DE102007039060A1 - Thermokraft element or Peltier elements made of sintered nanocrystals of silicon, germanium or silicon-germanium alloys - Google Patents

Abstract

The present invention relates to a method for the production of thermoelectric components or Peltier elements and these components of printed silicon layers, prepared from dispersions of at least partially doped semiconducting particles, in particular of silicon, germanium or silicon-germanium alloys.

Description

object The invention relates to a method for producing thermoelectric Structural elements or Peltier elements as well as these printed elements Silicon layers made from dispersions of at least partially doped semiconducting particles, in particular of silicon, germanium or silicon germanium alloys. Here, the doping of the Particle targeted adjusted to a value above the defect concentration the layers. The good electrical conductivity of the layers is in addition to the doping by a suitable etch and Sintering guaranteed.

thermoelectric Elements, also called thermoelectric elements, allow the recovery of electrical energy from a temperature difference due to the Seebeck effect. The Seebeck effect states that two different metals that are connected to each other at their Junction create a thermoelectric voltage, if along the metals there is a temperature gradient. These Thermoelectric voltage, also called thermoelectric force, is temperature-dependent and has a size of a few microvolts per metal Kelvin temperature difference.

Vice versa Each thermoelectric element can also be used as a Peltier element use, d. H. the generation of a temperature difference when creating a electrical voltage due to the Peltier effect. The Peltier effect states that at the junction of two different metals when applying an electrical voltage, a temperature difference arises.

In practice, however, instead of metals, semiconductor materials are used to generate energy from such thermoelectric elements, whereby the efficiency can be significantly increased. Common semiconductor materials are for example Bi ₂ Te ₃ , PbTe, SiGe, BiSb or FeSi ₂ with real efficiencies between about 3 and 8 percent. In order to obtain sufficiently high voltages, several elements mounted between the cold and the hot side are electrically connected in series. Two typical constructions of thermoelectric elements are in 5 and 6 shown.

In the article "Thermoelectric Multitalented", Physics Journal 6 (2007), No. 5 , gives a good overview of the current state of the art and current research in the field of thermoelectric devices and the modern nanomaterials used in manufacturing. There it is also described that a good thermoelectric material in addition to a high Seebeck coefficient (S) must also have good electrical conductivity and low thermal conductivity. With these two variables and the temperature (T), the dimensionless index ZT can be determined. The efficiency of the thermoelectric materials themselves is indicated by this index, which is given by ZT = S ² σT / κ with the specific electrical conductivity sigma (σ) and the specific thermal conductivity kappa (κ). This shows the physical dilemma for the improvement of thermoelectric materials. In some cases, the electrical and thermal conductivity can not be optimized independently: a) In metals, the free electrons mainly contribute to the transport of heat. The thermal and the electrical conductivity are directly proportional to each other as expressed in the Wiedemann-Franz law κ = T × L × σ. Where L is the Lorenz constant. b) In insulators, however, the heat transfer takes place exclusively via lattice vibrations. Some electrical insulators, eg. As sapphire or diamond, have comparable thermal conductivities as transition metals, eg. As titanium, zirconium on. c) In semiconductors, both free charge carriers (holes or electrons) and phonons contribute to the thermal conductivity. The contribution of the phonons to the thermal conductivity is decoupled from the electrical conductivity and numerous attempts to optimize the thermoelectric efficiency ZT are based on a reduction of the heat transfer by phonons.

The article points out that, in particular, semiconductors with high carrier concentrations which have the highest ZT values are suitable for thermoelectric applications. Although semi-metals have high carrier densities and low thermal conductivities, both holes and electrons are simultaneously present in high concentration, and since these contribute to the Seebeck coefficient with different signs, the thermoelectric effects cancel each other out. The semiconductors most commonly used for thermoelectrics are compounds of the V and VI main groups in the periodic table based on Be ₂ Te ₃ . These materials are suitable for room temperature applications. For medium temperatures up to about 400 ° C IV to VI compounds (PbTe) are used, while SiGe alloys are usually used at high temperatures.

The review article also describes that, in addition to the successful approaches to epitaxial superlattice, cost-effective manufacturing methods are at the center of interest in order to be able to introduce nanostructured thermoelectrics to the market in the medium term. One approach is z. For example, to circumvent the costly Epitaxiverfahren to increase over chemical and structurally self-organizing systems in the nano-scale range to the desired superlattices with to arrive at the ZT values. For this purpose, it would be possible in the nanometer range to coat element layers of suitable thickness to the stoichiometry of the V ₂ VI ₃ compounds alternately as a periodic layer sequence consisting of Bi-Te-Sb-Te deposits and by increasing the temperature in the associated alternating layers Bi ₂ Te ₃ / Sb ₃ Te ₃ convict.

One Another way to reduce thermal conductivity be thermoelectric nanocomposites. Located in these materials nanoparticles or nanocrystalline precipitates become more or less less ordered in a thermoelectric matrix. Try for now especially researchers in the US and China, thermoelectric nanoparticles, which can be produced in the kilogram range, at high pressure and to compress temperatures to nanocomposites that differs in conventional Use thermocouples.

On the field of thermoelectrics is still looking for materials which have a large figure of merit ZT, which means that in the same material at the same time large values of Thermoelectric and electrical conductivity at the same time low thermal conductivity needed become. Here, particulate materials can be great Advantages over volume material unfold, since the thermal conductivity is significantly reduced. During the state of the art itself Mainly served by powdery materials through ceramic processes are sintered, the ones used here Dispersions of nanoparticles have the advantage of being both large-area printing in thin layers as well as for other molding processes (injection molding etc.) are usable.

To the Purpose of electrical energy recovery from temperature gradients If a thermoelectric element is to be realized, the large area consists of printable material. The substrate used is a material which shows little heat conduction and is insulating behaves. Alternatively, thermocouples from self-supporting molded parts be structured.

The dispersions used here preferably contain nanoparticles of crystalline silicon, germanium or silicon germanium. These can be manufactured in a plasma reactor from precursors Ar, SiH ₄ , GeH ₄ and H ₂ . The possibility of doping is the direct addition of z. B. B ₂ H ₆ or PH ₃ to the reaction gases in a corresponding dilution to achieve a p- or n-type conductivity of the particles. It was found that approximately 100% of the boron is incorporated in the particles, while only 5-10% of the doping efficiency was observed with phosphorus. The concentration of doping atoms must reach a critical value of typically at least 1 × 10 ¹⁸ cm ^{-3 in} order to achieve a significant influence of the doping. In this case, the concentration of doping atoms is to be understood as an average concentration since, because of the small particle size, not necessarily every particle contains dopants. Furthermore, the procedure is as described below to produce conductive layers. Among other things, the observed thermopower depends on the doping of the nanoparticles and can be adapted to the respective requirements (high thermoelectric voltages or high thermal current) by choosing the doping. By suitable structuring of the material then thermoelectric components can be produced.

essential Component of thermoelectric devices are semiconducting P-type and n-type elements. In the prior art are semiconductive layers and field effect transistors made therefrom known, the particles from the group of semiconductor compounds PbSe, PbTe, PbS, CdSe, CdTe or CdS. These are going out a dispersion or suspension applied to a substrate. The associated with the use of such heavy metal compounds problem of environmental hazard and its allergenic, irritating or even toxic effects on humans are also have long been known and represent an increasingly serious obstacle to the economic use of such materials.

Talapin in Science, 310, p. 86 (2005) discloses transistors consisting of PbSe, CdSe. Cadmium and its compounds are classified as "toxic" or "very toxic". There is also a reasonable suspicion of a carcinogenic effect in humans. Ingested cadmium-containing dust causes damage to the lungs, liver and kidneys. PbTe, PbS and PbSe are also toxic. Use of these materials in printed electronics is therefore not advisable. The systems presented at Talapin have been extensively modified by chemical treatment steps. The functioning of the created components was sensitive to the arrangement of the particles whose distances between them must be in the range of below one nanometer.

Sun in "Nano Letters", 5, p. 2408 (2005) discloses transistors made of particulate ZnO. In contrast to silicon, ZnO has the distinct disadvantage that it is not possible in the current state of the art to carry out both p-type and n-type doping for the production of both hole-conducting and electron-conducting structures. Although it has been possible to produce first p-type layers and pn-junctions from ZnO, the latter is still considered controversial, and one is still far from a high, homogeneous, reproducible and stable p-doping removed ( Klingshirn in "Physik Journal", 5, No. 1 (2006) ).

The Use of silicon or germanium as semiconductor material and the provision of electronic components made from these materials requires a high degree of purity of these semi-metals, as well a precisely observed doping, and because of the requirement of State of the art on macroscopic crystallinity very complex production processes for wafers to the desired Semiconductor properties, especially in the production of miniaturized Circuits, to be able to control.

Although silicon or germanium in the form of particles is technically much easier to produce. Roth et. al., Chem. Eng. Technol. 24 (2001), 3 , disclose the production of particulate silicon from the gas phase. The possibility of obtaining doped particulate silicon is disclosed in DE 10 353 996 discloses, wherein the silicon powder has a BET surface area of more than 20 m ² · g ^-1 . However, such particles of silicon and also germanium have impurity of transition metals, as also known in the literature and for example in Sze in Physics of Semiconductor Devices, John Wiley and Sons, 1981 , is discussed. Furthermore, for example, silicon has deep impurities due to surface states. Such surface conditions can in principle by special methods, eg. B. passivate by means of hydrotreating and / or functionalization. EP01760045A1 discloses that such methods are applicable to silicon particles to significantly reduce the number of surface defects. For layers of silicon particles, the z. In DE 10 2005 022383 are disclosed, these methods are only conditionally applicable and do not lead to a significant reduction of the defect concentration. The conductivity is limited by these surface defects, which prevents the applicability of these layers for electronic applications.

Volker Lehmann describes in "Electrochemistry of Silicon," ISBN 3-527-29321-3, Wiley-VCH Weinheim, 2002 in that surface defects in porous silicon adversely affect the conductivity of a porous semiconductive layer by trapping charge carriers. Disclosed is a porous silicon having a BET surface area of 43 m ² .g ^-1 , which was prepared by electrochemical etching. The resistivity is between 10 ⁴ and 10 ⁷ Ω · cm, which is too high for electronic applications. Another disadvantage of this material is the production route. Such a material is produced costly and processively from crystalline wafer silicon.

Out porous silicon material is for electronic applications thus very limited or not usable at all.

Out the prior art is also known that it is monocrystalline Semiconductors and volume semiconductors a connection between the Concentration of dopants and the conductivity gives. Furthermore, the skilled person gains the activation of the dopants Information about the density of free charge carriers in a voluminous semiconductor. This density will for example, in a field effect transistor by the influence an external electric field is modulated. The density must be based on the doping can be controlled. Furthermore become high demands on the crystalline purity of the carrier material posed.

task The present invention was therefore a thermoelectric To provide a component or Peltier element, which is based on semiconducting structures that are large in area and in thin layers and at the same time in a cost-effective Process can be produced. At the same time in the production of semiconducting structures disadvantages according to the prior overcome the technology.

Surprised was found to accomplish this task through a thermoelectric Component (thermoelectric element) or Peltier element is loosened, comprising one or more p-type elements and / or one or more n-type elements, characterized in that the p and / or the n-type elements are porous semiconductive structures that from dispersions of at least partially doped semiconducting Particles are made.

In a preferred embodiment of the thermoelectric element or Peltier element, the n-type and p-type elements are interconnected via conductive bridges or by direct conductive contact according to the scheme p _x - [np-] _z n _y , where x = 0 or 1, y = 0 or 1, z> 0, n = n-type element, p = p-type element, connected to one another, wherein the indices x, y and z are the number of the respective p-, n- or Specify np element. The conductive bridges may be made of metal, preferably of aluminum, silver, gold, nickel, iron, cobalt, titanium, tungsten, manganese, chromium, platinum or molybdenum.

In a further preferred embodiment of the thermoelectric element or Peltier element, the semiconducting structures have an electrical conductivity of 5 × 10 ^-8 to 1 × 10 ³ S · cm ^-1 , preferably 1 × 10 ^-5 to 100 S · cm ^-1 preferably 1 × 10 ^-3 to 10 S · cm ^-1 .

In a further preferred embodiment of the thermoelectric element or Peltier element, the semiconducting structures have an activity tion energy of the electrical conductivity of 0.1 to 700 meV, preferably 5 to 250 meV, more preferably from 10 to 50 meV, on.

In order to determine the conductivity, the semiconducting structure is contacted electrically according to the van der Pauw method ( LJ van der Pauw, A method of measuring specific resistivity and Hall effect of discs of arbitrary shape, Research Reports 13 (1), 1958 ). Thereafter, at a constant voltage, the conductivity is measured as a function of the temperature. The current-temperature curve is adjusted with an exponential compensation curve of the form exp (-E / kt) in the range 250 to 300 K, whereby the activation energy E is obtained.

In a further preferred embodiment of the thermoelectric element or Peltier element, the semiconducting structures have a pore size of 1 to 500 nm, preferably 5 to 250 nm, more preferably 10 to 100 nm, up.

The Pore size is within the scope of the present invention in a scanning electron micrograph (SEM) in 80,000 times Magnification determined.

The Semiconducting structures of the invention Thermoelectric element or Peltier element preferably have one solid fraction of 25 to 60% by volume, preferably 30 to 45% by volume, on.

Of further have the structures of the invention Thermoelectric element or Peltier element at least partially crystalline doped components with sizes from 5 nm to 500 nm up.

Of the Advantage of the present invention is that for the thermoelectric device according to the invention or Peltier element used porous semiconducting structure in a simple, apparatusless way in any size Areas can be provided. Another advantage of Present invention is that the used porous Semiconducting structure contains no binder. Binder are while substances that do not consist of the semiconducting material and fill in gaps between the semiconducting material and connect the particles mechanically. In particular contains the structure used does not contain inert binders. Inert binders are those which itself is not a semiconductor or conductor known to those skilled in the art Represent definition.

The for the inventive thermoelectric Component used structure has the basis of the sintered necks interconnected components an electrically percolating Network on whose electrical conductivity through the Doping the semiconducting components can be adjusted.

In a particularly preferred embodiment of the invention The thermo-power element or Peltier element comprises at least one Substrate, wherein a thermoelectric element as described above or Peltier element is arranged on this Substart. As a substrate All known in the art substrates are those in the field the thermoelectric elements or Peltier elements are used. In a particular embodiment, for. B. also provided be that the substrate is a flexible substrate, for. B. from textile Materials, is. So z. B. the gradient between Body heat and ambient temperature for energy production be exploited to z. B. in clothing integrated electronic Power components (watches, cell phones, etc.).

• a) Erzeugung von dotierten Halbleiterpartikeln des p-Typs,
• b) Erzeugung von dotierten Halbleiterpartikeln des n-Typs,
• c) Erzeugung einer ersten Dispersion mit den nach Schritt a) erhaltenen Halbleiterpartikeln,
• d) Erzeugung einer zweiten Dispersion mit den nach Schritt b) erhaltenen Halbleiterpartikeln,
• e) zumindest teilweises Beschichten eines Substrats mit einer der nach Schritt c) oder
• d) erhaltenen ersten oder zweiten Dispersion,
• f) Behandlung der nach Schritt e) erhaltenen Schicht durch eine Lösung enthaltend Fluorwasserstoff und Wasser,
• g) thermische Behandlung der nach Schritt f) erhaltenen Schicht, so dass eine erste poröse halbleitende Struktur des p- oder n-Typs erhalten wird,
• h) Verwendung der in Schritt e) nicht verwendeten Dispersion zur Herstellung einer zumindest partiellen zweiten Beschichtung auf dem Substrat, auf einem weiteren Substrat oder – unter Bildung eines Schichtsystems – auf der in Schritt e) erhaltenen ersten Beschichtung,
• i) Behandlung der nach Schritt h) erhaltenen zweiten Schicht durch eine Lösung enthaltend Fluorwasserstoff und Wasser,
• j) thermische Behandlung der nach Schritt i) erhaltenen zweiten Schicht, so dass eine zweite poröse halbleitende Struktur des anderen Typs erhalten wird.

Another object of the present invention is a method for producing a device for further use in the production of a thermoelectric element or Peltier element comprising the steps:

• a) Generation of doped semiconductor particles of the p-type,
• b) Generation of doped semiconductor particles of the n-type,
• c) production of a first dispersion with the semiconductor particles obtained according to step a),
• d) production of a second dispersion with the semiconductor particles obtained after step b),
• e) at least partially coating a substrate with one of the after step c) or
• d) obtained first or second dispersion,
• f) treatment of the layer obtained according to step e) by a solution containing hydrogen fluoride and water,
• g) thermal treatment of the layer obtained after step f), so that a first porous semiconducting structure of the p or n type is obtained,
• h) using the dispersion which is not used in step e) to produce an at least partial second coating on the substrate, on a further substrate or, on formation of a layer system, on the first coating obtained in step e),
• i) treatment of the second layer obtained after step h) with a solution containing hydrogen fluoride and water,
• j) thermal treatment of the second layer obtained after step i), so that a second porous semiconducting structure of the other type is obtained.

One Advantage of the thermoelectric for the invention Component used method is the production of porous Semiconducting structures from the dispersion by inexpensive Coating and printing process.

It can be beneficial if the stock parts of the structure used are partially interconnected. Preferably, these compounds may have mainly sintered bridges. Preferably, the interior of the sintering bridges of the structure used is mainly, but not exclusively, of a semiconducting material. More preferably, within the sintering bridge, the crystalline structure of the constituent continues without there being a crystallographic dislocation or grain boundary.

Optionally, the surface of the constituents of the structure used may at least partially contain further elements. Particularly preferably, these elements can be oxygen, hydrogen, and / or carbon. Optionally, the surface of the sintered bridges of the structure used may at least partially contain further chemical elements. Particularly preferably, these elements can be oxygen, hydrogen, and / or carbon. Optionally, the compounds between each two constituents of the structure according to the invention may contain further elements in addition to the semiconductive element. Particularly preferably, these elements can be oxygen, hydrogen, and / or carbon. The surfaces of the constituents of the structure used may preferably comprise SiO _x , Si-H, Si-OH groups, and / or adsorbates of carbon. Optionally, the surface targeted, z. By hydrosilylation.

Farther It may be advantageous if the ingredients of the used Structure having sizes of 100 to 500 nm, and the ingredients, the sizes below 100 nm, on a size scale of 2 microns up to 10 μm on and in the structure used with each other are mixed. Particularly preferably, the structure used can a size scale below 2 microns on and in the structure used components of 100 nm to 500 nm size. Furthermore, particularly preferred the structure used on the surface of the constituents, which have a size of 100 nm to 500 nm, Have components ranging in size from 5 to 100 nm, preferably from 5 to 90 nm, more preferably from 5 to 80 nm, most preferably from 5 to 50 nm.

The structure used may have a BET surface area of 10 to 800 m ² .g ^-1 . The structure used may preferably have the BET surface area of the semiconductor particles from which the structure used was obtained according to the method according to the invention. The BET surface area of the structure used can particularly preferably be from 30 to 500 m ² .g ^-1 , more particularly preferably from 40 to 200 m ² .g ^-1 , very particularly preferably from 50 to 100 m ² .g ^-1 .

The used structure may preferably have a thickness of 20 nm to 50 microns, particularly preferably from 50 nm to 10 μm, very particularly preferably from 100 nm to 3 microns.

It may be advantageous if in the generation of the doped semiconductor particles of the p- or n-type continuously (a) at least one vapor or gaseous Semi-metal hydride selected from silane, germanium hydride, corresponding dimers and oligomers, or a mixture of these Half metal hydrides and at least one vapor or gaseous Dopant, and an inert gas transferred to a reactor and is mixed there, wherein the proportion of Halbmetallhydrids between 0.01 and 90 wt .-%, preferably between 0.1 and 20 wt .-%, based to the sum of half metal hydride, dopant and inert gas, and then (b) by energy input by means of electromagnetic Radiation in the microwave range at a pressure of 1 to 1100 hPa, preferably 10 to 300 hPa, a plasma is generated, and subsequently (c) allowing the reaction mixture to cool and the reaction product in the form of a powder of gaseous Material is separated, with predominantly crystalline doped semiconductor particles are obtained. Preferably, in the proportion of the Dopant from 1 ppm to 30,000 ppm based on the sum of half metal hydride and dopant amount.

In the process according to the invention, it is possible with preference to use silane for the preparation of the semiconductor particles disclosed in US Pat DE 10 353 996 , Preferably, the semiconductor particles may have chemical impurities which may be less than 100 ppm, preferably less than 20 ppm and most preferably less than 5 ppm in their total concentration. Here, "impurity" refers to all types of atoms that are not part of the pure semiconductor and are not used as a dopant.

Furthermore, it may be advantageous if doped semiconductor particles are used in the method according to the invention which have a doping concentration of 10 ¹⁹ doping atoms cm ^-3 to 10 ²⁰ doping atoms cm ^-3 .

The measurement of the doping concentration is known to the person skilled in the art and is carried out by electron spin resonance or electro-paramagnetic resonance, glow discharge mass spectroscopy or by mass spectroscopy on inductively coupled plasmas or conductivity measurements on samples obtained by melting. The doping concentrations provided in the prior art are, for example, in DE 10 353 996 and in DE 10 2005 022383 disclosed.

In the method according to the invention, it may be advantageous if, in the production of the do a dopant selected from lithium-metal or lithium amide (LINH ₂ ), from the group of hydrogen-containing compounds of phosphorus, arsenic, antimony, bismuth, boron, aluminum, gallium is used doped semiconductor particles of the p- or n-type , Indium, thallium, europium, erbium, cerium, praseodymium, neodymium, samarium, gadolinium, terbium, dysprosium, holmium, thulium, ytterbium, lutetium, lithium, germanium, iron, ruthenium, osmium, cobalt, rhodium, iridium, nickel, palladium , Platinum, copper, silver, gold, zinc, or a mixture of these dopants. Particularly preferred in the method according to the invention, a dopant can be used, which is selected from phosphorus, arsenic, antimony, bismuth, boron, aluminum, gallium, indium, most preferably selected from boron, phosphorus, and aluminum.

Furthermore, it can be advantageous in the process according to the invention if semiconductor particles doped in the production of the respective dispersion have a mass fraction of 0.1 to 60% by weight, preferably 1 to 30% by weight, particularly preferably 2 to 12% by weight. %, based on the total mass of the dispersion, are used, and the doped semiconductor particles are dispersed in a liquid phase, the water and / or at least one organic solvent or consisting thereof, wherein a dispersion is obtained which at 23 ° C a Viscosity of less than 100 mPas, more preferably less than 10 mPas, at 1000 s ^-1 , having. The features of the dispersion which can be used in the process according to the invention, as well as features of the process for preparing such dispersions, are described, for example, in US Pat DE 10 2005 022383 disclosed.

Preferably, in the first and / or second or any further coating of the method according to the invention, the dispersion by spin coating, knife coating, dropping, screen printing, dip coating, or spray coating applied to the substrate and then the liquid phase are expelled by the entry of thermal energy to obtain a layer having a conductivity of 5 × 10 ^-8 to 1 × 10 ³ S · cm ^-1 , preferably 1 × 10 ^-5 to 100 S · cm ^-1 , more preferably 1 × 10 ^-3 to 10 S · cm ^-1 . Particularly preferably, in the first and / or second or any further coating of the process according to the invention, the liquid phase can be expelled in an oven or under hot inert gas. This procedure has the advantage that a coating resistant to rinsing can be obtained. Regardless of the doping of the particles used, it has a conductivity below less than 10 ^-11 S · cm ^-1 . The conductivity is measured by a method known to the person skilled in the art.

Advantageous In the process according to the invention, a substrate can be used can be used, which is selected from organic, inorganic material, or a composite of at least one organic and / or at least one inorganic material, wherein the substrate preferably resistant to hydrofluoric acid. In the process according to the invention, preference may be given a substrate can be used which has a total thickness of less than 1 cm, especially preferably less than 1 mm, most preferably less than 0.4 mm. In the context of the present invention is under a resistant substrate a substrate understood in step D of the invention Procedure neither dissolved, etched nor damaged becomes.

Preferably becomes a substrate in the process according to the invention used, which is selected from the materials sapphire, Glass, quartz or ceramics, by chemical vapor deposition applied diamond on a support, selected made of glass, quartz or ceramics, or is selected from Polymers, particularly preferably selected from polyethylene terephthalate, Polyimide, polyvinyl chloride, polystyrene, polyamide, polyethyletherketone (PEEK), polypropylene, polyethylene, polymethyl methacrylate (PMMA), or a combination of these substrates. extraordinarily in the process according to the invention, preference may be given a substrate can be used which is selected from Polymeric film selected from polyethylene terephthalate, polyimide, polyvinyl chloride, Polystyrene, polyamide, polyethyletherketone (PEEK), polypropylene, Polyethylene, polymethylmethacrylate (PMMA), or a combination of these Substrates.

In the method according to the invention, it may be advantageous if in the respective treatment step with hydrogen fluoride, a solution containing hydrogen fluoride and water in a concentration of hydrogen fluoride from 0.01 wt .-% to 50 wt .-%, based on the total mass of the solution containing hydrogen fluoride and water, by dipping, dripping, spraying or rinsing, and at a temperature between 10 ° C and 70 ° C, and applied for a period between 1 s and 1 hour, to obtain a layer having an electrical conductivity of 5 · 10 ^-11 to 10 ^-7 S · cm ^-1 , preferably 10 ^-10 to 5 · 10 ^-8 S · cm ^-1 .

Particularly preferably, in the process according to the invention in the respective treatment step with hydrogen fluoride, a solution containing hydrogen fluoride and water in a concentration of hydrogen fluoride from 1 to 30 wt .-%, most preferably from 5 to 15 wt .-%, and at one temperature from 15 ° C to 40 ° C, more preferably between 18 ° C to 30 ° C, and for a duration preferably from 5 seconds to 5 minutes, more preferably from 10 seconds to 60 s be used.

It may also be advantageous if in the inventive Process each obtained after one of the treatment steps with hydrogen fluoride Layer dipped in a further step in water or rinsed and then blown off with dry nitrogen is, or subsequently at a temperature of 60 to 200 ° C is dried.

In the method according to the invention, the respective thermal treatment may preferably be effected by introducing heat energy, more preferably by absorbing light energy, most preferably by absorbing flash lamp light, with a power density of 10 kW · cm ^-2 to 65 MW · cm ^-2 the particles of the layer obtained after one of the treatment steps with hydrogen fluoride are at least partially sintered.

In the method according to the invention, the input of heat energy by laser light with a wavelength of 500 to 1100 nm, more preferably from 505 to 850 nm, particularly preferably from 510 to 650 nm, very particularly preferably from 530 to 600 nm, can preferably take place. If pulsed laser light is used, the heat energy can be at an energy density of 1 mJ · cm ^-2 to 500 mJ · cm ^-2 , preferably 10 mJ · cm ^-2 to 200 mJ · cm ^-2 , more preferably 50 mJ · cm ^-2 to 150 mJ · cm ^-2 , and the laser light having a pulse duration of 8 ns to 20 ns and a number of 10 to 100 pulses at a pulse repetition rate of 0.01 · s ^-1 to 10 s ^{-1 is used} with a pulse train ,

Preferably, the energy input is from 10 to 30 pulses of increasing energy density, starting at an energy density of 10 mJ cm ^-2 to 30 mJ cm ^-2 and increasing up to an energy density of 60 mJ cm ^-2 to 200 mJ cm ^-2 . The maximum energy density introduced in this way essentially determines the properties of the layer. The energy density accumulated with the pulses is thus from 400 to 5000 mJ cm ^-2 .

In the method according to the invention, particularly preferably, the input of heat energy by laser light can take place with an energy density of 20 to 100 mJ cm ^-2 . In the method according to the invention, the heat energy can be registered by an Nd: YAG laser with the wavelength 532 nm extremely highly. Most preferably, the thermal energy can be supplied to the sintering of the layer obtained according to step D of the method according to the invention.

The The following illustrations are intended to illustrate the invention Explain structure in more detail without the present Invention limited to these embodiments should be.

1 shows in the SEM image (a), which is labeled with "as deposited", a recording of a layer of doped silicon particles on polyimide film, which was obtained after the treatment steps with hydrogen fluoride and prior to the thermal treatment of the inventive method. The SEM image (b) labeled "60 mJ cm ^-2 " shows the structure used after the thermal treatment of the process according to the invention, at a pulse energy density of 60 mJ cm ^-2 and the input of heat energy by a Nd: YAG laser at a wavelength of 532 nm was obtained.

2 (Prior Art) shows the resistivity as a function of the doping concentration of bulk silicon known in the art doped with the dopant boron and phosphorus, respectively. The picture was taken from the tower of Sze from "Physics of Semiconductor Devices, John Wiley and Sons, 1981" taken from. The resistivity is the reciprocal of the specific conductivity. In contrast to the behavior of the structure used, the boron- or phosphorus-doped bulk silicon shows a uniform decrease in the resistance with the doping concentration. There was no non-linear relationship between the conductivity and the doping concentration.

3 : Dependence of the thermo-power of sintered layers on the doping for boron and phosphorus doping (full or open symbols) with a temperature difference of 320 K. The lower scale of the boron concentration simultaneously indicates the real doping concentration.

4 : Sketch of a thermoelectric element or Peltier element made of sintered printed silicon layers of different polarity in a lateral configuration.

5 : Sketch of a thermoelectric element or Peltier element made of sintered printed silicon layers of different polarity in a vertical configuration.

in the The following will illustrate the subject matter of the present invention described, without the invention, the scope of which is apparent from the Claims and the description is limited thereto should be.

Examples:

example 1

Possible process steps: Specified Method steps a) to j), which serve to produce a thermoelectric Bauelemenst. Steps k) to l) serve for electrical contacting and characterization.

a) substrate

• Polyimide film Kapton ^® (DuPont ^TM) 125 microns thick
• Clean with acetone / isopropanol
• Dry with dry nitrogen

b) structuring

• Spinning up positive photoresist S1818 (Microposit) 40 s @ 4000 rpm
• Drying for 10 min. in oven @ 90 ° C
• Exposure 10 s through lithographic mask
• Develop for 15 s in developer (Microposit)
• Rinse in water, dry bubbles with stock
• Alternatives: Structuring with printing process
• Structuring by laser ablation etc.

c) silicon layer

• Dispersion 6% wt. in ethanol
• Dispersing in shaker with ZrO ₂ beads @ 1000-1400 rpm 4 h-18 h
• (Eppendorf Thermomixer compact)
• Spinning 40 s @ 1000-2000 rpm
• Dry 5-10 min @ 90 ° C oven
• Layer thickness typically 0.5-1.0 microns

d) possibly lift-off for structuring:

• Immerse in acetone for 10 sec, rinse with acetone, dry bubbles with nitrogen

e) possibly renewed application of photoresist w. o. For structuring a second layer before the laser sintering the first, a) -d)

f) wet chemical etching

• Immerse in 5-10% HF for 20 s
• Rinse in water, blow dry with nitrogen

g) laser sintering

• Frequency-doubled Nd: YAG laser at 532 nm: 10 pulses of increasing energy density starting at 20 mJ / cm ² , increasing to 80-140 mJ / cm ² , the best results between 100-120 mJ / cm ² .
• Most samples show a strong decrease in film quality at energy densities above 140-160 mJ / cm ²

h) possibly structuring after laser sintering by lithography b), suitable etching step z. B. reactive Ion etching (RIE), lift-off d)

i) possibly applying the second layer and Structuring after laser sintering of the first layer b) -d), f) -h)

j) applying metallic contacts

• Thermal vapor deposition or sputtering of contacts made of Al, Au, Ag etc. by shadow mask or after structuring b)
• Layer thickness about 80-500 nm
• subsequently lift-off d)
• Alternative: printing of conductive / metallic pastes

k) contacting / wiring

• Connection of suitable plug-in connections with the ones used Measuring instruments are compatible

l) measuring / characterizing

• Measurement of the conductivity at room conditions and between 80 and 450 K in vacuum as a function of temperature with a voltage source / ammeter z. Keithley K6517.
• Applying a fixed temperature gradient of z. B. 100 K-320 K under room conditions, z. B. by a hot tip contact. Measuring the occurring thermal voltage with a high-resistance electrometer, z. Keithley K6517.
• Or: applying a variable temperature gradient with variable average temperature in the temperature range between 80 K and 450 K in vacuum by partial contact with two separate temperature controllers. Measuring the occurring thermal voltage with a high-resistance electrometer, z. Keithley K6517 as a function of mean temperature.

Example 1a

One Substrate a) is purified by the method a) and dried.

To the Application of structured, strip-shaped n-doped Silicon layers on the substrate, the method c) is used, wherein, deviating from method c), a dispersion of 13% by weight n-doped silicon is used in ethanol, which differs from method c) applied by screen printing instead of by spinning becomes.

The dispersion is applied in the form of rectangular strips which are 6 mm wide and 5 cm long. The strips are parallel with each other with their long edge and have a distance of 10 mm from each other, with the short edges on one Line lie.

To the Application of structured, strip-shaped p-doped Silicon layers on the substrate, the method c) is used, wherein, deviating from method c), a dispersion of 13% by weight p-doped silicon in ethanol is used, which is different from method c) applied by screen printing instead of by spinning becomes.

The Dispersion is applied in the form of rectangular strips which 6 mm wide and 5 cm long. The stripes are with their long ones Edge parallel to each other and have a distance of 10 mm from each other, where the short edges are in line.

These Dispersion thus becomes centered between the n-doped silicon stripes applied that they are parallel to each other and their short edges all lie in one line.

The Thickness of the p-doped and the n-doped layer is 5 μm.

After that Laser sintering of the silicon strips is carried out according to the method g).

After that become by the method j) metallic contacts by thermal Evaporation of aluminum applied through a shadow mask. The Contacts are arranged in the form of two rectangular fields, from the first, arranged lengthwise in a row, rectangular contacts, each with a length of 14 mm and whose width is 5 mm. The distance between adjacent contacts in this row is 2 mm each. The second field exists also arranged lengthwise in a row, rectangular contacts, each with a length of 14 mm and their Width is 5 mm, except for the first and last Contact this second series. These two contacts each have only a length of 6 mm with a width of 5 mm. Of the Distance between adjacent contacts in this second row is also 2 mm each.

These both rectangular fields are parallel to each other, with their flush short edges with each other. The distance between the two fields is 40 mm.

The Shadow mask is aligned with the silicon strip so that the long sides of the contacts perpendicular to the silicon strips aligned, the outer edges of the contacts each flush with the outer edges complete the short edges of the silicon strips and so each overlap between the contacts and the ends the silicon strip on areas of 5 mm × 6 mm arises.

The Thickness of the aluminum contacts is 500 nm.

Example 1b

One Substrate a) is purified by the method a) and dried.

To of method j) are metallic contacts by thermal evaporation of aluminum applied by a shadow mask on the substrate. The contacts are arranged in the form of two rectangular fields, of which the first is arranged in a row lengthwise, rectangular contacts, whose length is 14 mm each and whose width is 5 mm. The distance between adjacent ones Contacts in this row is 2 mm each. The second Field also consists of lengthwise in a row arranged, rectangular contacts whose length each 14 mm and its width is 5 mm, with the exception of the first and the last contact of this second series. These two contacts each have only a length of 6 mm at a width of 5 mm. The distance between adjacent contacts in this second Row is also 2 mm each.

These both rectangular fields are parallel to each other, with their flush short edges with each other. The distance between the two fields is 40 mm.

The Thickness of the aluminum contacts is 500 nm.

To the Application of structured, strip-shaped n-doped Silicon layers on the aluminum partially vaporized substrate method c) is used, which differs from the method c) a dispersion of 13% by weight of n-doped silicon in ethanol which is different from the method c) with screen printing instead of being applied by spinning.

The Silicon strips are aligned with the aluminum contacts, that they are strip-shaped, rectangular, parallel to each other and perpendicular to the long sides of the aluminum contacts, wherein the silicon strips are 6 mm wide and 5 cm long and from each other have a distance of 10 mm. The long outer ones Edges of the silicon strips lie in a line with the short edges of the aluminum strips. The short edges of the silicon strips close each flush with the outer edges the contacts off, so that each overlap between the Ends of the silicon strips and the contacts on surfaces of 5 mm × 6 mm is formed.

For applying structured, strip-shaped p-doped silicon layers on the with Aluminum partially vapor-deposited substrate and with n-doped silicon partially coated substrate, the method c) is used, which deviates from the method c) a dispersion of 13% by weight of p-doped silicon in ethanol is used, which differs from the method c) is applied by screen printing instead of by spinning.

These Dispersion thus becomes centered between the n-doped silicon stripes applied to the n-doped and the p-doped silicon strips are parallel to each other and their short edges all in one line lie.

The Thickness of the p-doped and the n-doped layer is 5 μm.

After that Laser sintering of the silicon strips is carried out according to the method g).

Example 2

• 1. Flexible film substrates (polyimide, eg Kapton (Dupont)) are laterally structured with a dispersion of silicon particles in ethanol (6% by weight) in a layer thickness of 1 μm. The doping of the particles is selected so that a doping concentration of 10 ¹⁹ cm ^-3 is present (corresponding to an addition of 0.01% diborane or 0.4% phosphine in relation to the silane used or silane / German when produced in a plasma reactor).
• 2. By a wet-chemical etching step (10% hydrofluoric acid in water, 30 s), the oxide of the particles is removed.
• 3. To realize layer systems becomes a second layer made of silicon particles, partially coated with silicon the first layer overlaps. The dopings of both layers in particular have different polarities.
• 4. With a pulsed Nd: YAG laser (532 nm, 8 ns pulse duration), the layers are thermally treated. Pulse sequences of 10 laser pulses each are used at a pulse repetition rate of 1 / s. The energy density of the laser radiation is increased stepwise within each pulse sequence, up to an energy density of 80-120 mJ / cm ² .

Of the ideal value for the doping of the layers for thermoelectric applications was determined by deliberately different high doping concentrations were used for layer production, and the size of the thermoelectric power thereafter experimentally was determined.

3 shows the result of these measurements. It is clear that the thermoelectric force has a pronounced maximum, which is just above the critical doping in the range of the defect concentration of the material at 10 ¹⁹ cm ^-3 . Below this value, the thermoelectric power disappears due to lack of conductivity of the material. Above this value, the thermoelectric power decreases again continuously. It also turns out that accurate knowledge of the doping efficiency of the material in question is needed to produce efficient bipolar thermoelectric elements. In the figure, values for phosphorus and boron-doped samples are plotted as a function of doping. Since both the electrical conductivity and the thermoelectric force depend critically on the doping concentration, an optimum doping concentration must be set for each of the two polarities.

4 shows a sketch of a thermoelectric element or Peltier element in a lateral configuration on a substrate. To realize this configuration, layers of particles are prestructured on the substrate and then sintered. The mutual metallic connections can be realized either by overlapping the p- and n-layers or by applying metallic contacts to the structure. For tapping a thermoelectric voltage at the end contacts, a lateral temperature gradient across the substrate is necessary.

Alternatively, a thermocouple after 5 will be realized. Here again areas of different doping are structured, but here the temperature gradient is impressed in the vertical direction. Such a thermocouple can be realized either by multiple processing of thin layers in the vertical direction or by an alternative sintering process of self-supporting structures. While the contacts on one side can be made overlap again, on the other side alternating bridge contacts must be applied over the cuts. This structure has the advantage of a significantly reduced electrical resistance due to the larger cross-sectional area.

QUOTES INCLUDE IN THE DESCRIPTION

This list The documents listed by the applicant have been automated generated and is solely for better information recorded by the reader. The list is not part of the German Patent or utility model application. The DPMA takes over no liability for any errors or omissions.

Cited patent literature

• - DE 10353996 [0016, 0041, 0043]
• EP 01760045 A1 [0016]
• - DE 102005022383 [0016, 0043, 0045]

Cited non-patent literature

• - "Thermoelectric Multitalents", Physik Journal 6 (2007), No. 5 [0005]
• Talapin in "Science", 310, p. 86 (2005) [0013]
• - Sun in "Nano Letters", 5, p. 2408 (2005) [0014]
• - Klingshirn in "Physik Journal", 5, No. 1 (2006) [0014]
• - Roth et. al., Chem. Eng. Technol. 24 (2001), 3 [0016]
• - Sze in Physics of Semiconductor Devices, John Wiley and Sons, 1981. [0016]
• - Volker Lehmann describes in "Electrochemistry of Silicon", ISBN 3-527-29321-3, Wiley-VCH Weinheim, 2002 [0017]
• - LJ van der Pauw, A method of measuring specific resistivity and Hall effect of discs of arbitrary shape, Research Reports 13 (1), 1958 [0025]
• - Sze "Physics of Semiconductor Devices, John Wiley and Sons, 1981" [0058]

Claims (28)

Thermoelectric component (thermoelectric element) or Peltier element comprising one or more p-type elements and / or one or more n-type elements, characterized in that the p- and / or the n-type elements are porous semiconducting structures prepared from dispersions of at least partially doped semiconductive particles. Thermoelectric element or Peltier element according to claim 1, wherein the n-type and p-type elements with each other via conductive bridges or by direct conductive contact according to the scheme p _x - [np-] _z n _y , where x = 0 or 1 , y = 0 or 1, z> 0, n = n-type element, p = p-type element, are interconnected, where the indices x, y and z are the number of the respective p-, n- or Specify np element. Thermoelectric element or Peltier element after a of the preceding claims, wherein the conductive bridges of metal, preferably of aluminum, silver, gold, nickel, iron, Cobalt, titanium, tungsten, manganese, chromium, platinum or molybdenum, consist. Thermoelectric element or Peltier element according to one of the preceding claims, wherein the semiconducting structures an electrical conductivity of 5 · 10 ^-8 to 1 · 10 ³ S · cm ^-1 , preferably 1 · 10 ^-5 to 100 S · cm ^-1 , continue preferably 1 × 10 ^-3 to 10 S · cm ^-1 . Thermoelectric element or Peltier element after a of the preceding claims, wherein the semiconductive structures a pore size of 1 to 500 nm, preferably 5 to 250 nm, more preferably 10 to 100 nm. Thermoelectric element or Peltier element after a of the preceding claims, wherein the semiconductive structures an activation energy of electrical conductivity from 0.1 to 700 meV, preferably from 5 to 250 meV. Thermoelectric element or Peltier element after a of the preceding claims, wherein the semiconductive structures a fixed proportion of 25 to 60% by volume, preferably 30 to 45 Vol%, have. Thermoelectric element or Peltier element after a the previous claims, wherein the fixed portion of the structures over Sinter necks is interconnected. Thermoelectric element or Peltier element after a of the preceding claims, wherein the structures at least partially crystalline doped components with sizes from 5 nm to 500 nm. Thermoelectric element or Peltier element after a of the preceding claims, wherein the semiconductive structures the elements silicon, germanium or an alloy of these elements contained and at least partially doped. Thermoelectric element or Peltier element after a of the preceding claims, wherein the doping is selected is made of lithium, phosphorus, arsenic, antimony, bismuth, boron, aluminum, Gallium, indium, thallium, europium, erbium, cerium, praseodymium, neodymium, Samarium, gadolinium, terbium, dysprosium, holmium, thulium, ytterbium, Lutetium, lithium, germanium, iron, ruthenium, osmium, cobalt, rhodium, Iridium, nickel, palladium, platinum, copper, silver, gold, zinc, or from a mixture of these dopants. Thermo-power element or Peltier element according to one of the preceding claims, wherein the p-type semiconducting structure is doped with 1 x 10 ¹⁸ to 1 x 10 ²¹ , preferably 1 x 10 ¹⁹ to 1 x 10 ²⁰ boron atoms cm ^-3 . Thermo-power element or Peltier element according to one of the preceding claims, wherein the n-type semiconducting structure is doped with 1 · 10 ¹⁸ to 1 · 10 ²¹ , preferably 1 · 10 ¹⁹ to 1 · 10 ²⁰ phosphorus atoms · cm ^-3 . Thermoelectric element or Peltier element according to one of the preceding claims, wherein the semiconducting structures have a BET surface area of 10 to 800 m ² · g ^-1 , preferably 30 to 500 m ² · g ^-1 , more preferably 40 to 200 m ² · g ^-1 , more preferably from 50 to 100 m ² · g ^-1 . Thermoelectric element or Peltier element after a of the preceding claims, wherein the semiconductive structures a thickness of 20 nm to 50 microns, more preferably from 50 nm to 10 microns, most preferably from 100 nm to 3 microns, have. Comprising thermoelectric element or Peltier element at least one substrate and a thermo-power element arranged thereon or Peltier element according to one of the preceding claims. A method of making a device for further use in the manufacture of a thermoelectric or Peltier device according to any one of claims 1 to 16, comprising the steps of: a) generating p-type doped semiconductor particles, b) producing n-type doped semiconductor particles . c) production of a first dispersion with the semiconductor particles obtained according to step a), d) generation of a second dispersion with the semiconductor particles obtained according to step b), e) at least partial coating of a substrate with one of the first or obtained after step c) or d) second dispersion, f) treatment of the layer obtained according to step e) by a solution containing hydrogen fluoride and water, g) thermal treatment of the layer obtained after step f), so that a first porous semiconducting structure of the p or n type is obtained, h) use of the dispersion not used in step e) to produce an at least partial second coating on the substrate, on a further substrate or - to form a layer system - on the first coating obtained in step e), i) treatment of the after step h ) obtained by a solution containing hydrogen fluoride and water, j) thermal treatment de r after step i) obtained second layer, so that a second porous semiconducting structure of the other type is obtained. The method of claim 17, wherein in the production the doped semiconductor particles of the p- or n-type continuously a) at least one vapor or gaseous metalloid hydride, selected from silane, germanium hydride or a mixture from these half metal hydrides and at least one vapor or gaseous corresponding dopant, and transferred an inert gas in a reactor and mixed there, wherein the proportion of Halbmetallhydrids between 0.01 and 90 wt .-%, preferably between 0.1 and 20 wt .-%, based to the sum of half metal hydride, dopant and inert gas, and subsequently b) by energy input by means of electromagnetic radiation in the microwave range at a pressure of 1 to 1100 hPa, preferred 10 to 300 hPa, a plasma is generated, and subsequently c) Allow the reaction mixture to cool and that Reaction product in the form of a powder of gaseous Material is separated, with predominantly crystalline doped semiconductor particles of the corresponding type can be obtained. Method according to one of claims 17 to 18, whereby in the production of the doped semiconductor particles of the p-type or n-type dopant is used, the selected is made of lithium metal or lithium amide (LiNH2), from the group the hydrogen-containing compounds of phosphorus, arsenic, antimony, Ice-muth, boron, aluminum, gallium, indium, thallium, europium, erbium, Cerium, praseodymium, neodymium, samarium, gadolinium, terbium, dysprosium, holmium, Thulium, ytterbium, lutetium, lithium, germanium, iron, ruthenium, Osmium, cobalt, rhodium, iridium, nickel, palladium, platinum, copper, silver, Gold, zinc, or a mixture of these dopants. Method according to one of claims 17 to 19, wherein in the production of the respective dispersion doped semiconductor particles having a mass fraction of 0.1 to 60 wt .-%, preferably 1 to 30 wt .-%, particularly preferably 2 to 12 wt .-% based on the total mass of the dispersion, and the doped semiconductor particles are dispersed in a liquid phase comprising or consisting of water and / or at least one organic solvent to give a dispersion having a viscosity at 23 ° C of less than 100 mPas, more preferably less than 10 mPas, at 1000 s ^-1 . Method according to one of claims 17 to 20, wherein in the first and / or second coating, the respective dispersion by spin-coating, knife coating, dropping, screen printing, dip coating and spray coating applied to the substrate and then expelled the liquid phase by the entry of thermal energy to obtain each a layer having a conductivity of 5 × 10 ^-8 to 1 × 10 ³ S · cm ^-1 , preferably 1 × 10 ^-5 to 100 S · cm ^-1 , more preferably 1 × 10 ^-3 to 10 S · cm ^-1 . Method according to one of claims 17 to 21, wherein one or more substrates are used, each and optionally independently selected made of organic, inorganic material, or a composite at least one organic and / or at least one inorganic Material, wherein the respective substrate resistant to hydrofluoric acid is. Method according to one of claims 17 to 22, wherein in the respective treatment step with hydrogen fluoride, a solution containing hydrogen fluoride and water in a concentration of the hydrogen fluoride of 0.01 wt .-% to 50 wt .-%, preferably from 1 to 30 wt. %, more preferably from 5 to 15 wt .-%, based on the total mass of the solution containing hydrogen fluoride and water, by immersion, dripping, spraying or rinsing, and at a temperature of 10 ° C to 70 ° C, preferably 15 ° C. to 40.degree. C., more preferably from 18.degree. C. to 30.degree. C., and over a period of from 1 second to 1 hour, preferably for a period of from 5 seconds to 5 minutes, more preferably from 10 to 60 seconds, respectively obtaining a layer having a conductivity of 5 × 10 ^-11 to 10 ^-7 S · cm ^-1 , preferably 10 ^-10 to 5 × 10 ^-8 S · cm ^-1 . A method according to any one of claims 17 to 23, wherein each of the after treatment Steps with hydrogen fluoride layer is immersed in a further step in water or rinsed and then blown off with dry nitrogen, or is then dried at a temperature of 60 to 200 ° C. Method according to one of claims 17 to 24, wherein the respective thermal treatment is carried out by introduction of heat energy with a power density of 10 kW · cm ^-2 to 65 MW · cm ^-2 , wherein the particles of the obtained after the treatment with hydrogen fluoride layer at least partially sinter. The method of claim 25, wherein the entry of Heat energy by laser light with one wavelength from 500 to 1100 nm, preferably from 505 to 850 nm, more preferably from 510 to 650 nm, more preferably from 530 to 600 nm. The method of claim 26, wherein the input of heat energy by pulsed laser light and the heat energy with an energy density of 1 mJ · cm ^-2 to 500 mJ · cm ^-2 , preferably 10 mJ · cm ^-2 to 200 mJ · cm ^-2 , more preferably 50 mJ · cm ^-2 to 150 mJ · cm ^-2 , and the laser light having a pulse duration of 8 ns to 20 ns and a number of 10 to 100 pulses at a pulse repetition rate of 0.01 · s ^-1 to 10 s ^{-1 is used} with a pulse train. Use of dispersions of at least partially doped semiconductive particles for producing a thermoelectric element or Peltier element, preferably according to one of the claims 1 to 16.