HK1073720B - Method for producing a device for direct thermoelectric energy conversion - Google Patents

Description

Method for producing a device for direct conversion of thermoelectric energy

PRIORITY INFORMATION

This application claims priority from U.S. provisional patent application No. 60/317,692, filed on 6/9/2001 and from U.S. utility application No. unknown on 5/9/2002, both of which are incorporated herein by reference in their entirety.

Background

1. Field of the invention

The present invention is directed to a process of manufacturing a device for direct conversion of thermoelectric energy, thereby significantly increasing the conversion efficiency from thermal energy to electric energy or from electric energy to thermal energy, and a synthetic substance used in the production manufacturing of the device for direct conversion of thermoelectric energy.

2. Description of the prior art

The use of powder metallurgy as a means of producing synthetic materials as described above requires careful attention to the recent trends of the National Institute of Standards and Technology (NIST). In the publication "Thermoelectric News" (Thermoelectric News) of the international society for thermoelectricity, 6 months, 1996, j.terry Lynch published a new technical development program or invention entitled "synthesis of ultrafine Powder Polycrystalline bismuth-selenium-tellurium, bismuth-antimony-tellurium and bismuth-antimony-selenium-tellurium alloys for Thermoelectric Applications" (synthetic Fine Powder-Powder Polycrystalline Bi-Se-Te, Bi-Sb-Te, and Bi-Sb-Se-Te alloys for Thermoelectric Applications). The alloy starting precursors with the general synthetic species bismuth-selenium-tellurium, bismuth-antimony-tellurium and bismuth-antimony-selenium-tellurium were synthesized by aqueous coprecipitation (aquouus co-precipitation) and organometallic complex methods. The hydrogen reduction of the original precursor causes the alloy to be in the form of a polycrystalline body of very fine powder. The process is simpler than traditional melt processing methods and yields 88% -92% in laboratory-wide testing. The new process reduces equipment, material and labor costs by directly producing fine powders, thereby eliminating the necessary rolling and screening steps after the melt process. The original precursor synthesis was performed by ordinary chemical methods in aqueous solution at 100 degrees celsius. Alloy synthesis at 300-400 ℃ below the melt processing temperature can produce 88% more product than theoretical. It is possible to increase the continuous throughput using conventional chemical continuous reactor technology. This new development or invention improves the efficiency and cost effectiveness of producing solid state thermoelectric cooling and freezing devices. It is therefore highly desirable to further develop this new technology with the goal of adapting or extending it to the composition, which constitutes a basic embodiment of the present invention. This will significantly eliminate one of the fundamental drawbacks of the powder metallurgy technology, in particular the undesired impurities or dopings of the composition, i.e. the iron Fe impurities coming from the steel balls and steel containers of the planetary ball mill. This is because a planetary ball mill is not used since rolling and grinding of the composition or alloy are no longer required. Moreover, this new technology developed by NIST, if successfully adapted to composites, as defined and claimed herein, would help overcome or eliminate the previously described major drawbacks associated with melt metallurgy technology. These disadvantages are the need to stir or vibrate the constituent components during melting in order to ensure the production of a single phase alloy, and also the need to maintain the molten components in argon or helium gas while applying a relative pressure of 2 to 30 physical atmospheres to them in order to suppress the loss of magnesium and thus ensure that a stoichiometric alloy is obtained.

Thermoelectric or thermoelectric, as it is known today, occurs due to the first thermoelectric effect of johnston massbeck (Thomas Johann Seebeck) in 1821, which since then has also been referred to as the Seebeck (Seebeck) effect or the Seebeck (Seebeck) coefficient. In 1833, the Peltier (Peltier) found a second thermoelectric effect, which was also called the Peltier effect since then. Seebeck (Seebeck) found that when the compass needle is placed near a closed loop made of two different metals and the temperature of one of the two connection points is higher than the other, the compass needle deflects. This reveals the fact that a voltage difference exists or is generated whenever there is a temperature difference between the two junctions. This also depends on the nature of the metal used. Peltier finds that whenever current passes through a junction of dissimilar metals, a temperature change occurs at the junction, with an accompanying heat absorption or heat dissipation. In 1838 Lenz explained that the heat absorption or dissipation at the connection point depends on the direction of the current. Furthermore, William Thomson jazz (later known as Kelvin (Kelvin)) together with german physicist rutuff julies mannukellaaus (Rudolf Julius emmanuuelclausius) are known in the middle of the 19 th century because of their formulas of the first and second laws of thermodynamics and the concept of entropy they found and established, which has likewise made thermoelectric senseAn important contribution. He found the third thermoelectric effect, the Thomson effect (Thomson effect), which involves heating and cooling of a uniform conductor subject to temperature gradients. He also established four important formulas that relate all three effects, namely Seebeck (Seebeck), Peltier (Peltier) and Thomson coefficients. These are known in the art as Kelvin relations (Kelvin relations) and are described in any standard textbook on thermoelectrics or direct energy conversion. Furthermore, thermoelectrics was mainly developed in 1885, when Seebeck (Seebeck) effect was considered and suggested to be used by Rayleigh (Rayleigh) gazz for power generation. In 1911, Altenkirch established a milestone of our general concept of thermoelectricity, specifically how it is best used for direct thermoelectric or electrothermal conversion. He created a suitable thermoelectric theory for power generation and cooling. He theorizes that for optimum performance, the Seebeck (Seebeck) coefficient or the thermoelectric power currently being called must be as high as possible, that the electrical conductivity must likewise be as high as possible, and that the thermal conductivity should be as low as possible. Thus, we obtain the power factor: PF is S²σ＝S²Where S is the Seebeck (Seebeck) coefficient or thermoelectric power, σ is the electrical conductivity, ρ is the electrical resistivity, the value of the power factor must be increased or maximized as much as possible, and k is the thermal conductivity, the value of which must be decreased or minimized as much as possible. Thus, Altenkirch (Altenkirch) establishes the following formula:

where Z is referred to as the thermoelectric figure of merit, having K^-1The dimension of (2). This formula can be made dimensionless by multiplying by some absolute temperature T, which can be the temperature at the thermal connection point of the thermoelectric device. This yields another quantity: i.e., dimensionless thermoelectric figure of merit, ZT, where Z can also be used for evaluation of the performance and energy conversion efficiency of any thermoelectric material or device.

Contemporary thermoelectrics is now being developed as engineers and scientists focus more and more on semiconductors. Semiconductors refer to those substances or materials whose electrical conductivity lies between a metal and an insulator. A comparison of so-called minerals and metals, which were known at the time and are referred to as semiconductors, shows that metals have ductility, relatively constant properties, i.e. temperature-independent properties, and chemical stability, while minerals or semiconductors, when moderately or highly doped, have a relatively high Seebeck (Seebeck) coefficient S and thus a moderate thermoelectric quality factor Z. The drawbacks of the metals found are the low Seebeck (Seebeck) coefficient S, the low thermoelectric figure of merit Z, and the limitations imposed by the law of vedman-Franz (Wiedemann-Franz) on the ratio between thermal and electrical conductivity, mainly electronic. The law states that when the ratio is plotted against absolute temperature T, it is a straight line or linear relationship to the metal, with a slope of the lorentz (Lorenz) number L. Thus, the Wiedemann-Franz law for metals can be expressed as follows:

wherein k is_elIs the electron thermal conductivity.

For metals, k-k is negligible or very small lattice thermal conductivity_elIs the total thermal conductivity.

The mineral or semiconductor has disadvantages of brittleness, temperature-dependent properties and lack of chemical stability. In fact, the dependence of semiconductors on temperature makes the theoretical analysis of their performance, figure of merit, energy conversion efficiency, coefficient of performance, power generated or consumed, heat absorbed or dissipated at cold junctions, heat dissipated, absorbed or transferred at hot junctions much more complex than for metallic materials when used as thermoelectric materials or elements. Thus, metals are more suitable for thermocouple wires, while semiconductors are more suitable for making small modules, constituting the basic thermoelectric elements, pins or branches of a thermoelectric device. It should be emphasized that many of the technical difficulties encountered in thermoelectrics stem from the fact that thermoelectric devices comprise modules or thermoelectric elements made of semiconductors, which generally do not have the flexibility, elasticity and chemical stability of metals.

In the 1930 s, thermoelectrics has further advanced when synthetic or compound semiconductors were first investigated. In 1947, Maria siellas (Maria talk) developed and constituted a thermoelectric power generator with an energy conversion efficiency of 5%. In addition, in 1949, semiconductor thermoelectric theory was established by royalty (a.f. ioffe). He has written two original books "semiconductor Physics" and "semiconductor Thermoelectric elements and Thermoelectric Cooling" (semiconductor Thermoelectric elements and Thermoelectric Cooling). A semiconductor is actually a substance or material with electrical conductivity between a metal and an insulator. By increasing the free charge carriers, the conductivity of the semiconductor can generally be increased. This can be accomplished by introducing an appropriate amount or ratio of atoms, compounds or materials of a suitable foreign element, called a dopant or impurity, into the semiconductor. The process of incorporating atoms or impurities of a foreign element into a semiconductor is called doping. Thus, the doping is carried out so that the free charge carrier concentration in the room temperature semiconductor is between 1 × 10 per cubic centimeter¹⁸To 5X 10²⁰Between the individual carriers. Having a thickness of 10 per cubic centimeter¹⁸Doped semiconductors with free charge carrier concentrations of the order of magnitude, referred to as "lightly doped" semiconductors, have a concentration of 10 per cubic centimeter¹⁹Doping of free charge carrier concentration of order of magnitudeThe semiconductor is referred to as a "moderately doped" semiconductor and has a thickness of 10 per cubic centimeter²⁰Doped semiconductors with free charge carrier concentrations of an order of magnitude are referred to as "highly doped" semiconductors. It should be noted that when the free charge carrier concentration is 10 per cubic centimeter¹⁹Order of magnitude, power factor or S²σ max. This is an approximate rule of thumb that generally applies to all semiconductors, but with subtle differences for different semiconductors.

Most semiconductors are either non-single element or synthetic, i.e., compounds, and typically have low to moderate band gaps. Most early semiconductors contained elements of high atomic number and atomic mass. This is intended and intended in order to select elements with as low a thermal conductivity as possible in order to optimize the thermoelectric figure of merit. Therefore, the rule applied here is that the higher the atomic number and atomic mass of an element, the lower its thermal conductivity. This leads to a "heavy element selection criterion" without any doubt. Thus, elements of high atomic mass, i.e. heavy elements, should be prioritized and selected over other light elements, since this is the conclusion above that such elements will have the lowest possible thermal conductivity. This will thus result in the highest possible thermoelectric figure of merit. This type of reasoning is well known and proved to be highly effective in the thirty, forty and fifty years and pioneered at an agreement fee (a.f. ioffe) without any doubt. This undoubtedly initiates the reaction of bismuth telluride Bi₂Te₃And lead telluride PbTe are established as the research and development work for the two most famous thermoelectric materials most frequently used today. Since then, the former has been widely used for thermoelectric freezing or cooling, while the latter has been successfully used for thermoelectric cooling and thermoelectric power generation. However, this notion or concept of a lower thermal conductivity of an element being of higher atomic mass or atomic number is not true for all elements in the periodic table. This is only partially effective. The effectiveness of starting from the column representing the group IVB elements, moving downwards towards the lower rows and to the right to the group VB and VIB elements in the periodic table becomes more apparent and evident. Thus, despite it being inEarly thirty, forty and fifty years were successful, but in the selection of good thermoelectric elements and compounds, the heavy element selection criteria or concept is universally applicable to all elements in the periodic table. Furthermore, this early observation, concept or criterion, in addition to helping to identify and develop the two best materials to date in the field of thermoelectrics, also helps to identify or discover a total of five major heavy elements simultaneously, namely: lead, bismuth, antimony, tellurium and selenium. All five of these elements also have low thermal conductivity, which are the main contributors to the success achieved in thermoelectric technology in the thirty, forty and fifty years, i.e. in thermoelectric cooling and thermoelectric power generation. Thus, due to the aforementioned standards, more synthetic or compound semiconductors have emerged and eventually developed. Examples are (to illustrate only a portion): lead selenide, lead antimonide, lead telluride (lead telluride selenide), lead antimonide (lead antimonide selenide), bismuth antimonide, bismuth selenide, antimony telluride, silver antimonide (silver antimonide), bismuth selenide telluride (bismuth telluride selenide), and bismuth antimonide (bismuth selenide).

In summary, since the electrical conductivity of semiconductors must generally be increased in order to increase the thermoelectric power factor PF ═ S²σ＝S²The/p is maximized and the semiconductor is typically moderately or highly doped. In addition, in order to make the thermoelectric quality factorAt its maximum, the thermal conductivity must also be reduced or diminished as much as possible. To achieve this, it is necessary to observe the periodic table of elements and to consider the possibility of using five elements occupying the seventh or last row of the periodic table and belonging simultaneously to groups IVB, VB, VIB, VIIB and VIIIThe "heavy element selection criterion for a.f. ioffe" described in the preceding section of the present specification is fully utilized. These five elements have the highest possible five atomic numbers in the periodic table, namely 100, 101, 102, 103, and 104, with corresponding atomic masses 257, 258, 259, 262, and 261, respectively. The corresponding names of these elements are fermium Fm, mendeleave Md, nobelium No, lawrencil Lr, and No. 104 element Unq, respectively. These are names recommended by the international union of theory and applied chemistry IUPAC and modified according to the recommendations of berkeley (us) researchers. The five elements have the highest atomic number and atomic mass in the periodic table, but are not optimal for our purpose, thermoelectric energy conversion. They are metallic, synthetic, radioactive, short lived and must be discarded. Attention must therefore be directed to the five elements in the sixth row immediately above the five elements (i.e., Fm, Md, No, Lr, and Unq). Accordingly, five new elements are discovered or determined from which to select the best or desired thermoelectric semiconductor material. These are lead, bismuth, polonium, astatine and radon. Radon Rn is a heavy gas radioactive element and must therefore be excluded. Astatine At is a highly unstable radioactive element and must therefore be excluded. Polonium Po is a naturally radioactive metal element and must also be excluded from possible choices. Leaving only bismuth Bi and lead Pb with atomic numbers 83 and 82, respectively, and atomic masses 208.98 and 207.2, respectively, as our ideal semiconductor thermoelectric element or material. At this point, theoretically or experimentally or theoretically in connection with experiments, which are evident to any scientist working on thermoelectrics, and which is very likely to involve the a.f. ioffe itself, further alloying or reacting of bismuth or lead with the non-metallic semiconductor element tellurium will clearly produce the semiconductor compound. And bismuth or lead is alloyed or reacted with tellurium to produce bismuth telluride Bi respectively₂Te₃A compound or a lead telluride PbTe compound, which will further reduce the thermal conductivity of the resulting compound and make it an intermediate value between the thermal conductivity values of the original components. Thus, the alloy of bismuth and tellurium reduces the thermal conductivity of the former to a value somewhere between the thermal conductivity of bismuth and the thermal conductivity of telluriumAn intermediate value. Although lead behaves more like a metal than a semiconductor, unlike bismuth, making it relatively difficult to identify or consider as a possible thermoelectric material initially, it produces another excellent synthetic or compounded semiconductor with extraordinary or excellent thermoelectric properties, which is lead telluride, PbTe, after being alloyed or reacted with tellurium again. Although bismuth telluride is more famous for its widespread or widespread use in thermoelectric freezing, although it comes from the silicon germanium alloy Si_0.7Ge_0.3But lead telluride is one of the best materials for thermoelectric power generation to date. Thus, before the sixties, the two synthetic materials or compound semiconductors, i.e., Bi, came into existence₂Te₃And PbTe, have undoubtedly led to great success and success in thermoelectrics. In summary, the first thermoelectric chiller or heat pump was built in 1953, while in 1947 Maria zelkox (Maria talk) constructed the first thermoelectric power generator with 5% efficiency.

Most semiconductors have low to moderate band gaps. The band gap is the only most important factor to consider in developing, designing or synthesizing any new semiconductor material that may be used to control thermoelectric energy conversion. The width of the bandgap is critical for thermoelectric materials because the width of the gap is a measure of the energy required to move electrons away from the localized key orbitals and up to the conduction level. Materials with narrow band gaps are undesirable because they mean that such materials become degenerate or intrinsic at relatively low temperatures. According to the formula given in pierce ergland (pierce Aigrain), the narrower the band gap of a material, the lower the temperature at which the material becomes intrinsic or degenerate, and thus is not useful for thermoelectric energy conversion. The reason for this is that as the material becomes degenerate, its electrical and thermal conductivity increases, but its thermoelectric power (up to power 2) decreases considerably, which has a detrimental effect on the quality factor. Also, as can be inferred from the formula of Eagland (Aigrain), the wider the band gap of a material, the higher will be the maximum thermal junction temperature at which a device comprising such a material can operate, while maintaining a high thermoelectric figure of merit. A device with a relatively high maximum hot junction temperature and thermoelectric figure of merit will also have a high total energy conversion efficiency. On the other hand, a very wide band gap is also not desirable, since it means that it will become more difficult to move electrons from the localized bond orbitals to the conduction band. Therefore, a moderate band gap, i.e., about 0.6ev, is suitable for direct thermoelectric energy conversion. This index is proposed by pierce aigran (pierce Aigrain) as one of the characteristics of a good thermoelectric material. The following table shows the band gaps of various semiconductor intermetallic compounds or synthetic semiconductors and the elements of the semiconductor and the metal in proportion.

Compounds or elements	Band gap eV	Compounds or elements	Band gap eV	Compounds or elements	Band gap eV
						CaSi	1.9	PbS	0.37	α-LaSi	0.19
CaSn	0.9	InSb	0.27	OsSi	1.4
						CaPb	0.46	InAs	0.47	OsSi	2.3
MgSi	0.78	AlSb	1.6	RuGe	0.34
						MgGe	0.70	GaSb	0.8
MgSn	0.36	ReSi	0.12
						MgPb	0.10	FeSi	0.9
BaSi	0.48	RuSi	0.9
						MnSi	0.67	Si	1.1
CrSi	0.35	Ge	0.60
						SiGe	0.7	Sn	0.10

To recapitulate, most semiconductors, especially those used for thermoelectric applications, typically have low to moderate band gaps and are selected or produced to have high atomic mass in order to have low thermal conductivity. Many semiconductors are soft or brittle, have covalent chemical bonds, are somewhat chemically unstable or react with atmospheric oxygen and moisture, and have low to moderate melting points.

In 1956, the concept of alloying or forming a solid solution of isomorphic semiconductor compounds was conceived at about cost (a.f. ioffe) in order to reduce the thermal conductivity of thermoelectric materials. This is due to phonon-phonon interaction, resulting in phonon-phonon scattering due only to more phonons surrounding, the scattering rate of which increases with temperature. In the quantum mechanical concept of phonons, this type of phonon-phonon scattering is described as one phonon being absorbed or emitted by another phonon. Thus, in phonon-phonon interaction, an incident or incoming phonon increases in energy due to its interaction with an obstacle and absorption of one phonon. Phonon emission is similar except that incident or incoming phonons lose energy, where an obstruction is represented by an emitted phonon.

Another important source of phonon scattering is due to point defects. Point defects in a simple sense mean that one of the atoms constituting the crystal is different from the other atoms. By definition, point defects are very small, having little or no effect on long wavelength or low energy phonons. However, point defects can cause extreme scattering of short wavelength, high energy phonons. Any type of defect scatters phonons, but the most important type of point defect in thermoelectric materials is typically an atom with a mass very different from that of most atoms.

When the primary difference between a point defect atom and a majority atom is the mass of the atom, scattering is often referred to as "alloy scattering", "mass concentration fluctuation scattering", or "mass concentration fluctuation alloy scattering". For the same reason, when the main difference between the point defect atom and the majority atom is the volume of the atom, the scattering is referred to as "volume concentration fluctuation scattering" or "volume concentration fluctuation alloy scattering". In general, the main difference between a point defect atom and a majority atom relates to the mass and volume of the atom. Thus, mass concentration fluctuation scattering and volume concentration fluctuation scattering generally occur simultaneously. Thus, the term "alloy scattering" generally means point defect phonon-phonon scattering due to fluctuations or differences in mass and volume concentrations between a point defect and a plurality of atoms. The term "mass and volume concentration fluctuation scattering" or "alloy scattering" is generally preferred over the term "point defect scattering" when the atoms of the point defect represent a substantial proportion of the mixture or alloy consisting of the defect and the majority of the atoms. However, the concept or principle remains the same, i.e. if the lattice is indeed homogeneous, the phonons propagate with little scattering. While when the crystal lattice has many defects, phonons are strongly scattered.

Summary of The Invention

According to one embodiment of the present invention there is provided a process for producing a device for the controlled thermoelectric energy conversion, the device consisting of a P-type branch, an n-type branch, a hot junction and a cold junction, the process comprising using a composition in the manufacture of the n-type branch and/or the P-type branch of the device, wherein the composition comprises magnesium, silicon, lead and barium, and optionally one or more other doping materials.

These four basic constituent components of the composition, i.e., Mg, Si, Pb, and Ba, are mixed together to chemically react with each other to produce a compound. Thus, according to another embodiment of the invention, there is provided a process for producing a device for controlling thermoelectric energy conversion, the device consisting of a P-type branch, an n-type branch, a hot junction and a cold junction, the process comprising using a composition in the manufacture of the n-type branch and/or the P-type branch of the device, wherein the composition comprises magnesium silicide Mg₂Si, wherein part of the magnesium is replaced by barium and part of the silicon is replaced by lead. Whereby the composition is an alloy or solid solvent comprising an intermetallic compound of magnesium silicide, magnesium plumbate, barium silicide and barium plumbate, wherein the composition has the general formula:

Ba_2rMg_2(1-r)Si_1-xPb_x

wherein r, (1-r), (1-x) and x represent the atomic proportions of barium, magnesium, silicon and lead, respectively, in the alloy, and wherein the composition optionally comprises one or more other dopant materials. The composition may also contain no other doping materials.

According to another embodiment of the present invention, the n-type and p-type branches of a device for controlling thermoelectric energy conversion are fabricated according to thin film technology, wherein the thickness or length of the branches is sufficiently reduced, resulting in a sufficient reduction in the overall size of the device, and an improvement in energy conversion efficiency.

According to another embodiment of the invention, the n-type and p-type thermoelectric elements or branches are encapsulated in or covered or surrounded by a very thin layer of material which is a poor conductor of heat and electricity, wherein the thin layer or encapsulation is not in contact with the hot or cold connection points, is in little contact with the sides of the respective thermoelectric element and extends over its entire length, wherein the contacts are very close to the hot and cold connection points, wherein the encapsulation is of circular, quasi-square or rectangular cross-section, wherein the material does not chemically interact or diffuse with the composition constituting the branches immediately or in long-term operation, said encapsulation material having a very high chemical and mechanical stability and being highly resistant to acids, corrosion and high temperatures.

According to another embodiment of the invention, thin film technology, integrated circuit technology and packaging technology are combined during the manufacturing and assembly of the device for controlling thermoelectric energy conversion comprising said composition.

Brief description of the drawings

FIG. 1 is a flow diagram of the basic components of an apparatus for direct conversion of thermoelectric energy; and

fig. 2 is a periodic table of elements highlighting the basic concept of the invention.

Detailed description of the preferred embodiments

As shown in fig. 1, the present invention relates to a process or method for producing a device for controlling thermoelectric energy conversion, thereby significantly increasing the efficiency of conversion from thermal energy to electric energy or from electric energy to thermal energy. Thermal energy sources include solar radiation, nuclear elements or batteries, fossil fuel combustion, waste heat from boilers, gas turbine or automotive exhaust, and biological waste, or life forms.

The invention also relates to a composite material for use in the manufacturing process of a device for controlling thermoelectric energy conversion.

The present invention relates to an apparatus for efficient direct thermal-to-electrical or electrical-to-thermoelectric conversion.

The present invention relates to a method for preparing a thermoelectric energy direct conversion composition.

According to an embodiment or aspect of the invention there is provided a method for producing a device for controlling thermoelectric energy conversion, the device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, the method comprising using a composition in the manufacture of the n-type branch and/or the P-type branch of the device, wherein the composition comprises magnesium silicide Mg₂Si, wherein a portion of the magnesium is replaced by barium and a portion of the silicon is replaced by lead, whereby the composition is an alloy or solid solution of intermetallic compounds comprising magnesium silicide, magnesium plumbate, barium silicide and barium plumbate, wherein the composition has the following general formula:

Ba_2rMg_2(1-r)Si_1-xPb_x

wherein r, (1-r), (1-x) and x represent the atomic proportions of barium, magnesium, silicon and lead, respectively, in the alloy, and wherein the composition optionally comprises one or more other dopant materials.

By carefully adjusting the r and x parameters in the above formula, it is possible to obtainCompositions having very low thermal conductivity, the minimum of which should be approximately 0.002Wcm^-1K^-1. The atomic or molecular proportion of the dopant or impurity and the concentration of free charge carriers in the composition should preferably be at 10^-8To 10^-1And 1 x 10 per cubic centimeter¹⁵To 5X 10²⁰Within a range of carriers. By carefully controlling the doping level and the free charge carrier concentration, it is possible to make the thermoelectric power factor S²σ max, the maximum thermoelectric power factor associated with about 0.002Wcm by using the composition^-1K^-1Together cause or produce 10 a minimum thermal conductivity^-2K^-1An order of magnitude thermoelectric figure of merit Z. For thermoelectric power generators, this should help to bring the energy conversion efficiency to about 43%.

In addition to barium, one or more other elements may be substituted for magnesium. Also, one or more other elements may be substituted for silicon in addition to lead. This contributes to compositions having a more general chemical formula. These other elements, especially those used as substitutes for magnesium and/or silicon, cause an increase in the average band gap and average melting temperature of the resulting composition. Such an increase generally results in a corresponding increase in the maximum thermal junction temperature at which the thermoelectric energy conversion device can operate. Thus, the Carnot (Carnot) energy conversion efficiency of the device will increase as well as the overall energy conversion efficiency. On the other hand, other alternatives to magnesium and/or silicon will stop reducing the exact or minimum atomic proportions of barium and lead, which would otherwise be required to result in absolute minimum lattice thermal conductivity as well as overall thermal conductivity. Thus, the thermal conductivity of the resulting composition will increase, which is undesirable. The less barium and lead in the composition, the higher the thermal conductivity will be. This will adversely affect the thermoelectric figure of merit as well as the overall energy conversion efficiency. Therefore, the respective minimum atomic proportions of barium and lead have been set to 10% in all the general formulae. This will ensure that the thermal conductivity of the composition defined by the general formulae does not increase significantly whilst utilising any possible increase in the working thermal junction temperature, thermoelectric power and thermoelectric power factor caused by other elements as substitutes for part of the magnesium and/or part of the silicon.

Other elements that partially replace magnesium and/or silicon may be considered as simple substitutes for possible improvement of thermoelectric power factor and quality factor as described above, and may also be considered as doping materials or dopants specifically for producing n-type or p-type compositions.

A detailed description will now be given of how to prepare the composition by melt metallurgical method or powder metallurgy. With certain precautions, the melt metallurgy process is more likely to produce single crystal materials, although this is quite difficult. In this regard, the best opportunity to obtain single crystal material is to use a heat exchanger method known in the art as HEM. It may not be as important to produce a single crystal material. For example, magnesium silicide Mg is produced by powder metallurgy techniques₂Si produces a material with excellent thermoelectric properties and quality factor. Since the composition consists essentially of magnesium silicide, it clearly involves powder metallurgy techniques and is therefore the most recommended method for the manufacturing process. However, certain precautions must be strictly taken during the preparation phase and also during the long-term working of the materials produced by powder metallurgy technology. Precautions include avoiding all types of exposure to oxygen by preparing and controlling the composition in absolute vacuum or, preferably, in an inert gas maintained at a pressure above atmospheric pressure, preferred inert gases include argon. The above-mentioned precautionary measures are met in part by another embodiment of the invention, which includes packaging.

The performance and efficiency of devices for controlling thermoelectric energy conversion, including composites, may be improved by using functional graded material technology or FGM methods. Alternatively, cascaded or segmented FGM techniques may be used, where the number of cascades, segments or stages may be 3 or 4. Likewise, integrated circuit technology known in the art as i.c. technology can be used in the manufacture of devices for controlling thermoelectric energy conversion including composites, where a large number of p-type and n-type thermoelectric element pairs are connected in series or in parallel to produce any intensity of current and voltage, and thus any power in the case of thermoelectric power generators, and any cooling or heating capacity in the case of thermoelectric chillers and thermoelectric heat pumps.

According to another embodiment or aspect of the invention, the other doping material of the n-type branch of the device as described in the previous first embodiment comprises one or more elements selected from the group consisting of nitrogen, phosphorus, arsenic, antimony, bismuth, oxygen, sulphur, selenium, tellurium, chlorine, bromine, iodine, magnesium, barium, lithium, gold, aluminium, indium, iron and/or combinations thereof.

According to another embodiment or aspect of the invention, the further doping material of the p-type branch of the device as described in the previous first embodiment comprises one or more elements selected from the group consisting of copper, silver, sodium, potassium, rubidium, cesium, boron, silicon, lead and/or combinations thereof.

According to another embodiment or aspect of the invention, as described in the previous three embodiments, r ranges from 0.1 to 0.4, (1-r) ranges from 0.6 to 0.9, x ranges from 0.1 to 0.3, and (1-x) ranges from 0.7 to 0.9, the atomic or molecular ratio of the dopant material in the alloy ranges from 10^-8To 10^-1And the free charge carrier concentration ranges from 1 x 10 per cubic centimeter¹⁵Current carrier up to 5 x 10 per cubic centimeter²⁰A carrier.

According to another embodiment or aspect of the invention there is provided a process for producing a device for controlling thermoelectric energy conversion, the device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, the process comprising using a composition in the manufacture of the n-type branch and/or the P-type branch of the device, wherein the composition comprises magnesium silicide Mg₂Si, wherein a portion of the magnesium is replaced by barium and a portion of the silicon is replaced by lead, whereby the composition is an alloy or solid solution of intermetallic compounds comprising magnesium silicide, magnesium plumbate, barium silicide and barium plumbate, wherein the composition has the following general formula:

Ba_2rMg_2(1-r)Si_1-xPb_x

wherein r, (1-r), (1-x) and x respectively represent the atomic ratio of barium, magnesium, silicon and lead in the alloy.

According to another embodiment or aspect of the invention, in the above embodiments, r ranges from 0.1 to 0.4, (1-r) ranges from 0.6 to 0.9, x ranges from 0.1 to 0.3, and (1-x) ranges from 0.7 to 0.9.

According to another embodiment or aspect of the invention there is provided a process for producing a device for controlling thermoelectric energy conversion, the device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, the process comprising using a composition in the manufacture of the n-type branch and/or the P-type branch of the device, wherein the composition comprises magnesium silicide Mg₂Si, wherein a portion of the magnesium is replaced by one or more elements selected from the group consisting of beryllium, calcium, strontium, and barium, and a portion of the silicon is replaced by one or more elements selected from the group consisting of germanium, tin, lead, antimony, bismuth, selenium, and tellurium, wherein the composition has the following general formula:

(Be，Ca，Sr，Ba)_2rMg_2(1-r)Si_1-s(Ge，Sn，Pb，Sb，Bi，Se，Te)_swherein the composition has a more specific form of the general formula:

Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_g

wherein r ═ u + v + w + z, denotes the sum of the atomic proportions of the elements used for the partial replacement of magnesium, s ═ a + b + c + d + e + f + g, denotes the sum of the atomic proportions of the elements used for the partial replacement of silicon, and wherein the composition optionally comprises one or more further doping materials.

According to another embodiment or aspect of the invention, in the above-described embodiments, the further doping material of the n-type branch of the device comprises one or more elements selected from the group consisting of nitrogen, phosphorus, arsenic, antimony, bismuth, oxygen, sulfur, selenium, tellurium, chlorine, bromine, iodine, magnesium, barium, lithium, gold, aluminum, indium, iron and/or one or more compounds of these elements.

According to another embodiment or aspect of the invention, in the seventh embodiment described above, the further doping material of the p-type branch of the device comprises one or more elements selected from the group consisting of copper, silver, sodium, potassium, rubidium, cesium, boron, silicon, lead and/or one or more compounds of these elements.

According to another embodiment or aspect of the invention, in the above three embodiments, r ranges from 0.1 to 0.4, (1-r) ranges from 0.6 to 0.9, u, v and w ranges from 0 to 0.3, u + v + w ranges from 0 to 0.3, z is not less than 0.1, s ranges from 0.1 to 0.3, (1-s) ranges from 0.7 to 0.9, a, b, d, e, f, g ranges from 0 to 0.2, (a + b + d + e + f + g) ranges from 0 to 0.2, c is not less than 0.1, and the atomic or molecular ratio of the doping material or materials in the alloy ranges from 10^-8To 10^-1And the free charge carrier concentration ranges from 1 x 10 per cubic centimeter¹⁵Current carrier up to 5 x 10 per cubic centimeter²⁰A carrier.

In another embodiment or aspect, there is provided a process for producing a device for controlling thermoelectric energy conversion, the device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, the process comprising using a composition in the manufacture of the n-type branch and/or the P-type branch of the device, wherein the composition comprises in its most general form magnesium silicide Mg₂Si, wherein a portion of the magnesium is replaced by one or more elements selected from the group consisting of beryllium, calcium, strontium, and barium, and a portion of the silicon is replaced by one or more elements selected from the group consisting of germanium, tin, lead, antimony, bismuth, selenium, and tellurium, wherein the composition has the following general formula:

(Be，Ca，Sr，Ba)_2rMg_2(1-r)Si_1-s(Ge，Sn，Pb，Sb，Bi，Se，Te)s

wherein the composition has a more specific form of the general formula:

Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_g

where r ═ u + v + w + z denotes the sum of the atomic proportions of the elements used for the partial substitution of magnesium, and s ═ a + b + c + d + e + f + g denotes the sum of the atomic proportions of the elements used for the partial substitution of silicon.

According to another embodiment or aspect of the invention, in the above embodiments, r ranges from 0.1 to 0.4, (1-r) ranges from 0.6 to 0.9, u, v and w each ranges from 0 to 0.3, (u + v + w) ranges from 0 to 0.3, z is not less than 0.1, s ranges from 0.1 to 0.3, (1-s) ranges from 0.7 to 0.9, a, b, d, e, f, g each ranges from 0 to 0.2, (a + b + d + e + f + g) ranges from 0 to 0.2, c is not less than 0.1.

According to another embodiment or aspect of the invention, the thermoelectric elements or branches of the device for controlling thermoelectric energy conversion as described in the above embodiments, whether n-type or p-type, are manufactured according to a functional graded materials technique called FGM method, wherein the chemical composition and/or the energy band gap and/or the doping level and/or the free charge carrier concentration continuously vary from the hot junction to the cold junction, wherein the electrical conductivity of the respective thermoelectric elements is maintained constant.

According to another embodiment or aspect of the invention, the thermoelectric elements or branches of the device for controlling thermoelectric energy conversion as described in the above embodiments are manufactured according to a cascaded or segmented FGM technique, wherein the number of cascaded, segmented or phase may be 3 or 4, and wherein the chemical composition and/or the energy band gap and/or the doping level and/or the free charge carrier concentration is kept constant in each segment or phase, but continuously varies from phase to phase of each thermoelectric element or branch, wherein the doping level or impurity concentration varies from a lower value at the cold junction to a higher value at the hot junction.

According to another embodiment or aspect of the present invention, the n-type and/or p-type thermoelectric elements or branches of the device for controlling thermoelectric energy conversion as described in the above embodiments are fabricated according to thin film technology, wherein the thickness or length of the n-type and/or p-type branches or thermoelectric elements is sufficiently reduced or decreased, ultimately resulting in a sufficient reduction or decrease in the overall size of the device and an increase in energy conversion efficiency.

According to another embodiment or aspect of the invention, the n-type and/or p-type thermoelectric elements or branches as described in the above embodiments are encapsulated in or covered or surrounded by a very thin layer of material which is a poor conductor of heat and electricity, i.e. a good thermal and electrical insulator, wherein the thin layer or encapsulation is not in contact with the hot or cold connection points, is in little contact with the sides of the respective thermoelectric element and preferably extends over its entire length, wherein the contacts are very close to the hot and cold connection points, wherein the encapsulation is of circular, or quasi-square or rectangular cross-section, wherein the material does not chemically interact or diffuse with the composition constituting the n-type and p-type branches immediately or in long-term operation, said encapsulation material having a very high chemical and mechanical stability and being highly resistant to acids, corrosion and high temperatures, the thin layer or encapsulation material comprising a material selected from the group consisting of beryllium, lead or lead-free from chemical interactions, At least one compound selected from the group consisting of carbides, nitrides and oxides of magnesium, calcium, strontium, barium, titanium, zirconium, hafnium, vanadium, niobium, tantalum, scandium, yttrium, chromium, molybdenum, tungsten, lanthanum and other elements of the lanthanide series between lanthanum and hafnium in the periodic table of the elements.

According to another embodiment or aspect of the invention, the n-type and p-type branches, pairs or couples thereof, constituting a single device for controlling thermoelectric energy conversion as described in the above embodiments are fabricated and assembled according to integrated circuit technology known in the art as i.c. technology, wherein the devices are connected in series, parallel or series-parallel to produce current and voltage of any amperage or strength and thereby produce any power in the case of a thermoelectric power generator, or to handle any cooling or heating load in the case of a thermoelectric chiller or a thermoelectric heat pump, the fabrication and assembly methods as described herein result in a substantial reduction in the overall size of future thermoelectric devices, as well as a further improvement in the overall energy conversion efficiency or coefficient of performance, regardless of their power generation, cooling load or heating load capabilities.

According to another embodiment or aspect of the present invention, all three methods described above, i.e., thin film technology, integrated circuit technology and packaging technology, are used in combination during the design, manufacture and assembly of a device for controlling thermoelectric energy conversion as described above, wherein the packaging method or technology or the configuration and profile of the package may be changed or modified to some extent so as to be able to use or adapt to both thin film and integrated circuit technology in the construction and assembly of the thermoelectric energy conversion device.

According to another embodiment or aspect of the present invention, there is provided a process for the preparation, production of a composition of the following two general formulae:

Ba_2rMg_2(1-r)Si_1-xPb_x (1)

and Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_g (2)

A convenient method of synthesizing a composition as defined in any one of the preceding embodiments, the method comprising mixing a predetermined proportion of starting elements, which must be as extremely pure as possible to avoid unwanted doping, wherein the starting elements comprise magnesium, silicon, lead and barium of formula (1) above and any other doping material if necessary, or one or more elements selected from the group consisting of beryllium, calcium, strontium and barium according to formula (2) above with Mg constituting a magnesium silicon compound₂Si together with one or more elements selected from the group consisting of germanium, tin, lead, antimony, bismuth, selenium and tellurium, and any thereofIt is doped with a material, wherein the starting element and possibly other doping materials are preferably in the form of particles or in the form of a very fine powder; the method further comprises pouring the starting elements and other doping materials into a vessel (vessel or receptacle), boat or crucible of appropriate size and shape, the containers are made of a material that does not react or contaminate the composition, alloy or solution of the composition, alloy or solution to be produced, thereby avoiding any unwanted or unintended doping, wherein the material is preferably composed of one or more elements selected from the group consisting of tungsten, rhenium, ruthenium, rhodium, palladium, platinum, gold, iridium, osmium, tantalum, hafnium, zirconium, titanium, molybdenum, chromium, vanadium and niobium, or the material may be composed of at least one compound selected from the group consisting of carbides, nitrides, and oxides of beryllium, magnesium, calcium, strontium, barium, titanium, zirconium, hafnium, vanadium, tantalum, lanthanum, and other elements constituting the lanthanides between lanthanum and hafnium in the periodic table of the elements; placing said container, boat or crucible concentrically within a suitable furnace, wherein said furnace operates according to a temperature gradient freezing technique (temperature gradient freeze), wherein said furnace and said technique are typically a Bridgman furnace and a Bridgman crystal growth technique, respectively, wherein in a standard version of the Bridgman technique the configuration of the furnace and boat or crucible is vertical, and in a modified version of the non-conventional Bridgman technique the configuration of the furnace and boat is horizontal, wherein the interior or outer shell of the furnace, where the crucible is placed vertically or the boat is placed horizontally, is completely isolated from air, the absolute pressure of which is preferably 10^-4To 10^-6mmHg, then filled with an inert gas, preferably helium or argon, maintained at a relative pressure of about 2 to 30 physical atmospheres (or 2 to 30 bar), and then sealed, thereby suppressing excessive loss due to the higher volatility of magnesium relative to barium, lead and silicon, since the boiling points of the basic components are 1363K, 2170K, 2022K and 3538K, respectively, and the melting point of silicon is 1687K, thereby heating the starting element and the doping material to a temperature about 15 to 30 ℃ above the melting point of silicon, which is the component with the highest melting point among the four elements, and the othersThe melting points of the three components magnesium, barium and lead are 923K, 1000K and 600.6K, respectively, and it is preferable to mix the starting elements: magnesium, barium, lead and silicon, and possibly doping impurities, are heated to 1700K to 1715K to first ensure complete melting of the silicon, and then held at this temperature for about 2 to 3 hours to allow sufficient time for the necessary chemical reactions to occur, i.e. between magnesium and silicon and lead and between barium and silicon and lead, and also for sufficient time to mix the resulting compounds and the formation of a single phase alloy or solid solution, wherein no or no chemical reactions between magnesium and barium should occur, wherein the electronegative difference between magnesium and barium is 0.42, the electronegative difference between silicon and lead is 0.43, the electronegative difference between magnesium and silicon is 0.59, the electronegative difference between magnesium and lead is 1.02, the electronegative difference between barium and silicon is 1.01, the electronegative difference between barium and lead is 1.44, wherein the first two electronegative differences (i.e.0.42 and 0.43) are greater than the next four electronegative differences (i.e.59, 0.43) 1.02, 1.01 and 1.44) are much smaller, which prevents any chemical reaction or formation of chemical compounds directly between magnesium and barium and between silicon and lead, which in turn allows chemical reactions and therefore formation of chemical compounds between magnesium and silicon, between magnesium and lead, between barium and silicon, between barium and lead, which conclusion can also be deduced entirely independently from the electronic structure of the above elements indicated in the periodic table of elements shown in fig. 2, that after a temperature of about 2 to 3 hours between 1700K and 1715K, the above composition, i.e. the magnesium barium silicon lead alloy or solution, can be doped or undoped, will cool down very slowly to room temperature, wherein the temperature of the furnace first drops from a temperature between 1700K and 1715K, preferably over a period of 12 to 24 hours, until the hottest part of the material or composition in the crucible or boat drops below the solidus temperature of the particular alloy composition being produced by about 5 c, wherein a solidification rate of about 1 to 5 mm per hour at which the isothermal solid-liquid interface moves should give satisfactory results, maintaining a linear temperature gradient along the entire length of the crucible and maintaining an arcuate solid-liquid interface, the solid-liquid interface being concave toward the liquid phase during crystal growth, generally resulting in the production of a crucible having relatively few crystal dislocations and seedsSingle crystal alloys that substantially reduce defects such as fine cracks and uneven crystal growth.

However, it cannot be guaranteed that a single-crystal solid solvent or alloy can be obtained, particularly with a material containing four elements whose atomic mass, atomic radius, density, specific heat and thermal conductivity vary widely. It is likely that polycrystalline material will end up due to atomic and physical properties as a result of the above. The main material that can be expected from the above preparation and crystal growth methods is a polycrystalline material with several grains, which are all quite large. This may be the one closest to obtaining a single crystal magnesium barium silicon lead alloy or solid solution defined by either of the two general formulas:

Ba_2rMg_2(1-r)Si_1-xPb_x

or Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_g

According to another embodiment or aspect of the invention, a convenient method for preparing and producing the composition of the first 11 embodiments described above comprises mixing a predetermined ratio of starting elements that must be as extremely pure as possible to avoid detrimental doping, wherein the starting elements comprise Ba_2rMg_2(1-r)Si_1-xPb_xMagnesium, silicon, lead and barium of defined composition, and any other doping material as necessary or desired, or including_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_gOf another composition as defined, constituting the magnesium-silicon compound Mg₂Magnesium and silicon elements of Si, and one or more elements selected from the group consisting of beryllium, calcium, strontium and barium for partially substituting magnesium elements, and germanium, tin for partially substituting siliconOne or more elements selected from the group consisting of lead, antimony, bismuth, selenium and tellurium, and any other doping material, wherein the starting element and the other doping material, if any, are preferably in the form of particles or in the form of a very fine powder; the method further comprises injecting the starting elements and other doping materials into a suitably sized and shaped container or crucible made of a material that does not react or contaminate the components of the composition, alloy or solution to be produced, thereby avoiding any unwanted or unintended doping, wherein the material is preferably comprised of one or more elements selected from the group consisting of tungsten, rhenium, ruthenium, rhodium, palladium, platinum, gold, iridium, osmium, tantalum, hafnium, zirconium, titanium, molybdenum, chromium, vanadium and niobium, or the material may be comprised of at least one compound selected from the group consisting of carbides, nitrides or oxides of these elements, beryllium, magnesium, calcium, strontium, barium, titanium, zirconium, hafnium, vanadium, tantalum, lanthanum, and other elements constituting the lanthanum group between lanthanum and hafnium in the periodic table; the crucible containing the ingredients is then evacuated to an absolute pressure of preferably 10^-4To 10^-6mmHg, then filled with an inert gas, preferably helium or argon, to a relative pressure of about 2 to 30 physical atmospheres (or 2 to 30 bar), then finally sealed, said crucible being then placed in a horizontal or vertical furnace and heated so that the composition contained therein reaches a temperature above the melting point of silicon (1687K), so that the molten composition is preferably maintained at a temperature of 1700K to 1735K for about 15 to 30 minutes to ensure complete melting of the silicon and Mg₂Formation of the Si compound, however, the melting temperature is gradually reduced to 1500K in the next 20 to 30 minutes, at least 20 minutes being maintained at that level, and then the composition components are maintained in a fully molten state for a period of time sufficient to ensure the formation of intermetallic compounds and the production of a mixture with a homogeneous composition, wherein the period of time may be referred to as a mixing period, typically lasting at least one hour, and when the contents of the crucible are in the liquid state, they are subjected to intense agitation so that they are intimately mixed together, thereby producing a single phase alloyThe stirring of the contents of the crucible is carried out by intermittently withdrawing the crucible with the crucible tongs, shaking and returning it to the furnace, wherein a rocking type furnace is also used for stirring the contents of the crucible, and after said mixing cycle, the obtained composition is cooled at a rate of about 2 ℃ to 20 ℃ per hour, which continues until the ambient temperature, alternatively at a rate of about 400 ℃, at which the cooling rate is preferably increased to 50 ℃ to 100 ℃ per hour, and finally the composition or alloy thus produced is withdrawn from the crucible as a generally polycrystalline material which can be used for the manufacture of thermoelectric energy conversion devices.

According to another embodiment or aspect of the invention, a convenient method for preparing and producing the composition of the first 11 embodiments described above comprises separately producing a composition according to the following two general formulas:

Ba_2rMg_2(1-r)Si_1-xPb_x (1)

and Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_g (2)

Each intermetallic compound necessary for any one of them by mixing and heating predetermined stoichiometric amounts of the components to a temperature of about 50 ℃ higher than the melting point of the respective compound, wherein if the desired formula is formula (1), the compounds are prepared by heating Mg and Si, Mg and Pb, Ba and Si, and Ba and Pb to appropriate temperatures, and when the compounds are to be prepared according to formula (2), the same or different combinations of elements may be required, wherein Mg is generated₂Si and BaSi₂A heating temperature much higher than the melting point of the compound is required to ensure complete melting of the silicon, the remaining steps including: the molten components are preferably agitated vigorously and in argon gas preferably having an absolute pressure of about 2 to 30 physical atmospheres (or 2 to 30 bar)Maintaining at a suitable temperature for about one hour, then gradually cooling the resulting compound to ambient temperature, preferably after granulation or comminution, mixing the thus obtained compounds together in the desired proportions and then injecting into a crucible of suitable size and shape, optionally introducing a suitable amount of a suitable doping material or dopant during the mixing of the intermetallic compound, wherein preferably a portion or all of the doping impurities or dopants are added during melting, the crucible containing the ingredients then being evacuated to preferably up to 10 deg.f^-4To 10^-6Mmhg absolute pressure, then the crucible is filled with an inert gas such as helium or argon, preferably at about 2 to 30 bar, or about 0.2 to 3Mpa or about 2 to 30 atmospheres absolute, preferably argon, and finally sealed, then the crucible is placed in a horizontal or vertical furnace, heated to a temperature a few degrees above the highest melting temperature of all the constituent compounds to ensure complete melting of all the constituents, while the constituents of the composition are in a molten state, they are vigorously agitated by any of the methods described in the above examples, the contents of the crucible are maintained at an appropriate temperature for about one hour to obtain a single phase alloy or solution, the composition or alloy is then cooled at a rate of about 2 to 20 ℃ per hour, which continues until ambient temperature, or the cooling rate can be maintained at a temperature of about 400 ℃, from this temperature, the cooling rate may be increased preferably to 50 to 100 ℃ per hour, and the composition or alloy thus produced is finally removed from the crucible.

The alloy or composition produced according to either of the two embodiments described above is generally single phase polycrystalline. It is strained and contains a large number of dislocations. In order to prevent or reduce the strain in the alloy obtained, it is preferred that the constituents are initially injected into and melted in a flexible mold made of a very thin, easily deformable platinum sheet or foil, rather than into a rigid crucible. Such a mold does not cause strain in the material when the molten composition expands during cooling. A more robust outer crucible made of graphite, stainless steel or any suitable refractory metal may be used to support the mold for additional strength. The composition or alloy may be converted into a single crystal or a single crystal material prior to use in the manufacture of a thermoelectric energy conversion device. The production of such alloys and materials can be accomplished by a variety of methods. One such method is the temperature gradient freezing technique, also known in the art as the bridgman method.

According to another embodiment or aspect of the invention, has the general formula Ba_2rMg_2(1-r)Si_1-xPb_xThe single crystal or single crystal barium magnesium silicon lead alloy or solid solution can be prepared by pouring the polycrystalline material prepared according to any of the three previous embodiments into an elongated horizontal crucible having an opening of suitable size and shape, such crucible often being referred to as a (boat-shaped) boat having a bottom surface and a pair of lateral end surfaces, the bottom surface being integrally joined to a pair of side surfaces, the boat or crystallization vessel then being suitably evacuated to 10 deg.f^-4To 10^-6An absolute pressure ampoule (ampoule) of mmhg, preferably made of stainless steel or made of one or more of the above-mentioned compounds, then preferably filled with an inert gas, preferably argon, of about 2 to 30 physical atmospheres or of about 0.2 to 3Mpa absolute pressure, and finally sealed, said horizontal crucible or boat preferably being made of a material consisting of at least one compound selected from the group consisting of carbides, nitrides or oxides of beryllium, magnesium, calcium, strontium, barium, titanium, zirconium, tungsten, hafnium, tantalum, lanthanum and other elements of the lanthanum group consisting of the elements of lanthanum and hafnium, intervening between the elements of the periodic table, or said horizontal crucible or crucible preferably being made of a material consisting of one or more elements selected from the group consisting of tungsten, rhenium, ruthenium, rhodium, palladium, platinum, gold, iridium, osmium, tantalum, hafnium, zirconium, titanium, molybdenum, chromium, vanadium and niobium, the flask was concentrically placed in an open-ended tubular heat conducting sleeve closed at the hot gathering end by a removable heat insulating plug, the sleeve being made of a material having a thermal conductivity higher than that of the boat and its contents, an array of tubular heat insulating sleeves being placed along the heat conducting sleeveAxially surrounding the heat conducting sleeve, the resulting assembly is then placed in a furnace having heating elements designed to induce a linear temperature differential across the furnace, heating of the furnace is then continued until the temperature of the coldest end of the ingot reaches a minimum temperature corresponding to the liquidus temperature of the particular alloy composition being produced, the furnace is maintained at this minimum temperature for at least one hour to ensure complete melting of the contents of the crucible, and the temperature of the crucible is then reduced over a period of about 12 to 24 hours until the temperature of the hottest portion of the implant in the boat reaches 5 ℃ below the solidus temperature of the particular alloy composition being produced, it being found that a solidification rate of about 1 to 5 mm/hour of isothermal movement of the solid-liquid interface gives satisfactory results.

The apparatus assembly described in the above embodiments includes an insulating sleeve, a heat conducting sleeve, a horizontal boat, a flask and a specially designed heating element to enable a linear temperature gradient to be maintained along the entire length of the crucible and to maintain an arcuate solid-liquid interface that is concave toward the liquid phase during crystal growth. The above measures generally result in the production of single crystal alloys having relatively few crystal dislocations and substantially reduced defects such as fine cracks and uneven crystal growth.

Further, it is to be understood that the steps described above, including mixing, heating and reacting the constituent components of the composition or alloy, and producing the single crystal or polycrystalline structure associated therewith, may all be performed together continuously in a single apparatus, such as the temperature gradient freezing apparatus components described above. In this case, a longer period of time should be maintained so that the necessary chemical reactions between the elements are completed with sufficient time and so that there is sufficient time for the formation of a single phase solid solvent or alloy. Excess magnesium over the stoichiometrically required amount is preferably added to the mixture prior to heating to compensate for excess losses due to evaporation due to the higher volatility of magnesium relative to the three elements barium, lead and silicon. The amount of excess magnesium added is adjusted so that a stoichiometric composition or alloy is ultimately obtained.

The high volatility of magnesium stems from the fact that the melting point of silicon is 1687K, while the boiling points of the above four elements, i.e. magnesium, silicon, lead and barium, are 1363K, 3538K, 2022K and 2170K, respectively. Since silicon has the highest melting point of these four elements, i.e. 1687K, which is thus about 300K higher than the boiling point of magnesium, this temperature difference causes a high volatility of this element magnesium.

Preferably, the composition or alloy so produced is ultimately subjected to either of two processes known in the art as zone refining and zone melting. This final step or process, together with the strong agitation of the molten components during the preparation of the solid solvent, ensures the production of a sufficiently homogeneous single phase alloy.

The purity of the starting elements (i.e. magnesium, silicon, lead and barium) required to produce such a composition should be higher than 99.999 per cent by weight for each element. For silicon, lead and barium, purity levels well above this index are preferred.

The heat exchanger process known in the art as HEM can still be used to produce or prepare the composition or alloy. While HEM has not been widely commercially available to date, it offers the potential for substantial cost reduction in large-scale manufacturing. HEM is a directional solidification technique that has been applied to the growth of large area square cross-section silicon ingots obtained by melting.

HEM technology incorporates a furnace for material growth under a reducing or neutral gas atmosphere. The furnace has a graphite heating zone supported by a graphite insulating layer. The assembly was placed in a vacuum-tight water-cooled stainless steel chamber. Heat is provided by a grid-type graphite heater powered by a suitable three-phase power supply. A high temperature heat exchanger is inserted through the bottom of the chamber and the heating zone. The heat exchanger is a closed-end tube with an injection tube for the flow of helium as a coolant. There are no moving parts in the HEM furnace to minimize the required seal. Furthermore, the solid-liquid interface is submerged below the melt, thus providing only a small viewing port at the top of the furnace. Other ports in the furnace are used for evacuation, as well as for control and measurement of the pyrometer. These features create a well-insulated hot zone design. The control instrument relies on a standard dual channel microprocessor that can be simply programmed for heat input and heat removal.

The hot zone is designed such that no significant gradients are created in the furnace without coolant flowing through the heat exchanger. This is achieved by thermal symmetry, multi-layer insulation around the hot zone, and minimization of the viewport. For example at the edges of the heating element, some natural temperature gradient is expected. In a HEM furnace, the temperature along the crucible wall is approximately constant. This feature makes HEM different from temperature gradient freezing techniques.

The heat exchanger method, HEM, has been developed for growing large, high quality crystals. The seed crystal is placed at the bottom of a crucible, which is located on a high temperature heat exchanger. A feedstock or charge containing the basic components of the composition to be produced, i.e. magnesium, silicon, lead and barium, is then loaded into the crucible on top of the seed crystal. After evacuation, the furnace enclosure is filled with an inert gas, preferably argon, up to a relative pressure preferably between 2 and 30 atmospheres, to suppress the excessive loss of magnesium that may occur, due to its high volatility relative to the other three components. The graphite heater then provides heat to melt the charge. Seed melting is prevented by flowing a minimal amount of gaseous helium through the heat exchanger. After melting in addition to the seed crystal, growth is driven by increasing the flow of helium and thereby lowering the temperature of the heat exchanger.

Essentially, the process involves directional solidification by melting, the temperature gradient in the solid being controlled by the heat exchanger, and the temperature gradient in the liquid being controlled by the furnace temperature. After solidification is complete, the gas flow through the heat exchanger can be reduced to allow the temperature of the crystal to equilibrate during the annealing and cooling stages.

The technique is unique in that the liquid temperature gradient can be controlled independently of the solid temperature gradient without moving the crucible, hot zone or silicon ingot. The most notable feature is the submerged interface, which is stabilized by the surrounding liquid. It is not affected by hot spots, mechanical vibrations and convection. Thus, no rotation of the crucible is required to achieve thermal symmetry.

Growth at submerged interfaces makes HEM ideally suited for low purity silicon, where many of the regeneration phase impurities, such as carbides and oxides, tend to float at the surface of the melt, away from the growth interface. The melt acts as a buffer layer, protecting the submerged solid-liquid interface during most growth cycles. Thus, in a HEM, there is minimal temperature, concentration, and fluctuations at the interface because of the surrounding liquid. During growth, the cooler material is at the bottom and the hotter melt is at the top. This minimizes convection and thus allows growth to be carried out at a steady temperature gradient. Minimization of temperature, concentration, fluctuations, and stable temperature gradients minimize compositional overcooling and promote uniform growth. This results in a high degree of crystal perfection and chemical homogeneity. This remarkable feature is attributed to the exceptional ability of HEM to grow nearly single crystal silicon ingots in one solidification using commercially available metallurgical grade silicon as the molten feedstock.

As the silicon ingot growth progresses, the size of the interface also increases. Therefore, a large-sized silicon ingot can achieve a high growth rate. The linear movement of the interface slows as the distance between the interface and the heat exchanger increases. However, the volume growth rate is still increased due to the large size of the interface. This feature is significant when directionally solidifying low purity molten feedstock by HEM. As growth proceeds, impurities are rejected into the liquid due to the segregation effect. However, this effect is reduced as the interface becomes progressively larger. As more and more impurities accumulate in the liquid, the slower linear growth inhibits the composition from overcooling.

In HEM, the stability of submerged solid-liquid interfaces is evident by the fact that growth continues around the particles while they are embedded on the interface without destroying the structure. The absence of a local high gradient at the interface ensures that growth from the interface takes precedence over growth from the grain. This is in contrast to the georchirosky (Czochralski) process, where such embedding causes spurious nucleation and thus polycrystalline growth.

The controllable heat exchanger of the HEM allows precise control of the temperature and temperature gradient at the bottom of the crucible. This precise control of the interface also allows for high growth rates at low temperature gradients. This reduces the solidification stress that causes deformation of the defect. Furthermore, since the ingot does not leave the hot zone during solidification, the ingot can be annealed in-situ after growth is complete. This is achieved by lowering the temperature of the furnace just below the freezing temperature and then reducing the helium flow. Thus, the entire ingot can be subjected to high temperatures and then uniformly cooled at a controlled rate. This further reduces internal stresses and eliminates a costly separate annealing step. This annealing and controlled cooling prevents cracks caused by thermal shock and thus allows the production of large silicon ingots.

The heat exchanger process or HEM is adapted to be constructed from the basic chemical formula Ba_2rMg_2(1-r)Si_1-xPb_x(1) Defined or defined by a more general, broader range of chemical formula Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_g(2) The defined composition is grown or produced as a single crystal or polycrystalline material. If this method is used, there is no need to vibrate or agitate the molten ingredients in the crucible. Nor does a changing temperature gradient be required. However, the following problems must be noted:

(1) precautions must be taken so that the melting of the essential components of the composition produced according to either of the two formulae described above in the crucible is carried out in an inert gas, preferably composed of helium or argon. When the composition is formed according to formula (1), this prevents magnesium from having magnesium due to its presence relative to the other three componentsHigher volatility and excessive loss, which prevents excessive loss of magnesium, selenium, tellurium and possibly reduction to a lesser extent of strontium, for similar reasons as described above, when the composition is produced according to formula (2) above, by maintaining the gaseous environment at a relative pressure preferably between 2 and 30 bar or 0.2 and 3MPa, wherein the crucible has been evacuated to preferably 10 prior to filling with an inert gas^-4To 10^-6Absolute pressure of mmhg.

(2) The crucible used to melt the basic ingredients of the composition should be made of a material that does not contaminate or react with the ingredients. Thus, the crucible should contain the materials described in the other embodiments of the invention described in the previous sections of the specification of the invention. For example, crucibles made of quartz or graphite should be completely excluded. They are absolutely not and not allowed for the production or preparation of the compositions proposed in the examples of the invention described in the previous part of the description of the invention.

The composition can still be prepared and produced using powder metallurgy techniques. The advantages of powder metallurgy over melt metallurgy are that excessive loss of magnesium, and possibly selenium, tellurium and strontium when one or more of these three elements are included in the composition, due to their relatively high volatility and high vapor pressure and the consequent difficulties in producing perfectly stoichiometric compositions and solid solutions, is avoided or overcome. Another advantage of powder metallurgy over melt metallurgy is that the alloy produced does not lose homogeneity when it contains elements with widely varying atomic masses or densities. In such circumstances, however, the melt metallurgical technology or process requires strong vibration or agitation of the molten components to ensure complete homogeneity of the solid solvent produced. When powder metallurgy technology is used for production and growth of Ba_2rMg_2(1-r)Si_1-xPb_xDefined or defined by a broader range of general chemical formulae Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_gWhere a composition is defined, any of the following alternatives may be taken or followed:

(1) the basic components, i.e. the elements, are mixed and melted together. The resulting solid solvent or alloy is then ground and milled, typically in a planetary ball mill. The powder thus obtained is subsequently subjected to hot pressing in uniaxial hot pressing, or to cold pressing, then to sintering;

(2) grinding and milling the basic components or constituent elements in a planetary ball mill, then hot pressing under uniaxial hot pressure, or cold pressing, then sintering without the need for initial melting; and

(3) each intermetallic compound is prepared by mixing and melting together each of the basic elements. The compound thus produced is subsequently ground and milled in a planetary ball mill and then hot pressed under uniaxial hot pressure or cold pressed and then sintered.

Regardless of which powder metallurgy process is selected, care must be taken that the grinding and milling of the components can only be performed once. This is necessary so that the degree of iron contamination or adulteration of the produced composition by the stainless steel grinding balls, which are usually derived from planetary ball mills, is reduced to an absolute minimum. This doping or contamination must be eliminated altogether. The method of accomplishing this elimination consists in making the grinding balls from a material that does not mechanically affect the composition of the composition ground with the planetary ball mill. For example, a much harder type of steel may be selected for the grinding ball. Special attention is given to the metallurgical composition or constitution and the necessary heat treatment and resulting microstructure to be able to solve this problem.

An alternative solution to avoid or eliminate contamination due to grinding is to select other materials than steel for the grinding balls. This step is not required when much harder steel is found or selected for the manufacture of the grinding balls. Otherwise, it would be absolutely necessary to select other materials that do not cause significant corrosion or wear due to mechanical effects on the components undergoing grinding.

If powder metallurgy techniques are used according to any of the three processes described above, the following is a worthwhile guideline:

(1) if the basic components (whether the starting elements or the intermetallic compound itself) are initially mixed and melted together, then uniaxial cold pressing and subsequent sintering are preferred;

(2) if the base components (whether the starting elements or the intermetallic compound itself) are not initially mixed and melted together, then it is appropriate to heat press the mixture under uniaxial hot pressure;

(3) to avoid further adulteration or contamination during the powder metallurgy practice (whichever of the two processes is employed), it is preferable to use a platinum cylinder and a platinum piston for hot or cold pressing of the milled or ground components followed by sintering;

(4) preferably, the powder metallurgy techniques, in particular sintering and hot pressing processes, should be carried out in an argon atmosphere. In other words, direct contact between the composition being produced and the air oxygen and moisture or, in general, air must be completely avoided in the implementation of powder metallurgy techniques. The same protective measures apply also in the long-term operation of thermoelectric energy conversion devices comprising a composite, i.e. a base material for producing the n-type, or n-type and p-type branches thereof. These measures are necessary in order to prevent the degradation of the set of thermoelectric properties of the composition that can certainly occur during the initial stages of the powder metallurgical manufacture and during the long-term use of the composition for the direct conversion of thermoelectric energy;

(5) for preparing a compound represented by the general formula Ba by mechanical alloying_2rMg_2(1-r)Si_1-xPb_xOr a broader range of the formula Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_gThe compound defined, the stoichiometric elements are charged in bulk (< 5 mm) in a vessel of about 500 ml capacity, preferably made of very hard special alloy steel or other suitable material, containing about 100 grinding balls, also preferably made of very hard special alloy steel or other suitable material, each ball having a diameter of about 10 mm, and 150 ml of n-hexane in the vessel. The vials were sealed under an argon atmosphere. Preferably, the milling or grinding process is carried out in a suitable planetary ball mill for 8 to 150 hours or any other suitable time period. The consolidation of the powder preferably corresponds to an absolute pressure p < 10^-4In a vacuum of mbar, at a uniaxial hot pressure, preferably at a pressure of 50Mpa, at a temperature preferably between 1073K and 1123K. Alternatively, consolidation of the powder may be carried out in an inert gas, preferably argon. Alternatively, solidification of the powder or ground ingredients may be carried out by cold pressing under uniaxial cold pressure, followed by sintering at a temperature preferably between 1073K and 1200K, preferably at a temperature corresponding to p.ltoreq.10^-4In vacuo or in an inert gas, preferably argon; and

(6) in order to further ensure that no contamination or undesired adulteration occurs during the comminution of the milled components, in particular Fe (iron), the container and the grinding balls used for this purpose as basic components of the planetary ball mill should comprise the same special alloy steel with a very high hardness. If this proves to be infeasible, other materials with sufficiently high hardness must be sought or selected for this purpose. In other words, the steel alloy currently used in the manufacture of the aforementioned containers and grinding balls must be replaced by a more rigid material, whether it be another steel alloy or a completely different material.

When used for thermoelectric energy conversion, with magnesium silicide Mg₂Recent experimental investigators relating to the preparation of Si, temperature dependence of Seebeck (Seebeck) coefficient, resistivity and thermoelectric power factor, and long-term performance reliabilityShow that:

(1) mg prepared by powder metallurgy techniques, i.e. by cold pressing and subsequent sintering in argon at a temperature in the range 1073K to 1200K₂The thermoelectric properties of samples of Si are much better than those of samples prepared by melting, which are also exposed to atmospheric oxygen for different periods of time. In other words, Mg is produced by a conventional method of casting or melt metallurgy₂Si, by exposing it to atmospheric oxygen, which results in obtaining a material with greatly deteriorated Seebeck (Seebeck) coefficient, resistivity and thermoelectric power factor; and

(2) the thermoelectric properties of the samples initially prepared by cold pressing and subsequent sintering in argon deteriorate significantly after the samples are exposed to air for various periods of time due to sublimation and oxidation of magnesium.

Thus, it is necessary to separate from air, i.e. to prepare and use magnesium silicide Mg in an inert gas, preferably argon₂And (3) Si. Furthermore, powder metallurgy techniques, whether cold pressing and subsequent sintering or hot pressing, are much better than traditional melt metallurgy methods for preparing or producing compounds. As previously mentioned, this also depends on whether the above-mentioned compounds are remote from air or atmospheric oxygen. This means that the preparation and use of magnesium silicide Mg must be carried out in absolute vacuum or in an environment preferably consisting of argon₂Si。

For a chemical formula of Ba_2rMg_2(1-r)Si_1-xPb_xThe arguments and facts set forth in the two paragraphs above apply equally to the defined composition.

Although in the above formula a portion of the magnesium is replaced by barium and a portion of the silicon is replaced by lead, the composition is essentially Mg silicide₂And Si. Accordingly, all of the above mentioned references to magnesium silicide Mg₂The description and measures of the preparation of Si, the temperature dependence of the Seebeck (Seebeck) coefficient, and the long-term performance reliability can be applied virtually equally to the composition defined by the above basic formula. Also, it is entirely possible to use the sameThe description and measures extend to cover or are substantially equally applicable to a device defined by the broader general formula Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_gIn the compositions defined.

Alloy scattering can be used as a powerful method to reduce the lattice thermal conductivity of thermoelectric materials. This leads to an increase in the thermoelectric figure of merit for these materials, since for semiconductors the lattice thermal conductivity is very close to the overall thermal conductivity at relatively low temperatures. Therefore, the most useful thermoelectric materials are alloys or solid solutions because their lattice thermal conductivity is reduced due to alloy scattering. But at the same time, alloying or mixing also generally reduces electrical mobility and conductivity. But the formation of alloys or solid solutions is successful for thermoelectric materials because generally the reduction in lattice thermal conductivity is significantly more than the reduction in electrical conductivity. However, for the thermoelectric power factor S²Electrical properties per se, for example σ, generally speaking, pure materials, either elemental or compound, are much better than alloys or solid solutions.

Optimizing the thermoelectric figure of merit for any material is a very complex and difficult task. With particular reference to semiconductors, two basic and most practical ways to do this are by doping with foreign impurities and forming alloys or solid solutions. The only practical way to control the thermoelectric power or Seebeck coefficient S is to vary the free charge carrier concentration. This means that the doping level is changed. Thus, increasing the doping level causes a decrease in the Seebeck (Seebeck) coefficient and vice versa. The exact opposite is the conductivity. Increasing the doping level increases the number of free charge carriers, either electrons or holes, which increases the conductivity. In terms of thermal conductivity or heat flow, we must recognize that heat is conducted by phonons and electrons. Therefore, the thermal conductivity must contain two components: a lattice or phonon component, and an electron component. In fact, the contribution of electrons to thermal conductivity is approximately the same as the contribution to electrical conductivityAnd (4) in proportion. This proportionality between electrical and thermal conduction due to charge carriers or electrons is known as the Vildman-Franz (Wiedemann-Franz) law. Electron component k of thermal conductivity_e1The proportionality coefficient with the conductivity σ is called the lorentz (Lorenz) number L. This law is very important to theoretical solid physicists. The key here is that the above law, although originally derived or established for metals, is still applicable to semiconductors or any other material used for this purpose. This applicability is valid or correct, bearing in mind the fact that the thermal conductivity of non-metallic materials is composed of an electronic component and a lattice or phonon component. Thus, the total thermal conductivity of a semiconductor material can be expressed as:

k＝k_{lattice of the product}+k_{Electronic device}＝k_{Lattice of the product}+σLT

Where T is the absolute or thermodynamic temperature in kelvin.

In general, the Seebeck (Seebeck) coefficient and the lattice or phonon component of thermal conductivity decrease with increasing doping levels, which increase the number of free charge carriers. On the other hand, as the doping level increases, the electron components of the electrical and thermal conductivities increase. Thus, the optimum doping level is the value at which the thermoelectric figure of merit is at its maximum, i.e., between 10 per cubic centimeter¹⁹To 10²⁰The range of carriers.

When an alloy or a solid solvent is formed using two or more semiconductor elements or compounds, the following effects generally occur because the Seebeck (Seebeck) coefficient is small in variation with the composition of the alloy. This is especially true for semiconductors, but not for metals. Furthermore, due to alloy scattering, both the electrical and thermal conductivities will be less than the simple linear average of the electrical and thermal conductivities corresponding to the two or more constituents of the alloy. In fact, alloy scattering tends to affect thermal conductivity, especially the lattice component that affects thermal conductivity, much more significantly than electrical conductivity. In fact, the thermal conductivity resulting from mixing two or more semiconductors together is determined only by those components or fractions having the greatest difference in atomic mass and atomic volume (covalent bond volume). Therefore, the thermal conductivity obtains a certain minimum value at some intermediate composition between x-0 and x-1, and is generally much smaller than the thermal conductivity corresponding to either of x-0 and x-1.

It is generally possible to determine the lattice thermal conductivity obtained by mixing or alloying any two semiconductors together. This is based on the theory that clemens (p.g. klemens) originally developed in 1955, although this theory is commonly referred to as the kalavir (Callaway) theory. When point defects scatter phonons, primarily due to their mass difference, professor Klemens derives the following formula resulting in a change in lattice thermal conductivity:

where k is the lattice thermal conductivity, ω, due to scattering of point defects. Is the phonon vibration angular frequency at which the mean free path of point defect scattering is equal to the mean free path of intrinsic scattering,is the phonon vibration Debye frequency, which is equal to k theta_DK is Boltzmann (Boltzmann) constant and v is the speed of sound or phonon. For pure or unalloyed semiconductors, without point defects, the intrinsic lattice thermal conductivity can be defined as:

wherein B is a constant. The formula can be finally written as:

in the extreme case of scattering of strong point defects,

and therefore the number of the first and second electrodes,

based on the work of kalapart (Callaway) and beijer (Von Baeyer), boschkiski (Borshchevsky), caja (cailat) and fleriell (fleuri) can put the theoretical results of the above-mentioned cleimers (p.g. klemens) into the following formula, i.e. a formula more generally applicable to the actual calculations:

wherein the content of the first and second substances,

and the number of the first and second groups,

wherein, theta_DIs the Debye temperature, delta³Is the average volume per atom in the crystal, v_sIs the average acoustic or phonon velocity, h is the Planck (Planck) constant, u is the alloy scattering ratio parameter, and Γ is the alloy scattering parameter. The above formula applies to all types of alloys or solid solutions, especially those involving chemical or intermetallic compounds. The speed of sound v is preferably obtained by direct measurement_s. The scattering parameter includes a quality fluctuation term gamma_A ^MAnd the volume fluctuation term Γ_A ^VThey are defined as follows:

wherein the content of the first and second substances,

is the relative proportion of each atom in a particular position a,

ε_{A，B，C，D}is an adjustable strain parameter that is,

m is the total average mass of the alloy

p_iIs the atomic ratio of A atoms in the compound

Wherein A is_aB_bC_cD_dIs a chemical formula of a specific alloy or a solid solution agent, A, B, C and D represent each element.

Total alloy scattering parameter is given byAnd (4) defining.

The above theoretical analysis is used to indicate that mass and volume fluctuations must be maximized in order to minimize the lattice thermal conductivity of the alloy or solid solution. However, there is no way to control these mass and volume fluctuations individually. They are also determined by the characteristics of the elements chosen to constitute the alloy or the solid solution, which we are now concerned with or trying to develop. Furthermore, mass fluctuations or mass differences between elements at each position have a greater effect on reducing the thermal conductivity of the crystal lattice than volume fluctuations or volume differences. Furthermore, the mass fluctuation parameter can generally be calculated more accurately than the volume fluctuation parameter. This is because an accurate value of the strain parameter epsilon is required so that the volume fluctuation parameter can be accurately determined. Since reliable data for the strain parameter ε is generally not available, particularly for such novel materials as the alloys or solid solutions that comprise embodiments of the present invention, experimental measurements on samples of these alloys or solid solutions are needed. Moreover, an additional mechanism, namely phonon-electron interactions or phonon scattering by charge carriers, can cause a further reduction in lattice thermal conductivity. This additional scattering is very pronounced or significant, especially in highly doped n-type semiconductor materials, i.e. those having a 1 x 10 per cubic centimeter¹⁹To 5X 10²⁰A material with a free charge carrier concentration in the carrier range.

In order to develop or select the desired thermoelectric material, i.e., the material with the highest possible thermoelectric figure of merit, the thermal conductivity should be minimized. For example, in a thermoelectric power generation pair or thermocouple, high thermal conductivity means that heat will be transferred or short-circuited directly from the hot connection point to the cold connection point without being converted to electrical energy. In the previous description of this specification relating to the use of the "heavy element selection criteria" of the order of costs (a.f. ioffe) to minimize the thermal conductivity of the ideal thermoelectric material we expect, the minimization is by selecting bismuth or lead to make up the ideal thermoelectric material. Since these two elements have nearly the same atomic mass, Pb is 207.2 and Bi is 208.98, the chances of both being chosen to make up the composition are the same. On the other hand, Bi has a lower thermal conductivity than Pb, while their melting points are almost the same.

Since the materials being developed are semiconductor in nature, the second option is silicon. In practice, silicon and germanium are the most semiconducting elements throughout the periodic table. However, since silicon is also classified as a nonmetal or a semi-metal, silicon has an advantage over germanium. This is confirmed by the fact that the electrical conductivity of silicon is 2.52X 10 at 20 deg.C^-6(ohm cm)^-1And the conductivity of germanium is 1.45X 10^-2(ohm cm)^-1. This gives preference to silicon over germanium.

Since the thermal conductivity of silicon is relatively high, i.e., about 1.49Wcm at room temperature^-1K^-1And therefore must be reduced or minimized as much as possible. One way of doing this is to alloy silicon and magnesium, or more precisely to form a chemical compound between silicon and magnesium. This results in the formation of the compound magnesium silicide Mg₂Si, its thermal conductivity is about 0.08Wcm at room temperature^-1K^-1. Thus, by reacting magnesium and silicon, the thermal conductivity of silicon is reduced by a factor of about 19, which is quite significant, without any serious degradation of the silicon's excellent semiconductor properties, in particular its high thermoelectric power. Thus, magnesium Mg is the third choice.

The three elements selected, namely bismuth or lead, silicon and magnesium, are the basic constituents of the composition. It is now further necessary to select the fourth element in order to complete the invention. Since the main objective is to minimize thermal conductivity, the fourth element is chosen to have a sufficient reduction in thermal conductivity due to its "alloy scattering" interaction with magnesium. For maximum effect, the element should preferably have the same electronic structure, i.e. belong to the same group of elements as magnesium. We therefore focused attention on group IIA, which includes beryllium, calcium, strontium, barium, and radium, in addition to magnesium. Again, applying the "heavy element selection criteria," radium was selected because it has the highest atomic mass among all elements of group IIA, 226. However, the use of radium must be excluded because it has a high radioactivity. Leaving only four elements to choose from, specifically beryllium, calcium, strontium, and barium. Since barium has the highest atomic mass of these four elements, barium is selected as the fourth, i.e., last, element. This ensures that the interaction of "alloy scattering" or more correctly speaking "mass and volume fluctuation scattering" will be the largest possible between Mg and Ba. This is because the expected lattice thermal conductivity of the composition may be minimal or minimal.

Returning to the first choice of elements that make up the intended composition, lead or bismuth may be chosen. Since the "mass and volume fluctuation scattering" between Si and Pb or Bi is almost the same, the decisive criterion or factor is the similarity of the electronic structure between silicon and these two elements. This results in a preferred Pb because lead and silicon belong to the same column or group of the periodic Table of the elements, i.e., group IVB, and bismuth belongs to group VB. It therefore has an electronic structure different from that of silicon. Thus, bismuth is expelled or eliminated and the first element selected is lead Pb. Thus, there are four elements: lead, silicon, magnesium and barium to exactly constitute the composition. This is the basic embodiment of the present invention. Looking at the periodic table of elements, it can be seen that these four elements occupy the four corners of a rectangle. As described above, the lattice thermal conductivity of the composition is reduced by the double interaction effect, i.e., "mass and volume fluctuation scattering" between silicon atoms and lead atoms, and another "mass and volume fluctuation scattering" between magnesium atoms and barium atoms. This double or dual "mass and volume fluctuation scattering" results in a significant reduction in the lattice thermal conductivity of the resulting composition. This can be clearly appreciated by observing the following table:

from the above table we can conclude that:

(1) there is very strong mass fluctuation scattering between Mg atoms and Ba atoms and Si atoms and Pb atoms. This is because of the large difference in atomic mass between Mg and Ba and between Si and Pb.

(2) There is a certain amount of volume fluctuation scattering between Mg atoms and Ba atoms and between Si atoms and Pb atoms. This is because of the difference in atomic radius and atomic volume between Mg and Ba and between Si and Pb.

(3) Because Mg and Ba occupy a dominant electronegativity, they are susceptible to chemical reactions with Si and Pb, respectively, and to formation of compounds. Thus, the composition will consist of an alloy or solid solution comprising an intermetallic compound of magnesium silicide, magnesium plumbate, barium silicide and barium plumbate.

(4) Mg and Ba and Si and Pb will be less likely to form chemical compounds because the difference in electronegativity between them is significantly lower compared to the differences in electronegativity between Mg and Si, between Mg and Pb, between Ba and Si and between Ba and Pb.

The composition thus obtained is therefore defined by the following general chemical formula:

Ba_2rMg_2(1-r)Si_1-xPb_x

as is clear from the above formula, the composition consists essentially of magnesium silicide Mg₂Si, in which part of the magnesium is replaced by barium and part of the silicon is replaced by lead. This is evident in order to significantly reduce or minimize the thermal conductivity of the composition, particularly the lattice thermal conductivity. The composition so produced should have the lowest or smallest possible lattice thermal conductivity. The overall thermal conductivity is also minimal. On the other hand, the thermoelectric power factor S²σ should be maximal. This can be achieved by finely doping the composition with appropriate known atoms or impurities in appropriate amounts. The dopant or impurity may beIncluding one or more elements and/or compounds thereof. Incorporation of dopants or impurities in the composition is effected by obtaining a charge carrier concentration of 1X 10 per cubic centimeter¹⁵To 5X 10²⁰A carrier. The atomic or molecular proportion of dopants or impurities may be approximately 10^-8To 10^-1Within the range of (1). The lower limit of the free charge carrier concentration and the lower limit of the atomic or molecular proportion of the dopant mentioned above relate in practice to the limit case in which the composition is substantially undoped. In practice, the composition may be at least lightly or moderately doped, which corresponds to 1 × 10 per cubic centimeter¹⁸To 1X 10¹⁹Free charge carrier concentration of carriers. This generally results in a significant increase in electrical conductivity and hopefully increases the thermoelectric power factor and correspondingly the thermoelectric figure of merit. Preferably, heavy doping is also performed if the thermoelectric power or Seebeck (Seebeck) coefficient is not seriously deteriorated. This means that the free charge carrier concentration can be kept at 1 x 10 per cubic centimeter¹⁹To 5X 10²⁰A carrier. This will result exactly in the thermoelectric power factor S²Sigma maximization, which, together with the aforementioned minimization of thermal conductivity in the present specification, undoubtedly leads to a maximization of the thermoelectric figure of merit. Thus, the free charge carrier concentration in the composition is preferably at 1 × 10 per cubic centimeter¹⁸To 5X 10²⁰The atomic or molecular ratio of the corresponding dopant or impurity in the carrier range is preferably 10^-5To 10^-1Within the range of (1).

It should be noted that all of the above analyses relate to thermoelectric properties and characteristics of the composition at relatively low temperatures (i.e., not above room temperature). It should also be emphasized that the composition, even undoped, tends to behave as n-type at low temperatures. As the temperature of the material increases, the charge carrier concentration will increase due to thermal activation and the n-type characteristics will become more pronounced. For example, for Mg which is preferably undoped₂Si samples, preferably prepared by powder metallurgy techniques involving uniaxial cold pressing and subsequent sintering without exposure to atmospheric oxygen, reach a maximum or maximum level as the temperature rises from about 300K, i.e.About 800K, the thermoelectric power and thermoelectric power factor increase significantly. This indicates that no doping at all is required in the fabrication or production of the n-type thermoelectric elements or branches of the thermoelectric devices composed of the above-described composition. The n-type doping of the composition is still optional, only if necessary. This is particularly true for operating temperatures much higher than room temperature.

This is not the case when the composition is used to form a p-branch or thermoelectric element of a thermoelectric energy conversion device. Doping with an acceptor or p-type impurity or dopant is certainly necessary for the production of such p-type materials. The method of doing so has been described clearly in the preceding part of the specification and in the corresponding embodiments in the preceding paragraphs. In general, the production of p-type thermoelectric materials is much more complex than the production of n-type thermoelectric materials. This is particularly true for materials composed of several elements with significantly different atomic masses and atomic volumes, such as the magnesium-barium silicon lead alloy or solid solvent to which we are referring, which constitute the basic embodiment of the present invention. They all tend to be n-type and as the temperature rises, rising much higher than room temperature, this tendency becomes more and more pronounced and stronger. Also, p-type materials generally have poorer thermoelectric properties than n-type materials. This is because the mobility of holes is generally smaller than that of electrons. Careful attention to the doping process or method may help alleviate these problems. This situation is further improved if the p-type composition is not used at all for power generation purposes, but in thermoelectric heat pumps and thermoelectric refrigeration devices operating at much lower temperatures. In thermoelectric power generation devices, additional attention is given to doping techniques or methods such as the type of doping material, the doping level to be used, and the use of FGM or functionally graded material techniques described in the previous embodiments of the present description, which will help to ameliorate this situation. If there is still a problem with p-type compositions, i.e. with regard to their thermoelectric properties or the possibility of production, and in particular with regard to their p-type behaviour being retained at high temperatures, it is possible to replace the p-type compositions by passive Golds (Goldsmid) branches consisting of high-critical-temperature superconductors. In this case, the composition defined in the present specification may be used to manufacture only an n-type branch or a thermoelectric element of a device for controlling thermoelectric energy conversion. Thus, in an expected ideal thermoelectric device comprising an n-type branch or element composed of a composite, and instead of a p-type branch passive Golds (Goldsmid) branch or element, the overall performance of the device is determined entirely by the performance of the n-type branch. In practice, the passive branch is only used to complete or close the circuit. It does not contribute to any increase or decrease in the thermoelectric performance or energy conversion efficiency of the device. However, it is indirect because it helps us avoid the use of poorly performing p-type branches that can cause some degradation in thermoelectric performance and thermoelectric conversion efficiency.

Alternative embodiments of the invention are also based on the compound magnesium silicide Mg₂Si, except that a part of magnesium is replaced with at least one element selected from the group consisting of four elements of beryllium, calcium, strontium and barium, and a part of silicon is replaced with at least one element selected from the group consisting of seven elements of germanium, tin, lead, antimony, bismuth, selenium and tellurium. Alternative compositions so produced have the following general chemical formula:

Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_g (1)

it is emphasized that the compound represented by the general formula Ba_2rMg_2(1-r)Si_1-xPb_x(2) The basic or central embodiment of the invention as defined is only one particular case of the above general broader general formula (1), i.e. u, v, w, a, b, d, e, f, g are all set to 0. Comparing these two formulas, the following comments are noteworthy:

(1) the alloy or solid solution prepared according to the basic example or general formula (2) has the absolute minimum or lowest possible thermal conductivity, in particular its lattice component.

(2) Alloys or solid solutions prepared according to alternative examples or general formula (1) tend to have higher thermal conductivities than alloys or solid solutions prepared according to general formula (2). However, formula (1) helps to obtain alternative compositions with higher average band gap and melting temperature, which is desirable in high temperature applications.

(3) Despite the statements made in item (2) above, barium and lead should still be present in the composition, i.e. in a lesser amount or proportion, so as not to deviate too much from the minimum lattice thermal conductivity or to be much higher than it.

(4) Bismuth can still be used as a partial substitute for silicon according to the above formula (1). However, it is less effective than lead in reducing the lattice thermal conductivity of the resulting composition to an absolute minimum. This is because it is less compatible with silicon than lead because bismuth and silicon have different electronic structures because they belong to different groups of the periodic table. However, bismuth can improve the thermoelectric power factor of the composition. Thus, bismuth may be used in place of silicon in part, but preferably is replaced not by bismuth alone, but by a compound of bismuth and lead. This is preferable so as not to deviate from the minimum lattice thermal conductivity at all.

(5) The above general formula (1) represents an alternative broader embodiment of the present invention, and it is not specified that the atomic ratio of any other element other than bismuth and lead for partially substituting magnesium and silicon cannot be set to 0.

(6) Regardless of the analyses, the above general formula (2) still represents the basic central embodiment of the present invention.

In 1957, the following standards or principles for obtaining good thermoelectric materials with high quality factor and high energy conversion efficiency have been established for standard strip-type semiconductors for the cost of approx (a.f. ioffe):

(1) the ratio of charge carrier mobility to lattice thermal conductivity must be maximized. Since the mobility of electrons and holes is always susceptible to degradation when the material is not single crystalline, and when the temperature is much higher than room temperature, when several compounds are mixed together to form an alloy or a solid solution, the only way to achieve a maximum ratio of charge carrier mobility to lattice thermal conductivity is to drastically reduce the lattice thermal conductivity.

(2) Forbidden band gap E_gMust be greater than 4k_BT_intWherein k is_BIs the Boltzmann constant, T_intIs the inherent or maximum hot junction operating temperature expressed in degrees kelvin. Let T be_intIs 800 ℃ or 1073K, then E_g0.73eV (electron volts).

(3) The semiconductor must be dopable to foreign substances.

The above criteria are further elaborated by Pierre Aigran, who places them in the following definite more realistic form: for near room temperature heat sinks (junctions) to work, a good thermoelectric engine (power generation device) should utilize materials with the following characteristics:

(1) the temperature of the working heat source (connection point) is about 700-800 ℃.

(2) A solid solvent.

(3) If possible, an anisotropic material.

(4) Band gap E_gOn the order of 0.6 eV.

Remember Mg₂The band gap of Si is about 0.78eV, then, for example, according to Ba_0.4Mg_1.6Si_0.85Pb_0.15The prepared alloy or solid solution can approximate the average band gap, assuming a linear relationship between the band gaps of the respective compounds and their atomic or molecular ratios. The calculation thus made revealed that the average bandgap of this alloy was about 0.63 eV. This shows that the compositions associated with embodiments of the present invention substantially meet the Pierre Aigrain criterion (4), which requires an energy bandgap of about 0.6 eV. Basic embodiment and broader aspects of the inventionMore general alternative embodiments of the scope also substantially satisfy the above criteria (1) and (2). Criterion (3) can only be met when the material is monocrystalline, since it is known that monocrystalline is anisotropic. If the composition is produced as a single crystal, then any advantages of anisotropy should be fully exploited.

In summary, it should be emphasized that a primary desire and objective of the present invention is to develop or produce compositions or materials having greatly reduced or extremely low lattice thermal conductivity. This is achieved as follows:

(1) silicon is selected as one of the basic constituents of the composition. Thus, silicon is our first element.

(2) Reacting silicon with magnesium to form magnesium silicide Mg₂And (3) Si. Thus, magnesium is the second element.

(3) Part of the silicon was replaced with lead. Thus, lead is the third element.

(4) Part of the magnesium was replaced by barium. Thus, barium is the fourth element.

Each of the four steps described above results in a significant reduction in the lattice thermal conductivity of the resulting composition, alloy or solid solution. Of course, this also results in the composition of the basic embodiment representing the invention being represented by the general formula Ba_2rMg_2(1-r)Si_1-xPb_xDefined, and has extremely low lattice thermal conductivity. The lattice thermal conductivity is not exactly equal to 0, but should be very close to 0. This makes the present invention a central objective.

Furthermore, it is also desirable that the thermoelectric power factor PF of the composition is very high. This is based on the fact that:

(1) undoped n-type magnesium silicide Mg prepared by powder metallurgy techniques involving cold pressing in platinum cylinders and subsequent sintering in argon at a temperature in the range 1073K to 1200K₂The thermoelectric power of the Si samples has been experimentally measured to be about 230 μ VK at about 330K^-1And rises to about 1000 μ VK over a temperature range of 700K to 800K^-1. The maximum was found at 760K. The experimentally measured power factor range for the same sample was 0.3X 10 at about 330K^-3Wm^-1K^-2And about 5.4X 10 to about 760K^-3Wm^-1K^-2Is measured.

(2) Therefore, the thermoelectric figure of merit of the magnesium silicide samples prepared as above is quite high. At higher temperatures of 760K, the value may be based on a known value of its thermal conductivity, i.e., about 0.08Wcm^-1K^-1Calculated to obtain the thermoelectric quality factor of 5.4 multiplied by 10^-3×10^-2/0.08＝6.75×10^-4K^-1。

(3) Comprises a material formed by Ba_2rMg_2(1-r)Si_1-xPb_xThe basic embodiment of the invention of the defined composition should have a magnesium silicide Mg₂Si the same or even higher thermoelectric power. This is confirmed by the fact that it is known to have a band gap E as one component of the composition_g0.48eV or more properly barium disilicide MgSi₂Having a thermoelectric power S of 600 mu VK at room temperature^-1. At the same temperature, this is compared to pure Mg²Si is much higher. Thus, the above compound BaSi₂The relatively high thermoelectric power values result in a noticeable increase in the total thermoelectric power of the composition. This is further evidenced by the fact that semiconductors whose valence and/or conduction bands are dominated by the d-band characteristic can combine high Seebeck (Seebeck) or thermoelectric power values typical of transition metal elements with the ability to achieve optimal doping levels typical of conventional thermoelectric materials. Some metal silicon compounds appear to have this combination of properties, barium silicide BaSi₂And more suitably barium disilicide, is the defining one.

(4) As described in (1) above for magnesium silicide Mg₂The conductivity of a composition prepared as described for Si can be increased by lightly, moderately or highly doping with suitable hetero atoms as dopants. However, extreme care must be taken to dope to ensure that thermoelectric power S is not thereby generatedBut deteriorates or is adversely affected as the temperature rises much higher than room temperature.

(5) By a broader range of alternative general formulae Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_gCompositions are prepared to further improve thermoelectric power factors in which Ba and Pb are still present in relatively high atomic proportions, not less than about 80% of the appropriate or required stoichiometry, and part of the Ba and/or part of the Pb may be replaced by one or more elements indicated in the above formula. This can result in some increase in thermoelectric power, thermoelectric power factor, average band gap, and average melting temperature of the resulting composition. However, a significant reduction in the atomic proportion of Ba and/or Pb in the composition will certainly cause a corresponding increase in lattice thermal conductivity. This must be avoided absolutely if possible. Therefore, Ba and/or Pb must be finally, even partially, replaced by another element or elements to an absolute minimum so as not to adversely affect the lattice thermal conductivity. It is desirable to absolutely avoid the substitution of barium and lead by any element, which will result in or ensure the absolutely lowest lattice thermal conductivity, and the absolutely lowest total thermal conductivity. This is why magnesium silicon lead barium alloy or solid solution constitutes the basic or central embodiment of the invention.

Thermoelectric figure of merit for the above composition:can be calculated from:

(1) according to the basic formula Ba_2rMg_2(1-r)Si_1-xPb_xA composition is prepared wherein r is in the range of 0.1 to 0.4 and x is in the range of 0.1 to 0.3, and should have a minimum total thermal conductivity k-total 0.002Wcm at approximately room temperature^-1K^-1. It can be assumed that it is approximately equal to the minimum lattice thermal conductivity.

(2) Since the composition as defined by the above basic formula consists essentially of magnesium silicide Mg₂Si, which we assume has the same thermoelectric power and power factor as magnesium silicide. Experimentally measured Mg₂The power factor of Si at 760K is PF ═ S²σ＝5.4×10^-3Wm^-1K^-2. Therefore, the thermoelectric figure of merit is Z-5.4 × 10^-3/(0.002×10²)＝2.7×10^-2K^-1The dimensionless thermoelectric figure of merit is ZT 2.7X 10^-2×760＝20.52。

Due to the best thermoelectric materials known or used today, e.g. Si_0.7Ge_0.3Mg₂Si_xSn_1-xIt cannot far exceed ZT ═ 1, so the ZT values of the above compositions represent a breakthrough in thermoelectrics.

Current analysis cannot be done without calculating the thermoelectric energy conversion efficiency of devices containing the above-described compositions. The following equation, known in the art, can be used to calculate efficiency:

wherein:

wherein S is²σ depends on the electronic properties, and if the electronic thermal conductivity is neglected, k almost completely depends on the lattice properties. By substituting equations (2) and (3) into equation (1) and performing mathematical calculations, the following results can be obtained:

for the limiting case, when T_cApproach to T_hThen, we get:

and

then obtaining:

the first term between the small brackets on the right side of equation (7) is the maximum temperature T according to the second law of thermodynamics_hAnd a minimum temperature T_cThe energy conversion efficiency of the heat engine. This is also known in the art as Carnot (Carnot) efficiency. It is assumed that a thermoelectric energy conversion device, i.e. a power generator, operates between a hot junction temperature of 800K and a cold junction temperature of 300K, including a q-factor Z of 2.7 × 10^-2K^-1And assuming that the above value of Z remains at 800K instead of 760K, which is approximately correct, we can obtain:

the above energy conversion efficiency index can be preferably compared with the energy conversion efficiency of the best conventional power plant including a well-designed boiler, steam turbine and gas turbine and diesel engine used today. If we replace 800K with the average temperature between the hot and cold junctions, i.e. 500K, in the above formula, the energy conversion efficiency index can be calculated more accurately. Thus, a more accurate calculation using equation (4) can be obtained:

this is a higher thermoelectric conversion efficiency than the previous value of 0.408.

In summary, the embodiments of the present invention are as follows:

the basic embodiment of the present invention is a composition comprising magnesium, silicon, lead and barium. The composition can be used to make p-type and/or n-type thermoelectric elements or branches, including positive branches, negative branches, hot junctions, and cold junctions, of devices that control thermoelectric energy conversion. Since magnesium and barium chemically react with each of silicon and lead, respectively, and form a compound, the composition can be considered as an alloy or a solid solution containing an intermetallic compound of magnesium silicide, magnesium plumbate, barium silicide, and barium plumbate. Since the alloy or solution contains the stoichiometric atomic ratio of magnesium and siliconSubstantially 2:1, and thus the composition can be considered to consist essentially of magnesium silicide Mg₂Si, in which part of the magnesium is replaced by barium and part of the silicon is replaced by lead. Thus, represented by the following general formula Ba_2rMg_2(1-r)Si_1-xPb_xTo define the composition.

Alternative or broader embodiments of the invention are essentially based on the above general formula except that a portion of the magnesium is replaced by one or more elements selected from the group consisting of beryllium, calcium, strontium and barium, and a portion of the silicon is replaced by one or more elements selected from the group consisting of germanium, tin, lead, antimony, bismuth, selenium and tellurium, wherein the resulting composition is represented by the general formula Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_gWherein r is u + v + w + z, s is a + b + c + d + e + f + g.

The composition can be used in the undoped state, i.e. without any dopant or impurity added, whether defined by the basic formula containing only four elements or by an alternative broader formula containing 13 elements. However, doping can still be performed if necessary. This means that the addition of doping materials or impurities to the composition is only optional and is only carried out when required. Doping is a very delicate and complex problem and therefore requires great care. This is absolutely necessary if the thermoelectric properties and performance of the composition are to be optimized. It is clear that the above description is directed to n-type doping. Generally, semiconductors are typically doped to maximize thermoelectric figure of merit, especially when used in thermoelectric materials. The composition of the present invention is a semiconductor and is therefore no exception. However, for magnesium silicide Mg prepared by powder metallurgy technique₂Recent experimental work carried out on Si samples resulted in extremely high power factors, i.e. 5.4 x 10^-3Wm^-1K^-1When used in the composition and having a very low thermal conductivity therewith, i.e., 0.002Wcm^-1K^-1When combined, giveGive out Z ═ 2.7X 10^-2K^-1And ZT at 760K is 20.5. This sample is an n-type doped sample, making the above results for Z and ZT more specific.

The point is that these particular results can be achieved without doping any substance. However, since we are concerned here with an entirely new composition, i.e. magnesium silicon lead barium alloy or a solid solution, we cannot absolutely determine whether doping is required. The composition may still require slight or moderate doping. However, the above experimental evidence for pure magnesium silicide leads to an initial conclusion that n-type doping may not be required, especially for high temperature applications. In this regard, any definitive conclusions must be based on specific experimental evidence relating to the composition itself. Regardless of the method of preparation or production of the composition, the composition always tends to be n-type even when undoped. Furthermore, this trend towards n-type behavior is more and more pronounced with higher and higher temperatures, up to far above room temperature. This again confirms the initial conclusion that n-type doping may not be required for the composition, especially for those applications where the operating temperature is relatively high. However, the production or preparation of p-type compositions requires doping with p-type hetero atoms or impurities, which are generally classified as acceptors. In this case, a high doping with p-type doping elements and/or compounds thereof is required or suggested. Due to the illustrated tendency of the composition towards n-type behavior, the preparation of p-type materials is generally more difficult than the preparation of n-type materials, even without doping anything. This is why a high degree of doping is absolutely necessary in order to obtain a p-type composition. This is even more pronounced as the temperature rises much higher than the room temperature level of 298K, in order to prevent the p-type material from transforming into the n-type behavior.

For the reasons mentioned above, and to ensure more efficient doping, it is generally preferred to select the doping element from the group located to the left of the group containing Be, Mg, Ca, Sr and Ba (i.e., group IA). Thus, the elements that can be selected as preferred p-type doping materials or dopants are lithium, sodium, potassium, rubidium, cesium and francium. Lithium is expelled out of the acceptor or p-type dopant because it actually behaves as an n-type feature, i.e., it behaves as a donor element. The reason for this is that lithium atoms, because of their relatively small size, fill the gaps between the host atoms of the composition rather than being replaced. Also, francium is undesirable because it is unstable and radioactive. Thus, the four elements that can be selected for the composition as the most effective of the recommended p-type doping materials or acceptor impurities are sodium, potassium, rubidium, and cesium. Since these elements are highly electropositive, when they are introduced into the host material or composition, a vigorous chemical reaction occurs. If this is inconvenient, these elements, i.e. Na, K, Rb and/or Cs, can be reacted or formed into compounds with other elements, preferably silicon and/or lead. In this case, one or more of these elements may serve as a partial substitute for magnesium and/or barium. The chemical formula of the resulting composition is thus:

Na_2uK_2vRb_2wCs_2yBa_2zMg_2(1-r)Si_1-xPb_x

wherein r is u + v + w + y + z, which represents the sum of the atomic proportions of the elements replacing part of the magnesium, wherein r ranges from 0.1 to 0.4 and (u + v + w + y) ranges from 10^-8To 10^-1Wherein u, v, w and y are all from 0 to 0.1, z is not less than 0.1, and x ranges from 0.1 to 0.3. It is assumed that the above p-type doping elements, i.e. Na, K, Rb and Cs, form chemical compounds with Si and/or Pb in the correct stoichiometric ratio of 2: 1. Specifically, in order to make the above general formula correct, the above four doping elements must be formed such as Na₂Si、K₂Si、Rb₂Si and Cs₂A compound of Si. However, the compounds which may be naturally formed are NaSi, KSi, and the like in practice. Thus, for example, a mixing should take place between the natural compounds NaSi and Na, so that the final product corresponds to Na₂And (3) Si. For example, suppose Mg is in magnesium silicide₂In the Si compound, only Na is used instead of Mg, then:

rNaSi+rNa+(1-r)Mg₂Si＝Na_2rMg_2(1-r)Si

and for alloys or solid solutions: mg (magnesium)₂Si_1-xPb_xAnd then:

r(NaSi)_1-x+rNa_1-x+(1-r)Mg₂Si_1-x+r(NaPb)_x+rNa_x+(1-r)Mg₂Pb_x＝Na_2rMg_2(1-r)Si_(1-x)Pb_x

this gives a general idea or hint as to how to dope any other doping element. For elements or compounds used for n-type doping, a similar approach can be used. Thus, doping with one or more elements of Na, K, Rb and/or Cs, whether in pure elemental form or as a compound with other elements, preferably Si and/or Pb, as described above, will ensure more efficient and robust p-type doping than any element belonging to groups IIIA to IIIb of the periodic table. Some of these elements (groups IIIA to IIIb) are p-type or acceptor elements, dopants, some are donor elements, and due to lack of experimental evidence, many elements are still unpredictable, so exclusion of these elements is preferred. For example, in the above-described region of the periodic table, the ascertained p-type doping elements are only Cu and Ag. While other elements such as Fe, Al, Ga and In are n-type dopants. Boron B is an amorphous element that sometimes behaves as a donor element and sometimes as an acceptor element, depending on the doping level or charge carrier concentration. It was found to be suitable for controlling the concentration of P-type charge carriers. Which generally gives a higher p-type charge carrier concentration. Boron, by itself or in combination with other dopants, can thus be used to increase the effectiveness of p-type doping.

The preparation of n-type compositions requires one of three ways: lightly doped material, or moderately doped material, or simply left unchanged without doping anything, i.e., without doping at all. Thus, the high doping is excluded because the composition exhibits n-type behavior and characteristics even in the undoped state. Generally, if n-type doping is desired or necessary, this can be achieved by incorporating or introducing hetero atoms, generally classified as donor elements or n-type dopants, into the composition. For more powerful and efficient doping, in general, the dopant or material preferably comprises one or more elements selected from the group located to the right of the group comprising Si and Pb in the periodic table, i.e. groups VB, VIB and VIIB. Thus, the elements that can be used as n-type doping materials or dopants are preferably nitrogen, phosphorus, arsenic, antimony, bismuth, oxygen, sulfur, selenium, tellurium, polonium, fluorine, chlorine, bromine, iodine, and astatine. Polonium and astatine are excreted because they are radioactive. Fluorine is also emitted because it is the most electronegative element of the periodic table and is therefore highly reactive. Thus, a list of 12 elements remains to be selected as the most effective and recommended n-type dopant material or donor impurity for the composition. Thus, preferred n-type dopant materials include one or more elements selected from the group consisting of nitrogen, phosphorus, arsenic, antimony, bismuth, oxygen, sulfur, selenium, tellurium, chlorine, bromine, and iodine. In addition to the gaseous elements nitrogen, oxygen and chlorine, these elements may be used in the form of pure elements or in the form of compounds with other elements, preferably Mg and/or Ba. If the n-type doping material or dopant is introduced in the form of a compound with Mg and/or Ba, the chemical formula of the resulting composition is:

Ba_2rMg_2(1-r)Si_1-sPb_aN_bP_cAs_dSb_eBi_fO_gS_hSe_iTe_jCl_kBr₁I_m

since the reaction of the gaseous elements nitrogen, oxygen and chlorine with magnesium and/or barium to form compounds is a very delicate and complex chemical process, these three elements are excluded from the list of doping materials, thus obtaining the following simpler, more practical general formula:

Ba_2rMg_2(1-r)Si_1-sPb_aP_bAs_cSb_dBi_eS_fSe_gTe_hBr_iI_j

wherein s is a + b + c + d + e + f + g + h + i + j, which represents the sum of the atomic proportions of the elements replacing part of the silicon, where s ranges from 0.1 to 0.3 and (b + c + d + e + f + g + h + i + j) ranges from 10^-8To 10^-1B, c, d, e, f, g, h, i, j are all from 0 to 0.1, wherein a is not less than 0.1 and r ranges from 0.1 to 0.4.

All the above analysis of how to produce a mixture of actual compound and element while doping p-type so as to maintain the correct stoichiometric ratio of 2:1 is equally applicable here to doping n-type. Doping according to the above formula ensures more efficient n-type doping than using any of the elements belonging to groups IIIA to IIIB of the periodic table. Whereas for these elements (group IIIA to group IIIB), it was experimentally determined that only Au, Al, In and Fe are elements that can serve as n-type dopants. Thus, the above formula containing 9 doping elements (i.e., phosphorus, arsenic, antimony, bismuth, sulfur, selenium, tellurium, bromine, and iodine) provides more efficient and more powerful n-type doping than using only a single pure element of Au, Al, In, and Fe. However, a combination of a dopant composed of a simple element and/or other elements and a compound preferable as Mg and/or Ba may be used without any limitation or limitation.

However, the composition may be doped by introducing an excess of magnesium, silicon, lead or barium over the stoichiometrically required amount. In principle, excess Mg or Ba functions as an n-type material, i.e. n-type doping, while excess Si or Pb forms a p-type material, i.e. p-type doping. Thus, the doping can be generally achieved by making any one of the four basic constituent components Mg, Si, Pb, or Ba excessive or deficient, or by introducing a hetero atom or element. Doping with a foreign element is preferred as this enables better control of the concentration of free charge carriers in the composition and the type of conductivity, i.e. p-type or n-type. The amount of dopant or impurity added to the composition as described previously should be adjusted so that the free charge carrier concentration is preferably in per cubic centimeter1 × 10 meter¹⁵To 5X 10²⁰The concentration of carriers. The doping element or dopant can be doped in its pure elemental form or in the form of a compound with Mg and/or Ba, or in the form of a compound with Si and/or Pb, depending on whether the doping is p-type or n-type. Alternatively, more than one doping element or compound may be used in order to obtain better results. This applies to n-type and p-type doping and becomes even more important because the composition is essentially composed of four elements that differ very much in atomic mass, atomic volume, electronegativity and electronic structure. Thus, in general, for p-type and/or n-type doping, the ideal chemical formula is one combining the above-mentioned defined formulas corresponding to p-type and n-type doping. Thus, the general doping formula can be written as:

Na_2uK_2vRb_2wCs_2yBa_2zMg_2(1-r)Si_1-sPb_aP_bAs_cSb_dBi_eS_fSe_gTe_hBr_iI_j

wherein the subscripts denote the atomic proportion of the relevant elements, wherein r ═ u + v + w + y + z, and ranges from 0.1 to 0.4, wherein (u + v + w + y) ranges from 10^-8To 10^-1U, v, w and y are all from 0 to 0.1, z is not less than 0.1, wherein s ═ a + b + c + d + e + f + g + h + i + j, and ranges from 0.1 to 0.3, and wherein (b + c + d + e + f + g + h + i + j) ranges from 10^-8To 10^-1B, c, d, e, f, g, h, i and j are all from 0 to 0.1, and a is not less than 0.1. The above formula defines the entire type spectrum for p-type and n-type doping, especially when the doping element is introduced as a compound with one or more of the basic constituent elements Mg, Si, Pb and Ba. Using the broad spectrum formula above, the type of conductivity ultimately obtained, whether p-type or n-type, depends on the relative proportion of the doping element to the left of Ba and the relative proportion of the doping element to the right of Pb. Furthermore, it should be emphasized that neither Ba nor Pb is considered as a doping element. Rather, they are the basic building blocks of the composition.

The composition may be single crystal or polycrystalline. The single crystal material has a high electron mobility because there is no grain boundary, thereby having a high conductivity. Undoped n-type polycrystalline materials or samples prepared by powder metallurgy techniques involving uniaxial cold pressing at a pressure of about 10Mpa followed by sintering at a temperature of 1073K to 1200K in argon have very high thermoelectric power and very high thermoelectric power factor. In fact, it was found that the thermoelectric power factor measured was 10 times higher than another sample prepared by the melt metallurgy technique. The data above are for pure magnesium silicide Mg₂And (3) a Si sample. However, this data can be extended to the compositions described herein, as the compositions are substantially comprised of magnesium silicide. Producing the composition in the form of a single crystal would be difficult to achieve because it is essentially a quaternary alloy or solid solution composed of four elements that differ significantly in atomic mass, atomic volume, valence and electronegativity, and the four intermetallic compounds produced, magnesium silicide, magnesium plumbate, barium silicide and barium plumbate, have a limited solubility detrimental effect on each other in the possibility of obtaining such a single crystal. For example, polycrystalline materials consisting of a large number of large grains can be produced at best using Bridgman crystal growth techniques. It is likely that the best way to obtain the composition in single crystal form would seem to be to use the heat exchanger method known as HEM described earlier in this specification.

The preparation of the composition by the melt metallurgical technique must be carried out in an inert gas, preferably argon, in order to completely avoid or eliminate exposure to atmospheric oxygen. The pressure of the argon must preferably be maintained between 2 and 30 atmospheres in order to suppress or prevent loss of magnesium due to its high volatility and, therefore, deviation from the stoichiometry of the final material. Furthermore, the preparation of the composition by powder metallurgy techniques is preferably carried out in such a way that exposure to air or atmospheric oxygen is completely avoided or eliminated. Thus, the powder metallurgy preparation process is preferably carried out in an inert gas environment, preferably argon. Furthermore, when used in the form of pins, branches or thermoelectric elements constituting a device for controlling thermoelectric power conversion, the long-term operation of the composition also requires complete elimination of exposure to atmospheric oxygen, regardless of whether the production method is powder metallurgy or melt metallurgy. Thus, as a primary option or minimum requirement, it is highly recommended to work in an environment of absolute vacuum.

We are dealing with a composition that is extremely susceptible to damage or oxidation when exposed to atmospheric oxygen. This is confirmed by the fact that, in particular, the two constituent components of the composition, magnesium and barium, have a great affinity for oxygen due to the large electronegativity difference with oxygen. Of course, the higher the working temperature, the stronger the above-mentioned affinity of the composition for oxygen becomes. It is desirable to operate in an argon atmosphere maintained at a certain pressure, i.e. several physical atmospheres, in order to prevent oxidation and to suppress the eventual loss of magnesium due to gradual sublimation that may occur, especially when the composition is at high operating temperatures. If both of the above options are not feasible in practice, encapsulation of the thermoelectric elements, branches or pins must be carried out in order to prevent direct contact of the thermoelectric elements with air or oxygen, and in order to suppress any possible gradual loss of magnesium due to sublimation. Such a package has been described in detail in an earlier part of the specification. This may be the best alternative.

Generally, in order to improve the performance of devices for controlling thermoelectric energy conversion comprising compositions and to make industrial production more efficient and faster, the following means are required:

(1) in the manufacture of new devices, thin film and integrated circuit technologies and packaging are used. Operation of these devices in an argon environment of absolute vacuum and/or relative pressure of 2 to 5 physical atmospheres is still a viable alternative if packaging is not practical or difficult to implement. Moreover, the use of surface treatments such as covering, spraying or painting is completely excluded. The additional covering material applied to the surface is likely to diffuse into the pins, branches or thermoelectric elements of the thermoelectric devices over long periods of use, especially when these devices are operated at relatively high temperatures. This diffusion can cause unwanted doping, can lead to degradation of the thermophysical properties, and is very likely to lead to degradation of the thermoelectric performance of the device. Thus, the only three operational options are: absolute vacuum, argon maintained at a pressure, i.e., preferably 2 to 5 atmospheres, or packaging.

(2) In the manufacture of new devices, FGM or functionally graded material technology is used. The technique is based on the concept that the temperature T is measured from the hot junction_hTemperature T to the cold junction_cThe free electron concentration through the entire thermoelectric element or branch of the thermoelectric device should be adjusted such that the electrical conductivity σ remains constant at the temperature in the respective part of the thermoelectric element or branch. In semiconductors, the conductivity generally increases as the temperature decreases. Therefore, in order to meet the requirement of constant conductivity, the thermoelectric elements or branches must be manufactured with variable impurity content or doping level, or be constituted by several parts, each part having a constant but not identical impurity content. In the low temperature region, the impurity content or doping level is lower than that in the high temperature region. For this reason we keep the thermoelectric power S constant. Therefore, the thermoelectric power factor S²σ also remains approximately constant. Furthermore, the thermoelectric figure of merit Z may also remain substantially constant between the cold and hot junctions, since the overall thermal conductivity may not experience significant changes with temperature between the cold and hot junctions. This is what FGM methods are about: in general, we must maintain a constant figure of merit Z across the entire thermoelectric element or branch, generally maximizing the overall performance of the thermoelectric energy conversion device. To better understand this, note that the thermoelectric power S is at a given effective mass m^*Time by ratio T^2/3And/n is dominant. This can be derived from the following Seebeck (Seebeck) coefficient formula, which is known in the art as:

wherein m is^*Is the effective mass of the charge carrier, n is the concentration of the charge carrier, T is the absolute temperature, q is the electronic charge, k is the Boltzmann constant, h is the Planck constant, r is a constant that depends on the type of scattering experienced by the carrier as it moves through the material, where r is 0 for an excellent covalent lattice and 2 for the presence of impurities. From this formula it is clear that the thermoelectric power can be increased or increased by substituting impurities into the crystal lattice or to a lesser extent selecting substances of large effective mass. On the other hand, increasing the number of charge carriers results in a decrease in thermoelectric power, and conversely for temperature, increasing temperature results in an increase in thermoelectric power. This is because metal (n ═ 10)²²Carrier/cubic centimeter) has a specific semiconductor (n 10 ═ c) ratio¹⁸To 10¹⁹Carriers/cubic centimeter) of thermoelectric power.

FGM technology is fundamentally based on keeping the conductivity constant from the hot connection point to the cold connection point on the thermoelectric elements, branches or pins of the thermoelectric device. Thus, by changing the electron concentration, impurity content or doping level from a lower value in the cold temperature region to a higher value in the hot temperature region, the electrical conductivity is the only thermoelectric property that is actually controlled to remain constant. As a result, the thermoelectric power, which may rise significantly from the cold connection point to the hot connection point, will experience much less change or variation. The same applies to thermoelectric power factor and thermoelectric figure of merit. Thus, it cannot be ensured that the two parameters mentioned above, namely PF and Z, remain exactly constant over the entire thermoelectric element, branch or pin. This is mainly because the electrical conductivity and thermoelectric power follow different physical laws in terms of their dependence on charge carrier or impurity concentration and temperature.

In this regard, the basic concept is as follows: as the temperature increases, the thermoelectric power increases and the conductivity decreases. Conversely, as the temperature decreases, the thermoelectric power decreases and the conductivity increases. Thus, as control is applied to the electrical conductivity to prevent it from decreasing from the cold to the hot regions of the thermoelectric element, the same is true for thermoelectric power, which will be expressed as follows due to the different laws of physics governing thermoelectric power:

wherein k is_e1And k_phIs the electronic and phonon components of the total thermal conductivity. Thus, despite the fact that the electrical conductivity is kept constant as required by FGM technology, thermoelectric power factor, and thermoelectric figure of merit will actually experience some degree of variation, albeit much less than in reality. Maintaining a constant power factor and constant quality factor across the entire leg, or thermoelectric element of a thermoelectric device represents an ideal situation that will undoubtedly result in the greatest benefit being obtained by the FGM method. However, this method or technique does also result in a definite improvement in the performance of the thermoelectric energy conversion device. Therefore, is notTheir use is often recommended.

(3) If necessary, HEM is used in preference to the conventional Bridgman or other crystal growth techniques in the production of single crystals of the composition.

(4) In the production of the composition, powder metallurgy techniques are used in preference to conventional melt metallurgy or casting methods.

(5) Any exposure of the composition to atmospheric oxygen is to be avoided, whether during manufacture by any method or during long term use as a thermoelectric energy conversion material.

(6) During the preparation and production of the composition by any technique, any unwanted doping or contamination by foreign impurities is avoided.

(7) Deviations from stoichiometry, mainly due to loss of magnesium by evaporation or sublimation, are to be avoided, both during production of the composition and during its long-term use.

(8) If powder metallurgy technology is used, the powders required for powder metallurgy technology are produced, using a new technology involving the synthesis of alloy precursors with the general composition Ba-Mg-Si-Pb, by aqueous coprecipitation and organometallic complexation methods, followed by hydrogen reduction of the original precursors to produce alloys in the form of fine powders in polycrystalline form. Thus, there is no need to use a planetary ball mill to manufacture necessary powder and contamination by iron is avoided.

(9) If the above techniques are not feasible for producing the powders required to practice the powder metallurgy process, it is preferred to use gas atomization to prepare the powders, as is well known in the powder metallurgy art. Gas atomization is generally less costly than the rotating electrode process REP and produces spherical particles about 100 microns in diameter, which are smaller than the particle size produced by the plasma rotating electrode process PREP and the water atomization process. Furthermore, the oxide staining was about 120ppm, almost negligible. Thus, gas atomization produces particles with good packing and flow characteristics and exhibiting apparent and tap densities in the theoretical range of 60% to 65%. Thus, gas atomization is the best alternative to the original parent synthesis process described in item (8) above.

According to another embodiment or aspect of the invention there is provided a process for producing a device for controlling thermoelectric energy conversion, the device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, the process comprising using a composition in the manufacture of one or both of the two branches of the device, wherein the composition in its complete form comprises magnesium silicide Mg₂Si, wherein a portion of the magnesium is replaced by one or more elements selected from the group consisting of sodium, potassium, rubidium, cesium, beryllium, calcium, strontium, and barium, and a portion of the silicon is replaced by one or more elements selected from the group consisting of boron, germanium, tin, lead, nitrogen, phosphorus, arsenic, antimony, bismuth, oxygen, sulfur, selenium, tellurium, chlorine, bromine, and iodine, wherein the composition has the following general formula:

Na_2qK_2tRb_2uCs_2vBe_2wCa_2xSr_2yBa_2zMg_2(1-r)Si_1-xB_aGe_bSn_cPb_dN_eP_fAs_gSb_hBi_iO_jS_kSe₁Te_mCl_nBr_oI_p

where r is q + t + u + v + w + x + y + z, which represents the sum of the atomic proportions of the elements replacing part of the magnesium, and s is a + b + c + d + e + f + g + h + i + j + k + l + m + n + o + p, which represents the sum of the atomic proportions of the elements replacing part of the silicon, where r ranges from 0.1 to 0.4 and (q + t + u + v + w + x + y) ranges from 10^-8To 10^-1Q, t, u, v, w, x and y are all from 0 to 0.1, z is not less than 0.1, wherein s is a + b + c + d + e + f + g + h + i + j + k + l + m + n + o + p in the range of 0.1 to 0.3 and (a + b + c + e + f + g + h + i + j + k + l + m + n + o + p) in the range of 10^-8To 10^-1A, b, c, d, e, f, g, h, i, j, k, l, m, n, o and p are all from 0 to 0.1, and d is not less than 0.1. The above formula defines p-type andthe entire type spectrum of the n-type doping, in particular the doping elements are introduced in the form of compounds with one or more of the basic constituents Mg, Si, Pb and Ba. Using the broad spectrum formula above, the type of conductivity ultimately obtained, whether p-type or n-type, is determined by the relative proportions of the doping element to the left of Mg and the doping element to the right of Si. This obviously excludes Ba and Pb as they are essential components of the composition. As such, they cannot be considered as doping elements or dopants. Again, the free charge carrier concentration is preferably at 1 × 10 per cubic centimeter¹⁵To 5X 10²⁰The range of carriers in order to optimize thermoelectric performance. This applies to all other formulae in this specification. In the present embodiment, the free charge carrier concentration range is obtained again by adjusting the relative atomic ratio of the element (other than barium) for replacing a part of magnesium and the relative atomic ratio of the element (other than lead) for replacing a part of silicon.

In the previous embodiment of the invention, and throughout the entire specification, it is stated that the atomic proportions of barium and lead, respectively, cannot be lower than 0.1 or 10%, regardless of how much of the element is introduced into the composition to replace part of the magnesium and/or part of the silicon. This is intended to ensure that the thermal conductivity of the composition approaches an absolute minimum. In practice, such a minimum thermal conductivity can be obtained if the atomic proportion of barium is approximately in the range of 20% to 25% and the atomic proportion of lead is approximately in the range of 15% to 20%. Furthermore, since the thermal conductivity, and in particular the lattice component thereof, will start to drop very rapidly when a small atomic percentage of barium and lead is introduced, the above-mentioned minimum atomic ratio of the two elements, i.e. 10%, will ensure that the thermal conductivity of the composition does not deviate much from the absolute minimum.

In all the general formulae in the present specification, the atomic ratio of each element other than Mg, Si, Pb, and Ba may be 0 in a limiting case, whether broad or complex. This ultimately leads to the following general formula:

Ba_2rMg_2(1-r)Si_1-xPb_x

this formula represents the basis of the present invention as previously described.

According to another embodiment or aspect of the invention there is provided a process for producing a device for the controlled thermoelectric energy conversion, the device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, the process comprising using a composition in the manufacture of the n-type branch and/or the P-type branch of the device, wherein the composition comprises magnesium Mg, silicon Si, lead Pb and barium Ba, and optionally one or more other doping materials.

According to another embodiment or aspect of the invention there is provided a process for producing a device for controlling thermoelectric energy conversion, the device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, the process comprising using a composition in the manufacture of the n-type branch and/or the P-type branch of the device, wherein the composition comprises magnesium Mg, silicon Si, lead Pb and barium Ba.

According to another embodiment or aspect of the invention, with respect to the previous embodiment, the atomic ratio of barium relative to the maximum atomic stoichiometric ratio of magnesium in the absence of barium is 0.1 to 0.4 and the atomic ratio of lead relative to the maximum atomic stoichiometric ratio of silicon in the absence of lead is 0.1 to 0.3.

According to another embodiment or aspect of the invention, for the first of the three preceding embodiments, the atomic ratio of barium relative to the maximum atomic stoichiometric ratio of magnesium in the absence of barium is 0.1 to 0.4, the atomic ratio of lead relative to the maximum atomic stoichiometric ratio of silicon in the absence of lead is 0.1 to 0.3, wherein the atomic or molecular ratio of the doping material in the composition is 10^-8To 10^-1And the free charge carrier concentration is in the range of 1 x 10 per cubic centimeter¹⁵To 5X 10²⁰A carrier.

According to another embodiment or aspect of the invention, the other doping material for the n-type branch of the device comprises one or more elements selected from the group consisting of nitrogen, phosphorus, arsenic, antimony, bismuth, oxygen, sulfur, selenium, tellurium, chlorine, bromine, iodine, magnesium, barium, lithium, gold, aluminum, indium, iron and/or compounds thereof, as defined in the previous embodiment.

According to another embodiment or aspect of the invention, the further doping material for the p-type branch of the device comprises one or more elements selected from the group consisting of copper, silver, sodium, potassium, rubidium, cesium, boron, silicon, lead and/or compounds thereof, as defined in the first of the two aforementioned embodiments.

Various modifications may be made to the embodiments herein chosen for purposes of disclosure without departing from the spirit and scope of the invention.

Claims (6)

1. A process for producing a device for controlling thermoelectric energy conversion, the device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, the process comprising using a composition in the manufacture of the n-type branch and/or the P-type branch of the device, wherein the composition comprises magnesium, silicon, lead and barium, wherein the atomic ratio of barium relative to the maximum atomic stoichiometric ratio of magnesium in the absence of barium is 0.1 to 0.4 and the atomic ratio of lead relative to the maximum atomic stoichiometric ratio of silicon in the absence of lead is 0.1 to 0.3, wherein the device is made up of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, wherein the composition comprises magnesium, silicon, lead and bariumThe composition preferably comprises one or more further doping materials, wherein the further doping material for the n-type branch of the device comprises one or more elements selected from the group consisting of nitrogen, phosphorus, arsenic, antimony, bismuth, oxygen, sulphur, selenium, tellurium, chlorine, bromine, iodine, magnesium, barium, lithium, gold, aluminium, indium, iron and/or compounds thereof, and the further doping material for the p-type branch of the device comprises one or more elements selected from the group consisting of copper, silver, sodium, potassium, rubidium, caesium, boron, silicon, lead and/or compounds thereof, wherein the doping material comprises one or more elements selected from the group consisting of copper, silver, sodium, potassium, rubidium, caesium, boron, silicon, lead and/or compounds thereof, wherein the atomic or molecular ratio of the doping material in the composition is 10^-8To 10^-1Free charge carrier concentration of 1X 10 per cubic centimeter¹⁵To 5X 10²⁰And a carrier.

2. A process for producing a device for controlling thermoelectric energy conversion, said device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, said process comprising the use of a composition in the manufacture of the n-type branch and/or the P-type branch of said device, wherein said composition comprises magnesium, silicon, lead and barium, wherein said composition preferably comprises one or more further doping materials, the further doping materials for the P-type branch of said device being selected from the group consisting of sodium, potassium, rubidium and cesium, wherein said doping materials or elements are chemically reacted with silicon and/or lead prior to the introduction of a host material so as to form a compound, said composition having the following general formula:

Na_2uK_2vRb_2wCs_2yBa_2zMg_2(1-r)Si_1-xPb_x

wherein r is u + v + w + y + z, representing the sum of the atomic proportions of the elements replacing part of the magnesium, wherein r ranges from 0.1 to 0.4 and (u + v + w + y) ranges from 10^-8To 10^-1Wherein u, v, w and y all range from 0 to 0.1, z is not less than 0.1, x ranges from 0.1 to 0.3, whereinThe range of charge carrier concentration is 1 x 10 per cubic centimeter¹⁵To 5X 10²⁰And a carrier.

3. A process for producing a device for controlling thermoelectric energy conversion, said device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, said process comprising the use of a composition in the manufacture of the n-type branch and/or the P-type branch of said device, characterized in that said composition comprises magnesium, silicon, lead and barium, wherein said composition preferably comprises one or more further doping materials, the further doping materials for the n-type branch of said device being selected from the group consisting of phosphorus, arsenic, antimony, bismuth, sulphur, selenium, tellurium, bromine and iodine, said doping materials or elements chemically reacting with magnesium and/or barium before introduction into the main material so as to form a compound, said composition having the general formula:

Ba_2rMg_2(1-r)Si_1-sPb_aP_bAs_cSb_dBi_eS_fSe_gTe_hBr_iI_j

wherein s is a + b + c + d + e + f + g + h + i + j, represents the sum of the atomic proportions of the elements replacing part of the silicon, where s ranges from 0.1 to 0.3 and (b + c + d + e + f + g + h + i + j) ranges from 10^-8To 10^-1B, c, d, e, f, g, h, i, j all range from 0 to 0.1, a is not less than 0.1, r ranges from 0.1 to 0.4, wherein the free charge carrier concentration ranges from 1 x 10 per cubic centimeter¹⁵To 5X 10²⁰And a carrier.

4. A process for producing a device for controlling thermoelectric energy conversion, said device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, said process comprising the use of a composition in the manufacture of the n-type branch and/or the P-type branch of said device, characterized in that said composition comprises magnesium, silicon, lead and barium, wherein said composition preferably comprises one or more further doping materials, wherein said further doping materials for the n-type and P-type branches of said device are combined together to form a composition having the following general formula:

Na_2uK_2vRb_2wCs_2yBa_2zMg_2(1-r)Si_1-sPb_aP_bAs_cSb_dBi_eS_fSe_gTe_hBr_iI_j

wherein the further doping material comprises one or more elements selected from the group consisting of sodium, potassium, rubidium and caesium and one or more elements selected from the group consisting of phosphorus, arsenic, antimony, bismuth, sulphur, selenium, tellurium, bromine and iodine, wherein the subscripts denote the atomic proportions of the relevant elements, wherein r ═ u + v + w + y + z, in the range 0.1 to 0.4, wherein (u + v + w + y) is in the range 10^-8To 10^-1U, v, w and y all range from 0 to 0.1, z is not less than 0.1, wherein s ═ a + b + c + d + e + f + g + h + i + j, and ranges from 0.1 to 0.3, wherein (b + c + d + e + f + g + h + i + j) ranges from 10^-8To 10^-1B, c, d, e, f, g, h, i and j all range from 0 to 0.1, and a is not less than 0.1, wherein the finally obtained conductivity type, whether p-type or n-type, is determined by the relative atomic proportions of the elements located to the left of barium and the elements located to the right of lead in the formula, wherein the other doping elements are in the same atomic proportion as the basic constituent: introduced into the host material in the form of a compound of one or more of the elements Mg, Ba, Si and Pb, wherein the free charge carrier concentration is in the range of 1 x 10 per cubic centimeter¹⁵To 5X 10²⁰And a carrier.

5. A process for producing a device for controlling thermoelectric energy conversion, said device consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, said process comprising the use of a composition in the manufacture of the n-type branch and/or the P-type branch of said device, characterized in that said composition comprises magnesium silicide Mg₂Si, wherein part of the magnesium is replaced by barium and part of the silicon is replaced by lead, the composition being a packageAn alloy or solid solvent comprising an intermetallic compound of magnesium silicide, magnesium plumbate, barium silicide and barium plumbate, wherein the composition has the general formula:

Ba_2rMg_2(1-r)Si_1-xPb_x

wherein r, (1-r), (1-x) and x represent the atomic proportions of barium, magnesium, silicon and lead, respectively, in the alloy, wherein the composition preferably includes one or more other dopant materials, wherein r ranges from 0.1 to 0.4, (1-r) ranges from 0.6 to 0.9, x ranges from 0.1 to 0.3, and (1-x) ranges from 0.7 to 0.9, wherein the atomic or molecular proportions of the dopant materials in the composition ranges from 10^-8To 10^-1The free charge carrier concentration is in the range of 1 × 10 per cubic centimeter¹⁵To 5X 10²⁰Carriers, other doping materials for the n-type branch of the device include one or more elements selected from the group consisting of nitrogen, phosphorus, arsenic, antimony, bismuth, oxygen, sulfur, selenium, tellurium, chlorine, bromine, iodine, magnesium, barium, lithium, gold, aluminum, indium, iron and/or compounds thereof.

6. Process for producing a device for controlling thermoelectric energy conversion, consisting of a P-type branch or thermoelectric element, an n-type branch or thermoelectric element, a hot junction and a cold junction, comprising the use of a composition in the manufacture of the n-type branch and/or the P-type branch of the device, characterized in that said composition comprises magnesium silicide Mg₂Si, wherein a portion of the magnesium is replaced by one or more elements selected from the group consisting of beryllium, calcium, strontium, and barium, and a portion of the silicon is replaced by one or more elements selected from the group consisting of germanium, tin, lead, antimony, bismuth, selenium, and tellurium, wherein the composition has the following general formula:

Be_2uCa_2vSr_2wBa_2zMg_2(1-r)Si_1-sGe_aSn_bPb_cSb_dBi_eSe_fTe_g

wherein r is u + v + w + z, representing the sum of the atomic proportions of the elements used for partial substitution of magnesium, and s is a + b + c + d + e + f + g, representingA sum of atomic ratios of elements for partial substitution of silicon, wherein the composition preferably includes one or more other dopant materials, wherein r ranges from 0.1 to 0.4, (1-r) ranges from 0.6 to 0.9, u, v, and w each ranges from 0 to 0.3, (u + v + w) ranges from 0 to 0, 3, z is not less than 0.1, s ranges from 0.1 to 0.3, (1-s) ranges from 0.7 to 0.9, a, b, d, e, f, g each ranges from 0 to 0.2, (a + b + d + e + f + g) ranges from 0 to 0.2, c is not less than 0.1, atomic ratios or molecular ratios of dopant materials in the composition range from 10^-8To 10^-1And the free charge carrier concentration ranges from 1 x 10 per cubic centimeter¹⁵To 5X 10²⁰Carriers, other doping materials for the n-type branch of the device include one or more elements selected from the group consisting of nitrogen, phosphorus, arsenic, antimony, bismuth, oxygen, sulfur, selenium, tellurium, chlorine, bromine, iodine, magnesium, barium, lithium, gold, aluminum, indium, iron and/or one or more compounds of these elements, and other doping materials for the p-type branch of the device include one or more elements selected from the group consisting of copper, silver, sodium, potassium, rubidium, cesium, boron, silicon, lead and/or one or more compounds of these elements.