Surface defects from the adsorption of bulk nanobubbles and solutions to prevent it

In semiconductor manufacturing, the fight against defectivity is as old as the creation of the first transistor. Cleaning of metallic impurities and particles has always been necessary to achieve the high production yield leading to economic gains. In this PhD, a new potential source of defects will be studied. Bulk nanobubbles have a diameter of about 100 nm, are generated by the agitation of gas-containing water and can survive for months, while the prediction is less than one msec, based on the size. Moreover, the mechanism of stabilization of bulk nanobubbles is still under discussion.

The formation of drying marks by surface nanobubbles has been shown in the literature. Hence, surface defects are thought to be caused by the drying of adsorbed bulk nanobubbles. Bulk nanobubbles are negatively charged and will adsorb on positively charged surfaces. Also, drying of wafers is accompanied in the final step by the evaporation of a film of about one micron, forcing all impurities present in the rinsing water to deposit on the surface.

The goals of the proposed PhD are fourfold. (1) determine the role played in the generation of bulk nanobubbles by some representative impurities present in the ultra-pure water (UPW) used in semiconductor manufacturing, (2) study the defects caused by nanobubbles on different types of surfaces, (3) develop solutions for the annihilation of nanobubbles, and (4) study the influence of bulk nanobubbles on the properties of water that are relevant for wet etching and cleaning.

1. Impurities have been considered as playing a role in the stabilization of bulk nanobubbles. Results were negative, but only a few impurities were tested. It is the purpose of this first work package to test whether impurities typically encountered in UPW, play a role in the generation of bulk nanobubbles. Characterization will be performed using different techniques, to determine the concentration and size of nanobubbles, the viscosity of suspensions, their colour, etc. The starting point will be a method of nanobubble generation developed at imec. Tests will be performed to screen a few selected impurities, looking at the rate of nanobubble generation. Do impurities, and how, contribute to the rate of generation and long lifetime of bulk nanobubbles ?
2. Bulk nanobubbles are expected to contribute to the formation of drying marks. This hypothesis will be tested on different types of substrates, to be selected based on surface charge and chemistry. The characterization of surface defects will rely on the broad experience of imec SIP group in this respect. The mechanisms of generation of surface defects will be studied.
3. This is the most creative part of this proposal. On one hand, the generation of bulk nanobubbles could be prevented by further purification of UPW would impurities play a role in their generation. In this proposal, a method to destroy nanobubbles will be developed instead. Characterization methods learned in part (1) will be used to prove the disappearance of nanobubbles. The developed method will be implemented on a wafer spin processor, and the process recipe will be modified to prevent the formation of new nanobubbles during processing.
4. In semiconductor manufacturing, water is the preferred solvent for wet etching and cleaning. Water properties are being modified when dealing with nanoconfined spaces, while bulk nanobubbles cannot penetrate in the nanoconfinements because of their size. The influence of bulk nanobubbles on water properties will be tested by comparing results obtained with standard UPW and nanobubble-free water obtained with the method developed in part (3). Properties to be tested should be relevant to semiconductor manufacturing, such as surface tension. The number of properties to be tested will depend on the time left for this last part.