Patent Application
20140332795
The present invention relates to a transparent supported electrode, comprising a stack of thin layers of silver and of metal oxides, an organic light-emitting diode (OLED) optoelectronic device containing at least one such electrode, preferably as an anode, and to a process for manufacturing such a device.
Transparent conductive oxides (TCOs) and in particular ITO (indium tin oxide) are widely known and used as transparent material for forming transparent thin electrodes for electronic devices and in particular optoelectronic devices.
In the field of OLEDs (organic light-emitting diodes), ITO is used as an anode material since it is characterized by a high work function, generally between 4.5 and 5.1 eV. For large-area OLEDs, the sheet resistance (R□) of the ITO is however too high and, in order to obtain a good uniformity of light emission, it is necessary to coat the ITO layer with one or more conductive thin layers, such as silver layers.
The use of stacks of thin layers comprising one or more silver layers for increasing the conductivity of TCO-based anodes is also known. An anode for an OLED comprising both a layer of ITO and one or more silver layers is described, for example, in international application WO 2009/083693 in the name of the applicant.
In order to obtain a good crystallinity of the silver layer, the latter is, as is known, deposited on a crystalline underlayer of zinc oxide (ZnO), generally doped with aluminum (AZO). This crystalline underlayer of ZnO or of AZO is deposited, in turn, on a relatively more amorphous layer of mixed tin zinc oxide (SnZnO) which makes it possible to limit the RMS roughness of the following layers to a value generally of less than 1 nm.
Finally, each silver layer is generally covered with a thin metallic layer, referred to as a “blocker” or “overblocker”, typically of 0.5 to 5 nm, intended to protect the silver against oxidation during the step of depositing the following layer. These protective layers are also sometimes described as sacrificial layers because they are consumed by reacting with the oxygen against which they must protect the underlying silver layer.
The processes for manufacturing optoelectronic devices containing electrodes with such stacks of silver layers generally comprise at least one step of heating at high temperature (150° C.-350° C.) with a view to the etching, cleaning or passivation of the electrode.
The applicant has observed that the optical and electrical properties of the silver stacks were modified by this, often inevitable, annealing step. Annealing at a moderate temperature certainly improves the crystallinity of the silver layers and consequently the sheet resistance and the absorption of the electrode, but the applicant has observed that, unfortunately, at higher annealing temperatures, typically above 200° C., an increase in the sheet resistance and the absorption (a reduction of the light transmission) was observed.
The applicant has furthermore observed the appearance, during the annealing, of undesirable surface imperfections, referred to hereinbelow as “dendrites”. Dendrites are local depletions of silver which create, at the surface of the electrode, depressions having a depth of around 5 to 10 nm and a diameter ranging from about ten nanometers to about ten micrometers approximately. In the center of such a “well”, a protruding part is often observed.
This local increase in the roughness risks resulting in an increase in short-circuit currents.
After many experiments, the purpose of which was to understand the mechanisms of formation of dendrites and to reduce, or even prevent, their appearance, it has turned out that increasing the thickness of the metallic overblocker and/or inserting an underblocker makes it possible to reduce, but not to completely eliminate, the formation of dendrites. Furthermore, such measures inevitably result in an undesirable reduction in the light transmission (LT) of the electrode.
Although the applicant, after many tests, has not completely elucidated the mechanism of formation of dendrites, it was able to establish that the problem came from the layer of SnZnO, because a stack with an underlayer of ZnO, in the absence of SnZnO, did not give dendrites. it is likely that the presence of an excess of oxygen in the SnZnO layer is the cause of these defects. Without wishing to be tied to any one theory, the applicant theorizes that oxygen present in excess in the amorphous layer of SnZnO diffuses, during the annealing, into the thickness of the electrode and, when it arrives at the silver layer, oxidizes the latter. The formation of silver oxide could result in an increase in the local stresses causing the dendrites.
The present invention is based on the idea of protecting the silver layer or layers by inserting a protective layer, which is assumed to operate as a barrier to oxygen, between the silver layer and the SnZnO layer or layers of the stack. This insertion must not of course take place between the silver layer and the directly underlying crystalline layer of ZnO (AZO) which is essential for good crystalline growth during the deposition of the silver layer.
The applicant discovered that silicon nitride (Si3N4) and silica (SiO2), even at a small thickness, made it possible to play this protective role and to effectively reduce, or even eliminate, the formation of dendrites without their presence resulting in a degradation of the electrical and optical properties of the electrode before and after annealing. As will be shown below in the example, it was also observed that the presence of Si3N4 or SiO2 resulted in an advantageous decrease in the sheet resistance and in the absorption.
It is important to also note that the presence of the silicon nitride or silica layer between the silver layer and the SnZnO layer has no significant impact on the RMS roughness (measured by AFM on 5 μm×5 μm) of the sample, which increases by at most around 0.2 nm.
One subject of the present invention is consequently a transparent electrode for an organic light-emitting diode (OLED), comprising, on a transparent support made of mineral glass, n individual stacks of thin layers, each individual stack successively comprising, starting from the glass support,
The layer (a) is preferably an essentially amorphous layer of SnZnO. The ratio of the number of Sn atoms to the number of Zn atoms is preferably between 20/80 and 80/20, in particular between 30/70 and 70/30. Preferably, the percentage, by total weight of metal, of Sn preferably ranges from 20% to 90% (and preferably from 80% to 10% for Zn) and in particular from 30% to 80% (and preferably from 70% to 20% for Zn), in particular the Sn/(Sn+Zn) weight ratio preferably ranges from 20% to 90% and in particular from 30% to 80%. And/or it is preferred that the sum of the weight percentages of Sn+Zn is at least 90%, by total weight of metal, better still at least 95% preferably and even at least 97%. It is also preferred that it is devoid of indium or at least has an indium percentage, by total weight of metal, of less than 10% or even less than 5%. It is preferred that the layer (a) essentially consists of tin zinc oxide.
In order to do this, it is preferred to use a metal target made of zinc and of tin, for which the percentage by weight (total weight of the target) of Sn preferably ranges from 20% to 90% (and preferably from 80% to 10% for Zn) and in particular from 30% to 80% for Sn (and preferably from 80% to 30% for Zn), in particular the Sn/(Sn+Zn) ratio preferably ranges from 20% to 90% and in particular from 30% to 80% and/or the sum of the weight percentages of Sn+Zn is at least 90%, better still preferably at least 90% and even at least 95%, or even at least 97%. The metal target made of zinc and tin may be doped with a metal, preferably with antimony (Sb).
As indicated above, the role of the layer a) is to smooth, that is to say to limit the roughness of, the thin layers (AZO and Ag, or GZO and preferably Ag) deposited subsequently. It may be doped with a metal, for example with antimony (Sb).
In the present application, when mention is made of a “succession of layers”, “successive layers”, or else a layer located on top of or underneath another layer, reference is always made to the process for manufacturing the electrode during which the layers are deposited one after the other on the transparent substrate. The first layer is therefore that which is the closest to the substrate, all the “following” layers being those located “on top of” this first layer and “underneath” the layers deposited afterwards.
The expression “electrode for an OLED” used in the present application implies, inter alia, that the present invention does not encompass similar multilayer structures for which the last layer (the outermost layer) is a non-conductive layer, such as a layer made of silicon carbide, or preferably at the very least a non-conductive layer that is thick enough to prevent the vertical conduction from silver to the layer containing a light-emitting organic substance. Indeed, such structures would be unsuitable for use as an electrode.
In the present invention the SnZnO layer is either designated (a) or a), the ZnO layer is either designated (b) or b), the Ag layer is either designated (c) or c) and the Si3N4 or SiO2 layer is either designated (d) or d).
The electrode of the present invention preferably comprises from 1 to 4 individual stacks with a silver layer, that is to say n is preferably an integer between 1 and 4, in particular between 2 and 3, and is in particular equal to 2.
Naturally, according to the invention, the expression between a limit A and a limit B includes the limits A and B.
These silver layers preferably have a thickness between 4 nm and 30 nm, in particular between 5 and 25 nm and particularly preferably between 6 and 12 nm.
Preferably, the total thickness of the electrode is less than 300 nm, and even less than 250 nm.
Preferably, a thin layer is a layer having a thickness of less than 150 nm.
The protective layer is preferably a layer of Si3N4 or of SiO2 that is “doped”, for example with aluminum or zirconium. As is known, silicon nitride is deposited by reactive sputtering from a metallic (Si) target with use of nitrogen as reactive gas.
And, as is known, silica is deposited by reactive sputtering from a metal (Si) target with use of oxygen as reactive gas. Aluminum and/or zirconium are present in the target (Si) in relatively large amounts, generally ranging from a few percent (at least 1%) to more than 10%, typically up to 20%, ranging beyond a conventional doping, and intended to give the target a sufficient conductivity.
In the present invention, a layer of silicon nitride doped with aluminum (especially the dendrite barrier layer) preferably comprises a weight percentage of aluminum to the weight percentage of silicon and aluminum, therefore Al/(Si+Al), ranging from 5% to 15%. The aluminum-doped silicon nitride corresponds more precisely to a silicon nitride comprising aluminum (SiAlN).
In the present invention, preferably, in a layer of silicon nitride doped with aluminum or even zirconium (especially the dendrite barrier layer) the sum of the weight percentages of Si+Al or Si+Zr+Al is at least 90% by total weight of metal, or preferably 95% by weight or even at least 99%.
In the present invention, a layer of silicon nitride doped with aluminum and with zirconium corresponds more precisely to a silicon zirconium nitride comprising aluminum. The weight percentage of zirconium in the layer may be from 10% to 25% by total weight of metal.
In the present invention, the aluminum-doped silicon oxide layer (dendrite barrier) preferably comprises a weight percentage of aluminum to the weight percentage of silicon and aluminum, therefore Al/(Si+Al), ranging from 5% to 15%. The aluminum-doped silicon oxide corresponds more precisely to a silicon oxide comprising aluminum.
Preferably, in the layer of silica doped with aluminum or even with zirconium (dendrite barrier layer) the sum of the weight percentages of Si+Al or Si+Zr+Al is at least 90% by total weight of metal, or preferably at least 95% or even at least 99%.
As already mentioned in the introduction, the silica and the silicon nitride have proved to be effective protective layers, even at a small thickness. The thickness necessary for reducing or preventing the formation of dendrites increases with the annealing temperature and time. For annealing temperatures below 450° C. and annealing times of less than 1 h, thicknesses of the layers of less than 15 nm appear to be sufficient.
The thickness of the layer of Si3N4 or of SiO2 (especially in each individual stack, and between each individual stack) is preferably between 1 and 10 nm, in particular between 2 and 9 nm, and particularly preferably between 3 and 8 nm.
Each silver layer of an individual stack according to the invention is protected by the layer of Si3N4 or SiO2 not only from the SnZnO layer located underneath, but also from the SnZnO layer of the optional next individual stack by a layer of Si3N4 or SiO2.
Preferably, each silver layer of the electrode according to the invention is protected by a layer of Si3N4 or SiO2 that is especially between 1 and 10 nm in thickness, preferably between 2 and 9 nm in thickness and in particular between 3 and 8 nm in thickness, from an SnZnO layer located underneath, which layer of Si3N4 or SiO2 optionally makes contact with the silver layer, and also from an SnZnO layer located above by a layer of Si3N4 or SiO2 that is especially between 1 and 10 nm in thickness, preferably between 2 and 9 nm in thickness and in particular between 3 and 8 nm in thickness.
At least one of the stacks of layers, preferably each stack, also comprises, on top of the metallic silver layer, generally in contact therewith, a sacrificial layer comprising a metal chosen from titanium, nickel, chromium, niobium or a mixture thereof. As explained in the introduction, the use of such layers, better known under the name of blockers or overblockers, is known and serves mainly to protect the silver layer against a possible chemical or thermal degradation during the process for manufacturing the electrode. These layers may be partially oxidized. They are preferably very thin (generally less than 3 nm, for example of the order of 1 nm) so as not to adversely affect the light transmission of the stack.
Titanium (Ti, TiOx), which protects the silver layer(s) during the steps of processes for manufacturing the OLED and absorbs little, especially after heat treatment, is very particularly preferred.
The electrode may comprise at least two (preferably two) metallic silver layers, and, only on top of the last metallic silver layer, preferably in contact with the latter, is arranged a sacrificial layer comprising a metal chosen from titanium, nickel, chromium, niobium or a mixture of said metals.
For an electrode which is a silver bilayer, it turns out that a single overblocker, preferably made of titanium, on the second silver layer may sometimes be sufficient to protect the silver layers during the steps of processes for manufacturing the OLED.
For example, the electrode comprises the following (preferably strict) sequence, for n=2 or more, starting from the glass support:
Preferably, each individual stack comprises only one layer of SnZnO.
Preferably, for n equal to 2 or more, between two silver layers, there are only two layers of SiO2 or Si3N4.
The ZnO layer (under the silver layer) may preferably be made of doped zinc oxide, preferably doped with Al (AZO), Ga (GZO) as indicated above, or even with B, Sc, or Sb, or else with Y, F, V, Si, Ge, Ti, Zr, Hf and even with In in order to facilitate the deposition and a lower electrical resistivity.
It is also possible to choose a crystalline layer predominantly made of zinc and containing a very small amount of tin which may be likened to a doping, referred to as ZnaSnbO, preferably with the following weight ratio Zn/(Zn+Sn)>90%, better still ≧95%. In particular, such a layer having a thickness of less than 10 nm is preferred.
As already indicated, these crystalline layers are preferred to amorphous layers for a better crystallization of the silver.
It is optionally possible to use a silicon nitride layer under a silver layer, optionally forming a layer protecting from the underlying SnZnO, in particular (at least) under the first silver layer. This Si3N4 layer is preferably between 1 and 15 nm in thickness, in particular between 2 and 9 nm in thickness, and in a particularly preferred way between 3 and 8 nm in thickness. Its thickness may also be dependent on optical criteria. It may be thicker than in the case of an individual stack according to the invention.
When the electrode according to the invention is used as an anode of an OLED, the outermost layer, that is to say the one in contact with the hole-transport layer (HTL), must preferably have a certain work function. Certain transparent conductive oxides are known for their relatively high work function. ITO, for example, has a work function which is generally greater than 4.5 eV, sometimes greater than 5 eV.
The electrode according to the invention consequently comprises on top of the last silver layer (especially of the nth stack)—which is generally a silver layer or a blocker layer—a layer of a transparent conductive oxide (TCO), preferably a layer of ITO (tin-doped indium oxide).
This layer, described as a work function matching layer, may also be the second-to-last layer of the electrode (the anode), the last layer then being a relatively thin layer so as not to interfere with the work function matching role of the second-to-last layer and so as to preserve the vertical conductivity from the silver to the layer containing a light-emitting organic substance.
This TCO layer preferably has a thickness between 5 and 100 nm, in particular between 10 and 80 nm and particularly preferably between 10 and 50 nm.
This TCO layer is preferably directly on the (sole) overblocker of the last silver layer—the overblocker preferably being made of titanium.
Therefore, in one preferred embodiment, this TCO layer is at least one of the following metal oxides, optionally doped: indium oxide, optionally sub-stoichiometric zinc oxide, molybdenum oxide (MoO3), tungsten oxide (WO3), vanadium oxide (V2O5), indium tin oxide (ITO), indium zinc oxide (IZO or even IAZO or IGZO).
However, as ITO layer, MoO3, WO3 and V2O5 are preferred as the last, and even sole, layer on top of the overblocker.
A range of preferred proportions for the ITO is from 85% to 92% by weight of In2O3 and from 8% to 15% by weight of SnO2. Preferably, it comprises no other metal oxide or less than 10% by weight of oxide out of the total weight.
As explained in the introduction, for obvious reasons the protective layer (d), located between the silver layer and the nearby SnZnO layer or layers, must not be inserted between the silver layer (c) and the underlying ZnO crystalline support layer (b).
It is therefore preferably inserted between the amorphous SnZnO layer (a) and the crystalline ZnO layer (b).
In a first advantageous embodiment, the layer positioned between each silver layer and each of the SnZnO layers closest to the silver layer is a layer of silica (SiO2).
Each individual stack is therefore composed of, or consists of, the (preferably strict) sequence of the following layers:
Moreover, in the case of at least 2 individual stacks, it is recalled that an Si3N4 layer is also under each SnZnO arranged between two silver layers, preferably directly below the SnZnO. This Si3N4 layer is preferably between 1 and 10 nm in thickness, in particular between 2 and 9 nm in thickness, and in a particularly preferred way between 3 and 8 nm in thickness.
Thus, in this first embodiment, a layer of Si3N4 is chosen for all the protective layers.
In a second advantageous embodiment, the layer between each silver layer and each of the SnZnO layers closest to said silver layer is a layer of silica (SiO2).
Each individual stack is composed of, or consists of, the (preferably strict) sequence of the following layers:
Moreover, in the case of at least 2 individual stacks, it is recalled that a layer of SiO2 is also under each SnZnO arranged between two silver layers, preferably directly below the SnZnO. This SiO2 layer is preferably between 1 and 10 nm in thickness, in particular between 2 and 9 nm in thickness, and in a particularly preferred way between 3 and 8 nm in thickness.
Thus, in this second embodiment, a layer of SiO2 is chosen for all the protective layers.
Furthermore, naturally, two individual stacks (one after the other) may be separated only by the SiO2 or Si3N4 layer. This SiO2 or Si3N4 layer is preferably between 1 and 10 nm in thickness, in particular between 2 and 9 nm in thickness, and in a particularly preferred way between 3 and 8 nm in thickness. Therefore, for example, the electrode comprises the following (preferably strict) sequence for n=2 (or more) starting from the glass support:
In one preferred embodiment of an electrode with two metallic silver layers, comprising in this order a first silver layer, a SiO2 or Si3N4 layer (preferably as the layer d)) and the individual stack comprising the layers a)/d)/b)/c), c) which corresponds to the second and preferably last metallic silver layer, at least 60%, preferably at least 80%, of the thickness of the layers separating the two silver layers is formed from the thickness of the layer a) and/or its thickness is preferably greater than or equal to 50 nm, and better still greater than or equal to 60 nm and preferably less than or equal to 100 nm.
Naturally, for n equal to 2 or more, two individual stacks (for which the last layer is preferably a metallic silver layer or a sacrificial layer, of overblocker type) may be separated by the SiO2 or Si3N4 layer and by one or more other layers, and preferably by a single layer other than the SiO2 or Si3N4 layer, for example ZnO or AZO or GZO.
In one embodiment, a crystalline layer of ZnO (or AZO or GZO) separates the last layer of one stack (which is preferably a metallic silver layer or a sacrificial layer, of overblocker type) from the first layer of the following stack. The (first) protective layer SiO2 or Si3N4 (as the layer d) preferably) is then inserted between this layer of ZnO (or AZO or GZO) and the SnZnO layer (layer (a)) of the following individual stack.
This ZnO layer on the silver layer may preferably be made of doped zinc oxide, preferably doped with Al (AZO), Ga (GZO) or with B, Sc, or Sb, or else with Y, F, V, Si, Ge, Ti, Zr, Hf or even with In in order to make deposition easier and obtain a lower electrical resistivity. Preferably, its thickness is smaller than 30 nm, better still smaller than 15 nm and even better still smaller than or equal to 10 nm.
Consequently, in one preferred embodiment of the electrode of the present invention, each protective layer of SiO2 or Si3N4 is in contact, on one side, with a layer of ZnO, preferably doped with aluminum, and on the other side with a layer of SnZnO.
It results from the foregoing that, when n is at least equal to 2, that is to say when the electrode of the present invention comprises at least two individual stacks of thin layers as described above, it preferably comprises between the last layer of one stack (which is preferably a metallic silver layer or a sacrificial layer, of overblocker type) and the first layer of the following stack, successively:
Therefore, the electrode comprises the following (preferably strict) sequence for n=2 or more, starting from the glass support:
The metallic silver layer may be pure, alloyed or doped, for example with Pd, Cu, Sb, etc.
For n equal to 2, preferably the electrode comprises, starting from the glass, the following (preferably strict) sequence: (p layer(s))/a)/d)/b)/c)/(q layer(s))/SiO2 or Si3N4 layer/a)/d)/b)/c)/ and preferably p is an integer preferably less than or equal to 2, better still equal to 1 or even equal to 0 and q is an integer less than 3.
The added layer or layers preferably:
As a layer, in particular for the thin layer closest to the glass (referred to as the base layer), it is possible to use oxides such as niobium oxide (such as Nb2O5), zirconium oxide (such as ZrO2), alumina (such as Al2O3), tantalum oxide (such as Ta2O5), tin oxide (such as SnO2), or silicon nitride (such as Si3N4).
For the layers (b), the thickness is preferably less than 10 nm.
For the first layer (a) starting from the glass, the thickness is preferably greater than 20 nm, preferably from 30 to 50 nm. For the second layer (a) starting from the glass, the thickness is preferably greater than 40 nm, preferably from 60 to 100 nm, or even from 60 to 90 nm.
More broadly, in order to optimize the optical performance of an electrode according to the invention which is a silver bilayer (therefore with two silver layers), it may be advantageous to adjust the thicknesses of the layers under the first silver and between the two silver layers. By considering the optical thickness L1 of all of the layers under the first silver layer it is possible to choose L1 greater than 20 nm, better still greater than or equal to 40 nm, and less than 180 nm and even better still ranging from 100 nm to 120 nm.
By considering the optical thickness L2 of all of the layers between the first silver layer and the second silver layer, it is possible to choose L2 greater than 80 nm, better still greater than or equal to 100 nm and less than 280 nm and even better still ranging from 140 nm to 240 nm and even 140 to 220 nm.
The judicious choice of the optical thicknesses L1 and L2 makes it possible firstly to adjust the optical cavity in order to optimize the efficiency of the OLED and also to significantly reduce the colorimetric variation as a function of the angle of observation.
For the SnZnO layer between the two silver layers of a silver bilayer, the thickness is therefore preferably greater than 40 nm, preferably from 60 to 100 nm, or even 60 to 90 nm.
For n=2, a few examples of particularly preferred stacks are given below (with optional doping not specified again for the layers other than the layers under silver):
And more preferably still:
It is understood that after annealing and/or depositing the overlying oxide layer each overblocker (titanium or NiCr, etc.) may be at least partially oxidized.
The electrode according to the invention may (preferably) form a silver bilayer or trilayer, therefore comprising at least two metallic silver layers, and, for example, n is equal to 1 and the individual stack comprises a SiO2 or Si3N4 protective layer/a)/d)/b)/c) which is located on top (preferably directly or on a sacrificial overblocker layer or even on a layer of ZnO) of a silver layer, which is the first silver layer starting from the glass support, and preferably arranged under the first silver layer is a multilayer chosen from the following:
The first added layer or layers preferably:
For n=1, a few examples of particularly preferred stacks are given below (with optional doping not specified again for the layers other than the layers under silver):
The electrode may preferably be directly on the support or else on a layer, for example a light extraction layer, in particular a layer having a higher refractive index than the support, and/or a diffusing layer.
The glass may comprise an outer light extraction element on the face opposite the face with the anode already known per se such as:
The electrode according to the invention may alternatively form a silver bilayer or trilayer (preferably a bilayer), therefore comprising at least two metallic silver layers, and n is equal to 1 and the individual stack comprises SnZnO/SiO2 or Si3N4/ZnO/Ag where Ag is the first silver layer starting from the glass support.
Furthermore, preferably between the first silver layer and the second silver layer, it comprises the following multilayer: sacrificial layer preferably of Ti/layer of ZnO (of the composition described above for b), tq AZO and GZO) preferably having a thickness adjusted for optical properties and for example of at least 50 nm and even between 60 and 110 nm, even of 60 to 100 nm.
However, the sequence described above, SiO2 or Si3N4/a)/d)/b)/c), is preferred between the two silver layers when the wet processing of the OLED is critical because the thick SnZnO layer provides a certain chemical resistance.
Another subject of the present invention is an organic light-emitting diode (OLED) optoelectronic device comprising at least one electrode according to the present invention such as described above. This electrode preferably plays the role of an anode. The OLED then comprises:
Preferably, the OLED device may comprise a thicker or thinner OLED system, for example of between 50 and 350 nm.
The electrode is suitable for tandem OLEDs for example described in the publication entitled “Stacked white organic light-emitting devices based on a combination of fluorescent and phosphorescent emitters” by H. Kanno et al., Applied Phys. Lett. 89, 023503 (2006).
The electrode is suitable for OLED devices comprising a highly doped HTL layer (Hole Transport Layer) as described in U.S. Pat. No. 7,274,141 for which the high work function of the last layer of the overlayer is not very important.
Another subject of the present invention is a process for manufacturing an optoelectronic device according to the invention. This process comprises of course the deposition of successive layers constituting the individual stack or stacks described above.
The deposition of all of these layers preferably takes place by magnetron sputtering.
In this process, a plasma is created under high vacuum in the vicinity of a metal or ceramic target comprising the chemical elements to be deposited. The cationic active species of the plasma are attracted by the target (cathode) and collide with the latter. They then pass on their momentum, thus giving rise to the sputtering of the atoms of the target in the form of neutral particles which condense on the substrate forming the desired thin layers.
This process is referred to as “reactive” when the thin layer formed consists of a material resulting from a chemical reaction between the elements extracted from the target, for example the atoms of a metal target, and the gas contained in the plasma, for example oxygen or nitrogen. It is referred to as “non-reactive” when the target has essentially the same chemical composition as the layer formed, for example when it is a ceramic target containing the metal in oxide or nitride form. When the deposition takes place by magnetron sputtering from a ceramic target, the latter is generally doped with at least one metal, for example aluminum, intended to give a sufficient conductivity to the target.
The process according to the invention also comprises a step of heating the transparent electrode at a temperature above 180° C., preferably above 200° C., in particular between 250° C. and 450° C., and ideally between 300 and 350° C., for a duration preferably between 5 minutes and 120 minutes, in particular between 15 and 90 minutes.
It is in the course of this heating (annealing) step that the electrodes of the present invention are distinguished by the absence of the formation of dendrites in the silver layer and by a remarkable improvement in the electrical and optical properties, as will be shown below with the aid of the application example.
In a first deposition series, prepared by magnetron sputtering are, on the one hand, a transparent electrode according to the prior art comprising two individual stacks of thin silver layers on a glass support (comparative E1), and, on the other hand, a transparent electrode according to the invention (E2) which differs from the comparative electrode E1 by the fact that it comprises three thin layers of silicon nitride having a thickness of 4 nm, separating each of the two silver layers from the SnZnO layers.
The conditions for the magnetron sputtering depositions of the layers for comparative E1 and for E2 are the following:
The first Ti overblocker layer may be partially oxidized after deposition of AZO on top. The second Ti overblocker layer may be partially oxidized after deposition of ITO on top.
Table A below summarizes the deposition conditions and also the refractive indices:
Alternatively, it is possible to choose a metal target of zinc and tin doped with antimony comprising, by weight, for example 65% of Sn, 34% of Zn and 1% of Sb, or comprising, by weight, 50% of Sn, 49% of Zn and 1% of Sb.
Table 1 below shows, in comparison, the chemical composition and the thickness of all of the layers forming these two electrodes.
The electrodes E1 and E2 are heated for 1 hour at a temperature of 300° C. (annealing).
The light transmission (LT) and the absorption (Abs) and the sheet resistance (R□) of each of the electrodes are measured before and after this annealing.
Table 2 below shows the results of these measurements, before and after annealing, for the electrode E2 according to the invention in comparison with the electrode E1 according to the prior art.
It is observed that the annealing results in a degradation of the properties of the comparative electrode E1, that is to say a reduction in the light transmission and an increase in the absorption and in the sheet resistance, whereas the electrode E2 according to the invention sees the same properties improved (increase in LT and reduction in Abs and R□).
a and 3b present the optical microscopy images respectively of the electrode E1 (according to the prior art) and of the electrode E2 (according to the invention) after annealing at 300° C. Whereas on the first image (E1) very many white points corresponding to dendrites are observed, these points are completely absent from the second image of the electrode according to the invention (E2).
The use of thin layers of Si3N4 as protective layers between an Ag layer and an SnZnO layer therefore makes it possible to completely prevent the formation of dendrites.
In a second deposition series, prepared by magnetron sputtering are, on the one hand, a transparent electrode according to the prior art comprising two individual stacks of thin silver layers on a glass support (comparative E1′) and, on the other hand, a transparent electrode according to the invention (E2′) which differs from the comparative electrode E1′ by the fact that it comprises three thin layers of silica having a thickness of 5 nm, separating each of the two silver layers from the SnZnO layers.
Table 1′ below shows, in comparison, the chemical composition and the thickness of all of the layers forming these two electrodes.
The electrodes E1′ and E2′ are heated for 1 hour at a temperature of 300° C. (annealing).
The light transmission (LT) and the absorption (Abs) and the sheet resistance (R□) of each of the electrodes are measured before and after this annealing.
Table 2′ below shows the results of these measurements, before and after annealing, for the electrode E2′ according to the invention in comparison with the electrode E1′ according to the prior art.
It is observed that the annealing results in a degradation of the properties of the comparative electrode E1′, that is to say a reduction in the light transmission and an increase in the absorption and in the sheet resistance, whereas the electrode E2′ according to the invention sees the same properties improved (increase in LT and reduction in Abs and R□).
Whereas very many white points corresponding to dendrites are observed on the electrode E1′, these points are completely absent from the electrode according to the invention (E2′).
The use of thin layers of silica as protective layers between an Ag layer and an SnZnO layer therefore makes it possible to completely prevent the formation of dendrites.
The SnZnO layer between the two silver layers has useful properties with regard to the resistance of the electrode to chemical treatments for the OLED, namely cleaning especially using the following procedure:
The detergent is “TFDO W” sold by Franklab SA. It is an organic antifoaming detergent containing ionic and nonionic surfactants, chelating agents and stabilizers. Its pH is about 6.8 at 3% dilution.
On inspection of the surface of the electrode E2 thus treated with an optical microscope at a ×10 magnification, no pitting or surface defects were seen.
The aluminum-doped silicon nitride barrier layer may also be replaced by a silicon zirconium nitride layer SiZrN:Al for example produced under a reactive atmosphere from a metallic target having the following constituents given in percent of total weight: Si 76 wt %, Zr 17 wt % and Al 7 wt %.
The following alternative stacks have also yielded satisfactory results (post-anneal).
The AZO of the first layer and/or the second layer and/or the layer on the first silver layer may be replaced (preferably in all these layers) by GZO for example produced from a ceramic target for example with 98 wt % zinc oxide and 2 wt % Ga oxide.
Therefore, the AZO layers were replaced by GZO layers to create the following stacks:
An electrode was produced by replacing in E2 the first SnZnO underlayer with a layer of titanium oxide (the thickness of which could also be decreased due to its higher optical index). The TiO2 layer was deposited by sputtering using a ceramic target made of titanium oxide in an argon atmosphere with oxygen. The deposition conditions are given in table B below:
| Number | Date | Country | Kind |
|---|---|---|---|
| 1162472 | Dec 2011 | FR | national |
| 1252238 | Mar 2012 | FR | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/FR2012/053093 | 12/27/2012 | WO | 00 | 6/20/2014 |
