Development of a semi-continuous centrifugal process for classifying colloidal particles

在过去的几年里，以亚微米粒子为基础的新型材料的开发取得了进一步的进展。然而，对于大多数工艺来说，获得通常所需的窄粒度分布是一个问题。为了中试或工业规模的生产，合成胶体的粒度分布往往太宽，不能用于尖端技术，如可印刷电子产品、半导体抛光介质(化学机械抛光浆料)和用于癌症治疗的复合颗粒。出于这个原因，在后续步骤中，必须将窄分数与主要广泛的分布分开。这个过程的术语是分类。分级是通过去除细级分或粗级分来实现的。中试规模的连续或半连续离心法对颗粒进行分级，在颗粒大小和颗粒与悬浮液周围液体的密度差异方面存在一定的限制。这些分离过程依赖于大于1微米的颗粒尺寸以及固体和液体之间至少两倍的密度差。到目前为止，亚微米颗粒的分类仅限于实验室规模。在中试和工业规模中，采用选择性低的深床过滤器去除CPM浆液中的粗组分。深床过滤过程的优化是不可能的，因为确定物性阻碍了选择性的增加。因此，应用项目的目标是使用管状碗离心机在纳米尺度上进行分级研究。在固体和周围液体之间的密度差为100公斤/立方米的情况下，期望切片尺寸为100纳米。对于更大的颗粒和液体之间的密度差异，似乎可以实现20纳米的切割尺寸。通过使用碳纤维转子和主动磁轴承，管式碗离心机可产生高达160000 G的胶体颗粒分级。离心机是在当前的研究项目中建造的，需要进行改装以满足预期的应用。研究结果将描述不同颗粒系统的沉降行为与颗粒特性和工艺参数的关系。另一项工作是在离心机中用激光多普勒测速仪表征流型。该技术已用于管式离心机内的流型测量。研究结果已于近期发表。所获得的流动特性知识将被用于优化进气几何形状。进一步测量是必要的，以了解离心机进料系统附近的颗粒和流动通道。这对于达到预期的切割尺寸是非常重要的。此外，实验确定的流型能够验证将在应用项目中进行的模型计算。

Within the last years further progress was made in development of novel materials based on submicron particles. However, obtaining the commonly required narrow particle size distribution is a problem for most processes. With a view to production in pilot or industrial scale, particle size distributions of the synthesized colloids are often too broad to be used in cutting edge technology such as printable electronics, polishing media for semiconductors (CMP slurries) and composite particles used for cancer treatment. For this reason narrow fractions must be separated from the primarily broad distributions in subsequent steps. The term for this process is classification. Classification is either achieved by removing the fine or the coarse fraction.A continuous or semi-continuous classification of particles by centrifugation in pilot scale shows certain limitations concerning particle size and difference in density of the particles and the surrounding liquid of the suspension. Those separation processes depend on particle sizes larger than 1 µm and density differences between solid and liquid of at least a factor of two. The classification of submicron particles is so far limited to the laboratory scale. In pilot and industrial scale, deep bed filters that are low in selectivity are used to remove the coarse fraction in CPM slurries. An optimization of deep bed filtration process is not possible, because the determining physics impede an increase of selectivity. Thus, the aim of the applied project is the investigation of classification in the nanometre scale using a tubular bowl centrifuge.With a density difference between solid and the surrounding liquid of 100 kg/m³ a cut size of 100 nm is aspired. For higher differences in density between particles and liquid a cut size of 20 nm seems achievable. By using carbon fibre rotor and active magnetic bearings, the tubular bowl centrifuge generates up to 160000 G for classification of colloidal particles. The centrifuge is built in a current research project and needs to be modified for the intended application. The results of the investigations will describe the sedimentation behaviour of various particle systems in dependence of particle properties and process parameters.A further work package is the characterization of the flow pattern by Laser Doppler Anemometry in a centrifuge. This technique had been employed for measuring the flow pattern in a tubular bowl centrifuge. The results have been published recently. The gained knowledge of the flow characteristics will be used to optimize the inlet geometry. Further measurements are necessary to understand the particle and flow channelling close to the feeding system in the centrifuge. This is highly important to achieve the intended cut size.Furthermore the experimentally determined flow pattern enables the validation of the model calculations which will be carried out within the applied project.