有机发光器件性能优化及激子扩散问题研究

【作者】 周叶春；

【导师】 侯晓远； 丁训民；

【作者基本信息】 复旦大学 ， 凝聚态物理， 2007， 博士

【摘要】 本文工作主要围绕有机发光器件展开，主要分为两大方面：1)优化薄膜结构，提高器件性能；2)深入研究激子在有机薄膜中的扩散问题。具体包括对以下四个问题的研究：1．高对比度OLED。制备了具有低反射率阴极（Black Cathode TM）的OLED器件，可以实现器件在工作时的高对比度。该阴极由半透明金属层、介质层、金属层构成。文中选用Al掺杂的Alq₃作为介质层，器件对环境光的反射率大大降低，在合适的掺杂比例下（体积比1：5），仅为13％。选用该介质层的优点在于：1)采用热蒸发方式制备，避免使用溅射方法，基本消除制备过程中对发光层分子的破坏；2)虽然低反阴极的引入增加了器件厚度约70％，但整个器件的电学性能几乎保持不变，例如1 mA时驱动电压仅增加了0.26V。从SEM照片可以看到，在掺杂层中，Al形成了一些纳米颗粒，而导电性的提高应归因于Al和氧化铝纳米颗粒。2．空穴传输层厚度研究。本文研究发现，在所研究的掺杂器件中，NPB的最佳厚度约为12 nm，远远小于常用的空穴传输层厚度（约40 nm）。最佳厚度的差别可以归因于电子容易陷在DCM掺杂分子上形成空间电荷区域。文中提出了空间电荷导致能带弯曲影响电荷注入的模型，可以解释实验中观察到的器件效率和发光光谱随NPB层厚度而变化的规律。当NPB厚度超过5 nm后，器件效率随NPB厚度继续增加主要可以归因于NPB对电子的阻挡作用。3．单线态激子扩散研究。讨论了一维扩散方程所用边界条件“有机-真空界面激子流为零”在薄膜厚度与激子扩散长度可比拟时所带来的较大误差。本文采用蒙特卡罗模拟的方法计算了激子在有机薄膜中的扩散过程，模拟结果很好地符合了原位光致方法测量的PL强度-Alq₃厚度关系，并得到在Alq₃中单线态激子扩散长度为26nm（三维），等效于一维情况下的15nm。该结果与模拟中所假设的分子大小无关。另外，进一步模拟了OLED工作时其中的激子浓度分布。4．三线态激子扩散研究。三线态主体材料（Host）通过将激子能量转移给客体分子（Guest），达到较高的发光效率。由于通常主体材料本身不发光，所以荧光材料中常用的原位光致发光方法（in situ Photoluminescence）不能用来测量其三线态激子的扩散长度。与光电流方法（Photo Current Spectra）相比，采用高磷光发光效率的三线态客体材料作为探测层（Sensing Layer）来研究主体材料中三线态激子的扩散行为，涉及的相关物理过程较少，容易建立模型并得出精确的结果。在以前的研究中，研究者通常采用简单的一层模型，忽略了掺杂分子对主体材料中三线态激子扩散产生的影响。在本文中，通过分析掺杂层对激子扩散的影响，建立双层模型，以系统地描述激子在主体材料层和掺杂层中的浓度分布。利用这个模型，可以得到常用的主体材料CBP中的三线态激子扩散长度为61.4 nm、常用的掺杂层Ir（ppy）₃：CBP中的三线态激子扩散长度为6.8 nm。更多还原

【Abstract】 The thesis is focused on improving the performance of OLEDs and investigating the exciton diffusion process in organic thin films. The four problems discussed are listed below:1．High-contrast organic light-emitting devices. High-contrast OLEDs with low-reflection cathodes are fabricated. The cathode consists of a semi-transparent metal layer, a phase-changing (PC) layer, and a reflective metal layer. With Al doped Alq3 as PC layer, devices exhibit the average reflectivity of the ambient light as low as about 13%. And its electrical characteristics are almost identical to that of a conventional device, although the thickness is increased by 70%. The improvement in conductivity could be attributed to the conductive Al cluster distributed in the organic matrix.2．Optimize the thickness of hole transport layer in doped OLEDs. Current-voltage (Ⅰ-Ⅴ) and electroluminescence (EL) characteristics of OLEDs with NPB of various thicknesses as hole transport layer and Alq3 selectively doped with DCM as electron transport layer have been investigated. A trapped charge induced band bend model is proposed to understand theⅠ-Ⅴcharacteristics. It is suggested that space charge changes the injection barrier and therefore influences the electron injection process in addition to the carrier transport process. Enhanced external quantum efficiency of the devices due to the electron blocking effect of an inserted NPB layer is observed. The optimal thickness of NPB layer is experimentally determined to be 12±3 nm in doped devices, a value different from that of undoped devices, which is attributed to the electron trap effect of DCM molecules. This is consistent with the result that the proportion of Alq3 luminescence in total EL spectra increases with NPB thickness up to 12 nm under a fixed bias. 3．Singlet exciton diffusion in organic thin films. Limitations of the analytical method for calculating the exciton distribution in organic thin films, attributed to the improper boundary conditions when the organic film approaches the exciton diffusion length, were analyzed by comparison with an exciton random walk simulation. The random walk simulation results are in better agreement with in situ photoluminescence (PL) measurements than predictions based on the one-dimensional (1D) diffusion equation, especially for thin films (< 15 nm). The three-dimensional (3D) exciton diffusion length in Alq3 is determined to be 26 nm, equivalent to 15 nm upon projection to 1D. The result is not sensitive to the molecular size, a parameter arbitrarily set in the simulation. In addition, the exciton distribution in operating organic light emitting devices (OLEDs) was also simulated.4．Triplet exciton diffusion in organic thin films. Measuring the luminescence of inserted sensing layers with high phosphorescent efficiency is an effective way to investigate the triplet exciton diffusion in non-emitting host materials. Researchers used to adopt simple models without taking the influence of doped layers into account. In this letter, to include the dopant effects, a "double-layer" model is proposed for describing the exciton diffusion in doped and pure layers separately. The triplet diffusion length in Ir(ppy)3-doped and pure CBP films are calculated to be 6．8 nm and 61．4 nm, respectively.更多还原