Introduction

Contamination in the form of chemical films, sometimes as thin as a single molecule, can cause yield losses due to changes in the chemical, electrical, optical, and physical properties of the product surfaces. In addition to damaging the final product, these chemical films can damage or impair costly optics and other tools used in the manufacturing process. Real-time monitoring of Airborne Molecular Contamination (AMC) interaction with surfaces, or Surface Molecular Contamination (SMC), on silicon, metal, and polymer surfaces can help reduce or eliminate the costs associated with this contamination.

The AiM® from Particle Measuring Systems uses surface acoustic wave (SAW) technology to measure the accumulation of contamination on critical surfaces such as silicon dioxide, metals, and polymers. Since monitoring SMC is a relatively new concept, few users know what rates of contamination are typical and/or acceptable. This paper provides guidelines on comparing surface contamination rates based on experience in a number of different semiconductor manufacturing facilities. Further, it compares different techniques for monitoring SMC, and illustrates the importance of real-time contamination monitoring.

The Importance of Monitoring Surfaces

Sensor surface coatings such as silicon dioxide and copper mimic the critical wafer surface, allowing the monitor to track and report wafer contamination events. Directly monitoring molecular contamination on a surface offers the advantage of being able to see the net effect of complex air chemistries, environmental conditions, and the chemical/physical nature of the surface. The quantity of SMC varies on time scales ranging from a few minutes to a few days. Real-time monitoring is required to identify contamination events and to distinguish trends in the deposition of low volatility or bonded contamination from the sometimes larger, transient deposition of medium and high volatility AMC. Real-time monitoring of the product surface complements and extends the value of traditional test wafer monitoring.

Guidelines for Contamination Rates

A wide range of contamination levels and behavior patterns are found when SMC is monitored in cleanroom environments. Examples of SMC on a silicon dioxide surface are shown below in Figures 1 through 4. Figure 1 shows contamination trend data in an exceptionally clean bay. There are no contamination events observed and the background contamination trend is 0.1 ng/cm2/day. Figures 2 and 3, depict large daily oscillations with occasional contamination events and moderate to high background contamination trends, ranging from 0.4 - 0.8 ng/cm2/day. In Figure 4, SMC accumulates on SiO2 at a very high rate (1.0 ng/cm2/day) and there are numerous contamination events occurring at irregular intervals. Some of these contamination events only last a few hours, but are in excess of 5 ng/cm2.

SMC fluctuations can make it difficult to interpret test wafer data. Without real-time data, it is not possible to understand if the SMC levels observed on the wafer represent stable background conditions, a contamination event, or a low or high phase of a daily contamination cycle.

A set of contamination monitoring guidelines has been developed, based upon many experiences monitoring SMC on silicon dioxide in semiconductor manufacturing environments. These guidelines are presented in Table 1.

Comparing Surface Monitoring Techniques

Directly monitoring molecular contamination on a surface offers the advantage of being able to see the net effect of complex air chemistries, environmental conditions, and the chemical nature of the surface. The main drawback with this direct monitoring approach is the need to use an additional analytical technique to identify the chemical species associated with the contamination. Table 2 summarizes the three SMC monitoring approaches.

Techniques available for monitoring molecular contamination on surfaces include the witness wafer, the quartz crystal microbalance (QCM), and the surface acoustic wave (SAW) monitor. The witness wafer, typically with a silicon dioxide surface, is a wafer that is exposed to ambient air in an area where product with the same or similar surface characteristics is exposed. A witness wafer will be exposed for 3 - 28 days, then analyzed by thermal desorption, gas chromatography/mass spectroscopy (TD-GC/MS), time of flight/secondary ion mass spectroscopy (TOF/SIMS), X-ray photoemission spectroscopy (XPS), or similar technique.

This approach provides a direct understanding of the chemical species that are interacting with the surface. Witness wafer analysis provides a "snap-shot" of the net impact of the various molecular contamination mixtures and environmental conditions to the wafer surface, from the moment of exposure until analysis. The lack of real-time data means that this approach cannot be used to alert contamination control personnel of the start of a contamination event or of transient molecular contamination events. Further, this approach cannot identify volatile compounds that have interacted with the surface.

The QCM monitor is based on a piezoelectric crystal that is supplied pulses of electric current at frequencies in the range of 3 - 25 MHz, creating an acoustic wave in the bulk of the crystal. As mass accumulates on the surface of the crystal, the frequency of the acoustic wave is reduced. The crystal surface is typically silicon dioxide or metallic. A monitor will be located where like surfaces are exposed to the ambient environment.

The QCM provides real-time data on the level of surface molecular contamination accumulated and the rate of accumulation, to alert the user to ongoing contamination events. After a period of monitoring or a contamination event, the sensor surface can be analyzed to identify the chemical species composing the SMC. The low sensitivity of this monitoring approach makes it well suited for environments with very large contamination events and extensive background molecular contamination (Tables 1 and 2). However, this technique is not well suited for semiconductor facilities or other clean applications.

The SAW monitor, Particle Measuring Systems AiM product, is based on a piezoelectric crystal sensor similar to the QCM in terms of its operational principles. The SAW device is different than the QCM in that the resonant acoustic wave only travels along the surface of the crystal instead of through the bulk of the crystal. Because the acoustic wave travels along the surface of the sensor, the SAW monitor can be operated at a much higher frequency, typically providing more than 100 times better mass sensitivity (sensitivity proportional to frequency squared) than the QCM.

The AiM monitor provides continuous real-time monitoring of SMC conditions with the sensitivity needed to detect contamination events in the early stages, before problems occur on the product surface. It is typically installed on tools, in air handling systems, by T/RH sensors, in photolithography areas, mini-environments, wafer pods, metrology tools, chemically filtered areas, or other areas where product surfaces are exposed to potential molecular contamination damage. It can be connected to a facility monitoring system to provide alerts or alarms based on contamination conditions, or it can be operated in a "stand-alone" mode, storing data internally and operating on battery power. The quartz crystal sensing element can be removed for TOF/SIMS analysis after a monitoring period or a contamination event.

Additionally, some customers choose to co-locate the AiM monitor and a witness wafer, using the real-time data from the SAW sensor to decide when to remove the wafer for analysis. This can help eliminate some of the uncertainty associated with interpreting wafer analysis results from locations where contamination levels vary strongly over the course of a day (Figure 3) or may be used to capture a transient event (Figure 4) on the wafer surface.

Guidelines for Using the AiM

It is recommended that the AiM monitor be used in a single location for one to two weeks to obtain the background SMC trend. Some locations may be very clean with only a slight upward trend in contamination mass over time, while others will have strong daily, weekly, or even longer term signatures (Figures 2 and 3). Once the background contamination rate(how many ng/cm2/day) and its time signature are understood, an abnormal contamination event can be readily distinguished (Figure 4). For guidelines on general classification of SMC contamination rates, refer to Table 1.

Summary

SAW sensor technology provides real-time data with the high sensitivity needed to detect SMC events and changes in background contamination rates. The real-time data can be automatically tracked through a facility monitoring software system ensuring alerts and alarms are generated when the SMC contamination rate exceeds acceptable limits. Real-time monitoring with the AiM allows the correlation of SMC data to events in the cleanroom, thereby providing the information necessary to eliminate or reduce the destructive molecular contamination.

Contact us if you need more information or have questions.

Learn more about our microcontamination monitoring services.

AiM® is a registered trademark of Particle Measuring Systems, In