DE102012003467A1 - Solar cell, has tunnel layer made of non-conductive rare earth oxides or rare earth oxy chalcogenides, and collector layer comprising energy difference between electron affinity and donor level, where difference is lesser than electron volt - Google Patents

Abstract

The cell has a tunnel layer (3) made of non-conductive rare earth oxides or rare earth oxy chalcogenides. The cell functions as an electron emitter and is made of yttrium compounds that are selected from a group consisting of mono sulfides, monoselenides, monotellurides, antimonides, bismuthides, germanides, stannides, plumbides, aluminides, gallide, indides or magnesides. A collector layer (4) comprises an energy difference between an electron affinity and a donor level near a surface beside the emitter, where the difference is lesser than an electron volt. The electron emitter functions as a photocathode, a low-light enhancer or LED.

Description

Photovoltaic cells according to the prior art consist of two semiconductors of opposite doping, which are contacted to one another. As with any contact between a p-type and an n-type semiconductor, electrons pass from the n-type semiconductor to the p-type semiconductor and pits from the p-type semiconductor to the n-type semiconductor, compensating for the charges, respectively. Overall, a zone forms in the interface, which is depleted of charge carriers. The N-type semiconductor charges positively on the P-type semiconductor negatively. As a result of the electric field building up, the charge flow is inhibited, the flow of the electrons into the P-type semiconductor and the holes into the N-type semiconductor comes to a standstill as a result of the applied field direction. The larger the charge carrier concentrations are, the faster this equilibrium state is achieved, the thinner the depletion zone.

When light enters the contact surface, incident photons form energy hole pairs there and in deeper layers of the semiconductor electron-hole pairs. Photons whose energy content is larger than the bandgap result in the formation of the electron-hole pairs within about 10 ^-14 seconds. The charge carrier pairs are separated by the existing electric field at the interface between n-type and p-type semiconductor. Electrons go into the donor level of the N-type semiconductor, holes occupy the acceptor level of the P-type semiconductor. The electrons travel through the N-type semiconductor, into a suitable arrester and from there through the external circuit. They recombine in the valence band of the P-type semiconductor (internal photoelectric effect). According to the state of the art, copper indium diselenide, in which indium can be partly replaced by gallium and selenium partly by sulfur, cadmium telluride and above all silicon are used as semiconductors. Silicon has a band gap of about 1.1 electron volts (eV), while the other semiconductors have somewhat higher band gaps. Incident photons with energies less than the bandgap can not cause charge carriers to lift from the valence band into the conduction band. However, energy exceeding the bandgap level can not be used. Higher energy carriers lose their increased energy within about 10 ^-12 seconds by collisions with the atoms of the crystal lattice before they reach the charge carrier levels and their surplus energy is usable; they become "thermalized".

Their increased energy over the band gap is lost as warming of the crystal lattice. This behavior is commonly known as the Shockley-Queisser-Limit. Thereafter, the theoretical efficiency at band gaps of 0.8 to 1.5 eV is limited to values between 30 and 33%. Additional effects, such as unwanted recombination of carriers, further reduce practical efficiency to levels of about 15% for commercial crystalline silicon based photovoltaic cells.

Photovoltaic cells based on crystalline silicon have a further disadvantage:
Crystalline silicon is an indirect bandgap semiconductor, which causes its light absorption to be greatly decreased compared to direct bandgap semiconductors. Therefore, in silicon cells, it is necessary to size the thickness of the photon absorbing layer to at least 200 micrometers, otherwise the absorption is too low and the efficiency drops to uneconomic levels. Thus, a silicon layer of 200 microns thickness absorbs about 90% of the intensity of the incident solar radiation.

To obtain higher efficiencies, various solutions have been proposed:
In tandem cells (multi junction cells), the incident photons successively stimulate up to four different semiconductor layers with different energy gaps, for each of which the Shockley-Queisser limit applies. So you can achieve efficiencies up to 45%. However, the tandem cells have up to 20 layers, which must be transparent and error-free for the respective wavelengths. These layers must be prepared in costly processes such as molecular beam epitaxy. Therefore, such cells are very expensive and hardly found in terrestrial applications.

As another way to increase the efficiency of photovoltaic cells, it has been proposed to either convert a high-energy photon into two lower-energy (down-conversion) or, conversely, two low-energy photons into a high-energy photon (up-conversion). This led to the possibility to achieve higher efficiencies with an optimally matched to the energy of each generated photon semiconductor by means of the internal photo effect. There would be more suitable photons available for use than the original solar spectrum.

Unfortunately, however, the "transformed" light is scattered in all directions; it only partially affects the active P / N transitions, resulting in high losses.

With the introduction of an intermediate band into the energy gap of the semiconductor to produce two different sized band gaps in a single semiconductor, one also tries to make better use of the energy distribution of the photons. But even this approach was previously unsuccessful in practice.

With so-called "optical rectennas" one tries to absorb light and heat as electromagnetic waves corresponding to their short wavelengths with a large number of short, submicrometer-long antennas arranged on one surface. The AC voltage generated at the antenna is rectified, comparable to the function of an early detector receiver. This principle works in the case of microwaves with frequencies up to the terahertz range with frequencies of 10 ¹² per second. However, fast diodes, Schottky diodes, are too slow to handle AC voltages of significantly higher frequencies of heat or even visible light, at frequencies of 10 ¹⁴ to 10 ¹⁵ per second. Therefore, only extremely low efficiencies of far below one percent have been achieved for the direct conversion of heat radiation through Rectennas. With very fast tunnel diodes with a metal-insulator-metal transition, an attempt is made to make the optical rectennas usable for visible light. ( IEEE Journal of Photovoltaics, Vol 1, No 1, July 2011, p. 78-83 ).

After registration DE 10 2010 010658 For example, low-work-function intermetallic compounds are used in photoelectric generators which do not use the internal photoeffect, as is done in the prior art, but the external photoeffect or photoemissive effect described by Einstein. According to this application, the high kinetic energy electrons emitted by the low work function emitters are not injected into a vacuum but into a semiconductor (solid state device).

Also the use of the photoemissive effect claims the application GB 246 8526 (Appl. Date 15.09.2010). There are used as emitter alkali metals with work functions around 2 eV.

The electrons emitted by the alkali metals have at most the energy of the photons of the incident light minus the work function. They are intended to tunnel through an oxidic insulator layer such as magnesium oxide, calcium oxide or barium oxide with layer thicknesses of less than 5 nanometers to form a metallic collector of higher work function. In contrast to thermionic generators, the kinetic energy of the emitted electrons should be used. The "hot" electrons should increase the kinetic energy of the electron gas of the collector, which should result in the usable potential difference. The insulator layer causes the electrical separation of emitter and collector; Energy bands between emitter and collector are not bent. This intrinsically attractive arrangement has the disadvantage that only photons with energies greater than 2 eV are used. In addition, it would be necessary for the metallic collector, here aluminum, itself to have the lowest possible work function. As with photovoltaic cells after the internal photoelectric effect, the electrons entering the collector also lose the proportion of energy corresponding to the difference between the work function of the collector material and the Fermi level.

The patent application DE 10 2011 102 886 describes generators for direct conversion of light and heat into electricity, also using low-work function intermetallic compounds. Here, however, the collector also has a higher work function than the emitter.

The use of the energy of "hot" charge carriers is a goal of current development of higher-efficiency photovoltaic cells (hot carrier cell). However, experimentally proven successes have not been reported so far. A good overview gives the patent application WO 2011/000055 , Pub. 06.01.2011 "Hot Carrier Energy Conversion Structure And Method Of Fabricating The Same".

It was therefore an object of the invention to provide an arrangement which does not have the disadvantage of thermalization of the charge carriers. Emitters or absorbers should be used as low as possible work function, if possible lower than 2 electron volts (eV). This arrangement should be producible by practicable, reproducible and economical processes for large areas. In addition, the emitter materials used should have the highest possible quantum yields.

It should also be dispensed with the use of toxic elements such as cadmium, mercury, thallium or arsenic. Expensive elements such as rare earth metals should only be used in very thin layers. In addition, layers with optimal properties were found, which allow the connection of the electron emitter to the entire system without significant loss of efficiency.

These tasks are solved by using rare earth compounds of very low work function below 2 eV as emitter. Depending on their electronic structure, these rare earth compounds may be metals, semimetals or semiconductors with a narrow band gap. Accordingly, upon impact of photons on the emitter, charge carriers can be generated either by the external photoemissive photoelectric effect as well as by the internal photoelectric effect. Due to the low operating temperatures, charge carriers from the collector, which would lower the efficiency, can hardly be emitted thermally. It is not possible for practical use but also not necessary to separate the individual processes, which lead in the emitter or on the emitter surface for generating the charge carriers, apart. Collisions between higher-energy and lower-energy electrons, or, to represent it in the wave pattern of electrons, interference between electron waves of different energy cause the broad distribution of the energy of the incident photons is desirably narrowed, whereby the energy distribution of the emitted electrons is narrower than that the incident photons.

Between emitter and collector, a very thin, because of their wide band gap transparent rare earth oxide layer is produced, which allows fast tunneling of hot electrons from the emitter to the collector, the adverse coupling and alignment of. Energy levels between emitter and collector is thus avoided.

As the collector, a semiconductor transparent to the incident light wide band gap is used, which has a low work function and has a donor level in the height of the work function of the collector surface. Semiconductors that meet these requirements have low electron affinities or negative electron affinities on their surface.

Their donor level is generated by a suitable n-doping and thus occupied with electrons. The donor level should be as close as possible to the vacuum level of the surface. With this arrangement of the levels, it is achieved that the electrons, which have passed through the tunnel layer and are injected into the vacuum level of the collector material, reach the donor level without any particular energy losses. The donor levels in n-doped semiconductors are usually below the lower band edge of the conduction band. If the vacuum level of the collector surface is to be close to the donor level, then this also means that it must also be below the band edge of the conduction band. In this case, there is a negative energy difference of vacuum level and lower conduction band edge and thus a negative electron affinity. Materials with surfaces of negative electron affinity are known to emit electrons very easily into the vacuum. Therefore, as a number of patent applications show, they are of great importance as particularly effective electron emitters for technical applications such as thermionic converters.

Conversely, surfaces with negative electron affinity are very good absorbers for electrons incident from the vacuum or from the energy comparable tunnel layers. Thus, N-type semiconductors used as a collector according to the invention preferably have a surface with negative electron affinity.

Sketch 1 schematically shows the working principle using an energy diagram:
The emitter E has the lowest possible work function W _E ; correspondingly high is his Fermi level E _FE . The collector K is a wide bandgap semiconductor above 3.5 eV between the valence band VB and the conduction band LB. Because of the wide band gap, the collector is continuous for the incident radiation. Accordingly, the conduction band in the energy diagram is relatively high, in the height of the vacuum level. The donor level of the collector E _FK is just below the minimum of the conduction band. If a photon with an energy content greater than the work function of the emitter W _E on the emitter, either an electron of energy E _photon - W _{E is} triggered photoemissively or one with the energy of the photon minus the band gap when it is at the emitter a semiconductor. This electron traverses the tunnel layer and hits the surface of the collector without any particular loss of energy. There it hits the level of the local vacuum level.

Under the loss of energy between the vacuum level of the surface, E _{Vak, Ob} , and the donor level E _D , this difference corresponds to the work function of the collector, it drops to the donor level of the collector, or if the donor level is higher than the vacuum level lies, overcome this gap while losing energy. At the level of the donor level, the external circuit is operated.

Collectors according to the invention behave like thermionic collectors, which due to no or very low work functions oppose the entering electrons no barrier. The electron enters the surface vacuum level of the collector with the energy content it has when exiting into the vacuum level of the emitter. Thus, the inventive arrangement combines the photoemissive and the photovoltaic release of electrons on the emitter side with the absorption of the emitted electrons of a thermionic collector.

The usable energy is E = E photon work function emitter work function collector.

If the difference between surface vacuum level and donor level of the collector and thus its work function is low, then a significant proportion of the energy of the electron and thus of the incident light can be used. Based on the energy of the incident photons corresponds to the energy loss of the applied work function of the emitter W _E plus the difference in energies between the vacuum level and donor level of the collector, the work function of the collector.

The exemplary energy scheme of sketch 1 is based on values published and accepted in the scientific literature:
Reference level is the vacuum level, whose energy corresponds to zero eV by definition. The values given are energy differences. The emitter is lanthanum monosulfide with a work function of 1.14 eV. Collector is a doped with phosphorus diamond film, which is deposited by means of chemical vapor deposititon from a methane / hydrogen plasma containing phosphine as a precursor for the dopant phosphorus. The band gap of the diamond film is 5.5 eV. The donor level generated by the doping is 0.6 eV below the conduction band minimum, the work function of the bulk material is 1.0 eV.

Thus, the vacuum level of the bulk material is 0.4 eV above the conduction band minimum. Due to the bonding of the hydrogen at the surface, the vacuum level is lowered there so that a negative electron activity of 0.8 eV is obtained. At the surface of the diamond film, the vacuum level is thus 0.8 eV below the conduction band minimum and 0.2 eV below the donor level. The energy of the usable electrons corresponds to the energy of the incident photons reduced by the amount of the work function of the emitter and the difference between the donor level and the surface vacuum level of the collector, ie the work function of the collector. In this example, the energy loss to the photons is 1.14 eV + 0.2 eV = 1.34 eV.

Assuming a work function of the emitter of 1 eV and the surface vacuum level of the collector optimally coincides with its donor level, the energy loss is only 1 eV. For this reason, it is important to use emitter materials with the lowest possible work function and to design the electronic states of the collector K such that the donor level is as close as possible to the surface vacuum level.

Sketch 2 shows the structure of the cells according to the invention schematically:
A metallically conductive back contact ( 1 ) with the emitter ( 2 ) coated very low work function. Photons incident on this emitter layer with an energy greater than their work function emanate electrons from the emitter surface whose energy corresponds to the difference between the photon energy and the work function of the emitter, in the case that the emitter is metallic or semimetallic. If the emitter is semiconducting, electron-hole pairs are formed close to the emitter surface, similar to prior art photovoltaic cells, which break down into electrons of the corresponding energy and holes. Because of the lower effective mass of the electrons compared to the effective mass of the holes, the majority of the energy of the electron-hole pairs is after their decay in the electrons.

The released electrons tunnel through the tunnel layer ( 3 ) and hit the collector ( 4 ) on. Through the tunnel layer emitter and collector are energetically separated from each other, there is no adverse band bending, by which the work function of the emitter can be increased.

The collector material has on its surface as low as possible or negative electron activity. Thus, the vacuum level is adjusted to the donor level, there is a slight transition from the vacuum level to the donor level. The electrons enter directly into the vacuum level of the collector surface without detour. At most, they only have to overcome the small distance between vacuum level and donor level.

Since the collector, like the emitter, should be very thin, it is used to improve the dissipation of the electrical charges with the layer of a thicker transparent electrically conductive oxide ( 5 ) coated. On this layer, arrester strips ( 6 ) applied according to the prior art. An antireflection coating ( 7 ) is also applied according to the prior art. From the arresters ( 6 ), the electrons flow through the consumer ( 8th ) and from this via the back contact ( 1 ) back into the emitter layer.

Particular care must be taken in the selection of emitter materials with a particularly low work function:
Recent work describes rare earth monosulfides as thermodynamically stable compounds with attractive low work functions. Very low work functions are measured on the rare earth-rich monosulfides of the SES (SE = rare-earth metal) type, as for NdS 1.36 eV, for CeS 1.05 eV or for LaS 1.14 eV ( PhD Thesis Yamini Modukuru, Univ. of Cincinnati, Electrical Engineering, 2003; Electrochemical Society Proceedings, Volume 2002-18, "Cold Cathodes", pp. 365-371, ISBN 1-56677-342-3 ).

The rare earth monosulfides are at the transition from semiconductors to semimetals with respect to their electronic properties. The lowest level of the conduction band is only slightly above the uppermost level of the valence band, or it overlaps slightly with the uppermost level of the conduction band. However, the conduction band is hardly occupied by electrons. This leads to optically induced electron transitions; the compounds are colored. Cermonosulfide shows as the lanthanum monosulfide a golden yellow color. The electrical resistances are pleasingly low in the range of 20 to 200 microohms by centimeters.

The melting points of the rare earth monosulfides are 2,000 ° C to 2,500 ° C (LaS: 2,200 ° C, CeS: 2,450 ° C). This results in very stable, highly refractory compounds.

A summary of the latest state of scientific work on rare earth monosulphides is available in the article "Rare-earth monosulfides as durable and efficient cold cathodes" dated 13.07.2011 of the Cornell University Library at http://arxiv.org/abs/1107.2547.

A possible explanation for the low work function of the rare earth monosulfides is that in the sulfides due to the thermodynamic equilibrium between monosulfide and sesquisulfide after 3SES ↔ SE ₂ S ₃ + SE an excess of rare earth metal is present. The phase diagrams of the rare earth sulfides allow a stoichiometric excess of rare earth metal and thus the atomic distribution of rare earth metal in the sulfide. In this case, a high "electron pressure" consists of the f-electrons of the rare earth metal in the sulfide, which raises the uppermost electron-occupied level accordingly close to the vacuum level and thus leads to a reduction of the work function ( DE 10 2011 102 886 ). The solubility of the rare earth metal in the sulfide, in addition to greatly reducing the work function, also has the advantage that the vapor pressure of the dissolved rare earth metal greatly decreases, resulting in increased creep strengths of the electron emitters.

Analogous conditions also apply to the selenides and tellurides of rare earth metals. They are deep-colored, so EuSe is dark brown in color. The melting points of the monoselenides are slightly lower than those of the sulfides, namely at 1,800 ° C (CeSe) to 2,000 ° C (LaSe). As expected, the melting points of the rare earth monotellurides are below those of the monoselenides. For example, CeTe has a melting point of 1,800 ° C, while LaTe has a temperature of 1,760 ° C. For reasons of analogy, it is to be expected that the work functions of the rare earth monoselenides and the rare earth monotellurides are below those of the rare earth monosulfides.

The combination of extremely low work functions and high thermodynamic stabilities make the rare earth monosulfides as well as the monoselenides and monotellurides effective electron emitters.

The rare earth monosulfides, rare earth monoselenides and rare earth monotellurides, with their low work function compared to superficially activated materials according to the prior art, have the advantage that they are extremely thermodynamically stable as 'bulk', characterized by concrete phase diagrams.

Thin layers of rare earth monosulphides, rare earth monoselenides or rare earth monotellurides or mixed compounds thereof are used as electron emitters. These correspond to the formulas SES SESe Sete SES _x Se _y SES _x Te _y with x + y = 1 SESe _x Te _y with x + y = 1 or SES _x Se _y Te _z with x + y + z = 1

Although yttrium does not belong to the rare earth metals, it behaves like rare earth metals, so that yttrium sulfide, YS, melting point 2,060 ° C, as well as yttrium selenide and yttrium telluride also belong to the electron emitters according to the invention.

For economic reasons, it may also be advantageous, instead of a defined rare earth metal, to use a mixture of rare earth metals, the so-called mischmetal, for producing the rare earth compounds used according to the invention.

The rare earth monosulfides are prepared by reacting the inexpensive rare earth oxides with sulfur compounds such as hydrogen sulfide or dithiocarbamates to the sesquisulfides in a first step. The sesquisulfides are reacted in the second step in solid phase as a powder with the corresponding rare earth metal at temperatures around 1800 ° C. For example, lanthanum monosulfide is obtained by the reaction La + La ₂ S ₃ ↔ 3LaS.

Formally, trivalent lanthanum with zero-valent, elemental lanthanum is reduced to divalent lanthanum. Further possibilities of producing the rare earth monosulfides, for example by reacting the sesquisulfides with the rare earth hydrides: Thesis by Kevin Gibbard, Univ. of Florida, 2005 "High Temperature Synthesis of Cerium Sulphides and Kinetic Modeling" (http://etd.fcla.edu/UF/UFE0010471/gibbard_k.pdf) ,

The rare earth monosulfides, selenides or tellurides can be pressed by hot pressing, for example, into sputtering targets, from the use of which thin layers of the rare earth compounds can be produced.

In many papers related to lanthanum monosulphide because of its low work function around 1 eV, it is stated that after deposition by pulsed laser deposition, crystalline lanthanum monosulphide in islands of more or less amorphous material or other crystalline orientations premises ( Journal of Nanomaterials, Vol. 2008 (2008) Article ID 682920 or as quoted above, Rare-earth monosulfides ..., http://arxiu.org/abs/1107.2547 ). While the monosulfide has the desirable low work function, the islands environment exhibits work functions of 2.8 to 3 eV. A simple explanation for these differences arises from the chemical behavior of the rare earth monosulfides:
At the high temperatures under the deposition conditions of the laser evaporation, the back reaction gets more weight according to the above equation, 3LaS → La + La ₂ S ₃ whereby one or a few monolayers of lanthanum and the sesquisulfide are formed on the surface. The analysis of lanthanum monosulfide evaporation supports this explanation: LaS vaporization involves 3 to 10% of the vapor as lanthanum and sulfur ( J. Phys. Chem., 1965, 69 (8), p. 2684-2689 ; J. Chem. Phys. 68, 3978 (1978) ; J. Chem. Phys. 65, 2400 (1976) ). Condensation on the relatively cold substrate not only forms the monosulfide, but also separates out the sesquisulfide, higher sulfides, and unreacted rare earth metal, which in combination form the undefined regions around the monosulfides.

After rapid cooling of the thin layer, the surface occupation is "frozen" with the sesquisulfide, it can not react back to the monosulfide with the elemental rare earth metal. The concentrations of sesquisulfide and rare earth metal are so low that they can not be detected by X-ray diffraction. However, they are so large as to determine the work function of the surface, namely, just the value of 2.8 to 3 eV, which corresponds to the values of the work function of the sesquisulfide at 2.8 eV and the lanthanum at 3.1 eV. The presence of sesquisulfide is also suggested by the reduction of oxy-sulfide form La ₂ O ₂ S by the presence of hydrogen sulfide during laser evaporation. However, such a reduction does not lead to the monosulfide, but to the sesquisulfide with its high work function. In addition, under the high temperatures presented by hydrogen sulfide, the formation of sesquisulfide diminishes 2LaS + H ₂ S → La ₂ S ₃ + H ₂ promoted.

According to this knowledge, it is advantageous to use coating processes for the separation of the mono compounds, which avoid the very high temperatures of the laser evaporation and thus the decomposition into the elements. Such a method is, for example, the sputtering (sputtering). Improved process conditions, which give rise to liquid or solid nanoclusters during the laser evaporation, are also expected to increase the yield with respect to area fractions of low work function by 1 eV by this deposition technique ( Rare-earth monosulfides ..., http://arxiv.org/abs/1107.2547 ). It is therefore necessary to ensure that the surfaces of the mono compounds are present with their low work function. In order to keep the surface free of sesquivolinkers, it is advisable to heat the emitter surface for a few minutes and, if appropriate, to coat in a second step with a small amount of rare-earth metallic metal and also to heat the layer for a short time in order to obtain a surface. which consists completely of the mono connection.

For economic reasons, it will be preferable to use the rare earth monosulfides over the selenides or tellurides; Selenium and tellurium are much more expensive than sulfur.

Further inventively usable electron emitter thermodynamically stable intermetallic compounds of rare earth metals with very low work function, as described in the German patent application DE 10 2010 004 061 are listed. These compounds contain a rare earth metal component as an electron donor which "squeezes" electrons into the acceptor component antimony, bismuth, germanium, tin, lead, aluminum, indium, gallium or magnesium to give particularly low leakage work.

These are the antimonides and bismuthides of rare earth metals (melting points in parentheses) such as CeSb (1760 ° C), La ₃ Sb ₂ (1690 ° C), SmSb (1875 ° C) or SbY (2310 ° C). Suitable low-work-function emitter materials are, for example, the bismuthides NdBi or YBi. These have melting points around 2,000 ° C.

Although yttrium is not formally one of the rare earth metals, it behaves very similar as stated above and has a comparable electronegativity. Therefore, the antimonide SbY is also suitable as an emitter.

The rare earth metals lanthanum, cerium, neodymium and samarium or mixtures of rare earth metals, so-called mischmetal, are also preferred here for economic reasons.

Further emitter materials for the purposes of this invention are the high-melting germanides, stannides or plumbides such as La ₅ Ge ₃ (1475 ° C.), LaGe _2-x (1500 ° C.), Ce ₅ Ge ₃ (1500 ° C.), CeGe _2-x ( 1.513 ° C), Ce ₅ Sn ₄ (1.530 ° C), La ₅ Sn ₄ (1600 ° C), Sm ₅ Sn ₃ (1.505 ° C), La ₅ Pb ₃ (1450 ° C), Sm ₅ Pb ₃ ( 1,580 ° C) or Ce ₂ Pb (1,380 ° C).

As further emitter materials, the aluminides, indides and gallides of rare earth metals can be used.

Such are aluminides such as LaAl ₂ (1450 ° C), GdAl ₂ (1480 ° C), CeAl ₂ (1480 ° C), SmAl ₂ (1490 ° C), NdAl ₂ (1460 ° C) or EuAl ₂ (1300 ° C ) as well as YAl ₂ (1.455 ° C),
Gallides such as CeGa ₂ (1.444 ° C), LaGa ₂ (1450 ° C), NdGa ₂ (1.488 ° C), SmGa ₂ (1380 ° C),
Indides such as SmIn ₃ (1,130 ° C), SmIn (1,210 ° C), NdIn ₃ (1,220 ° C), NdIn (1,230 ° C), LaIn ₃ (1,140 ° C), LaIn (1,125 ° C), La ₃ In ₅ (1,185 ° C).

Another very important class of compounds for the photoelectric generators of the present invention is negatively charged magnesium intermetallic compound, called magnesides, referred to in the English literature as 'Magnides'. With the magnesides is the extreme case of a really very electropositive acceptor, the magnesium, before. This atom, which seeks to form Mg ²⁺ ions by electron donation and thus minimize its energy, is forced by the strong donor still additional negative charges. The work functions are correspondingly low. Such magnesides are
CeMg ₃ (790 ° C), LaMg ₃ (800 ° C), NdMg (800 ° C), SmMg (796 ° C) or SmMg ₂ (744 ° C).

Many of these magnesides are known as 'Laves phases'.

The listed low-work-function rare earth compounds may be both semiconductors, semimetallic in character, or behave like metals.

The invention is not limited to binary connections. Both the rare earth component and the complementary binding partner may consist of several elements of similar work function. However, the binary compounds are preferred for their thermal stability, higher melting points, simpler phase diagrams, reproducible coatings, and easier accessibility.

The listed rare earth intermetallic compounds may contain an excess of rare earth component to further reduce their work function. This excess may be 0.01 to 10 atomic percent, preferably 0.1 to 1 atomic percent.

This procedure is possible because, according to the phase diagrams in these intermetallic compounds, an excess of rare earth metal dissolves atomically in the solid state. It is also advantageous to dissolve in the emitter materials a small excess of rare earth metal relative to the stoichiometry, because such an excess makes it possible to produce the insulating tunnel layer in a practicable manner.

A crucial property in the application as material for an emitter is its quantum efficiency. For metals, this is only 0.001 to 0.1%, which means that only one electron exits the material surface per thousand to one hundred thousand light quanta. However, the quantum yield increases with decreasing work function. The Alkaliantimonide, Cs ₃ Sb or K ₂ CsSb with work functions around 1.5 eV have as metallic materials after all quantum yields of 1 to 15%, Cs ₂ Te up to 20%.

As materials of very low work function, as well as because of the donor to acceptor linkage, the rare earth element aluminides, gallides, indides, and magnesides have significantly higher quantum yields. Because of the substantial amount of chemical bonding, there are much fewer free electrons compared to metals that reflect incident radiation or to transform the energy of radiation into useless heat by means of collisions. On the other hand, enough electrons are still available to allow a high electrical conductivity. The high proportion of electron binding can be seen from the melting points: The melting point of CeAl _{2 is} 1,480 ° C, the melting point of cerium at 795 ° C and that of aluminum at 660 ° C are much lower. For comparison, the cesium antimonide Cs ₃ Sb, which is used as an emitter in photomultipliers, has a melting point of 725 ° C, the cesium melting point is 28 ° C, that of antimony is 630 ° C, indicating that the chemical bond is in the Ceraluminid is stronger and its quantum yield is thus greater than that of cesium antimonide.

This applies in a special way also for the connection formation between rare earth metal and sulfur, selenium or tellurium. Thus, the rare earth monosulfides, rare earth monoselenides and rare earth monotellurides also have high quantum yields.

The high proportion of electron binding can be read from the high melting points of the compounds in the range of 1,800 to 2,500 ° C. The rare earth metals melt at relatively low temperatures of 800 to 1100 ° C, sulfur, selenium and tellurium melt below 450 ° C.

Particular care should be taken when choosing the material of the tunnel layer ( 3 ):
The tunneling layer should be as thin as possible in order to enable the tunneling effect even at low potential differences and also have a sufficiently high energy gap to be transparent. From the Microelectronics are tunnel layer thicknesses of 0.2 to 10 nanometers, preferably 0.5 to 5 nanometers for tunneling layers known. This layer thickness range is also used in the arrangements according to the invention. Is the voltage difference between emitters ( 2 ) and collector ( 4 ) For example, 0.1 volts and the distance between the two 1 nanometer, the applied field strength is 10 ⁸ volts per meter. Under such conditions, the hot electrons tunnel from the emitter ( 2 ) to the collector ( 4 ) without losing any of their energy significantly. The time for the tunneling of the electrons is 10 ^-14 to 10 ^-16 seconds (eg. Phys. Our time 2/2009 (40), pages 6, 7 ) and is thus much shorter than the time of the thermalization with about 10 ^-12 seconds. These ratios allow injection of the hot electrons into the collector ( 4 ).

As tunneling materials, the rare earth oxides are used, which are produced from the emitter layer by superficial oxidation. Rare earth oxides have energy gaps of 3.5 to 5 eV (eg La ₂ O ₃ : 4.3 eV); With melting temperatures of 2,300 ° C to 2,500 ° C, they are thermodynamically very stable. A disturbing electrical conductivity of the rare earth oxides by the transport of oxygen ions must not be feared at the temperatures of application far below 600 ° C. These properties led to the increased use of rare earth oxides as gate dielectrics in microelectronics ( s. N. Imanaka, "Physical And Chemical Properties of Rare Earth Oxides" in "Binary Rare Earth Oxides", pages 111 to 133, especially page 117, Kluwer Academic Publishers, 2004, ISBN 1-4020-2568-8 ).

After passing through the tunnel layer, the high-energy electrons hit the collector ( 4 ) on.

The collector ( 4 ) should not hinder the entry of the electrons into its conduction band or the donor level due to its electronic properties. This requires semiconducting materials with a very low energy gap between vacuum level and donor level. The value of the electron affinity, the energy difference between the vacuum level and the lower edge of the conduction band is low or negative in such semiconductors. Such materials are, above all, n-doped wide-band-gap semiconductors, as the following list shows by way of example (values in electron volts): semiconductor work function bandgap Elektronenaffintät n-doped Si 5 1.1 4.0 Ge 4.9 0.76 4.1 InSb 4.4 0.2 4.6 PbTe 4.3 0.3 4.6 GaN 4.1-4.3 3.4 3.3 AlN 3.7 6.2 1.9 Diamond, undoped 4.2 5.5 1.7 Diamond, n-doped 0.9-1.9 5.5 to -1.2

While narrow bandgap semiconductors such as silicon, germanium, indium antimonide or lead telluride have high electron affinities, those with wide band gaps show comparatively low electron affinities, especially because of the wide bandgap. The electron affinity of wide bandgap semiconductors can be lowered or brought to negative values by various known measures. Thus, the electron affinity of the surface of gallium nitride or aluminum nitride decreases to values around -0.7 eV after a surface adsorption of cesium / oxygen ( Applied Surface Science 162-163 (2000), p. 250-255 ). Thin films of diamond are the most widely studied in the scientific literature and can easily be deposited on substrates in hydrogen and methane plasmas using the Chemical Vapor Deposition (CVD) technique.

An overview gives the book "New Carbon Materials," Teubner, 2007, ISBN 978-3-519-00510-0 ,

Diamond films can be p-doped by the addition of boron-containing volatile compounds such as diborane in the deposition atmosphere. More interesting for use in the context of the invention is the N-doping, which leads to occupied donor levels just below the conduction band. N-type doping can be achieved both with nitrogen-containing compounds such as ammonia and with phosphorus-containing compounds such as phosphine.

The doping with nitrogen leads to a donor level which is 1.7 eV below the conduction band minimum, the work function is in the range of 1.4 to 1.9 eV and thus higher than the donor level by this amount of energy ( "Thermionic Electronemitters / Collectors Having a Doped Diamond Layer with Variable Doping Concentrations", USA 2011 0221 328, pub. 15/09/2011 ).

Cheaper values are achieved by doping with phosphorus. The work function is reduced to values of 0.9 to 1.2 eV. The occupied donor level created by doping is closer to the lower edge of the conduction band, its spacing is in the range of 0.6 eV (FIG. "Thermionic converter", USA 2011 001 7253, pub. 27.01.2011 ). Calculations show a minimum distance of only 0.2 eV ( "Thermionic emission from low work function phosphorus doped diamond films", Diamond and Related Materials 18 (2009) p. 789-791 ). Donor levels close to the lower edge of the conduction band can also be generated with other dopants. Arsenic and sulfur give a distance of 0.4 eV; It is preferable to use antimony with a distance of only 0.2 eV ( "Thermionic converter" USA 2011 0139 205, pub. 16/06/2011 ).

Finally, under the deposition conditions, surfaces are created to which hydrogen is bound. The bonding of the hydrogen leads to a significant lowering of the vacuum level and thus the electron affinity to values around -0.5 to -1.2 eV ( "Enhanced thermionic energy conversion and thermionic emission from doped diamond films through methane exposure". Diamond and Related Materials 20 (8), August 2011, p. 1229-1233 ).

The surface coverage of cesium also yields diamond surfaces with negative electron affinity.

The assignments of the diamond films with hydrogen or with cesium (melting point 28 ° C, boiling point 671 ° C) have the disadvantage that these surface assignments are less stable. On the other hand, modification with alkaline earth metals, especially with barium, gives diamond films which have a negative electron affinity even above 1000 ° C. ( "Photoelectron spectroscopy studies on diamond with respect to the modification of negative electron affinity characteristics", Diamond and Related Materials Vol 7, Issues 2-5, Feb. 1998, p. 651-655 ).

Accordingly, preferred collector films are n-doped diamond films with phosphorus or antimony, which are surface-doped with barium. These films are optically transparent due to the high band gap of 5.5 eV, have a stable negative electron affinity, and their donor level is very close to or coincident with the vacuum level.

It is of great importance for the function of the photovoltaic cells according to the invention that no disadvantageous chemical reactions take place between the individual layers with the formation of interfering compounds. Thus, the selection of the optimal material combination is also part of the object of the invention.

The structure of the cell according to the invention can be carried out as follows:
In the application of the listed electron emitters naturally the surface properties are of particular importance, so that one endeavors to produce these materials very cleanly and yet inexpensively in the form of very thin layers on inert substrates. The layers are preferably formed on the substrate by sputtering.

As back contact ( 1 ), a sheet or foil is selected from a material which does not form a compound with rare earth metals, that is, molybdenum, tungsten, titanium or chromium, or a material coated with these metals. On the oxide-free back contact, the emitter layer ( 2 ) applied. Because of the mostly semi-metallic or even metallic character of the emitter material, the transition of electrons from the back contact ( 1 ) in the electron emitter ( 2 ) facilitated.

The thickness of the emitter layer ( 2 ) is in the range of 1 to 100 nanometers. Higher thicknesses are not necessary, the electrons are released only from the uppermost molecule layers. Optionally, in a further coating step, a small amount of rare earth metal is additionally applied to dissolve it in the rare earth compound. With the low excess of rare earth metal, a particularly low work function is achieved. This requires a brief heating of the layer to temperatures around 300 to 1000 ° C.

In principle, two variants are possible for the reproducible formation of the tunnel layer:
The first variant is used to coat the emitter layer ( 2 ) with the tunneling material, preferably those rare earth oxides which correspond to the rare earth components of the emitter layer by suitable coating techniques.

According to the second, preferred variant, the deposited emitter layer ( 2 ) of a calculated amount of oxygen and heats it to accelerate the oxidation to temperatures around 200 ° C to 500 ° C. This can be done by adding to the inert gas or vacuum of the coating chamber exactly that amount of oxygen necessary to grow the desired thickness of the rare earth oxide layer on the emitter by the oxidation of chemically active rare earth metal. Thus, a small stoichiometric excess of rare earth metal is advantageous.

If the emitter layer consists of rare earth monosulphides, monoselenides or monotellurides, it is possible that rare earth oxides rather than rare earth oxides are formed, such as, for example, rare earth oxalcogenides LaS + O ₂ → LaO ₂ S.

This is not disadvantageous because even the rare earth chalcogenides of the form SEO ₂ S with their band gaps of 4.6 to 4.8 eV are good insulators and thus form a suitable transparent tunnel layer. Their melting temperatures of 2,000 ° C to 2,100 ° C indicate a likewise high thermodynamic stability.

Because the rare earth metals are extremely non-precious, the rare earth oxide or rare earth oxysulfide tunneling layer is not reduced to the rare earth metals under the deposition conditions. Also, to take advantage of this fact, rare earth oxides or rare earth oxysulfides are used as the tunneling material. Its layer thickness is 0.2 to 100 nanometers, preferably 1 to 10 nanometers.

The collector layer ( 4 ) is deposited in the prior art via a plasma process for the production of thin diamond layers, wherein the dopants such as phosphine, PH ₃ or alkylstibines, Sb (alkyl) ₃ such as trimethyl antimony, tributyl antimony or triisobutyl antimony are as gaseous precursor in the gas atmosphere. Subsequently, an evaporation can be done with barium. The layer thickness of the collector layer is 1 to 1,000 nanometers, preferably 10 to 500 nanometers.

Because of the small thickness of the collector layer ( 4 ) would result in too high a resistance on the way of the electric current to the arresters without further measures ( 6 ). Therefore, the collector layer ( 4 ) with the thicker layer of a transparent, electrically conductive material ( 5 ) coated. Suitable materials which meet these requirements are above all the transparent, electrically conductive oxides (TCO) such as indium-tin oxide, indium-zinc oxide, aluminum-zinc oxide, fluorine-doped tin oxide, niobium- or tantalum-doped titanium dioxides or, for example, electrides Mayenite-type, calcium aluminates of the basic formula 12CaO · 7A12O3, containing free electrons.

The transparent electrically conductive layer ( 5 ) are brought up by prior art techniques. The metallically conductive arrester strips ( 6 ) are printed in the form of a prior art grid via conductive pastes such as silver or aluminum-containing pastes or those of an aluminum-silicon alloy. Also the application of the antireflection coating ( 7 ) takes place according to the state of the art.

The antireflective layer ( 7 ) in addition to optimizing the incidence of light by eliminating the reflection task, the underlying layers to protect against environmental influences, from oxygen and moisture. It may consist of titanium dioxide, zirconium oxide, preferably of the SiO ₂ -type silicon oxides used as barrier materials in the packaging sector, or other suitable materials.

It is also possible to construct the arrangement according to the invention in a reverse order according to a second variant:
First, on a suitable glass plate, the antireflection coating ( 7 ), then the arresters ( 6 ) and then the transparent conductive layer ( 5 ). On this layer one separates the thin collector layer ( 4 ). On the layer ( 4 ), the thin tunneling layer of rare earth oxide or rare earth oxysulfide ( 3 ) evaporated. Thereafter, the emitter layer ( 2 ) and the back contact ( 1 ) deposited.

By comparison, it is not possible to obtain the hot charge carriers in the case of silicon cells according to the prior art. The work function of the silicon with around 5 eV is too high. The difference between the Work functions of p-type and n-type silicon are too low to allow carrier transfer which is faster than the rate of thermalization. In addition, the charge carriers arise because of indirect semiconductor silicon deep inside the silicon; the lifetime of the hot electrons is too low to be able to diffuse to the interface of the P / N junction.

By contrast, in the arrangement according to the invention the charge carriers are formed very close to the surface of the emitter of low work function; the time to transfer to the collector ( 4 ) is much lower than the time required for thermalization. The collector offers little resistance to the incoming electrons due to its electronic conditions, close proximity of donor level and surface vacuums, thus preventing thermalization of the electrons at that location.

With the materials used according to the invention, their arrangements and their material combinations, the use of hot charge carriers in the conversion of radiation into electrical energy is made possible in practice.

The material combinations according to the invention can also be used as emitters of electron guns and as photocathodes in photomultipliers, residual light amplifiers or generally as cold cathodes.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 102010010658 [0010]
• GB 2468526 [0011]
• DE 102011102886 [0013, 0038]
• WO 2011/000055 [0014]
• DE 102010004061 [0052]

Cited non-patent literature

• IEEE Journal of Photovoltaics, Vol 1, No 1, July 2011, p. 78-83 [0009]
• PhD Thesis Yamini Modukuru, Univ. of Cincinnati, Electrical Engineering, 2003; Electrochemical Society Proceedings, Volume 2002-18, "Cold Cathodes", pp. 365-371, ISBN 1-56677-342-3 [0034]
• "Rare-earth monosulfides as durable and efficient cold cathodes" dated 13.07.2011 of the Cornell University Library at http://arxiv.org/abs/1107.2547. [0037]
• Thesis by Kevin Gibbard, Univ. of Florida, 2005 "High Temperature Synthesis of Cerium Sulphides and Kinetic Modeling" (http://etd.fcla.edu/UF/UFE0010471/gibbard_k.pdf) [0046]
• Journal of Nanomaterials, Vol. 2008 (2008) Article ID 682920 or as quoted above, Rare-earth monosulfides ..., http://arxiu.org/abs/1107.2547 [0048]
• J. Phys. Chem., 1965, 69 (8), p. 2684-2689 [0048]
• J. Chem. Phys. 68, 3978 (1978) [0048]
• J. Chem. Phys. 65, 2400 (1976) [0048]
• Rare-earth monosulfides ..., http://arxiv.org/abs/1107.2547 [0050]
• Phys. Our time 2/2009 (40), pages 6, 7 [0069]
• s. N. Imanaka, "Physical and Chemical Properties Of Rare Earth Oxides" in "Binary Rare Earth Oxides", pages 111 to 133, especially page 117, Kluwer Academic Publishers, 2004, ISBN 1-4020-2568-8 [0070]
• Applied Surface Science 162-163 (2000), p. 250-255 [0073]
• "New Carbon Materials", Teubner, 2007, ISBN 978-3-519-00510-0 [0074]
• "Thermionic Electronemitters / Collectors Having a Doped Diamond Layer with Variable Doping Concentrations", USA 2011 0221 328, pub. 15.09.2011 [0076]
• "Thermionic converter", USA 2011 001 7253, pub. 27.01.2011 [0077]
• "Thermionic emission from low work function phosphorus doped diamond films", Diamond and Related Materials 18 (2009) p. 789-791 [0077]
• "Thermionic converter" USA 2011 0139 205, pub. 16.06.2011 [0077]
• "Enhanced thermionic energy conversion and thermionic emission from doped diamond films through methane exposure". Diamond and Related Materials 20 (8), August 2011, p. 1229-1233 [0078]
• "Photoelectron spectroscopy studies on diamond with respect to the modification of negative electron affinity characteristics", Diamond and Related Materials Vol 7, Issues 2-5, Feb. 1998, p. 651-655 [0080]

Claims (9)

Solar cells containing as electron emitter low work function of below 2 electron volts rare earth compounds and compounds of yttrium from the classes of monosulfides, monoselenides, monotellurides, antimonides, bismuthides, germanides, stannides, plumbides, aluminides, gallides, indides or magnesides, characterized in that they the electron emitter a tunnel layer of non-conducting rare earth oxides or rare earth oxyalkoxides and a collector containing as energy difference of near-surface work function and donor level has a value of less than one electron volt. Solar cells according to claim 1, characterized in that the rare earth compounds contain a stoichiometric excess of rare earth metal from 0.01 to 10 atomic percent dissolved in the crystal lattice, preferably 0.1 to 1 atomic percent. Solar cells according to Claims 1 to 2, characterized in that they have as the material of low work function an electron emitter of LaS, CeS, NdS, SmS, YS, LaSe, CeSe, NdSe, SmSe, YSe, LaTe, CeTe, NdTe, SmTe, YTe, CeSb, La ₃ Sb ₂ , SmSb, NdSb, YSb, YBi, NdBi, La ₅ Ge ₃ , Ce ₅ Ge ₃ , Ce ₅ Sn ₄ , La ₅ Sn ₄ , Sm ₅ Sn ₃ , La ₅ Pb ₃ , Sm ₅ Pb ₃ , Ce ₂ Pb, YAl ₂ , LaAl ₂ , CeAl ₂ , GdAl ₂ , SmAl ₂ , NdAl ₂ , EuAl ₂ , LaGa ₂ , NdGa ₂ , CeGa ₂ , SmGa ₂ , SmIn ₃ , SmIn, NdIn ₃ , NdIn, LaIn ₃ , La ₃ In ₅ , LaIn, CeMg ₃ , LaMg ₃ , NdMg, SmMg or SmMg ₂ , each of these compounds having an excess of a rare earth element. Solar cells according to claims 1 to 3, characterized in that a mixture of rare earth metals, so-called mischmetal, is used as the rare earth component. Solar cells according to claims 1 to 4, characterized in that the emitter layer has no through pores and has a thickness of 1 to 100 nanometers. Solar cell according to claims 1 to 5, characterized in that the tunnel layer has a thickness of 0.2 to 10 nanometers, preferably from 0.5 to 5 nanometers. Solar cells according to claims 1 to 6, characterized in that they contain as a collector dense diamond films with thicknesses of 1 to 1,000 nanometers, preferably from 10 to 500 nanometers, which are n-doped by phosphorus or antimony and their vacuum level on their surface by the Binding of hydrogen, cesium or barium is below the minimum of the conduction band, whereby the collector has superficially a negative electron affinity. Solar cells according to claims 1 to 7, characterized in that they use as a converter for the conversion of light into electrical energy so-called hot carriers by a semiconductor-insulator-semiconductor junction or a semimetal-insulator-semiconductor junction or a metal Insulator-semiconductor junction included. Photocathodes, residual light amplifiers, light emitting diodes and other devices whose function is based on low work function emitters, characterized in that they contain rare earth compounds according to claims 1 to 5 and a tunnel layer of rare earth oxides or rare earth oxyalkylene oxides.

Images