An Amorphous Mesophase Generated by Thermal Annealing for High-Performance Organic Photovoltaic Devices
An Amorphous Mesophase Generated by Thermal Annealing for High-Performance Organic Photovoltaic Devices
复制标题
DOI：
10.1002/adma.201200490
复制
发表时间：
2012-07-10
影响因子：
29.4
通讯作者：
Nakamura, Eiichi
中科院分区：
文献类型：
--
作者：
Tanaka, Hideyuki;Abe, Yoko;Nakamura, Eiichi
使用有机富勒烯受体的可印刷薄膜有机光伏（OPV）器件在能源和化学资源问题方面引起了极大的关注，[1，2]并且还因为三菱化学最近报道的10%功率转换效率（PCE）的创纪录性能。[3，4]器件制造通常需要退火工艺，其中器件在制造期间暴露于热或溶剂蒸汽以改善其性能。[5，6]热和溶剂退火的有益效果的可能来源是分别通过去除或引入溶剂而在由供体（p型）或受体（n型）组成的活性层中的形态变化。[7-19]然而，对这种效应的分子理解仍然不清楚，因为OPV器件的活性层通常定义不清，无法进行详细的研究。我们认为一个针装置（即由供体和受体的混合物组成的夹层），[20-22]使用四苯并卟啉供体形成（BP;图1中的绿色）[22-24]和甲硅烷基甲基[60]富勒烯受体（SIMEF;图1中的蓝色），将适合于退火的分子水平研究，因为高度不溶和热稳定的BP与富勒烯层完全相分离。[22]为了这项研究，我们开发了新的富勒烯衍生物，（o-anisylsilylmethyl）（phenylsilylmethyl）[60]富勒烯（2，SIMEF-Ph，o-An）及其同系物3和4（SIMEF-Ph，m-An和p-An），并研究了它们的晶体结构，热致性和光伏性能。我们发现，热驱动的去溶剂化的活性层含有甲苯或氯苯共晶体2创建一个新的无定形中间相，最大化的短路电流密度（JSC），以及整体设备的性能。另一方面，含苯装置的加热导致熔融相为2并且填充因子（FF）降低。发现由2-4的原始晶体制成的活性层较差。这一违反直觉的结果表明，富勒烯的非晶中间相可以最大化BP/富勒烯界面接触，从而最大化电荷载流子的产生，这似乎是我们的pin器件的性能限制因素，而熔融相破坏了富勒烯层的表面，降低了分流电阻（Rsh）。我们筛选了我们的富勒烯衍生物[25]，发现新的SIMEF化合物2是合适的底物，因为它能够与多种芳族溶剂共结晶。另一方面，m-An异构体3和p-An异构体4仅形成无溶剂晶体。如针对更简单的类似物所报道的那样合成这些分子（方案1）。[26][60]富勒烯在1，2-二氯苯中与3当量的[60]富勒烯反应，得到单甲硅烷基甲基富勒烯1，分离产率为86%。的二甲基苯基甲硅烷基甲基格氏试剂的反应。N，N-二甲基甲酰胺。第二个甲硅烷基甲基
Printable thin-film organic photovoltaic (OPV) devices using organofullerene acceptors are attracting tremendous attention with regard to energy and chemical resources issues,[1, 2] and also because of a record performance of 10% power conversion efficiency (PCE) reported recently by Mitsubishi Chemical.[3, 4] Device fabrication often calls for an annealing process, in which the device is exposed to heat or solvent vapor during fabrication to improve its performance.[5, 6] Possible origins of the beneficial effects of thermal and solvent annealing are morphological changes in the active layer, consisting of donor (p-type) or acceptor (n-type), by the removal or introduction of solvent, respectively.[7–19] However, the molecular understanding of such effects remains unclear, because the active layers of OPV devices are generally too ill-defined to carry out detailed studies. We considered that a pin device (the-imeaning an inter-layer consisting of a mixture of donor and acceptor),[20–22] formed using tetrabenzoporphyrin donor (BP; green in Figure 1)[22–24] and silylmethyl [60] fullerene acceptor (SIMEF; blue in Figure 1), would be suitable for molecular-level studies of annealing, because the highly insoluble and thermally stable BP is entirely phase-separated from the fullerene layer.[22] For this study, we developed new fullerene derivatives,(o-anisylsilylmethyl)(phenylsilylmethyl)[60] fullerene (2, SIMEF-Ph, o-An) and its congeners 3 and 4 (SIMEF-Ph, m-and p-An), and studied their crystal structures, thermotropic properties, and photovoltaic performance. We found that heat-driven desolvation of the active layer containing toluene or chlorobenzene co-crystals of 2 creates a new amorphous mesophase that maximizes the short-circuit current density (JSC), as well as the overall device performance. Heating of benzene-containing devices, on the other hand, resulted in a molten phase of 2 and lowering of the fill factor (FF). The active layer made of pristine crystals of 2–4 was found to be inferior. This counterintuitive result suggests that the amorphous mesophase of the fullerene can maximize the BP/fullerene interfacial contact and hence the charge-carrier generation, which appears to be the performance-limiting factor of our pin device, while the molten phase destroys the surface of fullerene layer and decreases the shunt resistance (Rsh).To study the effects of heat-induced desolvation of the active layer, we screened our fullerene derivatives [25] to find that the new SIMEF compound 2 is a suitable substrate because of its ability to co-crystallize with a wide variety of aromatic solvents. On the other hand, the m-and p-An isomers 3 and 4 form only solvent-free crystals. These molecules were synthesized as reported for a simpler analog (Scheme 1).[26] A monosilylmethylfullerene 1 was obtained in 86% isolated yield by the reaction of [60] fullerene in 1, 2-dichlorobenzene with 3 equiv. of a dimethylphenylsilylmethyl Grignard reagent in the presence of 30 equiv. of N, N-dimethylformamide. The second silylmethyl