Feasibility study of efficiency of wafer table cleaning liquids and alternative methods

T. Tajerian

Onderzoeksoutput: ScriptieEngD Thesis

Samenvatting

Customers of ASML are currently concerned about the availability and wafer edge yield performance of their immersion system. Both of these concerns can, at least in part, be traced back to contamination build-up on the top surface of the wafer table (WT) on which the wafer is clamped during scanning. When a contamination particle is trapped between the WT top surface and the wafer, the wafer will locally bulge up. This in turn leads to focus and overlay issues and a subsequently yield loss. Contamination on WT can be, either organic or inorganic. The organic type is coming from wafer material where the inorganic one is from the WT substrate and oxide particles mainly. For the organic contamination, photoresist materials are non-crosslinked where the BARC are crosslinked. In order to maintain a sufficient overlay and focus performance, the contamination on the WT needs to be removed by cleaning the WF periodically. This is done either by an in-situ cleaning method (called Spotless 2) or by shifting out the wafer stage and performing a manual WT clean (super soak). Both methods lead to machine down-time, but especially the manual WT cleaning creates a large availability hit. Currently cleaning WT contaminations causes roughly 2% availability hit, where the set goal by ASML is maximum of 0.5%. Understanding on the cleaning efficiency of various liquid agents (DI water, ethanol, IPA, RER and Acetone) in super soak procedures in order to remove the organic contamination including photoresist and BARC was the main aim of this project. In order to cover possible parameters for cleaning efficiency of the WT organic contamination, three main factors considered, i.e., materials, cleaning agents and cleaning methods. The first factor was materials to clean. It was known than the organic contamination is caused by several wafer processing both from the top side and back side of the wafers. Thus, the organic contaminates found in NXT immersion systems were addressed. Further the materials, should be reproducible for each and every experiment. Using WT’s for experimental purposes would cost enormously and also minimize the number of experiments and most importantly it was not conclusive due to non-homogenized distribution of contaminates on the WT surfaces. Therefore, two types of photoresists and BARCs polymer were used as experimental material. The second factor was cleaning agents. Over 628 liquids were scanned to find their best solubility match with the organic polymers. The Hansen model was used to compute the compatibility of the solvent vs polymers. Then, based on the solubility ratio, 66 liquids were in range. Lastly, due to availability, common use of liquid agents in both ASML and customer cleanroom and customer indicate five liquids that were selected. For each liquid, a dilution factor (100% or 25%) was examined as well. The third factor was acceleration methods, matrix combination of soaking time, temperature, ultra-sonication and mechanical force to each type of organic polymer. The results confirm that the Hansen Model is working properly to predict removal of the organic contamination, i.e., photoresist (non-crosslinked polymer), but BARC (crosslinked) followed different mechanism from the test samples. Removing non crosslinked material is much easier as the solvent will diffuse in the polymer irreversibly and in-homogenously. For removing BARC drastic solvent(s) are required such as sulfuric acid, to break the infinite polymeric network of the material. Also mechanical force will assist removal of the material. The experiments support acetone as a very useful agent to remove both type of contamination. For photoresist acetone directly diffuse of but out of place heterogeneously into the polymer. For BARC, it penetrates in to the polymeric network and makes it swells, thus internal stress increases that by rubbing and mechanical force the polymeric network peels off completely. The experiments also confirm; IPA can remove non crosslinked polymers with mechanical force. RER also is a very successful agent to remove photoresist. Ethanol have a random pattern for removing photoresist. DI water, was not successful to remove neither for photoresist nor for BARC. Sulfuric acid can break the BARC polymeric chain with in a second.
Originele taal-2Engels
Toekennende instantie
Begeleider(s)/adviseur
  • de With, G. (Bert), Begeleider
  • Cottaar, E.J.E. (Ward), Promotor
  • Klomp, A., Externe begeleider, Externe Persoon
  • Craus, C., Externe begeleider, Externe Persoon
Datum van toekenning14 jun. 2016
Plaats van publicatieEindhoven
Uitgever
StatusGepubliceerd - 2016

Bibliografische nota

PDEng thesis. - Confidential forever.

Vingerafdruk

Duik in de onderzoeksthema's van 'Feasibility study of efficiency of wafer table cleaning liquids and alternative methods'. Samen vormen ze een unieke vingerafdruk.

Citeer dit