 |
 |
|
 |
 |
|
 |
|
 |
 |
 |
 |
 |
 |
 |
 |
|
In-Situ Particle Monitoring - FabTronics First
|
|
|
 |
|
|
In-Situ Particle Monitoring is simply the application of a particle sensor within the semi-conductor process tool itself. In this way the process engineer can monitor particle levels close to the wafer in real time. Processing problems can be enormously costly if not identified early.
ISPM APPLICATIONS
ISPM applications exist within many wet processes such as etching, cleaning and stripping. Wet Bench chemical process tools can be of various designs but all fall into the category of either re-circulated or non re-circulated.
Non re-circulated or static baths can be monitored using a batch sampler such as the AFCS, Auto Flow Chemical Sampler.
This general applications note discusses ISPM applications that require a re-circulated bath. This means that the chemical in the bath is continually re-circulated via a pump and filter. Figure one depicts such a bath with a particle sensor installed.
BENEFITS
The implementation of ISPM has benefits in two distinct areas. When a sensor is first installed the process engineer will be able to optimize the performance of the bath. Secondly, once optimization has been achieved the ISPM system can be used as a process control tool.
|
|
 |
|
|
 |
|
|
|
 |
|
|
PROCESS OPTIMISATION
Lets look at the trend data from a 10:1 HF bath with a newly installed sensor. Wafer lots were being processed during data collection. See figure two.
The overall BASELINE LEVEL is high and the BASELINE STABILITY is poor.
The re-circulation rate of this particular bath was altered to optimize bath performance. Re-circulation rate is an important parameter. If the rate is to low then the bath will take a long time to recover from the addition of particles from normal wafer processing or chemical changes. However if the rate is to high then the filter performance can be compromised resulting in BASELINE INSTABILITY.
After optimization of the re-circulation rate the bath performance was much improved. See figure three.
The BASELINE is now low, stable and recovery is achieved after every wafer lot. Typically, most baths should be able to recover within 20 minutes if correctly optimized.
|
|
Figure 2
|
|
|
|
Figure 3
|
|
|
|

Figure 4
|
|
PROCESS
CONTROL
After optimization the ISPM system can be used as a process control monitor. An alarm threshold can be set at some level above the baseline and should be as close to the baseline as possible without resulting in false alarms. This will depend upon the level of baseline stability achieved during optimization. Normal process wafer lots will trigger the alarm threshold but this is useful in notifying the operator that the bath has recovered sufficiently to process the next wafer lot. Processing at to high a throughput will result in the baseline rising as recovery is never achieved before the start of the next lot. This effect can clearly be seen in the in figure four.
|
|
|
|
Figure 5
|
|
CHEMICAL OPTIMIZATION
Another benefit that can be found is during a chemical change. See figure six. When the bath is emptied and refilled contamination levels increase dramatically. A number of hours are usually spent re-circulating the chemical in the bath until processing can commence. ISPM will tell the operator exactly when the bath has recovered thus minimizing tool DOWN TIME.
|
|
|
|
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|