Extreme Silicon

Investigating the formation of time-dependent haze on stored wafers

Larry W. Shive, Richard E. Blank, and Karen H. Lamb, MEMC Electronic Materials

Study results indicate that time-dependent haze can be caused by humidity inside wafer packaging and by organic or ionic contamination on the wafer surface that exceeds typical levels.

Silicon wafer users expect that the surface of the wafers they receive from their suppliers will meet all requirements for light point defects (LPDs), metals, grown-on-film quality, and cleanability. Yet many users have observed a very large number of >0.12-µm LPDs on wafers that have experienced extended storage prior to use. This phenomenon has come to be called time-dependent haze (TDH). Its appearance may or may not trigger wafer rejection at incoming quality assurance or create a performance problem in the device manufacturing line, but it will certainly suggest to the user that something in the supply chain is out of control.

A typical LPD map of a hazed wafer is shown in Figure 1. There are at least several hundred more LPD counts on such wafers than on nonhazed wafers, and it is common to see a localized region of many random LPDs. It is the presence of such a random pattern or the detection of unexpectedly high counts that usually alerts wafer users to TDH. An atomic force microscopy (AFM) image of a patterned area may reveal as many as 10⁸ defects/cm², which are typically 5–25 nm high and <500 nm wide (as seen in the example in Figure 2).

 

Figure 1: A whole-wafer LPD map showing a localized pattern of defects, which is characteristic of TDH.

Figure 2: An AFM image of a 4 * 4-µm region of a wafer that exhibited TDH. The estimated defect density in this region is 108 defects/cm2.

The existence of TDH raises the questions of what actually changes on the surface of the affected wafers during shipping and storage and why only some wafers exhibit the phenomenon. TDH can usually be removed by rinsing the wafer with water or heating it to 200°C.¹ Individual defects also tend to disappear very quickly when a hazed wafer is exposed to x-ray beams. These observations suggest that TDH is composed of ionic or polar organic compounds with high vapor pressures at <200°C. Researchers have shown that organic and inorganic contamination can form a haze of particles on the wafer surface during storage. In one study, TDH was artificially created by contaminating wafers with inorganic ions and organic solvents and storing the wafers.² Another set of experimenters exposed wafers to organic compounds that are commonly observed on wafers; the TDH that formed was composed of small particles rather than a continuous film or large residues.³

The study presented in this article investigated the surface changes on silicon wafers during 6 or 18 months of storage and the relationship of those changes to TDH formation. The results indicate that although most wafers have the potential to form more than 1 million >0.12-µm surface particles if the wrong storage conditions exist, particle levels, oxide thickness, and surface organics, ions, and metals will remain very stable when conditions inside the wafer package are strictly controlled. Based on the data collected, a general mechanism for TDH formation was developed.

Experimental Procedures

Wafer Sampling and Storage. For more than 2 years, five prime 200-mm wafers were sampled weekly from the manufacturing line after final cleaning. The sampled wafers were then inspected, placed in Grade 1 packaging, and immediately stored in a local warehouse for either 6 or 18 months. The Grade 1 packaging consisted of a clean, dry polypropylene box; a polyethylene inner bag; a package of desiccant; and an outer, laminated-polyethylene bag with an aluminum layer. Other prime samples were pulled and packaged during the same period and then analyzed within 30 days to establish a "fresh-wafer" baseline.

During the study, the humidity in the warehouse was not controlled, and the temperature varied from 40° to 100°F, depending on the season. At the end of the designated storage time, the outer bags were inspected for tears or pinholes, and the sampled wafers were reinspected for LPDs. The wafers were then sent to the analytical laboratory, where measurements were taken of surface organics, ions, metals, and oxide thickness. (Because only five wafers were included in each sample, all of these analyses could not be performed for each set.) An accelerated TDH test also was performed using a different set of wafer samples to provide rapid feedback on conditions that could affect wafer quality.

Analytical Methods. The pre- and poststorage LPD measurements were taken using a CR80 surface inspection tool from ADE (Westwood, MA). A Dimension 5000 Nanoscope III AFM from Veeco Instruments (Plainview, NY) was used to image the defects. Desorbable surface organics were measured by thermal desorption/gas chromatography with a mass spectrometry detector before June 1998 and with an atomic emission detector after May 1998, as described elsewhere.^4,5 The instrument change was done to obtain more-stable and more-quantifiable results. Surface ion analysis was performed by obtaining a water extraction of the wafer surface and analyzing the extract by ion chromatography (IC) or capillary electrophoresis (CE).⁶ Average oxide thickness was determined using a variation of a method described by Vepa et al.⁷ Surface metals were determined using acid drop extraction and inductively coupled plasma mass spectrometry.

To conduct the accelerated TDH test, fresh prime test wafers were inspected for LPDs and then placed in slots 11, 13, and 15 of a storage box insert. After 1 ml of water was poured into the box, the box was closed and packaged normally, but without the desiccant used in the storage-sample packaging. The box was stored in a temperature-controlled chamber under a series of changing conditions: 16 hours at 17°C, 4 hours at 50°C, 4 hours at –10°C, and 16 hours at 17°C. The relative humidity remained steady at about 90% except during the 50°C step. After this 40-hour period, the wafers were unboxed and reinspected for LPDs, and the change in LPDs was charted.

Results and Discussion

Trend charts were developed depicting the changes in LPDs and whole-wafer averages for surface organics, surface ions, and oxide thickness that had occurred on wafer samples reinspected after 6- or 18-month storage intervals. The trend data for the change in >0.12-µm LPDs after 6 or 18 months are shown in Figure 3. The date the wafers were sampled from the manufacturing line is shown on the x-axis, and each data point represents the average LPD change for a five-wafer sample. The 18-month storage testing data points begin with the late-December 1998 wafers; all points before those data are for 6-month storage testing. The entire database was used to calculate the upper and lower control limits (UCL and LCL) of 48.9 and –44.8 LPDs, respectively. Data points that exceed the UCL are defined as TDH events.

 

Figure 3: Trend chart of the change in >0.12-µm LPDs after 6 or 18 months of storage. Each point represents the average of five wafers sampled on the manufacturing date shown on the x-axis.

As the figure shows, an increase in storage time from 6 to 18 months had no impact on the change in surface LPDs, which averaged only 2.0 per wafer for all samples. The variability around this average primarily reflects the long-term reproducibility of the inspection tools. These surface inspection instruments were serviced and recalibrated at least once between the pre- and poststorage inspections and long-term reproducibility was certainly affected. Before mid-July 1998, an organic cleaning agent was used in the manufacturing process, and that agent was thought to be the source of the TDH events shown. In any case, no TDH events were observed after that organic material was removed from the process.

One might expect that wafers stored for 6 months or longer would have significantly higher levels of common surface organics (listed in Table I) than those stored for only a few days. Freshly packaged wafers already typically have 0.5–1.5 x 10¹⁴ carbon atoms per square centimeter of these compounds, and researchers who have investigated contamination on wafers stored in new boxes for 30 days have found similar levels of the plasticizers and antioxidants.⁸ Thus, it is significant that the chart in Figure 4 depicting average total desorbable organics on wafers stored for 6 or 18 months reveals that the average for the test wafers, 1.2 x 10¹⁴ atoms/cm², did not differ significantly from that of freshly packaged wafers.

 

Figure 4: Trend chart of total desorbable surface organics after 6 or 18 months of storage. Each point represents the average for two wafers sampled on the manufacturing date shown on the x-axis.

Two typical chromatograms of wafers that were stored for 6 months are shown in Figure 5. Both show detectable flame ionization detector (FID) counts of antioxidants, plasticizers, and siloxanes. The dominant signals in the chromatogram at left are a cyclic methylsiloxane (10.3 min), a quinone (15.87 min), and an isobutyrate (24.12 min). Peaks from these compounds are also present in the right-hand chromatogram, but they are overshadowed by signals from the organic cleaning agent (8.46 min, 12.26 min, and 15.10 min) that caused TDH prior to mid-July 1998. Based on these data, it appears that up to 5 x 10¹⁴ carbon atoms per square centimeter of common plasticizers and antioxidants may be deposited during 6–18 months of storage but do not affect TDH if package integrity is maintained.

 

Figure 5: Chromatograms of organics desorbed from wafers after 6 months of storage. The dominant signals in the right-hand chart are from the organic cleaning agent.

The methods used in this study detect only desorbable organics. High-molecular-weight, low-volatility organics could easily have been present on the wafer surface at high levels but would not have been observed. If present, such contaminants would probably have been remnants of prior processes rather than contaminants transferred from the packaging materials.

The inorganic ions commonly found on silicon wafers are also listed in Table I. It is known that surface contamination by such ions, specifically sulfur-containing ions, causes TDH.² In this study, however, sulfur-containing compounds and ions were below the detection limits of the IC and CE methods, which are 10⁹ atoms/cm² and 5 x 10¹⁰ ions/cm², respectively. Fluoride and nitrate ions were also below the detection limits, but ammonium (NH₄+) and chloride (CL^–) ions were found in abundance. The 6- and 18-month trend data for these substances are shown in Figure 6. Freshly packaged wafers typically have as much as 3 x 10¹³ ions/cm² of NH₄+ and 1.9 x 10¹³ ions/cm² of Cl^–, levels that make these compounds among the most abundant contaminants on hydrophilic silicon wafers. The trend data indicated that these levels did not increase during storage, and TDH did not form when package integrity was maintained. However, the potential for particle formation is significant for these ionic contaminants.

 

Organics Inorganic Ions 2,2,4-trimethyl-1,3-pentane diol diisobutyrate (TXIB, plasticizer) Dibutyl phthalate (DBP, plasticizer) Dioctyl phthalate (DOP, plasticizer) 2,6-di-t-butyl-4-methylene-2,5-cyclohexadiene-1-one (oxidized BHT) 2,6-di-t-butyl-1,4-benzoquinone (oxidized form of antioxidant) cyclic polydimethylsiloxanes (-(Si(CH3)2O)n- [n = 5­10]) NH4+ CL­ SO4/ NO3­ NO2­ F­

Table I: Common surface contaminants found on cleaned and packaged wafers.

Figure 6: Trend charts for surface ammonium ions (top) and chloride ions (below) extracted from wafers after 6 or 18 months of storage. Each point represents the average of two wafers sampled on the manufacturing date shown on the x-axis.

Water is the well-recognized accelerating agent for TDH formation. When all other variables are kept constant, the total area of a wafer that is affected by haze can be related to the quantity of water added to a package, as shown in Figure 7. Therefore, a test was required to indicate whether traces of moisture had leaked into the Grade 1 packaging during storage. Tiny remote probes that are commercially available to measure relative humidity and temperature could not be used because they would contaminate the environment inside the package. However, an increase in oxide thickness can be a very sensitive indicator of a leak, because the wafers will continue to oxidize if exposed to humid air for extended periods, reaching a chemical oxide thickness of 1.5 nm compared with the 0.8–1.1-nm (<11-Å) level found on a freshly cleaned wafer.⁹ Therefore, if the results of an oxide measurement test showed that the oxide thickness of a wafer had reached 1.5 nm, it would indicate that the moisture barrier on the wafer's packaging had failed.

 

Figure 7: The area of a wafer affected by TDH as a function of water volume added to the wafer package in a rapid TDH test. Five wafers were equally spaced in slots 1­25 in each box.

A trend chart for average oxide thickness after 6 or 18 months of storage is presented in Figure 8. It is clear from the figure that no significant additional oxidation had occurred on the stored wafers. However, the data suggest a trend in time. There is currently no explanation for this trend, which is not echoed in the LPD changes on wafers sampled and reevaluated at the same time.

 

Figure 8: Trend chart for average oxide thickness after 6 or 18 months of storage. Each point represents the average of two wafers sampled on the manufacturing date shown on the x-axis.

Although the poststorage tests demonstrated that the stored wafers were in satisfactory condition, such testing does not provide the rapid feedback to process engineering that would allow corrective action to be taken if storage conditions are compromised. Therefore, a rapid TDH test that uses water and temperature as accelerants was developed. The rapid TDH test samples were pulled from the manufacturing line at the same time as the storage life samples. The test procedure was designed to create a weak but significant LPD increase on wafers with average levels of surface contaminants and extremely high LPD counts on very contaminated wafers. For example, the effect of 10¹⁵ atoms/cm² of the aforementioned organic cleaning agent is easily observable on a wafer map created after the rapid TDH test, as seen in Figure 9. The results of the accelerated TDH test are shown in the trend chart in Figure 10. Since the small increases in LPD that were observed following this test were not detected on the parallel sample sets used for poststorage evaluation, the rapid test may be slightly more sensitive than the poststorage testing.

 

Figure 9: A whole-wafer LPD map after an accelerated TDH test was performed on a wafer contaminated with 1015 atoms/cm2 of an organic cleaning agent.

Figure 10: Trend chart for the change in LPDs after an accelerated TDH test. Each point represents the average of three wafers sampled on the manufacturing date shown on the x-axis.

A Proposed TDH Formation Mechanism

Even given the low levels of surface contamination on commercially available silicon wafers, the potential to form millions of >0.12-µm particles is great. For example, 10¹² molecules/cm² of NH₄Cl could form 1000 particles/cm² or 300,000 particles/wafer under ideal conditions. Similarly, 10¹⁵ carbon atoms per square centimeter could form 100,000 particles per wafer. However, such formation usually does not occur without an accelerant such as water condensation.

The following general mechanism for TDH formation is proposed: (1) A packaged wafer is contaminated with water-soluble ions or water-soluble organic molecules; (2) other organic molecules also deposit on the wafer, rendering it more hydrophobic than a typical freshly packaged wafer; (3) changes in relative humidity caused by temperature changes or changes in total water content in the package cause water to condense on the wafer surface and dissolve the water-soluble contaminants; (4) the wafer's hydrophobic surface causes the water to form microscopic droplets; and (5) the microdroplets later evaporate, leaving individual TDH defects. The density and size of the microdroplets will be affected by both contamination levels and the relative humidity in the package. Given that organic molecules from plastic packages and some inorganic ions such as NH₄+ and Cl^– are ubiquitous, it may be necessary to take more-stringent measures to control relative humidity in the shipping package than are now used when wafer customers begin to inspect for 40–90-nm LPDs.

Conclusion

Study results indicate that silicon wafers stored for up to 18 months do not have significantly higher levels of >0.12-µm LPDs, oxide thickness, or surface organics, metals, and ions than freshly packaged wafers. Although the typical average levels of water-soluble inorganic ions on packaged wafers provide a huge potential to form TDH, this potential is seldom realized if humidity in the package is controlled, because condensed water is needed to accelerate the haze formation process.

However, organic or ionic contamination on the wafer surface that exceeds typical levels may cause TDH to form even if package humidity is properly controlled. For example, during this study, an organic cleaning agent caused TDH formation, even though package integrity (and therefore relative humidity inside the package) was maintained.

Acknowledgments

The authors would like to thank Gary Anderson, Marty Adams, Kenny Ruth, Phil Schmidt, Andrei Stefanescu, Peng Sun, and Hao Zhang for their contributions to analytical methods development and data collection during this research. Additional thanks go to Mike Tyler and Lillian Rose for their development of the accelerated test method and to Gianpaolo Mettifogo for his on-line testing of this procedure.

References

1. D Kinkead and J Higley, "Targeting Gaseous Contaminants in Wafer Fabs: Fugitive Amines," MICRO 11, no. 6 (1993): 37–40.
2. N Munter et al., "Preparation and Characterization of Time Dependent Haze on Silicon Surfaces," in Proceedings of the Fifth International Symposium on Ultra-Clean Processing of Silicon Surfaces (Leuven, Belgium: ACCO, 2000), 91–92.
3. K Vepa et al., "Role of Organics and Moisture on Silicon Wafer Surfaces," in Proceedings of the ECS Spring Meeting 93, no. 1 (Pennington, NJ: The Electrochemical Society, 1993), 1141–1147.
4. P Sun, M Adams, and T Bridges, "Monitoring Organics on Wafer Surfaces Using Thermal Desorption GC-MSD/AED," MICRO 18, no. 3 (2000): 59–71.
5. P Sun et al., "Molecular and Ionic Contamination Monitoring for Cleanroom Air and Wafer Surfaces" in In-Line Characterization Techniques for Performance and Yield Enhancement in Microelectronic Manufacturing, SPIE Vol. 3215, ed. DK DeBusk and S Ajuria (Bellingham, WA: International Society for Optical Engineering, 1997), 118–127.
6. P Sun and M Adams, "Demonstrating a Contamination-Free Wafer Surface Extraction System for Use with CE and IC," MICRO 17, no. 4 (1999): 41–46.
7. K Vepa, K Baker, and L Shive, "A Method for Native Oxide Thickness Measurement," in Cleaning Technology in Semiconductor Device Manufacturing IV, ed. R Novak and J Ruzyllo (Pennington, NJ: The Electrochemical Society, 1995), 358– 365.
8. F Sugimoto and S Okamura, "Adsorption Behavior of Organic Contaminants on a Silicon Wafer Surface," Journal of the Electrochemical Society 146, no. 7 (1999): 2725–2729.
9. L Shive, C Frey, and C Vitus, "A Probe of Chemical Oxide Growth Conditions," in Proceedings of the Fifth International Symposium on Ultra-Clean Processing of Silicon Surfaces (Leuven, Belgium: ACCO, 2000), 127–128.

Larry W. Shive, PhD, is an MEMC fellow in the epitaxial technology department at MEMC Electronic Materials (St. Peters, MO). During his 23 years at MEMC, he has performed research on crystal defects, surface cleaning, and silicon epitaxial defects. He has a PhD in chemistry from Texas A&M University (College Station, TX) and is a member of the American Chemical Society and the Electrochemical Society. (Shive can be reached at 636/474-5370 or lshive@memc.com.)

Richard E. Blank, PhD, is a senior engineer in the epitaxial technology department at MEMC Electronic Materials, where he has worked for five years. He has extensive experience in optical inspection and characterization of defects on polished and epitaxial silicon wafers. Before joining MEMC, Blank developed military and commercial night-vision systems and spectroscopic medical instruments. He has a BS in physics from the University of Rochester (Rochester, NY) and an MS and a PhD in physics from Michigan State University (East Lansing). (Blank can be reached at 636/474-7322 or rblank@memc.com.)

Karen H. Lamb is a research technician in the epitaxial technology department at MEMC Electronics Materials. She has a BS degree in psychology from St. Lawrence University in Canton, NY. (Lamb can be reached at 363/474-5534 or klamb@memc.com.)

MicroHome | Search | Current Issue | MicroArchives
Buyers Guide | Media Kit

Questions/comments about MICRO Magazine? E-mail us at cheynman@gmail.com.

© 2007 Tom Cheyney
All rights reserved.