NL8104155A - PROCESS FOR THE MANUFACTURE OF A HEAT REFLECTION FILTER. - Google Patents

Description

S ..si ψ% PHD 80,122 ^ N.V. Philips' Gloeilanpenfabrieken in Eindhoven.

"Method of Manufacturing a Heat Reflection Filter".

The invention is drawn to a method for the manufacture of a heat reflective filter, especially for light sources with a high infrared content, in which a tin-doped layer of indium oxide is applied to a light-transmitting support and the support during or after the coating cp is heated at a temperature between 300 ° C and the softening temperature of the support in a reducing atmosphere.

Such a filter, for instance known from German patent specification 23 41 647, behaves in respect of electro-magnetic radiation, such as visible light or infrared radiation, the wavelength of which is greater than the so-called plasma wavelength

P

of the material, as a metal, with respect to smaller wavelengths, however, as a dielectric, that is to say highly reflective in the first spectral region, but highly transparent in the other. Normal band absorption of the material only starts in the near ultraviolet spectral region.

The position in the spectrum of this more or less abrupt change in the optical property of the material, which is termed plasma side, is determined by the density of the free electrons (charge carrier density).

Such filters can then be used advantageously if heat radiation is to be either suppressed or to be radiated back to the radiation source, for example to increase its efficiency.

Applications of such heat-reflecting filters are, for example, the thermal insulation of sodium low-pressure gas discharge lamps, the reduction of radiation losses in solar collectors or the improvement of the thermal insulation of multi-glazed windows.

It has been found that heat-reflecting filters of the type mentioned should not only have a more or less high charge carrier density for the applications mentioned, but that the greatest possible mobility of the charge carriers in the filter layer is also desired. The mobility of the charge carriers can be determined via Hall-8104155 cl $ \ PHD 80,122 2 ....... effect measurements.

The heat filter produced by the known method with a filter layer of tin-doped indium oxide with a charge density (N) of about 1.3. 10 / cm has a charge carrier of about 30 cm / Vs. These values are not optimal for various applications, for example for the thermal insulation of solar collectors and also for a low-pressure sodium vapor discharge lamp with particularly high efficiency.

It is known that in semiconductors the mobility of the charge carriers decreases as their density increases. In "Physikalische

For example, Blatter "34 (1978), No. 3, p. 106 ff., Has shown that with a charge carrier density (N) of only about 5.10 / cm and 2, a charge carrier mobility (,, h) of about 40 cm / Vs can be achieved, however, the person skilled in the art cannot conclude from this pre-publication how independent of the primary manufacturing process of the filter layer for different charge carrier densities each time the optimum charge carrier mobility can be set afterwards. of Sn-doped In 2 O 2 layers, as indicated in German Patent 12 60 627.

The object of the invention is to improve the known method for the manufacture of a heat reflection filter of tin-doped indium oxide, which is highly transparent for visible radiation (light), but reflects infrared radiation, in such a way that a greater charge carrier mobility and in particular , at a given charge carrier density, an optimum value of charge carrier mobility can be set.

This object is achieved in a method of the type according to the invention described in the opening paragraph, because the indium oxide layer is doped with less than 7 mol% Sn, calculated on the amount of indium oxide and that for reducing a gas mixture consisting of 5 to 20 volume percent hydrogen with nitrogen as the remainder at 100 vol. %, with the addition of 0.5 to 3% oxygen or 2.5 to 15% air, each time calculated on the hydrogen part, is passed over the filter layer which is present for 10 minutes to 2 hours at the temperature at a temperature between 350 and 400 ° C heated support.

According to advantageous further embodiments of the invention, the indium oxide layer is doped with 0.5 to 6.5 mole% Sn, calculated on the amount of 8104 155 PHD £ 0.122 3 c indium oxide, and a gas mixture consisting of 5 to 20 vol% hydrogen with nitrogen as the residue at 100% by volume, with addition of 1% oxygen or 5% air each time carbon black with respect to the hydrogen part, passed over the filter layer for 15 to 30 minutes.

With this method, a way is indicated how the charge carrier mobility in heat reflective filter acting as tin-doped indium oxide layers can be increased and, moreover, an optimum value of the charge carrier mobility can be set for a given charge carrier density.

The invention is based on the insight that the heat-reflecting filters of the above-mentioned type have immanent optical properties, i.e. the greatest possible transmittance for the visible radiation part and the greatest possible reflection in the region of the infrared radiation, can be tuned very variably cp a random application.

In the known method for the production of net-doped ln 2 O 3 layers, be it by spraying at elevated temperature, sputtering or reactive evaporation, it must always be taken into account that free charge carriers are bound by excess oxygen, which then also the remaining remaining free charge carriers act as additional parasitic centers and greatly reduce their mobility. Surprisingly, this effect is avoided by the method in question.

The following example serves to clarify the operation of the method according to the invention:

An indium oxide layer with a date of 2 mol.% Sn is heat-sprayed with an aerosol of InCl3 in butyl acetate, doped with

SnCl., Made at about 500 ° C. At this layer, a * 20 2 charge carrier density (N) of 2.2 is first used. 10 / car and a charge carrier - beweg30 mobility (uu) of 40 cm / Vs. After reduction according to the invention, values for the charge carrier density N of 20 3 e o 5.5 were obtained. 10 / cm and for the load carrier mobility ^ u of 58 cm / Vs.

A further advantage which is obtained with the method according to the invention is the following: it has become possible with just the reduction gas-mixture used, the non-toxic (which is an important property for the manufacturing staff with a view to the operating personnel. ), however, both dynamic and dynamic at low temperatures, it is difficult to use hydrogen as a reduction gas. 8104155 PHD 80.122 4 3 '5 \. Due to the proven buffering of the hydrogen, oxygen is just .........

control the reduction process according to the method of the invention thermodynamically and dynamically in the same optimal manner as is known from reduction processes which work with a gas mixture of CO / CC 5. The decisive drawback for a technically large-scale manufacture is that CO is highly toxic and therefore should not be used in a manufacturing plant.

In a test series with different tin ratings (cc) in

; DU

In an indium oxide layer prepared according to the process of the invention, the corresponding values of tin doping (cc), 1 adherent density (N) oil 6, and charge carrier mobility (^ u), cited in the following table, were determined.

Table

15 cSn Ne ~ JU

(mol%) (1020 / cm3) (cmTVs) (10 ”4cm) 0.8 2.5 70 3.6 2 5.5 58 2.0 20 4 9.5 46 1.4 6 13 38 1, 3 9 14 34 1.3 12 13 27 1.8 * i 25 Cgn - tin doping

Ng = charge carrier density ^ u = charge carrier mobility ƒ = specific resistance of the layer.

Interpolation can be made between these values with the correct formulas. The charge carrier density (Ne) arises from

Ne ^ 3. 1020. cgn (1 - Cgn / 100) '1 cm "3; 1 = between 6 and 8.

The load carrier mobility (^ u) is given by 35 / U = a. (Ne / 1020) b cm2 / Vs a = between 100 and 200 b = between - 1/3 and - 2 / 3- 8104155

V

EHD 80.122 5

Familiarity with this relationship now makes it possible to calculate the optimal dating of an indium oxide layer for each intended application, following the following line of thought:

Depending on the application, the heat radiation to be reflected lies in 5 very different wavelength ranges, for example in the range of 5 to 50 µm with the 80 ° C radiation of a solar absorber, between 3 and 30 µm with the 270 ° C radiation, such as this for example in a low-pressure sodium vapor discharge lamp occurs or between 0.5 and 5 µm at the 3000 ° K radiation from an incandescent lamp.

On the other hand, great transparency is required in the visible spectral region. In thin layers, materials such as tin-dated indium oxide show the transition from high transmission power in the visible spectral region to high reflectivity in the infrared at their plasma wavelength (Λ, which with the charge carrier density 15 according to λ ^ = 4.0 / \ f ~ ^ Q (A in µm; NQ in multiples of 1020 / cm2) is related, so the charge carrier density must be selected according to the application, i.e. the appropriate amount of tin to be doped.

For theoretical considerations, for the losses in the visible region of the spectrum and in the infrared, which are based on absorption by the 10 layers on glass substrates, the following approximation terms arise: a = 1 "W * 0.526 · λ2 / ( “<Λ> * m82j 'He · <V / i and 25 A ^ = 1 - 6,6. 1 / (Ne. D. Yes) where A ^ g = absorption in the visible spectral region A ^ = absorption in the infrared T = transmission R = reflection 30 n (A) = refractive index of tin-doped indium oxide at the wavelength A, 2.0 for 0.05 µm X / m pi 0.35; effective mass of the free electrons in tin-doped indium oxide d = thickness of the layer.

The numerical values count as Nq in multiples of 1020 / αη2, d in yUm, yU. in cm / Vs and / l in yUm.

The variation of the charge carrier density (N) and layer thickness (d) 8104155 Ί, v PHD 80.122 6 ..... of the reflecting filter layer have qp losses in both spectral regions. The two losses are therefore interrelated, and there is a possibility. the following hyperbolic relationships are formulated; 5 \ ls3 (5 ^ 2 / H> J *] v /.

It is clear from this hyperbolic relationship that. in any case, high values of the charge carrier mobility (^ u) keep the losses of the heat reflector filter small. This means according to the table above always the smallest possible charge carrier concentrations, that is to say that according to small tin dopings should be chosen.

This requirement will continue to exist even if the two factors in the minimum requirement were not assigned the same weight, if, for example, a higher absorption was accepted for the sake of better heat reflection.

15

In order to optimize a heat-reflecting filter, Ne must therefore be chosen as small as is still compatible with the requirement that the transition from a high transmission power in the visible region to a high reflectivity in the infrared at

is such a wavelength (λ) that the wavelength range of the 20 P

reflect heat as completely as possible above Λ.

The absolute value of the heat refractive capacity must then only be influenced with the layer thickness (d), which is as a karprcm between a high value for R_r and a high value for A.

(both ΛΝ, d) should be chosen, taking into account the fact that, in general, a transmission maximum in the visible spectral region at a wavelength should be expected, which should be by / 1 2nd / m (m = 1, 2, 3 .....), wherein the refractive index n of the indium oxide layer can be reset to 2.

The information given also makes it possible to calculate the smallest layer thickness (d), with which a desired surface resistance R = J ° / d (inil), for example for a transparent electrically conductive electrode, can be achieved in a layer with about 6 up to 10 mol% tin (with which the highest charge carrier concentration can be achieved) is doped. The required layer thickness (d) is then 1.3 / µm. 35 /

According to a further advantageous embodiment of the invention, the gas mixture is passed over a catalyst which is at the same temperature as the carrier.

The catalyst advantageously consists of platinum. By switching on a catalyst through which the gas mixture is passed before it reaches the Ih 2 layer, the advantage is obtained that a layer of carbon black in the structure defined in the buffered atmosphere defined by the catalyst can be more homogeneously reduced and, moreover, that 5 the duration of the process becomes less critical. The indigenous structure of the layer to be reduced may be due to the manner of deposition of the layer, the morphology of the substrate surface, or an uneven temperature of the substrate.

For said reduction method, an H2 / N2 gas mixture should advantageously be used; however, another suitable reducing gas may also be used. The reduction process must be stopped before the indium oxide of the layer is decomposed, that is, reduced to indiim metal, which can be determined by blackening the layer.

15

This can be avoided in two ways: Firstly, the reduction can be interrupted in time by cooling the layer, whereby the electrical resistance R of the layer is preferably measured during the reduction and the oven is switched off to determine the time of the interruption. when this resistance has reached a minimum. At this point in time, the layer has reached the optimum properties: maximum charge carrier density (N) and maximum 6 carrier support motility (^, u), which, however, are different for each tin doping (cgn) chosen, RIZ 1 / iyyi (d = layer thickness), as can be seen from the above table. If the reduction is continued even longer, the layer resistance increases again. An over-reduction occurs, whereby the charge carrier mobility decreases substantially. A visible darkening of the layer due to over-reduction occurs only when the resistance has already risen about 10 to 301 above its optimum value.

30

The second possibility to prevent an over-reduction of the doped indium oxide layer is to buffer the reducing effect of the gas mixture. In doing so, a gas mixture is used that has a low oxygen partial pressure, which is not substantially greater than the decomposition pressure of the indium oxide at the reduction temperature. This decomposition pressure is known from thermodynamic data.

German patent specification 2,341,647 discloses the possibility of working with a reducing gas buffered by CC2 or ^ vapor, for example GO or H2. It was now determined that clear 8104155

HiD 80.122 έ 1 v ---------- better results can be achieved with regard to the mobility of the load carriers especially also in weakly doped In ^ layers, as the reducing gas mixture, in this case not just H20 vapor is buffered, but with oxygen, the gas mixture being either direct. or passed over a catalyst placed in front of the layer to be reduced, preferably of platinum. The catalyst must then have the same temperature as the support on which the layer is located. The temperature at which the reduction process must proceed, and at which both the support with the filter layer and the gas mixture is heated via the catalyst, must be maintained as uniformly as possible during the reduction process for all systems participating in the method.

An exemplary embodiment of the invention is described below.

To a solution of 100 g of InCl3 in 1000 ml of acetic acid-n. butyl-15 ester, 3 ml of SnCl 2 are added. The solution is atomized with air in an atomizer nozzle and the aerosol formed is passed on a flat transparent support, for example of glass. The support is heated to about 500 ° C by an oven. The aerosol jet is passed over the support until a layer having a thickness of about 0.3 microns is deposited thereon. The doping incorporated in the ln 2 O 3 layer is 6 mol% tin, based on the amount of indium oxide. The coated support is then heated to 350 ° C in a reactor. The reactor contains a site which passes the reducing gas mixture to be introduced into the reactor rather than the coated support, a catalyst also heated at 350 ° C, preferably of platinum. The reactor had been pre-purged with the reduction gas mixture

The amount of oxygen to be added to the H2 / N2 gas mixture depends on the reduction temperature, because both the discharge pressure of the indium oxide and the equilibrium composition of the gas mixture, which can be calculated from known thermodynamic data, depend on the temperature.

With the reducing gas mixture of 85% nitrogen (N ^ and 15% hydrogen (Ey), additional 2% oxygen (0 ^ calculated pp the hydrogen part is mixed in. This gas mixture is passed through the reactor at a flow rate of 6 l / h at normal pressure The reactor is brought to the temperature of about 350 ° C with the support on which the filter layer is present in about 30 minutes.

..... At high reduction temperatures, the reduction times are 8104155 Η © 80.122 9 9-~ -ff kart, which is generally a technical advantage, but there is also a great risk of rapid over-reduction if the gas composition is not adhered to accurately . At lower reduction temperatures, overreduction can be avoided for a relatively long time.

5 During the cooling of the indium oxide layer, either the oxygen supply must be switched off or the catalyst must not be cooled along, because otherwise the gas equilibrium shifts too strongly to oxygen, so that the optimally reduced layers in the cooling phase partially oxidize and their optimal properties, especially lose the TOo great mobility of the charge carriers.

15 20 25 30 35 8 104 155

Claims (4)

2. A method according to claim 1, characterized in that a gas mixture consisting of 5 to 20% by volume of hydrogen with nitrogen as the remainder at 100% by volume with addition of 1% oxygen or 5% air, each time calculated on the hydrogen portion , Is passed over the filter layer for 15 to 30 minutes. 3. Process according to claim 1 or 2, characterized in that the indium oxide layer is doped with 0.5 to 6.5 atomic% by weight of tin per pp. The amount of indium oxide. Process according to any one of the preceding claims, characterized in that the gas mixture is passed over a catalyst which is at the same temperature as the support with the indium oxide layer. 5. Process according to claim 4, characterized in that a platinum catalyst is used. Use of the heat reflective filter produced by the method according to any one of the preceding claims in the jacket of a low pressure sodium dan discharge discharge lamp. 30 35.810 4 115