In this study, airflow and airborne particle tests for a 3OO-mm loadport (IsoPortTM) are per- formed in an ISO Class 7-8 cleanroom per ISO 14644-1 standards, the objectives of these tests are to understand the effects of various factors on the overall cleanliness performance of the IsoPort. Such factors include the gap size between the FOUP (Front Open Unified Pod) shell and the IsoPort's port plate, the FOUPIFIMS (Front-opening Interface Mechanical Standards) door opening speed profile, and the differential pressure between the enclosure and the ambient environment. It was found that there are two particle sources during an IsoPort open/close cycle.

In this study, airflow and airborne particle tests for a 3OO-mm loadport (IsoPortTM) are per- formed in an ISO Class 7-8 cleanroom per ISO 14644-1 standards, the objectives of these tests are to understand the effects of various factors on the overall cleanliness performance of the IsoPort. Such factors include the gap size between the FOUP (Front Open Unified Pod) shell and the IsoPort's port plate, the FOUPIFIMS (Front-opening Interface Mechanical Standards) door opening speed profile, and the differential pressure between the enclosure and the ambient environment. It was found that there are two particle sources during an IsoPort open/close cycle. The first source is the ambient F AB air squeezed out of the space between the FOUP door and the FIMS door as the FIMS door advances towards the FOUP door in the docked FOUP position. Depending on the gap size, the squeezed air may either exit the system through the gap between the FOUP shell and port plate, or remain in the local area of the FOUP gap until the FOUP door is opened. In the latter case the contaminated ambient air may end up inside the FOUP and/or migrate into the enclosure as a result of FOUP/FIMS door movement. The second source is caused by ambient air infiltrating into the system through the gap as a result of negative differential pressure created behind the FOUP door during its opening stroke. The effects of gap size, FOUP door-opening speed profile, and mini environmental ambient air differential pressure on system cleanlinesswill be presented and discussed in this paper.

State-of-the-art chip manufacturing requires more stringent airborne particle cleanliness specifications. This is mainly due to shrinkage in line width feature size. As line width gets smaller, particles of smaller sizes become more detrimental to number of good die per wafer. The new ISO 14644-1 Standards therefore require equipment cleanliness certification to be done at a small- er particle size (0.1 micron) than the outdated FED-STD-209E Standards.

Loadports are part of the semiconductor equipment set used for automatic handling of silicon wafer~ through out the entire chip manufacturing cycle. Silicon wafers are isolated inside a sealed plastic enclosure (commonly referred to as SMIF POD for 200-mm wafers, FOUP for 300

This paper will focus on a 300-mm loadport (IsoPort). llle FOUP open/close sequence of events for the IsoPort are: 1) a FOUP is placed on the loadport, 2) the FOUP is latched to the advance plate, 3) the FOUP moves forward from its HOME (undock) position to the port plate, 4), before the FOUP reaches the port plate (dock) position, the FIMS door moves backward into the enclosure for a short distance, 5) the FIMS door moves forward towards the docked FO UP and the FIMS door latches to the FOUP door, 6) the FOUP door is opened, and the combined FOUPIFIMS door move into the minienvironnient enclosure, 7) the FOUPIFIMS doors move downward to STAGE position, 8) the FOUPIFIMS doors move towards docked FOUP, 9) FOUP is closed, 10) FOUP moves back to HOME position. The various steps have significant effects on the overall cleanliness of the IsoPort. The most critical step is the FOW door opening and the FOUPFIMS doors movement into the enclosure. In this paper, several parameters that detect the cleanliness of the IsoPort will be presented and discussed.

Experimental Setup

Figure 1 shows the experimental setup used throughout the study. It consists of an IsoPort and an IS0 Class 1minienviron- ment. The ambient surroundingthe setup is IS0 Class 7-8. The following equipment are used:

* Met One A2100B Airborne Particle Counter (calibratedon 9/18/02)

* ShortridgeADM-860 Air Data Multimeter (calibrated on lus/oz)

* ASHCROFT Pressure Sensor (0.1inch-water)

* Tektronk TDS3032 Digital Phosphor Oscilloscope

Air-borne particles are measured inside the enclosure locally at the gap between the FOrfP shell and port plate. Particles orig- inating from the space (ambient air) between the FOUP and FIMS doors, or particles brought in through the gap between FOUP shell and pod plate should be detected at the location shown in Figure 2a. Similar behavior will be seen if the sampling tube is placed at the opposite side of the port plate, or at the top or bottom sides, since the air "squeezing" effect is the similar in all directions.

In order to understand the airflow behavior behind the FOUP door during its opening stroke and at the gap between the FOUP shell and the port plate, a highly sensitive, fast acting differential pressure sensor is used (ASHCROm) which has a range of (0.1 inches-H20. The differential pressure output of the sensor is displayed on a Tektronk TDS 3032 Digital Phosphor OsciUosco~. Should a negative differential pressure signal be observed at the gap, it would indicate that particle laden ambient air is flowing into the enclosure through the gap. On the other hand, a positive differential pressure signal would indicate that clean, filtered air is flowing from the enclosure to the ambient through the gap. Figure 2b shows the two differential pressure sensing locations.

Several different pressure output signals for various conditions are recorded and analyzed. Table 1 is a summary of all cases studied. Case 1 is using a two-step doore opening speed profile with a small initial door opening accelerator. Case #2 is using a two step door opening speed profile with a nominal acceleration setting. The gap between FOUP shell and the port plate is set at 0.5mm for Case 3 and is 2mm for all other cases. Case 4 is using a one-step door opening with the default acceleration setting of 5.0 inch/sec2. For all cases studied, the door closing stroke is a one-step speed profile at an acceleration of 5 inch/sec2.

The speed profiles of the FOUP/FIMS doors used in each case are shown in Figure(3).

Results and Discussion Figure 4, Cases 1-4, display the differential pressure variation inside the FOUP (Eocation 1) during a full IsoPort cycle (i.e., FOUP advance, SMART port door motion, FOUP door-opening, downward motion, upward motion, FOUP door-closing, FOUP motion to HOME position). At the instant of FOUP door-opening, a sharp negative differential pressure is observed inside the FOUP. Conversely, when closing the FOUP a positive differential pressure is observed at the same location. Additionally, there are three negative differential pressure peaks during the door opening and two positive peaks during the door-closing steps, respectively. The value of the negative differential pressure peak (>O.1 inch-water) is about one order of magnitude greater than the differential pressure (0.01 inch-water) between the minienvironment and the ambient. The maximum negative pressure for location 1(behind the FOUP door) during the door-opening stroke for the various cases is shown in Figure 5. The magnitude of the peak negative pres- sure behind the FOUP door for a two-step speed profile is less than that for a one-step profile. Figure 5 also shows that the minimum negative differential pressure behind the FOUP door corresponds to case #3, which is the lowest door-opening acceleration case. The minienvironment differential pressure does not have significant effect on the differential pressure observed behind the FOUP door during a full IsoPort cycle.

Variations of differential pressure versus time at the gap between the FOUP shell and port plate (location 2) are shown in Cases 6a-6d. The differential pressure decreases at the gap during the door-opening stroke. As shown before, the negative pressure is formed at the instant of door opening. The negative pressure may cause ambient air to enter the FOUP and enclosure through the gap. For Case #1 and Case #2, the differential pressure dips twice during the two-step door opening as shown in Figure 6b due to the two-step door-opening speed profile. On the other hand, for a one-step door-opening speed profile (case # 4), the differential pressure dips only once as shown in Figure 6d. As the gap gets smaller (0.5 mm for case #3), the resistance to flow becomes much higher, hence there is no apparent differential pressure change as shown in Figure 6c. Differential pres- sure changes at location 2 are more when one-step rather than two-step door-opening speed profile is used.

At higher differential pressure (0.01inch-water) between minienvironment and ambient, no negative differential pres- sure is detected at the gap when a two-step door opening speed profile is used, as shown in Case %a. For the one-step door opening profile, a short duration negative differential pressure is observed as shown in Case #6c. When the differential pressure between the minienvironment and the ambi- ent is 0.005 inch-water, short duration negative differential pressure is observed during the two-step door-opening cycle Therefore, at higher differential pressure between the minienvironment and the ambient, little or no ambient air will be drawn into the minienvironment through the gap.

Airborne particle tests were performed at the location shown in Figure 1. Figure 7 shows the average particle concentration at 20.1 micron, under different conditions. FOI Case #3 with a very small gap between the FOUP shell anc the Port plate, ambient air may not enter the minienvironment because of the higher airflow resistance caused by 6 smaller gap size. The high particle concentration for this case indicates that it has been caused by entrapment of ambient ah between FOUP and FIMS doors. That is, when the FIMS door attaches to the FOUP door; the air volume between the two surfaces will be displaced outward in all directions. With very small gap, the displaced air may not be adequately flushed out of from the local area of the gap. When the FOUP door is opened, the negative differential pressure behind the FOUP door causes the ambient air to enter the FOUP. With higher minienvironment differential pressure and larger gap, most of the displaced air will be flushed out of the system. Also, higher minienvironment differential pressure can prevent the ambient air from entering the enclosure during door-opening. Therefore, low particle concentration can be achieved at higher minienvironment differential pressure and larger gap. It is also observed that the effect of the door-opening speed profile on particle performance is insignificant at the higher differential pressure, 0.01 inch-water, and 2 mm gap.

Particle per wafer per pass (F") experiments have been performed at two customer sites. The results clearly indicate that the mcdi6ed FOUPRIMS doors opening speed profile achieves an order of magnitude better WW results than does the standard speed profile. In addition, more studies are being performed to arrive at an optimized door-opening speed profile, which will achieve the best cleanliness performance, and shorter cycle time.

Conclusions

Differential pressure and airborne particle tests are performed for the IsoPort under different integration conditions. Based on the test results, the following conclusions can be drawn:

* Negative Uerential pressure is observed inside the FOUP behind the FOUP door during the door-opening stroke.

* The magnitude of the negative differential pressure peak is dependent on the door speed profile.

* At low minienvironment differential pressure (0.00sinch- water) and 2 mm FOUP-loadport gap, ambient air may enter the minienvironment through the gap between the FOUP shell and the port plate during the door-opening stroke.

* There are two possible sources of particle contamination from loadport, one is due to ambient air infiltration through the gap between the FOUP shell and the port plate. The other is attributable to the air displaced from between the FOUP and FIMS doors, during the FIMS door advancement towards the FOUP door.

* Higher minienvironment differential pressure can prevent ambient air from enteringthe enclosure dur- ing the door-openingstroke.

* Larger gap, between the FOUP shell and the port plate, can allow displaced ambient air (between FOUP and FIMS doors) to leave the system through the gap.

* Lower airborne particle concentration for the IsoPort can be achieved at higher minienvironment differen- tial pressure (0.01inch-water)and larger gap (2 mm).