Wafer charge damage
Reduce Wafer Starts
Wafer thinning
Resist stripping
Advanced materials processing
Bosch Process
Ion Plasma Immersion
Flip Chip
Plasma sources
- Situation: Company A needs to finalise process control for a MEMS product but is finding charge damage common.
- Problem: Charm wafers are used to diagnose this but are expensive at $4000 per wafer to monitor the potential damage of the process. These wafers have to be sent off to Charm for probe testing and the results are sent back 2 weeks later at further cost and time.
- Solution: Study charge build-up across the wafer with Quantemol-D, limit reactor time, the use of charm wafers and experimentation cycles.
- Situation: Company B is wasting 1000′s of wafers developing each process for a particular application. At the cost of materials, energy, tool time and personnel time.
- Problem: Trial and error approach to process development is in efficient (except for finalisation). Physics of variation between tools and devices is not well understood.
- Solution: Simulations with Quantemol-D to model and develop process changes up to final testing only cost computer time and will save £100,000′s
- Situation: Company C was plasma etching after a wafer thinning to remove the rough surface.
- Problem: Would like to use NF3 instead of SF6 for cost effective etch and would like to use existing equipment.
- Solution: Modelling using Quantemol-D discovered the benefits of using NF3 instead of a step wise degradation of the plasma when using SF6.
- Situation: Company D is trying to minimise resist stripping rates in some of their more sensitive processes.
- Problem: Chemical additions like N2, Ar or CF4 have a large impact on resist stripping rate which is also linked to surface material construction and RF choice.
- Solution: Using Quantemol-D to model the effects of these additions, surface recombination and different frequencies on the electron energy distribution function is of interest to companies investigating these effects.
- Situation: Company E wants to process new materials such as PZT, TIBr, Alsb, CdTe, CZT, CdHdTe, bulk GaN, ZnO, HfO2, InN and GesSi which are valuable but difficult to produce.
- Problem: Many issues including obtaining a high aspect ratio etch of low-K materials and resist stripping.
- Solution: Modelling of Hydrogen plasmas (which have shown best results so far) using Quantemol-D will reduce the experimental matrix considerably and therefore save on costs.
- Situation: Company F is developing a high performance Bosch process.
- Problem: Need to reduce gas cost and also enhance performance (i.e. same quality trench at higher speed).
- Solution: Modelling CF2 formation with Quantemol-D in the passivation step from the cyclic C4F8, C5F8, C3F6 and CF2=CF2 to optimise for low cost and high performance.
- Situation: Company G want to increase doping concentration and uniformity in a high bias plasma doping process.
- Problem: CMOS junctions are now becoming very shallow and ion implanters are very expensive with slow throughput. Plasma ion implantation offers a better solution.
- Solution: Simulation of BF2+ concentrations and surface fluxes with Quantemol-D in a BF3 plasma. Concentrations and across wafer uniformity can be simulated accurately. This would enable implementation of a BF3 plasma process and therefore save £100,000′s.
- Situation: Company H wants to improve chip contact, which is dependant on the Oxygen Plasma Clean and epoxy back fill.
- Problem: Poorly clean areas of the chip can lead to weak devices with poor shear strengths.
- Solution: Modelling the Oxygen radical concentration in the centre of the chip using Quantemol-D is required to understand this clean as the concentration is dominated by near chip surface interactions.
- Situation: Company I is a tool vender that wanted to improve its plasma source efficiency for a range of processes without sacrificing uniformity and anisotropy of etch.
- Problem: Building a new source just to test an idea is an incredibly expense process. Requiring months of work to take an idea to finalisation, where its usefulness may not be proved.
- Solution: The ability to experiment with new plasma sources with Quantemol-D in simulation rather than by building has saved this company over £150,000s so far. Simulations of new tools can be constructed and tested quickly, costing only computer time.
