Patent Application
20100139730
Moreover, the present invention relates to a thermoelectric material having a stoichiometry corresponding to the stoichiometric formula FeSb2, wherein all or part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu and a vacancy; and wherein all or part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: P, As, Bi, S, Se, Te, B, Al, Ga, In, Tl, C, Si, Ge, Sn, Pb and a vacancy; with the proviso that neither one of the elements Fe and Sb in the formula FeSb2 is fully substituted with a vacancy; characterised in that said thermoelectric material exhibits a power factor (S2σ) of 25 pW/cmK2 or more at a temperature of 150 K or less.
Furthermore the present invention relates to a process for the manufacture of the thermoelectric materials according to the invention, to thermocouples comprising such thermoelectric materials, use of thermocouples for the manufacture of a thermoelectric device, thermoelectric devices comprising such thermocouples.
Thermoelectric materials have been known for decades. We here define thermoelectric in to include both devices based on the Seebeck/Peltier effect and devices based on the Nernst/Ettingshausen effect.
Devices based on the Seebeck/Peltier effect are made by arranging a p-type thermoelectric material and an n-type thermoelectric material in couples, termed thermocouples, it is possible to convert heat into electric power or to create a temperature gradient by applying electric power.
A thermocouple accordingly comprises a p-type thermoelectric material and an n-type thermoelectric material electrically connected so as to form an electric circuit. By applying a temperature gradient to this circuit an electric current will flow in the circuit making such a thermocouple a power source.
Alternatively electric current may be applied to the circuit resulting in one side of the thermocouple being heated and the other side of the thermocouple being cooled. In such a set-up the circuit accordingly functions as a device which is able to create a temperature gradient by applying an electrical current. The physical principles involved in these above phenomena are the Seebeck effect and the Peltier effect respectively.
In order to evaluate the efficiency of a thermoelectric material a dimensionless coefficient is introduced. This coefficient, the figure of merit, ZT is defined as:
ZT=S
2σT/κ,
where S is the Seebeck coefficient, σ is the electrical conductivity, T is the absolute temperature, and κ is the thermal conductivity. The figure of merit, ZT is thus related to the coupling between electrical and thermal effects in a material; a high figure of merit of a thermoelectric material corresponds to an efficient thermoelectric material and vice versa.
The Seebeck coefficient, S of a material is defined as: S(T)=dV/dT, and thus expresses the ability of the material to respond to a temperature gradient by exhibiting a potential difference between points of the material having different temperatures. A material having a high Seebeck coefficient is able to respond to small temperature gradient by exhibiting a relatively large potential difference. This implies that in order to be able to utilise a thermoelectric material as an accurate temperature probe, it is imperative that the material exhibits a high Seebeck coefficient at the temperature of the measurement.
The techniques relating to the manufacture of thermocouples from thermoelectric materials as well as the manufacture of thermoelectric devices from such thermocouples are well documented in the art. See for example CRC Handbook of Thermoelectrics (ed. Rowe, M.), CRC Press, Boca Raton, 1995 and Thermoelectrics—Basic Principles and new Materials Developments, Springer Verlag, Berlin, 2001, which are hereby included as references.
Traditionally thermoelectric materials have been composed of alloys, such as Bi2Te3, PbTe, BiSb and SiGe. These materials have a figure of merit of approximately ZT=1 and operate at temperatures of 200 to 1300 K.
Further improvements appeared with the introduction of alloys of the Te—Ag—Ge—Sb (TAGS) type which exhibit ZT-values of approximately 1.2 in the temperature range of 670-720 K.
New types of materials were recently made with alloys of the Zn4Sb3 type. See for example Caillat et al. in U.S. Pat. No. 6,458,319 B1. These Zn4Sb3 type materials are suitable for applications in the temperature range of 500-750 K.
At lowest temperatures Bi—Sb based alloys are the best materials known to date with a maximum ZT=0.5 at approximately 150 K, which, however, is too low for commercial thermoelectric utilisations (Rowe D. M., MTS Journal, 27 (3), 43-48, 1997).
CsBi4Te6 is another recently discovered (Duck-Young Chung, et al. Science 287, 1024 (2000)) low-temperature thermoelectric with ZT=0.8 at approximately 225 K. Recently also so-called strongly correlated electron systems (e.g. strongly correlated semiconductors, Kondo insulators, heavy fermion systems etc.) are being considered as low temperature thermoelectric materials. However, so far none of these compounds have shown excellent thermoelectric properties (Mahan, G. D. in Solid State Physics, Vol. 51, p. 81-157 (1998) and Thermoelectrics Handbook, Macro to Nano, CRC Press, 2005).
Thin-film/superlattice Bi2Te3/Sb2Te3-based thermoelectric materials have recently shown very good thermoelectric properties with ZT=2.4 at room temperature. The enhancement of the ZT-value compared to the bulk material is mainly due to a reduction of the thermal conductivity caused by the superlattice (Venkatasubramanian et al. Nature 413, 597-602 (2001)).
In contrast to Seebeck/Peltier effect based devices the Nernst/Ettingshausen effect based devices are made of a single leg. The Nernst and Ettingshausen effects are only observed in the presence of a magnetic field.
For power generation a temperature gradient (ΔT) is applied to the material perpendicular to the magnetic field (B). A voltage difference (ΔV) perpendicular to both ΔT and B is observed and can be used for power generation. The Nernst coefficient is, in the case of rectangular shaped material, defined as N=(ΔV/ΔT)·(Δx/Δy) where Δx is the length of the material parallel to ΔT and Δy is the length of the material parallel to ΔV.
For cooling applications the Ettingshausen effect is exploited. An electrical current (I), perpendicular to B, is driven through the Ettingshausen element and a ΔT appears perpendicular to both I and B.
In both cases the efficiency and maximum ΔT depends on the figure of merit ZTN defined as:
ZT
N
=N
2
σT/κ,
κ and σ are perpendicular to each other and to B. N=(ΔV/ΔT)·(Δx/Δy) is measured with ΔV and ΔT parallel to σ and κ. For a device based on the Nernst/Ettingshausen effect the efficiency and maximum ΔT is the same as for a device based on the Seebeck/Peltier effect if ZTN=ZT.
The techniques relating to the manufacture of devices based in the Nernst/Ettingshausen effects are well documented in the art. See for example CRC Handbook of Thermoelectrics (ed. Rowe, M.), CRC Press, Boca Raton, 1995 and Thermoelectrics—Basic Principles and new Materials Developments, Springer Verlag, Berlin (2001) and Recent Trends in Thermoelectric Materials Research II—Semiconductors and Semimetals Vol. 70 (2001), Academic Press (ed. Terry M. Tritt).
In the literature the Nernst (and Ettingshausen) effect are not explored as thoroughly as the Seebeck (and Peltier effect) and no commercial devices based on the Nernst effect are available.
Bi—Sb based alloys are the best investigated materials see e.g. W. M. Yim et al. Solid-State Electronics 15, 1141 (1972) and Recent Trends in Thermoelectric Materials Research II—Semiconductors and Semimetals Vol. 70 (2001), Academic Press (ed. Terry M. Tritt). However, the ZTN values are too low for any commercial applications.
Various technical and scientific disciplines involve the features of setting up environments at very low temperatures, e.g. below 150 K. Such low temperature environments are inter alia necessary i) for devices/instruments based on super-conductor technology i.e. in the field of NMR technology that includes both NMR spectrometers at research institutions and MR scanners at hospitals. ii) photon detector technology where efficiency and sensibility increases upon cooling the detector material. It would be desirably to be able to utilise the special properties of thermoelectric materials in such low temperature environments. Such applications may comprise utilising the Seebeck/Nernst effect of a thermoelectric material for e.g. accurate temperature measurements at temperatures of 150 K or below; or utilising the Peltier/Ettingshausen effect for low temperature cooling (i.e. at a temperature of 150 K or below).
However, until now no efficient thermoelectric materials are known which are suitable for applications at a temperature of 150 K or less.
Hence, a need for thermoelectric materials, which are effective and suitable for low temperature applications (i.e. temperatures of 150 K or below), exists.
Accordingly it is an object according to one aspect of the present invention to provide a low-temperature use of a thermoelectric material.
Another object according to a second aspect of the present invention is to provide a thermoelectric material suitable for low-temperature uses.
Another object according to a third aspect of the present invention is to provide a process for the manufacture of such a thermoelectric material.
Yet another object according to a fourth aspect of the present invention is the provision of thermocouples comprising such thermoelectric materials.
Still another object according to a fifth aspect of the present invention is the use of such thermocouples for the manufacture of thermoelectric devices.
Yet a still further object according to a sixth aspect of the present invention is such thermoelectric devices per se.
The above objects are addressed according to:
In a first aspect by the use of a thermoelectric material for a thermoelectric purpose at a temperature of 150 K or less; said thermoelectric material is a material corresponding to the stoichiometric formula FeSb2, wherein all or part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu and a vacancy; and wherein all or part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: P, As, Bi, S, Se, Te, B, Al, Ga, In, Tl, C, Si, Ge, Sn, Pb and a vacancy; with the proviso that neither one of the elements Fe and Sb in the formula FeSb2 is fully substituted with a vacancy, characterised in that said thermoelectric material exhibits a power factor (S2σ) of 25 pW/cmK2 or more at a temperature of 150 K or less.
In a second aspect by a thermoelectric material having a stoichiometry corresponding to the stoichiometric formula FeSb2, wherein all or part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu and a vacancy; and wherein all or part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: P, As, Bi, S, Se, Te, B, Al, Ga, In, Tl, C, Si, Ge, Sn, Pb and a vacancy; with the proviso that neither one of the elements Fe and Sb in the formula FeSb2 is fully substituted with a vacancy, characterised in that said thermoelectric material exhibits a power factor (S2σ) of 25 pW/cmK2 or more at a temperature of 150 K or less.
In a third aspect by a process for the preparation of such thermoelectric material comprising the steps:
In a fourth aspect by a thermocouple comprising one or more thermoelectric materials according to the invention.
In a fifth aspect by the use of such a thermocouple for the manufacture of a thermoelectric device, and finally:
In sixth aspect by a thermoelectric device comprising one or more such thermocouples.
In a general aspect the present relates to the use of a thermoelectric material for a thermoelectric purpose at a temperature of 150 K or less, wherein said thermoelectric material corresponds to the stoichiometric formula FeSb2, wherein all or part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu and a vacancy; and wherein all or part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: P, As, Bi, S, Se, Te, B, Al, Ga, In, Ti, C, Si, Ge, Sn, Pb and a vacancy; with the proviso that neither one of the elements Fe and Sb in the formula FeSb2 is fully substituted with a vacancy, characterised in that said thermoelectric material exhibits a power factor (S2ρ) of 25 pW/cmK2 or more at a temperature of 150 K or less.
In one embodiment of the use according to the present invention, the thermoelectric material is a material having the formula FeSb2, wherein part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au and Hg and a vacancy; and wherein part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: P, As, Bi, S, Se, Te and a vacancy.
In another embodiment of the use according to the present invention, the thermoelectric material is a material having the formula FeSb2, wherein part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au and Hg and a vacancy; and wherein part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: B, Al, Ga, In, Tl, C, Si, Ge, Sn, Pb and a vacancy.
In yet another embodiment of the use according to the present invention, the thermoelectric material is a material having the formula FeSb2, wherein part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu and a vacancy; and wherein part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: P, As, Bi, S, Se, Te and a vacancy.
In still another embodiment of the use according to the present invention, the thermoelectric material is a material having the formula FeSb2, wherein part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu and a vacancy; and wherein part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: B, Al, Ga, In, Tl, C, Si, Ge, Sn, Pb and a vacancy.
Depending on the number of different elements substituted for Fe and Sb respectively in the formula FeSb2, different types of thermoelectric materials appear. Accordingly; in the low-temperature use according to the present invention, a thermoelectric material is employed which may be binary (i.e. consisting of two different elements), ternary (i.e. consisting of three different elements), quaternary (i.e. consisting of four different elements), quinary (i.e. consisting of five different elements), or even of higher order (i.e. consisting of more than five different elements).
It should be noted that the use according to the present invention does not comprise materials of the above type, wherein either one of the elements Fe and Sb in the formula FeSb2 are fully substituted with a vacancy.
The definition of a vacancy follows the normal definition used in the art and is defined as a missing atom in an otherwise periodically ordered 1-, 2- or 3 dimensional array of atoms. It is preferred that the amount of “vacancy substitution”, if present, is 10 mol % or less, such as 5 mol % or less, e.g. 1 mol % or less, such as 0.5 mol % or less, e.g. 0.1 mol % or less.
It should be noted that in the present description and in the appended claims, the term “a thermoelectric material exhibiting a power factor (S2σ) of 25 pW/cmK2 or more at a temperature of 150 K or less” not necessarily should be construed to mean that said thermoelectric material at all temperatures of 150 K or less exhibits a power factor (S2σ) of 25 uW/cmK2 or more. Rather, the term “a thermoelectric material exhibiting a power factor (S2σ) of 25 uW/cmK2 or more at a temperature of 150 K or less” shall be construed to mean that said thermoelectric material at least at one temperature of 150 K or less exhibits a power factor (S2σ) of 25 pW/cmK2 or more.
As stated above the low-temperature use according to the present invention may employ a binary thermoelectric material.
Hence, in one embodiment of the use according to the present invention, a binary composition is employed. Such binary composition may be any one selected from Table 1 below.
A preferred binary thermoelectric material for use in accordance with the present invention is a material comprising the combinations of elements selected from the group: Fe—Sb, Fe—Bi, Fe—As and Fe—P.
In another embodiment of the use according to the present invention the thermoelectric material employed, is a material having a ternary composition.
The compositions of such ternary compositions can be constructed from the formula FeSb2 by:
In one embodiment according to the use according to the present invention, the material employed is a ternary composition comprising Fe.
Table 2 below lists an array of combinations of constituent elements of a ternary thermoelectric material according to the use according to the present invention. The combinations listed in Table 2 are obtained by partly substituting Fe in FeSb2 with one element selected form the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu.
Alternatively, the use according to the present invention employs a thermoelectric material comprising a ternary composition which corresponds to a composition having the formula FeSb2 in which Sb is fully or partly substituted.
Table 3 below lists an array of combinations of constituent elements of a ternary thermoelectric material according to the use according to the present invention. The combinations listed in Table 3 are obtained either by partly substituting Sb in the formula FeSb2 with one element selected from the group comprising: P, As, Bi, S, Se, Te, B, Al, Ga, In, Ti, C, Si, Ge, Sn, Pb; or by fully substituting Sb in the formula FeSb2 with two different elements selected from the group comprising: P, As, Bi, S, Se, Te, B, Al, Ga, In, Ti, C, Si, Ge, Sn, Pb.
In the list of combinations of constituent elements appearing in Table 2 and 3, a term denoted “X-Y-Z” is to be understood as a composition composed of the constituents X, Y and Z. (for example the term “Fe—P—Bi” is to be interpreted as a composition consisting of Fe, P and Bi.
It should be noted that the list of combinations of constituent elements of the compounds for use according to the present invention as set out in Table 2 and 3 only relates to possible combination of constituent elements, and not to the exact stoichiometries of these constituent elements. The only restrictions on the exact stoichiometries of the constituent elements of the thermoelectric materials according to the use according to the present invention are those imposed as set out in the appended claims.
In another embodiment the thermoelectric material employed according to the use according to the present invention, may be any ternary combination of constituent elements obtained by fully substituting Fe in any of the combinations of elements listed in Table 2 and 3 with an element selected from the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au and Hg.
In a one embodiment according to the use according to the present invention, the ternary thermoelectric material is a material having the formula FeSb2, wherein part of or all Fe optionally being substituted by one or two elements selected from the group comprising: Mn, Co, and Ru; and wherein part of or all Sb optionally being substituted by one or two elements selected from the group comprising: Sb, Bi, As and P.
In a preferred embodiment according to the use according to the present invention, the ternary thermoelectric material is a material having a combination of constituents selected form the group of the combinations comprising: Fe—Ru—Sb, Fe—Mn—Sb, Fe—Co—Sb, Fe—Sn—Se, Fe—Pb—Te, Fe—Sn—Te, Fe—Sb—Te, Fe—Sb—Sn, and Fe—Sb—As.
In yet another embodiment of the use according to the present invention, the thermoelectric material is a material having a quaternary or quinary composition. The composition of such a material may be constructed by substituting part of or all of the Fe of one of the constituent combinations listed in Table 2 and 3 with a combination of the elements as set out in Table 4 below.
Hence, for use according to the present invention, any combination of two elements as set out in table 4 above can be substituted partly or in full with respect to Fe appearing in the combinations of constituents of the materials as listed in table 2 or 3, leading to quaternary thermoelectric compositions comprising no iron, or quinary thermoelectric compositions comprising iron.
By way of illustration, it is seen that the use according to the present invention relates to a material having the combination of constituents: Fe—Mn—Sb—Te: Table 3, first group, combination 6 lists the combination of the constituents: Fe—Sb—Te. By fully substituting Fe in this combination with the combination Fe—Mn (as appearing in Table 4, sixth group, first combination, the result is a combination of the constituents Fe—Mn—Sb—Te.
In one embodiment, the use according to the present invention relates to a quaternary thermoelectric material comprising a combination of four different constituent elements, wherein said combination being selected from the group of combinations comprising: Fe—Sb—C—S, Fe—Sb—C—Se, Fe—Sb—C—Te, Fe—Sb—Si—S, Fe—Sb—Si—Se, Fe—Sb—Si—Te, Fe—Sb—Ge—S, Fe—Sb—Ge—Se, Fe—Sb—Ge—Te, Fe—Sb—Sn—S, Fe—Sb—Sn—Se, Fe—Sb—Sn—Te, Fe—Sb—Pb—S, Fe—Sb—Pb—Se, Fe—Sb—Pb—Te.
In a preferred embodiment of such a quaternary composition, the element in the third position and the element in the fourth position are present in equal molar amounts.
In another embodiment, the use according to the present invention relates to a quaternary thermoelectric material comprising a combination of four different constituent elements, wherein said combination being selected from the group of combinations comprising:
Fe—Sb—B—S, Fe—Sb—B—Se, Fe—Sb—B—Te, Fe—Sb—Al—S, Fe—Sb—Al—Se, Fe—Sb—Al—Te, Fe—SbGa—S, Fe—Sb—Ga—Se, Fe—Sb—Ga—Te, Fe—Sb—In—S, Fe—Sb—In—Se, Fe—Sb—In—Te, Fe—Sb—Tl—S, Fe—Sb—Tl—Se, Fe—Sb—Tl—Te.
In a preferred embodiment of such a quaternary composition, the ratio of the molar amount of the element in the third position to the molar amount of the element in the fourth position is 1:2.
The present invention originates from the Inventors' surprisingly findings that the semiconductor FeSb2 exhibits an extremely high power factor at low temperature (the power factor S2σ is higher than 2000 pW/cmK2 at approximately 10 K).
In fact it is impossible to predict the low-temperature thermoelectric properties in respect of each and every of the combinations of constituents elements making up the thermoelectric material for use according to the present invention.
However, in the section below it is substantiated that based on the newly discovered low-temperature thermoelectric properties of FeSb2, it is highly plausible that substitution of part of or all Fe with an element selected form the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu and a vacancy; and/or substitution of part of or all Sb with one or more elements selected from the group comprising: P, As, Bi, S, Se, Te, B, Al, Ga, In, Tl, C, Si, Ge, Sn, Pb and a vacancy, will lead to thermoelectric materials having thermoelectric properties similar to those of FeSb2
A prerequisite for good thermoelectric properties is a low charge carrier (e.g. electrons or holes) density which is obtained in semiconductors or semimetals. All TX2, TXY and TY2, with the Marcasite, Pyrite or Arsenopyrite structure and with T=Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg and X=P, As, Sb, Bi and Y=S, Se, Te are, with a few exceptions, semiconductors (F. Hulliger and E. Mooser, J. Phys. Chem. Solids 26, 429 (1965); J. B. Goodenough, J. Solid State Chem. 5, 144 (1972)) and therefore potentially good thermoelectric materials.
Besides this new rare-earth (La, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu), actinide or transition metal (Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg) containing semiconductors have received renewed attention within the past 10-15 years.
Some of these compounds show temperature dependent electronic properties and are referred to as strongly correlated semiconductors or Kondo insulators (G. Aeppli and Z. Fisk, Comments Cond. Mat. Phys. 16, 155 (1992); P. S. Riseborough, Adv. Phys. 49, 257 (2000)). They are characterised by a small hybridisation energy gap at the Fermi level from mixing of a broad conduction band with a narrow d- or f-band (G. Aeppli and Z. Fisk, Comments Cond. Mat. Phys. 16, 155 (1992); P. S. Riseborough, Adv. Phys. 49, 257 (2000)).
Concerning thermoelectric applications, much of the interest in strongly correlated semiconductors comes from the fact that strong Coulomb repulsion between the d- or f-electrons leads to a large (and asymmetric) electronic density of states at the band edges of hybridisation gap. This favours a large thermoelectric power factor (G. D. Mahan, in Solid State Physics, Vol 51, 1998), Vol. 51, p. 81).
Theoretically, strongly correlated semiconductors have been treated by dynamical mean field theory using the periodic Anderson model (M. J. Rozenberg, G. Kotliar and H. Kajueter, Phys. Rev. B 54, 8452 (1996)) and calculations of S have recently appeared (J. K. Freericks, D. O. Demchenko, A. V. Joura and V. Zlatic, Phys. Rev. 68, 195120 (2003); T. Saso and K. Urasaki, J. Phys. Chem. Solids 63, 1475 (2002) and C. Grenzebach and G. Czycholl, Physica B 359, 732 (2005)). These results are model dependent but have shown that a large S combined with reasonably low p is possible at low temperatures (J. K. Freericks, D. O. Demchenko, A. V. Joura and V. Zlatic, Phys. Rev. 68, 195120 (2003)). It is the opinion of the present Inventors that FeSb2 is an example of a strongly correlated semiconductor.
It is well known that the electronic properties (e.g. Seebeck coefficient, Nernst coefficient and electrical resistivity) are very similar among isoelectronic compounds and it is very likely that isoelectronic compounds like FeSb2-xYx [Y=P, As, Sb, Bi], FeSb2-2xYxZx [Y=Si, Ge, Sn, Pb and Z=S, Se, Te], FeSb2-3xYxZ2x [Y=Al, Ga, In, Tl and Z=S, Se, Te] exhibit similar properties. This appear to be the case for Kondo insulators/strongly correlated semiconductors too (G. Aeppli and Z. Fisk, Comments Cond. Mat. Phys. 16, 155 (1992)). It is in principle possible to make an isoelectronic substitution of Fe by transition elements (Ru, Os) as well, however, the valence of transition and rare-earth elements is not as well-defined as for the non d- and f-electron elements, potentially making any transition or rare-earth element substitution isoelectronic.
Besides this it well known that thermoelectric properties can be optimized by doping/substitution for two reasons a) the charge carrier concentration can be varied and an optimum thermoelectric power factor can be found (G. A. Slack, CRC Handbook of Thermoelectrics (CRC Press LLC, 1995)), b) The lattice thermal conductivity can be reduced by increased disorder scattering (G. A. Slack, CRC Handbook of Thermoelectrics (CRC Press LLC, 1995)). Substitution of either Fe or Sb with any element has the potential to become a very good thermoelectric material.
Combining the facts that the TX2, TXY and TY2 compounds are i) almost all semiconducting, ii) transition-metal containing and iii) that FeSb2 is a strongly correlated semiconductor, all these compounds are potentially strongly correlated semiconductors with thermoelectric properties similar to FeSb2.
At temperatures below 100 K the umklapp phonon scattering starts to diminish and the thermal conductivity of low charge carrier density materials starts to be determined by impurities, imperfection etc. in the material. By reducing the thermal conductivity the ZT value can be further improved. This is done by introducing imperfections (e.g. vacancies), disorder (e.g. elemental substitution, alloys), spatially extended objects (e.g. synthesizing the material as a nano- and/or micro-sized composite or thin film/super lattice).
For devices based on the Nernst/Ettingshausen effect thin film/super lattice synthesis is a particularly attractive way to increase ZTN=N2σT/κ because κ perpendicular to the super lattice is reduced without impeding σ in the super lattice plane.
A material with large S2ρ−1 does often also exhibit a large N2ρ−1 and vice versa. This makes any material with good properties for a Seebeck/Peltier device a potential candidate for a material to a Nernst/Ettingshausen device.
With respect to all possible combinations of constituent elements making up the thermoelectric material for use according to the present invention, it is not possible to predict the maximum degree in which the Fe and/or Sb of the formula FeSb2 can be substituted and the thermoelectric properties of the compositions can only be verified by experimental measurements.
It is obvious that minute substitutions of Fe and/or Sb in FeSb2 will lead to only minute changes in the thermoelectric properties of a material having the formula FeSb2 and comprising such substitution(s), compared to the properties of the FeSb2-material itself. On the other hand, when Fe and/or Sb in FeSb2 is substituted in large amounts (e.g. 50-75 mol %) it more likely that the thermoelectric properties of the resulting material deviates to a larger extent, as compared to the properties of the FeSb2-material itself.
Accordingly, with respect to each specific combination of constituent elements of the materials for use according to the present invention, the exact composition which gives optimum performance as to thermoelectric properties must be determined experimentally.
In one embodiment of the use according to the present invention it is contemplated that the thermoelectric material is a material having the stoichiometric formula FeSb2, wherein total ratio of substitution of the Fe atoms is 0.1-50 mol %, such as 0.2-40 mol %, e.g. 0.3-30 mol %, e.g. 0.5-25 mol %, such as 1.0-20 mol %, e.g. 2-15 mol %, such as 3-10 mol %, e.g. 5-8 mol % in relation to the Fe content of FeSb2.
In another embodiment according to the present invention it is contemplated that the thermoelectric material is a material having the stoichiometric formula FeSb2, wherein the total ratio of substitution of the Sb atoms is 0.1-50 mol %, such as 0.2-40 mol %, for example 0.3-30 mol %, e.g. 0.5-25 mol %, such as 1.0-20 mol %, e.g. 2-15 mol %, such as 3-10 mol %, e.g. 5-8 mol % in relation to the Sb content of FeSb2.
The above stated possible substitution degrees of Fe and Sb in FeSb2 are applicable in respect of all stoichiometries of the thermoelectric materials for use according to claim 1.
It is preferred that the thermoelectric material of the use according to the present invention is a material having a structure corresponding or similar to that of pyrite, marcasite, or arsenopyrite.
It is preferred that the thermoelectric material of the use according to the present invention is a material having a single crystal structure.
In one embodiment according to the use according to the present invention the thermoelectric material comprises a nano- and/or micro-sized composite of two or more different materials mentioned in any of the claims 1-15 or 23-40. A nano- and/or micro-sized composite material is defined as consisting of two or more types of materials, mentioned in any of the claims 1-15 or 23-40 where the spatial extend of the grains of each material in the composite ranges from ˜1 nm (=1·10−9 m) to ˜10 μm (=10·10−6 m). A person skilled in the art of materials synthesis will know how to prepare such composites.
In another embodiment according to the use according to the present invention the thermoelectric material comprises a thin film/super lattice of two or more layers of any of the materials mentioned in any of the claims 1-15 or 23-40. A thin film/super lattice is defined as alternating layers of two or more types of the materials mentioned in any of the claims 1-14, where the thickness of each layer ranges from ˜1 nm (=1.109 m) to ˜10 μm (=10.10−6 m). A person skilled in the art of materials synthesis will know how to prepare such thin films and super lattices.
In a preferred embodiment according to the use according to the present invention, the thermoelectric properties of the thermoelectric material is utilised at a temperature of 125 K or less, most preferred at a temperature of 100 K or less, such as at a temperature of 50 K or less, for example at a temperature of 25 K or less, such as at a temperature of 15 K or less; or at 10 K or less.
In a preferred embodiment according to the use according to the present invention the thermoelectric material has a power factor (S2σ) of 25 pW/cmK2 or more at a temperature of 125 K or less, most preferred at a temperature of 100 K or less, such as at a temperature of 50 K or less, for example at a temperature of 25 K or less, such as at a temperature of 15 K or less; or at 10 K or less.
It is preferred that the thermoelectric material at least at one of the temperatures indicated above exhibits a power factor (S2σ) of 50 pW/cmK2 or more, such as 100 μW/cmK2 or more, for example 200 μW/cmK2 or more, such as 500 pW/cmK2 or more, e.g. 1000 μW/cmK2 or more, preferably 1500 pW/cmK2 or more, such as 2000 pW/cmK2 or more.
As mentioned above, the term “a thermoelectric material exhibiting a power factor (S2σ) of 25 uW/cmK2 or more at a temperature of 150 K or less” shall not necessarily be construed to mean that said thermoelectric material at all temperatures of 150 K or less exhibits a power factor (S2σ) of 25 pW/cmK2 or more. Rather, the term “a thermoelectric material exhibiting a power factor (S2σ) of 25 pW/cmK2 or more at a temperature of 150 K or less” shall be construed to mean that said thermoelectric material at least at one temperature of 150 K or less exhibits a power factor (S2σ) of 25 pW/cmK2 or more.
In conformity with this interpretation, an analogue interpretation shall be applied in respect of the other upper limits of the stated temperature as set out above.
In one embodiment according to the use according to the present invention the thermoelectric purpose relates to thermoelectric cooling utilising the Peltier effect or the Ettinghausing effect.
In another embodiment according to the use according to the present invention the thermoelectric purpose relates to thermoelectric temperature sensing utilising the Seebeck effect or the Nernst effect.
In a second aspect the present invention relates to a thermoelectric material per se. The thermoelectric material according to the present invention is a material generally described as having a stoichiometry corresponding to the stoichiometric formula FeSb2, wherein all or part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu and a vacancy; and wherein all or part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: P, As, Bi, S, Se, Te, B, Al, Ga, In, Ti, C, Si, Ge, Sn, Pb and a vacancy; with the proviso that neither one of the elements Fe and Sb in the formula FeSb2 is fully substituted with a vacancy, characterised in that said thermoelectric material exhibits a power factor (S2σ) of 25 pW/cmK2 or more at a temperature of 150 K or less.
The inventive thermoelectric material according to the present invention relates to all the compositions and/or compounds which are specifically mentioned in the section above relating to the use according to a fist aspect according to the present invention with the exception of binary compositions; and with the exception of non-alloy ternary compositions of the stoichiometric formula: TXY, wherein T is an element selected from the group comprising: Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg; and wherein X is an element selected from the group comprising: P, As, Sb, Bi; and wherein Y is an element selected from the group comprising: S, Se, Te.
Accordingly, the present invention relates to all the combinations of constituent elements satisfying the general definitions above and which are:
Hence, depending on the number of different elements substituted for Fe and Sb respectively in the formula FeSb2, different types of thermoelectric materials appear. Accordingly, the thermoelectric material according to the present invention may be ternary (i.e. consisting of three different elements), quaternary (i.e. consisting of four different elements), quinary (i.e. consisting of five different elements), or even of higher order (i.e. consisting of more than five different elements).
In one preferred embodiment, the thermoelectric material according to the present invention fulfils the proviso that it is not a binary composition; and fulfils the proviso that the thermoelectric material is not a non-alloy ternary composition of the stoichiometric formula: TXY, wherein T is an element selected from the group comprising: Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg; and wherein X is an element selected from the group comprising: P, As, Sb, Bi; and wherein Y is an element selected from the group comprising: S, Se, Te.
By non-alloy it is meant that the elements are arranged in a periodically ordered manner on atomic scale and where one specific site can be assigned to one specific element.
In one embodiment of the present invention, the thermoelectric material is a material of the general description above, wherein part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au and Hg and a vacancy; and wherein part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: P, As, Bi, S, Se, Te and a vacancy.
In another embodiment of the present invention, the thermoelectric material is a material of the general description above, wherein part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, La, Hf, Ta, W, Re, Os, Ir, Pt, Au and Hg and a vacancy; and wherein part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: B, Al, Ga, In, Tl, C, Si, Ge, Sn, Pb and a vacancy.
In still another embodiment of the present invention, the thermoelectric material is a material of the general description above, wherein part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu and a vacancy; and wherein part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: P, As, Bi, S, Se, Te and a vacancy.
In yet another embodiment of the present invention, the thermoelectric material is a material of the general description above, wherein part of the Fe atoms optionally being substituted by one or more elements selected from the group comprising: Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu and a vacancy; and wherein part of the Sb atoms optionally being substituted by one or more elements selected from the group comprising: B, Al, Ga, In, Tl, C, Si, Ge, Sn, Pb and a vacancy.
In a preferred embodiment of the present invention, the thermoelectric material is a material of the general description above, wherein the thermoelectric material comprises three different elements.
In one embodiment of the present invention, the thermoelectric material is a ternary composition of the general description above, wherein part of or all the Fe optionally being substituted by one or two elements selected from the group comprising: Mn, Co, and Ru; and wherein part of or all the Sb optionally being substituted by one or two elements selected from the group comprising: Sb, Bi, As and P.
In a preferred embodiment of the present invention, the ternary composition of the general description above is composed of a combination of 3 different constituent elements, said combination being selected from the group of combinations comprising: Fe—Ru—Sb, Fe—Mn—Sb, Fe—Co—Sb, Fe—Sn—Se, Fe—Pb—Te, Fe—Sn—Te, Fe—Sb—Te, FeSb—Sn, and Fe—Sb—As.
In one embodiment of the present invention, the thermoelectric material is a quaternary composition of the general description above, wherein the thermoelectric material is composed of a combination of 4 different constituent elements, said combination being selected from the group of combinations comprising: Fe—Sb—C—S, Fe—Sb—C—Se, Fe—Sb—C—Te, Fe—Sb—Si—S, Fe—Sb—Si—Se, Fe—Sb—Si—Te, Fe—Sb—Ge—S, Fe—Sb—Ge—Se, Fe—SbGe—Te, Fe—Sb—Sn—S, Fe—Sb—Sn—Se, Fe—Sb—Sn—Te, Fe—Sb—Pb—S, Fe—Sb—Pb—Se, Fe—Sb—Pb—Te.
In a preferred embodiment according to the present invention such quaternary composition has a stoichiometry wherein the element in the third position and the element in the fourth position are present in equal molar amounts.
In another embodiment of the present invention, the thermoelectric material is a quaternary composition of the general description above, wherein the thermoelectric material is composed of a combination of 4 different constituent elements, said combination being selected from the group of combinations comprising: Fe—Sb—B—S, Fe—Sb—B—Se, Fe—Sb—B—Te, Fe—Sb—Al—S, Fe—Sb—Al—Se, Fe—Sb—Al—Te, Fe—Sb—Ga—S, Fe—Sb—Ga—Se, Fe—Sb—Ga—Te, Fe—Sb—In—S, Fe—Sb—In—Se, Fe—Sb—In—Te, Fe—Sb—Tl—S, Fe—Sb—Tl—Se, FeSb—Tl—Te.
In a preferred embodiment according to the present invention such quaternary composition has a stoichiometry wherein the ratio of the molar amount of the element in the third position to the molar amount of the element in the fourth position is 1:2.
In one embodiment the thermoelectric material according to the present invention is a material having the stoichiometric formula FeSb2, wherein the total ratio of substitution of the Fe atoms is 0.1-50 mol %, such as 0.2-40 mol %, e.g. 0.3-30 mol %, e.g. 0.5-25 mol %, such as 1.0-20 mol %, e.g. 2-15 mol %, such as 3-10 mol %, e.g. 5-8 mol % in relation to the Fe content of FeSb2.
In another embodiment the thermoelectric material according to the present invention is a material having the stoichiometric formula FeSb2, wherein the total ratio of substitution of the Sb atoms is 0.1-50 mol %, such as 0.2-40 mol %, e.g. 0.3-30 mol %, e.g. 0.5-25 mol %, such as 1.0-20 mol %, e.g. 2-15 mol %, such as 3-10 mol %, e.g. 5-8 mol % in relation to the Sb content of FeSb2.
It is preferred that the thermoelectric material according to the present invention is a material having a structure corresponding or similar to that of pyrite, marcasite, or arsenopyrite.
It is preferred that the thermoelectric material according to the present invention has a single crystal structure.
In one embodiment according to the present invention the thermoelectric material comprises a nano- and/or micro-sized composite of two or more different materials mentioned in any of the claims 1-15 or 23-40. A nano- and/or micro-sized composite material is defined as consisting of two or more types of materials, mentioned in any of the claims 1-15 or 23-40 where the spatial extend of the grains of each material in the composite ranges from ˜1 nm (=1·10−9 m) to ˜10 μm (=10·10−6 m).
In another embodiment according to the present invention the thermoelectric material comprises a thin film/super lattice of two or more layers of any of the materials mentioned in any of the claims 1-15 or 23-40. A thin film/super lattice is defined as alternating layers of two or more types of the materials mentioned in any of the claims 1-15 or 23-40, where the thickness of each layer ranges from ˜1 nm (=1·10−9 m) to ˜10 μm (=10·10−6 m).
In a preferred embodiment according to the present invention the thermoelectric material has a power factor (S2σ) of 25 pW/cmK2 or more at a temperature of 125 K or less, most preferred at a temperature of 100 K or less, such as at a temperature of 50 K or less, for example at a temperature of 25 K or less, such as at a temperature of 15 K or less; or at 10 K or less.
It is preferred that the thermoelectric material at least at one of the temperatures indicated above exhibits a power factor (S2σ) of 50 pW/cmK2 or more, such as 100 pW/cmK2 or more, for example 200 pW/cmK2 or more, such as 500 pW/cmK2 or more, e.g. 1000 pW/cmK2 or more, preferably 1500 pW/cmK2 or more, such as 2000 uW/cmK2 or more.
As mentioned above, the term “a thermoelectric material exhibiting a power factor (S2σ of 25 pW/cmK2 or more at a temperature of 150 K or less” shall not necessarily be construed to mean that said thermoelectric material at all temperatures of 150 K or less exhibits a power factor (S2σ) of 25 pW/cmK2 or more. Rather, the term “a thermoelectric material exhibiting a power factor (S2σ) of 25 pW/cmK2 or more at a temperature of 150 K or less” shall be construed to mean that said thermoelectric material at least at one temperature of 150 K or less exhibits a power factor (S2σ) of 25 pW/cmK2 or more.
In conformity with this interpretation, an analogue interpretation shall be applied in respect of the other upper limits of the stated temperature as set out above.
In a third aspect the present invention relates to a process for the manufacture of a thermoelectric material according to the invention.
Single crystalline FeSb2 samples are prepared by a flux method. The binary Fe—Sb phase diagram (T. Massalski, Binary Alloys Phase Diagrams, 2nd ed. (ASM International, 1996)) shows that the liquidus curve decreases with increasing Sb content in the Sb-rich region. There exists two peritectic points that connected to the composition FeSb2 by a peritectoid at 88.8%, 99.2% Sb-molar % and at 738° C. and 628° C., respectively. This suggests that FeSb2 can be crystallized by mixing Fe and Sb in the molar ratio 11.2:88.8˜1:7.9 and by cooling slowly from 738° C. to 628° C. However, due to potential uncertainties in the phase diagram, the Sb content should be larger 1:7.9 (e.g. 1:11.5) in order to prevent other phases than FeSb2 to crystallize. Besides this the ranges with slow cooling rates should be started at higher temperatures e.g. 775° C. and ensures homogenous temperature conditions and cooling rates when the temperatune decrease below the temperature at which FeSb2 starts to crystallize. The minimum temperature of the slow cooling rates should be slightly larger than 628° C. (e.g. 640° C.) to prevent sudden crystallization of FeSb2 when the peritectic point at 628° C. is reached.
To prevent oxidation and unwanted reactions with air the elements should be placed in a sealed ampoule with an inert atmosphere (e.g. argon) or vacuum before heating. Remaining Sb-flux after the crystallization can be removed by heating the reactants to above the melting point of Sb (631° C.) but below the melting point of FeSb2 (738° C.) (T. Massalski, Binary Alloys Phase Diagrams, 2nd ed. (ASM International, 1996)) and decant the liquid or by other methods separate the solid from the liquid.
To ensure homogenous thermal conditions the crucible is isolated thoroughly with e.g. mineral wool.
An alloy where either Fe or Sb is substituted with one or more elements is made in a similar way. It can be expected that the pseudo binary phase diagram is similar to that of Fe—Sb (liquidus curves and peritectic points are at the same temperatures and cornpositions) if the substitutions are small. Otherwise other phase diagrams that reflect the compositions of the alloy can be used for obtaining good temperature- and compositions-parameters can be used.
The actual composition of the alloy may deviate from the nominal composition and the actual composition can be tuned by either reducing or increasing the nominal composition.
Single crystals can be grown by a gas-transport reaction method as described in A. K. L. Fan et al. J. Solid State Chem. 5, 136 (1972).
Single crystals can be pulled by a modified Czochralski method. Due to the incongruently melting FeSb2 phase an off-stoichiometry melt with a surplus of antimony is used. Pulling crystals from a non-stoichiometric melt is described in e.g. M. Burianek et al. Journal of Crystal Growth, 166 (1-4), 361 (1996).
A similar technique is the Bridgeman technique. A temperature gradient is applied along a horizontal reaction vessel, the position of which is kept fixed. The temperature gradient is maintained but the average temperature is decreased. Precipitation of FeSb2 will occur from Sb rich melts. This method was used to produce single crystalline RuSb2 as described in T. Caillat et al. J. Phase Equilibria, 14(5), 576 (1993).
A variant of the Bridgeman technique is the Stockbarger method where the temperature and gradient are fixed. The reaction vessel is slowly pulled towards the cold end. See e.g. S. Elliott. The Physics and Chemistry of Solids, Chichester (1998).
Polycrystalline samples can be synthesized as described in A. Bentien et al. Phys. Rev. B 74, 205105 (2006).
Single crystalline and homogeneous polycrystalline samples are prepared using the zone refinement technique described in S. Elliott. The Physics and Chemistry of Solids, Chichester (1998).
Composite materials can be made from two or more materials that e.g. can be synthesized as described above under the heading Single Crystals. The powdering can be done by ball-milling the materials in an inert atmosphere until the wanted average grain-size is obtained. The compaction of the composite is performed with both pressure and/or heat e.g. spark plasma sintering or hot pressing. A person skilled in the art of materials synthesis will know how to prepare such composites.
Thin films and super lattices are prepared by either sputtering, e-beam evaporation, pulsed laser deposition, thermal evaporation, electron beam evaporation, molecular beam epitaxy or similar methods (see e.g. Handbook of Thin-Film Deposition Processes and Techniques—Principles, Methods, Equipment and Applications (2nd edition), Ed. By K. Seshan, William Andrew Publishing/Noyes (2002)). A person skilled in the art of materials synthesis will know how to prepare such thin films and super lattices.
In a fourth aspect, the present invention relates to a thermocouple. According to this aspect, the obtained thermoelectric material is used as one out of two legs in a thermocouple. By cutting this material in suitable sizes, or in other ways adjusting the dimension to the application, and arranging and connecting this together with a dissimilar thermoelectric material, a thermocouple is obtained in a way known per se. See for example “Frank Benhard; Technische Temperaturmessung; Springer Berlin, 2003; ISBN 3540626727”.
In a fifth aspect according to the present invention the obtained thermocouple is used for the manufacture of a thermoelectric device.
Such uses are well-known for a person skilled in the art of thermoelectrics.
A sixth aspect according to the present invention relates to a thermoelectric device per se. The device is obtained by combining one or more thermocouples. A person skilled in the art of thermoelectrics will know how to obtain such a thermoelectric device once he has obtained the thermocouple according to the present invention.
Pure FeSb2 samples are prepared by a flux method. 0.95882 g bulk Fe (Alfa Aesar Puratronic® 99.995% metals basis) and 24.04118 g bulk Sb (Alfa Aesar Puratronic® 99.9999% metals basis) are mixed in an alumina crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 1050° C. and left there for 2 hours, followed by cooling to 775° C. over 14 hours and finally cooling to 640° C. over 15 days. The Sb-flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule. To remove any remaining Sb-flux on the FeSb2 samples they are cleaned in an ultra sonic bath of Aqua Regia for 3-8 minutes. Relatively large single crystals are obtained. The resulting samples can be seen in
The samples are prepared by a flux method using a melt with nominal stoichiometry Fe8Sb16.1Bi75.9. 0.61137 g bulk Fe (Alfa Aesar Puratronic® 99.995% metals basis) and 2.68264 g bulk Sb (Alfa Aesar Puratronic® 99.9999% metals basis) and 21.70598 g bulk Bi (Strem chemicals 99.999+% metals basis) are mixed in an alumina crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 1050° C. and left there for 2 hours, followed by cooling to 775° C. over 14 hours and finally cooling to 640° C. over 15 days. The flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule.
FeSb2-2xPbxSex, x=0.1, 0.5 is prepared by a flux synthesis. Bulk Fe (Alfa Aesar 99.98% metals basis), bulk Sb (ESPI metals 99.9999% metals basis), bulk Pb (ESPI metals 99.99 9% metals basis) and bulk Se (ESPI metals 99.999% metals basis) are mixed in an alumina crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 1050° C. and left there for 2 hours, followed by cooling to 800° C. over 14 hours and finally cooling to 600° C. over with a cooling rate of 1° Ch−1. The flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule. To remove any remaining flux on the samples they are cleaned in an ultra sonic bath of Aqua Regia for 3-8 minutes. Relatively large single crystals are obtained.
Compositions having the stoichiometries as set out in the table below were made.
The samples are prepared by a flux method. Bulk Fe (Alfa Aesar Puratronic® 99.995% metals basis), bulk Ru (Chempur 99.95% metals basis) and bulk Sb (Alfa Aesar Puratronic® 99.9999% metals basis) are mixed in an alumina crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 1050° C. and left there for 2 hours, followed by cooling to 775° C. over 14 hours and finally cooling to 640° C. over 15 days. The flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule. To remove any remaining flux on the samples they are cleaned in an ultra sonic bath of Aqua Regia for 3-8 minutes. Relatively large single crystals are obtained.
Compositions having the stoichiometries as set out in the table below were made.
The samples are prepared by a flux method. Bulk Fe (Alfa Aesar Puratronic® 99.995% metals basis) and bulk Sb (Alfa Aesar Puratronic® 99.9999% metals basis) and bulk Mn (Alfa Aesar 99.99% metals basis) are mixed in an alumina crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 1050° C. and left there for 2 hours, followed by cooling to 775° C. over 14 hours and finally cooling to 640° C. over 15 days. The Sb-flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule. To remove any remaining flux on the samples they are cleaned in an ultra sonic bath of Aqua Regia for 3-8 minutes. Relatively large single crystals are obtained.
Compositions having the stoichiometries as set out in the table below were made.
Pure FeSb2 samples are prepared by a flux method. Bulk Fe (Alfa Aesar Puratronic® 99.995% metals basis) and bulk Sb (Alfa Aesar Puratronic® 99.9999% metals basis) and bulk Co (Alfa Aesar Puratronic® 99.995% metals basis) are mixed in an alumina crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 1050° C. and left there for 2 hours, followed by cooling to 775° C. over 14 hours and finally cooling to 640° C. over 15 days. The Sb-flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule. To remove any remaining flux on the samples they are cleaned in an ultra sonic bath of Aqua Regia for 3-8 minutes. Relatively large single crystals are obtained.
Compositions having the stoichiometries as set out in the table below were made.
FeSb2-2xSnxSex, x=0.1, 0.5 is prepared by a flux synthesis. Bulk Fe (Alfa Aesar 99.98% metals basis), bulk Sb (ESPI metals 99.9999% metals basis), bulk Sn (ESPI metals 99.999% metals basis) and bulk Se (ESPI metals 99.999% metals basis) are mixed in an alumina crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 1050° C. and left there for 2 hours, followed by cooling to 800° C. over 14 hours and finally cooling to 600° C. over with a cooling rate of 1° Ch−1. The flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule. To remove any remaining flux on the samples they are cleaned in an ultra sonic bath of Aqua Regia for 3-8 minutes. Relatively large single crystals are obtained.
Compositions having the stoichiometries as set out in the table below were made.
FeSb2-2xPbxTex, x=0.1, 0.5 is prepared by a flux synthesis. Bulk Fe (Alfa Aesar 99.98% metals basis), bulk Sb (ESPI metals 99.9999% metals basis), bulk Pb (ESPI metals 99.999% metals basis) and bulk Te (ESPI metals 99.999% metals basis) are mixed in an alumina crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 1050° C. and left there for 2 hours, followed by cooling to 800° C. over 14 hours and finally cooling to 600° C. over with a cooling rate of 1° Ch−1. The flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule. To remove any remaining flux on the samples they are cleaned in an ultra sonic bath of Aqua Regia for 3-8 minutes. Relatively large single crystals are obtained.
Compositions having the stoichiometries as set out in the table below were made.
FeSb2-2xSnxTex, x=0.1, 0.5 is prepared by a flux synthesis. Bulk Fe (Alfa Aesar 99.98% metals basis), bulk Sb (ESPI metals 99.9999% metals basis), bulk Sn (ESPI metals 99.999% metals basis) and bulk Te (ESPI metals 99.999% metals basis) are mixed in an alumina crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 1050° C. and left there for 2 hours, followed by cooling to 800° C. over 14 hours and finally cooling to 600° C. over with a cooling rate of 1° Ch−1. The flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule. To remove any remaining flux on the samples they are cleaned in an ultra sonic bath of Aqua Regia for 3-8 minutes. Relatively large single crystals are obtained.
Compositions having the stoichiometries as set out in the table below were made.
The samples are prepared by a flux method. 0.95838 g bulk Fe (Alfa Aesar Puratronic® 99.995% metals basis) and 23.78979 g bulk Sb (Alfa Aesar Puratronic® 99.9999% metals basis) and 0.25183 g bulk Te (Alfa Aesar 99.999% metals basis) are mixed in an alumina crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 1050° C. and left there for 2 hours, followed by cooling to 775° C. over 14 hours and finally cooling to 640° C. over 15 days. The flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule.
The samples are prepared by a flux method. 0.95905 g bulk Fe (Alfa Aesar Puratronic® 99.995% metals basis) and 23.80650 g bulk Sb (Alfa Aesar Puratronic® 99.9999% metals basis) and 0.23445 g bulk Sn (Chempur 99.9999% metals basis) are mixed in an alumina crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 1050° C. and left there for 2 hours, followed by cooling to 775° C. over 14 hours and finally cooling to 640° C. over 15 days. The flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule.
The samples are prepared by a flux method. Bulk Fe (Alfa Aesar Puratronic® 99.995% metals basis), bulk Sb (Alfa Aesar Puratronic® 99.9999% metals basis) and bulk As (Chempur 99.9999% metals basis) are mixed in a glassy carbon crucible which is sealed inside an evacuated quartz ampoule. The ampoule is isolated with mineral wool and heated fast (over approximately 6 hours) to 850° C. and left there for 12 hours, followed by cooling to 775° C. over 14 hours and finally cooling to 600° C. over 15 days. The flux is removed by centrifuging at 690° C. on top of small broken quartz pieces inside an evacuated quartz ampoule.
Compositions having the stoichiometries as set out in the table below were made.
The crystal are oriented by a combination of Laue photos and measuring the reflection X-ray intensity on a powder diffractometer using θ, 2θ geometry (see
Due to the large thermal conductivity and relatively small sample dimensions, the thermal resistivity of the sample can be small compared with the thermal resistivity between the sample and sample holder. This leads to a sample heating (ΔT2) that can be considerable compared to the sample temperature difference (ΔTI) across the sample. To diminish ΔT2, a special sample holder has been constructed in which the sample can be very tightly screwed/mounted to the sample holder. This is seen in
Cernox thermometers and the heater are mounted directly with N-grease onto the thermometer and heater holders, respectively. 50 μm manganin wires are used for connecting the heater and thermometers, and they are approximately 20 cm long. This reduces the thermal conductance from heater and the thermometers to the sample holder by a factor of at least 10, compared to the original setup in the TTO (Thermal Transport Option) sample holder to the PPMS from Quantum Design.
Because of the large thermal mass of the sample-holder pieces (thermometer, holders etc.) compared to the sample, the approximations regarding the relaxation times used in the algorithm in the quasi-static measurement of S and κ (O. Maldonado, Cryogenics 32, 908-912 (1992)) in the PPMS no longer apply. Instead a steady state technique up to 45 K is used, and the data are analysed separately by software developed by the inventors.
The Nernst effect is measured similar to the measurement of S. The voltage contacts are mounted perpendicular to the thermal contacts and the magnetic field. To remove any S-component of the Nernst signal the measurement is done in two magnetic fields with opposite sign.
Pure FeSb2 has very promising thermoelectric properties.
One striking feature is the extremely large negative peak in S(T) observed at 10 K-20 K, the magnitude of S being 10-100 times larger than in classical semiconductors. The most important part of
However, for pure FeSb2 κ(T) is as large as 500 WK−1 m−1 and degrades the maximum ZT-value to 0.005 at 15 K. Nonetheless, in thin films, alloys and composites etc. κ˜-1 WK−1 m−1 and if the temperature range from 10 K-30 K thus emphasizing the potential of FeSb2 as thermoelectric material at the these low temperatures.
The thermoelectric properties have been measured on several samples along all directions. Along the c-direction data is shown for two different samples from the same batch, and represents the two samples with the largest differences. The Inventors believe that this difference is due to tiny differences in samples (e.g. tiny differences, of the order <1%, in composition, quality etc.). The data show that the maximum of |S(T)|(Smax), among all samples, increases with ρ(T) at anomalous bump in the ternperature range from 10 K to 30 K. This indicates anisotropy plays a minor role on the size of Smax and it is determined by the same parameters as the differences between the two samples measured along the c-direction.
However, at 300 K anisotropy in the Seebeck coefficient and lattice thermal conductivity (κL) is always observed where Sa˜Sb˜40 μVK−1, Sc˜0 μVK−1 and, κL,a˜κL,b˜4.5 Wm−1K−1, κL,c˜6.5 Wm−1K−1 where the subscript refers to the direction.
In the following the thermoelectric properties of some of these samples are presented.
For the x=0.003 samples S(T) is similar to the pure samples although |Smax| is smaller. As x increases it is seen that S(T) becomes more monotonic and the positive peak observed around 35 K is suppressed. At room temperature S(T) is anisotropic and similar for all samples. Since ρ(T) is decreased, S(T) becomes more monotonic and Smax is shifted to higher temperatures the PF has a tendency to become broader as x increases. For the x=0.01 sample the PF>100 μWcm−1K−2 below 50 K. As expected, κL(T) is reduced upon doping. The large variance of κL(T) for the x=0.003 samples could be erroneous due to geometry factor determinations or bad sample quality. However, for the x=0.01 sample a small bump in κL(T) appears around 15 K to 40 K, and for the x=0.1 sample a significant reduction is seen whole temperature range below 100 K.
Since both S(T) is reduced and ρ(T) is increased the PF is decreased compared to the pure samples. κL(T) has a tendency to decrease, but the Co amount is too small to make significant reduction.
Substitution with both Mn and Co appear to decrease |Smax|. Above approximately 30 K to 50 K they appear to have opposite effect. Mn substitution leads to a more negative S(T) while Co substitution leads to a more positive S(T). At room temperature S(T) is unchanged compared to the pure samples. For both Mn and Co substitution κL(T) is decreased and thereby favouring improved thermoelectric properties. However, only substitution with Mn appears reasonable since it can shift the maximum of the PF to higher temperature.
For the FeSb2-xTex the effect of substitution is different. In particular ρ(T) becomes metal-like with decreasing ρ(T) as the temperature decreases below approximately 150 K. A similar effect is seen in S(T) where the magnitude of the negative peak is decreased compared to the pure FeSb2 samples. The PF and ZT-value is broadened and the maximum is shifted to higher temperatures are of relatively large values. κL(T) is significantly reduced at lower temperatures and of the same magnitude as observed for the FeSb2-xSnx samples.
This type of substitution is also referred to as compensated doping or substitution, because the electronic configuration does not change in FeSb2-2xVxZx, where V=C, Si, Ge, Sn, Pb and Z=P, As, Sb, Bi, since Y has one electron less and Z has one electron more than Sb.
Compensated doping appears to be attractive in order to improve the thermoelectric properties of FeSb2. The above results (Fe1-xMnxSb2, Fe1-xCoxSb2, Fe1-xRuxSb2) have shown that normal substitution leads to a decrease of the PF because the magnitude of S(T) decreases without improving ρ(T). The compensated doping can reduce κL(T) significantly as seen from
| Number | Date | Country | Kind |
|---|---|---|---|
| 06125354.8 | Dec 2006 | EP | regional |
| 06126488.3 | Dec 2006 | EP | regional |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/DK2007/000530 | 12/4/2007 | WO | 00 | 12/7/2009 |
