Abstract Gas cluster cleaning is a method of dry cleaning avoiding the loss or damage of materials by forming aerosol clusters of tens to thousands of molecules via cooling and a pressure drops of a gas. Particle removal for dry cleaning techniques mainly relies on momentum transfer of the aerosol. Therefore, the generated aerosol size and velocity are key parameters for the aerosol cleaning method. To optimize and investigate gas cluster cleaning equipment, prediction of the cluster sizes is performed through indirect measurements, accomplished by measuring dent sizes on a photoresist (PR) film. An experimental analysis of variables such as the nozzle temperature, CO2 flow rate, gap distance between the surface and nozzle exit, and impaction angle was performed. To measure the cleaning performance of the equipment, CeO2 and SiO2 particles in the 30–300 nm range, which can occur in the semiconductor manufacturing process, were deposited on a Si wafer and cleaned. This paper studies gas clusters derived from gas cluster cleaning equipment. As a result, the particle removal efficiency (PRE) measurement showed a cleaning efficiency of more than 97% for CeO2 contaminants and greater than 92% for SiO2-based particles.

中文翻译：

---

通过 CO2 气体簇碰撞去除纳米级表面颗粒用于干洗

摘要 气体团簇清洗是一种干洗方法，通过气体的冷却和压降形成数以万计到数千个分子的气溶胶团簇，从而避免材料的损失或损坏。干洗技术的颗粒去除主要依赖于气溶胶的动量传递。因此，产生的气溶胶大小和速度是气溶胶清洁方法的关键参数。为了优化和研究气体团簇清洁设备，团簇大小的预测是通过间接测量进行的，通过测量光刻胶 (PR) 膜上的凹痕大小来完成。对诸如喷嘴温度、CO2 流速、表面和喷嘴出口之间的间隙距离以及撞击角度等变量进行了实验分析。测量设备的清洁性能，半导体制造过程中可能出现的 30-300 nm 范围内的 CeO2 和 SiO2 颗粒沉积在 Si 晶片上并进行清洁。本文研究了由气体团簇清洁设备衍生的气体团簇。因此，颗粒去除效率 (PRE) 测量表明，CeO2 污染物的清洁效率超过 97%，SiO2 基颗粒的清洁效率超过 92%。