Chapter 1 Introduction
1.4 Organization
The thesis is organized as follows: the principles and the features of the array light-enhanced layer and OLED panel will be presented in Chapter 2. In Chapter 3, the considerations of the fabrication technologies to realize the array light-enhanced layer are summarized. The major measurement equipments used to characterize the fabricated the array light-enhanced layer are illustrated. In Chapter 4, the simulated results including light enhancement, and various parameters of ALEL used to verify and optimize our design will be presented. In Chapter 5, the experimental results including several fabricated samples are demonstrated. Based on the measurement results, several critical parameters in fabrication process are discussed, and then the desired structure will be realized and modified. In Chapter 6, the summary of the dissertation and the future works are given.
Chapter 2
Principle
2.1 Introduction
The output coupling loss, one of various efficiency loss mechanisms of an OLED, has been introduced in devices. The potential for further improvements are also described.
The first loss process of various efficiency loss mechanisms of an OLED occurs due to that the charge carriers do not recombine in the organic layer to produce excitons. It is commonly referred as the charge balance, which is defined as γ, the probability of the balance between the number of positive and negative charges injected into the organic layer. The excitons formed of two types, singlets that have a radiative decay, triplets that decay through nonradiative processes, lead to efficiency loss. The ratio of singlets to triplets will be defined as γst. Another efficiency loss, the intrinsic photoluminescence efficiency of the organic material, is defined a quantity q. A large portion of the light generated in the device is unable to couple out. The loss is defined as ηc.
These four major losses described above can be combined to determine the external quantum efficiency (ηext) of OLED device, using (1)
(1)
2.2 Optical output coupling
A brief description is provided for the models that have been developed to calculate the output coupling of OLEDs. Models based on calculating the interference effects are due to emitted and reflected beams [15][16] to classical [17] [18].
2.2.1 Total internal reflection and trapped light
The large mismatch between the refractive index of the organic layers, glass substrate and air results in a large proportion of the light rays undergoing total internal reflection as lights pass from a high to a low refractive index material. Only light emitted at an angle smaller than the critical angle can escape from the surface, while all other light is internally reflected back into the device. The light trapped within the device will be waveguided, either absorbed or emitted from the edge of the substrate.
A simple expression can be derived for the maximum emission efficiency based on ray optics and Snell’s law. Cathode is set as a perfect reflector and the emission is isotropic distribution. Then, the output coupling efficiency is given by the relationship [19],
(2)
For emission from in-plane dipoles, the proportion of light travelling perpendicular to the surface is increased and thus light is emitted at angles smaller than the critical angle, the expression for the light out coupling approximates to the following
(3)
In small molecule-based OLEDs, there is no preferential arrangement of the dipoles and the emission is thus expected to be isotropic. However, in polymer based OLEDs, the polymer chains are found to be aligned preferentially in-plane [15][20][21] which leads to increase output coupling. A useful indication of the magnitude of the loss in efficiency due to poor light output coupling of the photons generated in the device could be obtained by using this simple model. Based on a typical value of the refractive index of n = 1.7, the output coupling is predicted to be only 17 % for isotropic emission and 26 % for in-plane emission.
2.2.2 Modeling of output coupling
To produce accurate modeling of the output coupling, the structures of OLEDs need to be taken into consideration. A typical OLED device consists of four major components, the cathode, organics layer where the light is generated, transparent ITO contact layer, and the glass substrate. The refractive indexes of the organic layer (n = 1.6 ~ 2.0 ), the ITO (n = 1.6 ~ 2.0 ) and the glass (n = 1.5 ) are larger than the air, which leads to three types of modes: direct transmission into the air, the glass total-internal-reflection mode, and the high index (ITO/organic) guided mode. These will either be absorbed by the various layers, emit from the edge of the substrate, or be reflected at other interfaces in the device. Thus, high efficiency requires the minimization of these modes.
Ji-Seon et al. [16] describes an optical model for the emission from a polymer-based OLED. The study includes the transmission through the various layers with appropriate values used for the complex refractive index of an actual device and takes into consideration the effect of interference from the metal cathode. The angular dependence of the emission intensity from the devices is also used to predict the output coupling in this model. An experimental relation is calculated for the output coupling ratio as a function of the refractive index of the organic layer,
(4) for isotropic emitting dipoles and
(5) for in-plane emitting dipoles.
For the refractive index of the organic layer (n = 1.7 ), the maximum output coupling based on the optimum location of the emission zone in the organic layer for isotropic and in-plane
dipoles corresponds to a value of 26 % and 42 %, respectively.
The out-coupling and emission spectrum is a function of the location of the recombination zone by considering the strong interference from the metal cathode. In addition, proximity of the emission zone to the metal results in the quenching of the luminescence at cathode, due to energy transferring to the metal [17][20]. Thus, the locating of the recombination zone away from the metal cathode and the constructive interference of the emitted and reflected light are optimized to obtain the maximum output coupling [15].
2.2.3 Fresnel equations
A more rigorous analysis by a simulation tool with the Fresnel equations assigned should be used to perform the simulation. We analyzed the transmittance at the interface between the top-glass substrate and the air. According to Fresnel equations, the coefficients of refraction for p-ray and s-ray, rpp and rss, satisfy: refractive index of the first medium and that of the second medium, respectively. Assume there is no loss in the mediums, the transmittances of p-ray and s-ray, then, can be derived as Tp=1-|rpp|2 and Ts=1-|rss|2, respectively. Practically, only the average transmittance, T=(Ts+
Tp)/2, will be measured in the experiment. The refractive index of top substrate was set as 1.46. The results, as illustrated in Fig. 2.1, show that the transmittance will decrease to less than 90%, when the incident angle is larger than 36 degrees. However, the critical angle
Fresnel equations is necessary for establishing an appropriate pyramidal-ALEL model.
42
o Critical angleCritical angle42
oFig. 2.1 Illustration of the transmittance at the interface between top-glass substrate and air.
2.3 Summary
A simple expression can be derived for the maximum emission efficiency based on ray optics and Snell’s law. Cathode can be set as a perfect reflector and the emission can be isotropic distribution. For the typical value of refractive index of the organic layer (n = 1.7), the maximum output coupling efficiencies for isotropic and in-plane dipoles correspond to a value of 26 % and 42 %, respectively [15]. Although, the critical angle analyzed from Snell’s law at the interface between the top-glass substrate and the air is about 42 degrees, the incident angle larger than 36 degrees causes the transmittance less than 90 % from Fresnel equations. Therefore, considering the Fresnel equations is necessary for establishing a simulation model of the appropriate pyramidal-ALEL.
Chapter 3
Fabrication and Measurement Instruments
3.1 Introduction
A preliminary structure will be used to confirm the features of the microlens array. The embodiment including several fabrication processes will be shown in the following sections, and all the fabrication process, technologies, instruments and a preliminary structure will be introduced in this chapter.
First, the semiconductor fabrication process including mask layout, spin coating, exposure and development were preceeded. Besides, the mold was subsequently filled with the thermal-cured elastomer to replicate the microlenses. Then, the characteristics and performance of the fabricated structure, such as similarity to the geometric design and light efficiency were measured by typical measurement systems, such as optical microscope (OM), scanning electron microscope (SEM), chromameter, and conoscope. The major features of the above mentioned instruments will be illustrated in this chapter.
3.2 Fabrication Process
The entire fabrication processes we utilized to fabricate the microoptical components include the typical VLSI fabrication processes and plastic modeling replication techniques.
First, we determined the feature parameters, like the height of each ALEL pixel (LEL_H), the tilt angle of pyramidal pixel (TILT) and the spacing between two adjacent pixels (LEL_S), need to be designed properly for practical component fabrication. Then, we transfer the desired structures into multi-pattern for mask layout. Third, the microoptical structure on
microoptical structure into thermal-cured elastomer to replicate the microlenses. The processes are shown schematically in Fig. 3.1.
Required parameters:
LEL_H, TILT, LEL_S, Pixel size
Generate the mask patterm in CADs sofeware
Lithographic mask
Lithography and Etching
Replication by plastic molding
Profile evaluation
Fig. 3. 1 Flow chart of array light enhancing layer fabrication process.
3.2.1 Semiconductor Fabrication Process
A prototype is fabricated to characterize the features of the microlens array structure. The semiconductor processes including cleaning, coating PR, UV exposure with G-Line (436 nm) stepper, develop, fixing, RIE etching and KOH etching will be proceeded to fabricate the desired structure on Si-wafer at semi-conductor research center (SRC).
The detail processes are shown in Fig. 3.2. First step is initial-cleaning. Wet-cleaning processes are necessary to obtain an ultra clean wafer surface for subsequent fabrication. Then a 0.5 µm thick layer of SiNx will be deposited by plasma enhanced chemical vapor on wafer.
Next, the wafer is placed on a vacuum chuck in the coater and the photoresist is dropped onto the center of the wafer. A uniform and thin photoresist layer can be coated on the wafer
surface after spinning the wafer. The following step is exposure. The mask pattern is transferred onto the wafer. The exposed wafer is loaded into the development system after exposure. Consequently the desired structure will show up in the photoresist. Etching is then applied to get the desired structures from the developed pattern to a wafer.
Non-photoresist-covered SiNx will be removed during RIE etching. Then the remained photoresist will be stripped by acetone. Finally, the regions of silicon will be etched by KOH and the fabricated microoptical structure will appear on the Si wafer.
(A) Initial Clean
H.M.D.S.
(C) Coating H.M.D.S.
SiNx
(B) Deposit ing SiNx
Photoresist (PR)
(D) Coating Photoresist
(F) Developing & Fixing
(G) RIE Etching
(H) PR removing
(E) Exposure
(I) KOH Etching 4" Dummy Wafer
Fig. 3.2 Semiconductor fabrication process of microoptical structure.
3.2.2 Replication
The plastic material for replication is considered to have a similar refraction index as optical glass. We use poly(dimethylsiloxane) (PDMS), a thermal-cured elastomer, as the material. Fig. 3.3 shows an overview of the major replication technology.
(a)
(b)
(c)
Fig. 3.3 ALEL replication process flow (a) filling the mold with PDMS (b) removing bubbles by vacuum process, and (c) curing.
The fabricated mold was pretreated by an anti-sticking layer (1H, 1H, 2H, 2H-perfluorodecyl trichlorosilane) to prevent the mold adhering to PDMS during the replication process. To fabricate the microlenses, the mold was subsequently filled with PDMS. Then we remove the bubbles from the microstructure by using vacuum chamber. The PDMS is cured at 50 °C for 3 hr leading to a flexible sheet that can easily be peeled off from the mold. The thickness of the PDMS layer ranges from 1 to 3 mm.
3.3 Measurement System
After the fabrication, we must make sure the fabricated components meet the design goal.
First, the geometric similarity to the designed structure is checked by the optical microscope (OM) and scanning electron microscope (SEM). Then, the optical performances of the fabricated components and designed ALEL on OLED panel are measured by the chromameter CS-100 and conoscope system, the detailed descriptions are shown below.
3.3.1 Scanning Electron Microscope (SEM)
A scanning electron microscope (SEM) is an essential instrument to measure the accuracy and fidelity of the fabricated microstructure, as shown in Fig. 3.4. The instrument scans electrons reflecting onto a fluorescent screen across the target where the image is captured by a camera and enlarged. Electrons are much smaller than atoms, so a scanning electron microscope paints a razor-sharp image of the target, and the feature variation of few angstrom can be observed. SEM is useful for mapping details of objects that optical microscopes can not resolve. Using the electromagnetic lenses to focus the accelerated electron beam, the diameter of electron beam can be converged to the dimension of 10 µm.
The secondary electrons are generated where the focused accelerated electrons bombard the sample. Detecting the secondary electrons can determine the location of bombardment.
Simultaneously, the focusing electron beam scans the surface of sample, with the aid of
-3
scanning coil, to map the feature of measured area.
HITACHI S-4000 SEM was utilized to measure the quality of our fabricated microstructure elements. The line width, etching depth, and aperture size can be measured accurately.
Fig. 3.4 Photograph of SEM.
3.3.2 Measurement Systems for light efficiency
3.3.2.1 Chromameter CS-100
The color and brightness of image are the important properties of system we want to evaluate. In order to get these information, a colormeter is necessary. We choose the Minolta Chroma Meter CS-100, which utilizes three high-sensitivity silicon photo cells, which are filtered to match Commission International de I’Eclairage (CIE) standard observer response. Chromaticity coordinates (x,y) and illumination (Y) as well as color temperature in Kelvin (K) are calculated from the three cells’ measurements. With this compact reliable color analyzer, we can easily get the Chromaticity coordinates and
3.3.2.2 Conoscope
Conoscope is a measurement system which utilizes Fourier transform lens to transfer the light beams emitted (or reflected) from the test area of different angles to the CCD array, as shown in Fig. 3.5. Every light beam emitted from the test area with a θ incident angle will be focused on the focal plane at the same azimuthal angle and at a position x = F(θ).
Therefore, the angular characteristics of the sample are thus measured simply and quickly, without any mechanical movement. Particularly, which kind of light source, i.e. collimated or diffuse illumination, is provided depending on the needs.Besides, by equipped with a fast photometer system and a high-sensitivity spectrometer, its functions are extended to comprise not only the simultaneous measurement of luminance and chromaticity versus viewing direction, evaluation of the data yields, i.e. luminance contras ratio, grey-scale inversion and reduction, color shift and many more characteristics, but also the spectra and temporal luminance variances.
C: cone of converging elementary parallel beams A: variable aperture Fig. 3.5 Schematic diagram of Conoscope.
Chapter 4
Simulated Results and Discussions
4.1 Introduction
In chapter 2, we had analyzed the transmittance at the interface between the top-glass substrate and air to confirm that the Fresnel equations are necessary for modelling a pyramidal ALEL. We established a simulation model to characterize the features of the array light-enhancing layer on OLED panel.
First, the parameters of pyramidal ALEL, including the height of each ALEL pixel, the tilt angle of pyramidal pixel and the spacing between two adjacent pixels, were designed and optimized. After that, the alignment issue of the array light-enhancing layer on OLED panel was also discussed.
4.2 Simulation Software
The optical simulator Advanced Systems Analysis Program (ASAPTM), developed by Breault Research Organization (BRO) was used to optimize the array light-enhancing layer on OLED panel and simulate its light efficiency.
4.3 Simulation description
4.3.1 Optical Model
In order to consider the whole effect of the array light-enhancing layer on an OLED panel, all the design and optimization were carried out under the complete system framework.
bottom glass substrates, as shown in Fig. 4.1. Ribs and insulators were set with 60 % transmittance. Since the organic layers, including the electron injection layer, the electron transport layer, the active layer (the emitting layer), the hole transport layer and the hole injection layer, are much thinner (~0.15 µm) than other layers, they are assumed as one “source layer” in our simulation, as shown in Fig. 4.2. Metal cathode was set with 90 % reflectance.
Fig. 4.1 Schematic diagram of OLED structure.
Glass Substrate
Cathode
Electron Transport Layer Electron Injection Layer
Hole Transport Layer Hole Injection Layer
Anode
Emitting Layer Organic Layers
Fig. 4.2 Schematic diagram for organic layers of OLED structure.
Lights from active layer was assumed to emit in the form of pixel array and was set as an isotropic light source. The array light-enhancing layer consists of the base substrate and microlens array. The pixel size of ALEL, as shown in Fig. 4.3, was set to fit one of OLED panel. The size of ALEL and OLED panel are W×L. In addition, a flat detector with the diameter of 3cm was located at a distance of 3.5cm above the OLED panel, as shown in Fig.
4.4.
1 Pixel size of OLED Panel 1 Pixel size of
OLED Panel
1 Pixel size of ALEL
1 Pixel size of
ALEL Alignment
Fig. 4.3 Schematic diagram for designing pixel size of light source and ALEL.
L
OLED panel ALEL
W
Detector 3 cm
Fig. 4.4 Schematic diagram of measurement.
Light efficiency is defined as the detected flux over the generated one. Furthermore, gain factor was defined as the ratio of the light efficiency of an OLED panel with ALEL to that of a panel without ALEL, which were served as the bases of the optimizations.
(1) = detected flux light efficiencyof OLED panel with ALEL
Gain Factor =
light efficiencyof OLED panel
light efficiencyof OLED panel with ALEL Gain Factor
light efficiencyof OLED panel
=
4.3.2 Parameters
In order to obtain to maximum gain factor, several variables, as shown in Fig. 4.5, such as the height of each ALEL pixel (LEL_H), the tilt angle of pyramidal pixel (TILT) and the spacing between two adjacent pixels (LEL_S), need to be designed properly. A simulation tool, ASAPTM, with the Fresnel equations assigned was used to perform the following simulation.
LEL_H
Fig. 4.5 Illustration of parameters for designing array light-enhancing layer.
4.4 Results and discussions
4.4.1 Tilt angle and spacing of pyramidal pixel
First, the relationship between the gain factor and the tilt angle (TILT) will be discussed.
Height and spacing are kept constant, and TILT is changed from 90° to the minimum angle where the two slopes meet at the intersection point, and then the relationship between the gain factor and TILT is formed. Besides, spacing is kept as different constants to represent the
relationship between the gain factor and spacing. The green sub-pixel light source is turned on.
From the simulated results, as illustrated in Fig. 4.6, the maximal gain factor can be derived at 50° to 55° and smaller spacing.
1.0 1.2 1.4 1.6 1.8 2.0
40 50 60 70 80 90
TILT (degrees)
Ga in Fa c to r
Spacing = 10 µm Spacing = 25 µm Spacing = 40 µm
Fig. 4.6 Illustration of the relationship between gain factor and TILT. The height is 125 µm.
4.4.2 Height of pyramidal pixel
Next, the relationship between the gain factor and the height is discussed. From the simulated results, as illustrated in Fig. 4.7, the curve of gain factor becomes stable when the height is higher than 100 µm. Since the lower height can bring higher yield rate during the demolding process, the optimized height is set as around 150 µm.
1.0 1.2 1.4 1.6 1.8 2.0
0 50 100 150 200 250
Height (µm)
G a in F acto r
TILT = 40 TILT = 50 TILT = 60
TILT = 40 TILT = 50 TILT = 60