Lab Report: Pod Box Cleaner

Cleaning Method Tested

This step-by step lab study details one group’s investigation into the optimal cleaning process for standard mechanical interface (SMIF) pods and front-opening universal pods (FOUPs). The results point to a cost-effective method for recycling these expensive wafer process containers.

BY: KEVIN SCHUMACHER

Recently, two groups got together to determine an optimal cleaning method for front-opening universal pods (FOUPs). (Figure 1) The joint investigation was done by Essex Products International, a cleaning equipment manufacturer in Chestnut Ridge, NY, and Sematech/I300I (International 300-mm Initiative), an international consortium of device companies designed to determine technical direction for 300-mm wafer technology. This testing was conducted at a mutually selected FOUP manufacturer.

At a unit cost of about $2000 each, effective cleaning and recycling of these pods has obvious cost-savings implications. Results of the investigation were based on current test methods used to examine the cleanliness of standard mechanical interface (SMIF) pods. The purpose of the test was to establish test procedures on FOUPs and to demonstrate the effectiveness of a new cleaning system over current aqueous cleaning methods.

New Versus Old Methods

The original system used to clean the SMIF pods incorporated an aqueous/air dry technology that required three independent steps:

(1) Process the SMIF pod through an aqueous clean line, then spinning and/or air drying the pods

(2) Ultrasonically cleaning the door in deionized (DI) water and wiping it with a polyester cloth with a 50/50 concentration of DI water and isopropyl alcohol (IPA), then blow drying the door with nitrogen gas (N2).

(3) Ultrasonically cleaning the gasket in DI water and again blow drying with N2.

Following the cleaning procedure, the pod was assembled and analyzed by liquid particle counting (LPC) testing.

EPI provided prototype equipment in the form of a compact solvent cleaning and drying system with an automatic cleaning sequence. The system incorporates robotic agitation, high-pressure spray, ultrasonics, heated immersion, and vapor-phase drying and allows the operator to clean complete FOUPs without disassembly. The main process bath uses ultrasonics, spray, and robotic agitation to remove any particles trapped in the FOUP and the door assembly. Once particles are removed, they are carried away from the product by a "focused overflow." In-situ filtration in the overflow recirculation loop ensures complete particulate removal from the process .

The next step in the processing is a vapor-phase rinse and dry. The process method allows the FOUP to remain in a sealed chamber during cleaning, which prevents exposure to air and airborne particles. The FOUP is surrounded by solvent during immersion and vapor phase drying. Testing has demonstrated that the process does not require sealing plugs, which were needed in the old process for the lid assembly.

The high density of the solvent—1½ times greater than that of water—assists in the removal and prevention of contamination redeposition. This results in more efficient displacement of particle soils, which are loosely held on the surface of the FOUP.

A critical factor in the drying process is that all drying occurs at the FOUP manufacturer's recommended temperature range, 60° –110° F. Unique fixturing prevents a shadowing effect, while proper positioning of spray nozzles enhances cleaning effectiveness and expedites the loading and unloading of the product. The combination of in-situ filtration and reprocessing helps to lower the cost of ownership since the cleaning agent is purified and recycled. This also eliminates the need for costly facility drains or exhaust.

Execution of the Project

The project was executed in three phases:

(1) An equipment baseline was established using LPC test methods and data collection currently in practice at the selected field testing facility

(2) Cleaning effectiveness was measured for the FOUPs

(3) Alternative test methods were utilized

The results of all three tests would be analyzed against previously established data in an attempt to show repeatability in various test methods and overall cleaning effectiveness.

The field testing facility had developed a database expanding over five years of cleaning results on 200-mm SMIF pods. This testing process consisted of multiple cleaning tests that utilize LPC analysis. The process offers consistent results as demonstrated by numerous pre- and post-test points in the database.

The first phase of analysis was to perform split-lot testing on 200-mm SMIF pods through the current clean line and the new system. The goal was to establish consistent, repeatable cleaning results on both lines and document the effectiveness of each process. LPC measurements were made before and after each cleaning cycle.

Phase One: Establishing Baseline

The first phase of testing was performed on the 200-mm SMIF pods. Two sample lots consisting of eleven pods each were used to perform split-lot testing through the two cleaning lines (original vs. prototype). Dirty pods were taken from the warehouse and measured before cleaning in the PMS model CLS-700 LPC system. They were then stored for 4 weeks in the shipping warehouse before testing and were not bagged or sealed.

Eleven pods each were cleaned via the original system and the prototype technique to collect comparative cleanliness data after preliminary LPC measurements were taken. The original system was utilized via the three-stage process discussed, and the new system operated via placement of the entire assembly on a basket into the load station.

Once the automatic cleaning cycle was complete, the SMIF pod assembly moved to the unload station, where it was sealed prior to exiting the cleaning system. The operator then removed the pod for LPC testing.

The data were collected and reviewed to determine consistency and cleaning effectiveness of each system. The field testing facility's analysis resulted in a confident determination that the data match their current results. Therefore, the cumulative data points were charted to determine if the new system demonstrated superior results over the original method. Particle removal per 200-mm SMIF pod is charted.

Measuring Cleaning Effectiveness

The second phase of testing was to measure the cleaning effectiveness of FOUPs. Testing was performed using the exact same analytical equipment, operator, and procedures used to develop the 200-mm baseline. The 200-mm SMIF data provided a high level of confidence that the prototype system would generate superior and consistent results for all aspects of cleaning. No changes were made in an attempt to limit any potential variables and ensure the validity of the collected data.

Since the new system required no equipment modifications to process 200-mm SMIF pods or 300-mm FOUPs, testing began on the 300-mm FOUPs. The field testing facility provided a quantity of eight 13-wafer, 300-mm FOUPs that were staged in the warehouse for 4 to 6 weeks to ensure excessive contamination. The boxes were loaded into the new equipment as complete assemblies and automatically processed with the exact same parameters and test protocol as the 200-mm SMIF pods.

Each 300-mm FOUP was checked by the LPC method before and after cleaning. Particles per box and removal ratio results are represented.

Alternative Verification Techniques

A review of the LPC data on 200-mm SMIF pods and 13 wafer, 300 mm FOUPs resulted in consistent superior results for the new cleaning system. As such, personnel were confident to begin testing alternative cleanliness verification methods.

The third phase of the test was to be carried out using Dryden QIII equipment (QIIIÒ Surface Particle Detector, Dryden Engineering, Fremont, Calif.). FOUPs were measured before and after cleaning. It was noted early in the collection of data that this measurement procedure was extremely operator-dependent and that new or inexperienced operators could cause surface scratches on the FOUPs. The field testing facility therefore requested Dryden to visit the site before testing to train operators on establishing test procedures.

After a technical review of the QIII test protocol, Dryden recommended that the tests be performed with the 90-degree-angle handle and a 2-inch vesbal head. Eight sites were selected to take measurements on the 13 wafer, 300 mm FOUP and door assembly.

The field testing facility provided five 13 wafer, FOUPs, which were stored in the warehouse for four weeks. The FOUPs were unbagged and left open to ensure excessive contamination. They were processed through the new system with the exact same process parameters as the 200-mm SMIF pods. Measurement with the QIII equipment followed cleaning.

The next set of testing was completed on the 25-wafer, 300 mm FOUPs. The testing facility provided six 25 wafer, FOUPs, which were split into two separate tests. Once again, FOUPs were stored in the warehouse for six weeks and, to ensure excessive contamination, were not bagged or sealed.

The first two FOUPs were run together in the new system, during one time frame. Again, the process sequence was identical to that used for the other pod tests. The LPC system was used to analyze before and after cleaning results.

The remaining four 25-wafer, 300 mm FOUPs were run randomly in the new system every two or three days during a period of 27 days. The processes sequence was the same as previously stated for the first two 25 wafer, 300 mm FOUPs. A mean average was 743 particles per FOUP was calculated based on the LPC recorded after cleaning the first two 25-wafer, 300-mm boxes. This average was given a zero value in an attempt to more easily monitor the performance of the new system.

Performance based on a mean average of 743 particles per FOUP.

Results Favor Prototype

As illustrated, the nonaqueous system tested in these trials offered several technological advantages over the old cleaning system. The primary advantage was its ability to clean and dry to a higher level than water-based system.

It has been successfully documented that humidity and residual surfactant can negatively affect the quality of a truly clean FOUP and the wafers that are staged inside. Comparative studies on spin drying and vapor-phase drying demonstrate superior results using vapor as a means to reduce static charges and particulate addition and show no change in humidity.

The prototype system offered another advantage in that it enabled movement of the FOUP from the liquid bath directly into the vapor region without exposing it to air. Air exposure can cause native oxide growth or airborne particle contamination on the surface of the FOUP. The vapor is clean, distilled solvent which quickly rinses across the FOUP and evaporates, leaving behind no traces of contamination.

The nonaqueous drying step was accomplished within the FOUP manufacture’s recommended temperature range, which was a critical Factor. Drying could be completed at a temperature of 105°F in 5 to 10 seconds, allowing the bulk of cleaning and drying time to concentrate on the removal of contamination from the FOUP.

Acknowledgments

EPI wishes to acknowledge the contributions by I300I for providing the direction necessary to attain the required industry testing. EPI would also like to thank Dave Staley, an investigator at the field testing facility, for his daily support during the equipment test period. Mr. Staley contributed the cleaning measurement data and personally performed the majority of cleaning tests and measurement analysis.

About the Authors

At the time this article was published Kevin Schumacher was the Vice President of Operation and Engineering at S&K Products International Inc. He is now President and owner of Essex Products International. He has developed several Patents for isopropyl alcohol vapor drying, isopropyl alcohol reprocessing and HFU megasonics; he has patents pending on a polymer removal system. He holds a BS degree in Engineering from Thomas Edison University.