to heat sink 6 to and/or from the thermoelectric elements 1 and 2.

(17) In applications involving thermoelectric generators two factors in particular are relevant, namely the efficiency of the thermoelectric generator and mechanical and thermal stability at the relevant operating temperatures and in temperature cycling.

(18) The achievable efficiency of a thermoelectric generator is limited by the maximum possible efficiency of a heat to electrical energy conversion process. This is set by the Carnot efficiency ηCarnot=ΔT/Th, where ΔT denotes the temperature difference between the hot and the cold side, i.e. between the first side 11 and the second side 12 in the embodiment shown, and Th denotes the temperature of the hot side, i.e. of the first side 11.

(19) The percentage of Carnot efficiency that can be exploited by a thermoelectric generator is influenced in particular by the thermoelectric efficiency of the thermoelectric materials (TE materials) used for the leg. At a temperature T high-efficiency materials have maximum possible Seebeck coefficient S, good electrical conductivity σ and low heat conductivity κ. This is summarised in the thermoelectric figure of merit ZT.

(20) According to the invention an alloy is provided that consists essentially of (Ti.sub.xTa.sub.yV.sub.zA.sub.cNb.sub.1-x-y-z-c)(Fe.sub.1-dMn.sub.d).sub.a(Sb.sub.1-eSn.sub.e).sub.b, where 0.06≤x≤0.24, 0.01≤y≤0.06, 0.05≤z≤0.4, 0.9≤(a, b)≤1.1, 0≤c≤0.05, 0≤d≤0.05 and 0≤e≤0.1. A denotes to one or more of the elements in the group consisting of Zr, Hf, Sc, Y and La. The alloy can contain up to 5 atom % impurities. This alloy has a positive Seebeck coefficient and a half-Heusler structure and can be used as a p-type in a thermoelectric module as shown in FIG. 1, for example.

Example 1—Comparative Example

(21) Samples of a compound with a nominal composition of Ti.sub.0.2Ta.sub.0.1V.sub.0.42Nb.sub.0.28FeSb were produced using a powder-metallurgical process. To this end the starting elements in the given composition were first melted by means of vacuum induction melting. The ingot was further processed by means of homogenisation at 950° C. for 24 hours in argon as a protective gas and then ground to a fine powder with a median grain size of between 1 μm and 10 μm. The powder was pressed into green bodies in a tool press at a pressure of 2 t/cm.sup.2 and sintered at 1080° C. for 8 hours in a vacuum (10.sup.−2 mbar) to form a dense body. The sintered samples were then homogenised again at 950° C. for 24 hours in argon as an inert gas. The density of the samples was 8.1 g/cm.sup.3.

(22) The microstructure of the samples produced in this manner was examined using scanning electron microscopy (SEM). The result is shown in FIG. 2. It primarily shows three different structural constituents, the composition of which was analysed by means of energy-dispersive x-ray spectroscopy (EDXS). According to the EDXS analysis, the uniformly grey areas 1 in the SEM image have a composition of approx. 33% (Ti+Ta+V+Nb), 33% Fe and the rest Sb (in atom %) and can thus be assigned to half-Heusler phases. Here the portion of Ta in the total Ti+Ta+V+Nb measured using EDXS is a maximum of 6%, i.e. the Ta added was not able to dissolve completely into the half-Heusler phase. The excess Ta can be identified in FIG. 2 in the light Ta-rich precipitations 2, which contain primarily Fe and smaller amounts of V and NB in addition to Ta. The light/dark structured regions 3 are Sb-rich foreign phases with the approximate composition 50% Sb, 25% Fe, rest V+Nb. Here the lighter regions have a higher percentage of Nb, the darker regions a higher percentage of V.

(23) Example 1 therefore shows that for NbFeSb compounds comprising Ti, Ta and a V content of more than 0.4, a Ta content of more than 0.06 leads to a high percentage of foreign phases 2 and 3. The thermoelectric properties measured for these samples with the molecular formula Ti.sub.0.2Ta.sub.0.1V.sub.0.42Nb.sub.0.28FeSb were poor, e.g. a ZT value at 500° C. of only 0.4.

Example 2—According to the Invention

(24) The influence of the Ta content in a composition comprising Ti, V and Nb was investigated in greater detail. Compounds with a reduced Ta content corresponding to the molecular formula Ti.sub.0.2Ta.sub.yV.sub.0.24Nb.sub.0.56-yFeSb where y=0; 0.03; 0.04 and 0.06 were produced. The compounds were produced as in Example 1; the sintering temperatures varied between 1070° C. and 1090° C.

(25) The microstructure of the compounds was examined using SEM. The results for Ta contents of y=0; 0.03 and 0.06 are shown in FIG. 3.

(26) In the Ta-free sample with y=0 the structure consists predominantly of half-Heusler phases 1. EDXS was able to demonstrate the presence of only a small number of Sb—Fe-rich foreign phases 3. The pores in this sample visible in FIG. 3 correspond to a porosity of approx. 3.7%, estimated by comparing the density of sample of 7.77 g/cm.sup.3 and the theoretical density of the compound of 8.07 g/cm.sup.3.

(27) Here the theoretical density is calculated using the lattice constants measured in the article “Are Solid Solution Better in FeNbSb-Based Thermoelectrics?”, Advanced Electronic Materials, p. 1600394 2016 by C. Fu, Y. Liu, X. Zhao, & T. Zhu.

(28) FIG. 3 again shows Ta-rich precipitations 2 as foreign phases in the sample with the highest Ta content y=0.06. The enlargement in FIG. 3 shows only the half-Heusler matrix 1 and Sb—Fe-rich precipitations 3 in the sample with the average Ta content y=0.03. Only at the higher magnification in FIG. 4 can the very finely distributed individual Ta-rich precipitations 2 be detected. A Ta content of the half-Heusler phase of approx. Ta/(Ti+Ta+V+Nb)=0.025 was measured in the sample using EDX.

(29) The thermoelectric properties of the samples produced in Example 2 were measured and the ZT values achieved evaluated. FIG. 5 shows the ZT values at 500° C. It was possible to achieve a clearly higher ZT value of 0.86 in the sample with a Ta content of y=0.03 than in the Ta-free compound, which has a ZT value of 0.82 at 500° C. As the Ta content increases further, so the ZT values falls due to the rise in foreign phases, though at y=0.045 it is still clearly above that of the Ta-free compound. Not until y=0.06 are ZT values lower than those of the Ta-free compound obtained.

(30) With a V content of 0.24, the highest ZT value was thus measured at a Ta content of 0.03.

Example 3—According to the Invention

(31) In Example 3 the influence of Ta content in a composition with a V content of 0.10 was investigated.

(32) Using the same method as in Examples 1 and 2, compounds with the molecular formula Ti.sub.0.2Ta.sub.yV.sub.0.10Nb.sub.0.70-yFeSb were produced with y=0; 0.02 and 0.03. Sintering temperatures of between 1060° C. and 1080° C. were required to achieve dense sinter samples. Following final homogenisation annealing at 950° C. for 48 hours the densities of the samples were determined at 8.03 g/cm.sup.3, 8.07 g/cm.sup.3 and 8.08 g/cm.sup.3. A relative density of 98.0% in relation to the theoretical density of 8.19 g/cm.sup.3 was therefore achieved for the Ta-free sample with y=0.

(33) The thermoelectric properties of the samples were measured. The data obtained for 500° C. dependent on the Ta content y are shown in FIG. 6. Substitution with Ta leads to an improvement in the thermoelectric properties. While a ZT value just under 0.8 is achieved for the Ta-free compound, this figure rises to 0.87 for the sample with a Ta content of y=0.03.

(34) The increase in the ZT value is due both to a decrease in electrical conductivity and to an improvement in electronic properties. An increase in the Seebeck coefficient overcompensates for a decrease in electrical conductivity, causing the power factor to rise as the Ta content increases.

Example 4—According to the Invention

(35) In Example 4 the influence of Ta content in a composition with a V content of 0.18 was investigated.

(36) The compounds with the molecular formula Ti.sub.0.2Ta.sub.yV.sub.0.18Nb.sub.0.62-yFeSb were once again produced using the method set out in the preceding examples with y=0 and 0.02. The density of the samples was 7.76 g/cm.sup.3 and 7.81 g/cm.sup.3. A relative density of 95.6% in relation to the theoretical density of 8.12 g/cm.sup.3 was therefore reached for the Ta-free sample with y=0.

(37) The temperature curve of the thermoelectric properties measured on the samples is shown in FIG. 7. As in the preceding Examples 2 and 3, improved thermoelectric properties were obtained by substitution with Ta. At 500° C. it was possible to increase the ZT value from 0.84 in the Ta-free sample to 0.93 in the sample with y=0.02. Once again, this increase is due to an improvement in the power factor and a drop in electrical conductivity.

Examples 5 to 7—According to the Invention

(38) In Examples 2 to 4 above, the Ti content of the compounds remained constant at x=0.2. In the examples below the positive influence of the Ta is also demonstrated for an extended Ti range where x=0.16 and x=0.12. The compositions are summarised in Table 1. To this end the following compounds were produced using the method set out in the preceding examples.

(39) TABLE-US-00001 TABLE 1 Example 5 Ti.sub.0,16Ta.sub.yV.sub.0,25Nb.sub.0,59−yFeSb y = 0 and 0.02 Example 6 Ti.sub.0,16Ta.sub.yV.sub.0,13Nb.sub.0,71−yFeSb y = 0 and 0.02 Example 7 Ti.sub.0,12Ta.sub.yV.sub.0,13Nb.sub.0,75−yFeSb y = 0 and 0.02

(40) The thermoelectric properties were measured and the results are shown in FIGS. 8 to 10. In all cases an increase in the ZT value was once again recorded due to the addition of Ta. Consistent with the previous examples, this was the result of an increase in the power factor coupled with a drop in electrical conductivity.