The effect of multilayer barrier materials on the lifetime of organic photovoltaic cells has been investigated. For thin film encapsulated cells a protective layer was used to prevent damage during barrier layer deposition. No post deposition effects developed after dry box storage. In accelerated temperature and humidity lifetime testing the degradation of the encapsulated cells can be related to the loss of effective cell area. An extrapolation of the lifetime at room conditions has been quantitatively determined by comparing the cell degradation with the loss of Ca in a Ca-oxidation test. The results indicate a barrier permeation rate of 10-4 gr/[m2* day] for these samples, corresponding to a lifetime of greater than 5000 hours. Routes to improvement of the OPV cell lifetime are discussed.
Thin film barrier coatings for protecting Organic Light Emitting Diode (OLED) displays against the environment are extremely difficult to fabricate. The coatings must have extremely low water/oxygen permeability, no defects, cover several microns of topography, and be applied at temperatures below 100°C in a process that does not compromise the performance of the display. Vitex Systems has succeeded in depositing such coatings using an organic/inorganic, thin film multilayer structure termed Barix encapsulation. In this paper results on encapsulation of OLED test pixels and passive matrix displays will be shown. Lifetime and permeability tests conducted at high temperature and humidity demonstrate that this thin film coating can meet the necessary performance requirements for commercial OLED displays. Processing parameters, layer architecture and manufacturing techniques are analyzed and discussed. Thin film encapsulated displays are used to demonstrate the utility of the encapsulation technique.
In order to apply polymer light-emitting diodes in commercial products a number of lifetime specifications have to be met. In this paper we report on the performance and stability of polymer light-emitting diodes based on dialkoxy-substituted fully conjugated PPV. Lifetime measurements have been performed on small and large area devices under different conditions, including variations in temperature, luminescence intensity and humidity. It will be shown that polymer LEDs can withstand extreme lifetime tests successfully. The result are compared with lifetime specifications for applications in consumer applications and are discussed in terms of the stability of the emissive polymer. Spectral measurements as a function of the operational lifetime are presented.
In this paper on polymer LEDs we discuss the formation of black spots, surface treatments of the anode, and photochemical degradation of the emissive polymer. We find that small pinholes in the cathode layer are the origin of the black spots. The black spots form when H2O or O2 diffuse through the pinholes and react with the cathode at the polymer-cathode interface.A model is presented that describes the growth of the spots. We find that for both indium-tin-oxide (ITO) and Au anodes, an UV/O3 or an O2 plasma cleaning treatment increases the work function by 0.8-0.9 eV. A higher work function may lead to a better hole injection and a reduction in the operating voltage. We present a method to measure the quantum yield for bleaching, (gamma) equals 1.6$MN4 and (gamma) equals 1.7 10-7 for bleaching of dialkoxy-PPV in air and vacuum, respectively, indicating that the polymer is 1000 times more stable in vacuum than in air.
This paper attempts to investigate the possible degradation of the ester formed in PLASMASK 200G photoresist during extended silylation times with HMDS (hexamethyldisilazane). The breakdown of the ester results in the formation of a phenolic moiety and an amide. This may lead to silylation of this newly released phenolic OH-group in the unexposed regions, and hence a decrease in the contrast might occur. This would have a detrimental effect on critical dimension (CD) control. It was found that under higher ammonia concentrations, an ester degrades to release a phenolic group and also an amide. The silylation times involved are much greater than that used in the DESIRE process, and as such degradation of the ester in the DESIRE process is itself quite minimal. It is, however, suggested that a silylating agent that does not form a base on cleavage be used instead to reduce the degradation effect.
Using a new program based on Maxwell''s equations which computes the latent image of relative inhibitor concentration in two dimensions we have investigated notching effects in a resist layer on non-planar substrates and were able to understand line-narrowing effects which were experimentally observed. The program properly takes into account oblique propagation of light rays (in particular reflections from nonplanar interfaces) effects due to defocus and the partial coherence of the illumination. Thicknesses and layout of the underlying layers were varied in order to determine which parts of the structure give rise to the notching problems. The results which would be prohibitively laborious to obtain experimentally can be used to narrow down the range of possible solutions to the reflection problem. Some situations are however amenable to experimentation. In order to find the optimum resist for making the desired structures with minimum line narrowing over topography the A and B parameters of the resist were varied. The simulation results are compared to experimentally obtained profiles.
The silylated image and etched profiles in the DESIRE process are simulated and compared with experimental results.
The simulations show sioped silylated profiles, due to the finite contrast of both the aerial image and the silylation.
As a consequence the etched profiles are positively sloped and the linewidth changes in the course of the etch process.
Experimental results reveal that the slopes of the silylated areas are even worse than simulated. This is attributed
to lateral swelling and lateral transport of the silylating reagent. In accordance with the simulation the linewidth
decreaseS during etching. The sidewall angles, however, remain near-vertical despite the worse slope in the silylated