PROCESS EQUIPMENT—TOOL PACKAGING

Comparing the molecular contamination contribution of clean packaging films

Steve Lin and Sarah Graves, Applied Materials

In the past few years, antistatic polyethylene (PE) films, also known as pink poly, have been used as an all-purpose packaging material for a wide range of cleanroom-compatible packaging applications. Intended to protect such products as components used in semiconductor process equipment, these films should not contribute to the degradation or recontamination of the package contents. During the film manufacturing process, however, a number of chemical additives are incorporated to meet various functional requirements; for example, an amine or amide may be added as an antistatic agent, calcium carbonate or calcium silicate as an antiblock,¹ and an antioxidant or high-temperature stabilizer for chemical stability.² The concern in the semiconductor industry is that outgassing of these additives can contaminate, and potentially corrode, tool components that come in direct contact with packaging films.³ Indeed, it has been confirmed that black residue found between antistatic PE and aluminum foil was closely related to the incorporation of chemical additives in the packaging film. Furthermore, outgassing measurements have shown that test coupons packaged in generic polymer films exhibited excessive levels of hydrocarbons.

Figure 1: Black residues from packaging materials collected on cleanroom wipes prior to component installation.

Figure 1 shows examples of black residue collected from the surfaces of a component packaged in a polymeric film. Because of such contamination, it is common practice to wipe critical components prior to their installation in process tools. This labor-intensive practice results in a significant productivity loss, and tool decontamination has become a critical quality issue, which can only worsen if no effective resolution is implemented soon.

At Applied Materials, various issues relating to clean packaging practices were clearly identified more than a year ago, and a number of strategic projects focusing on packaging material selection for contamination-free manufacturing have been completed. It was found that choosing the proper materials is critical to preventing packaging-related contamination problems. Although it is common practice to specify the particulate cleanliness levels of packaging films, molecular contamination levels are not specified. In the study reported in this article, two analytical techniques, outgassing per ASTM E 595-93 and ion chromatography (IC), were employed to quantify the levels of outgassing hydrocarbons and leachable ions of a variety of materials, ranging from a cleanroom wipe to expensive specialty polymeric films. These results were then compared with measurements of the cleanliness levels of representative metallic parts, which had been analyzed using a modified ASTM E 1559-93 technique and IC. The findings indicated that the proper choice of polymeric film can reduce the level of microcontaminants introduced by packaging materials. In addition, criteria for molecular cleanliness can be developed to facilitate the material selection process.

Experimental Methods

Outgassing Tests. After test specimens of polymeric films were randomly selected from various types of materials, each specimen was double bagged in a Class 10 cleanroom using film from the same batch. The packaged specimens were then sent to various analytical laboratories for designated tests and measurements. As mentioned above, the outgassing test for material specimens was performed in accordance with ASTM guidelines. In this test, a specimen is placed in a vacuum environment with a pressure of less than 1 x 10^—5 torr at 125°C for 24 hours. Three measurements—the total mass loss (TML), the mass of collected volatile condensable materials (CVCM), and the weight change of a previously outgassed specimen caused by water vapor recovery (WVR)—are obtained for each sample. These measurements, reported as weight percent (wt %), are the ratios of measured weight changes to the original weight of the test specimen. The collector for measuring CVCM is located downstream in the vacuum system and maintained at 25°C. At the end of each test, it is visually examined and the appearance of the condensed materials is recorded. CVCM is used frequently for research purposes and has been found an adequate indicator of the potential for recontamination from the hydrocarbon outgassing of polymeric films.^4,5

To quantify hydrocarbon outgassing of the metallic specimens, which included parts with surfaces of chemically cleaned 6061 T6 aluminum, electroless nickel plating, and electropolished 304 stainless steel, measurements were conducted using a method modified from ASTM E 1559.⁶ The protocol used is almost identical to the guidelines established in the ASTM standard with the exception of data manipulation and presentation. This test method requires placing a specimen in a temperature-controlled effusion cell, which is then inserted into an ultra-high-vacuum chamber at a pressure of about 1 x 10^—9 torr. The effusion cell with the specimen is evacuated for 2 minutes in a loadlock chamber prior to its insertion into the main chamber. The cell is then heated to 125°C and maintained at this temperature for 2 hours. The outgassing flux from the specimen leaves the effusion cell through a 3-mm-diam hole, and four quartz crystal microbalance (QCMs) detectors, maintained at temperatures of 80, 160, 220, and 298 K, respectively, are used to collect portions of the outgassing flux in situ. The four QCMs provide the mass gains of the outgassing species condensed at the various temperatures.

Category	Molecular Weight Range (amu)
Very high volatility	<50
High volatility	50—200
Medium volatility	200—400
Low volatility	>400

Table I: The four categories of volatility as referenced in ASTM E 1559.⁶ The primary species in the very-high-volatility category is water; most surface contaminants fall in the other three categories.

The use of four QCMs served as an integrator to quantify the mass loss of the specimens over the predetermined duration. The data acquired from each measurement were further manipulated to derive the final results in accordance with the volatility of the outgassing species. Each specimen's TML is the sum of the mass losses of very-high-, high-, medium-, and low-volatility species. Generally speaking, these four categories of volatility can be associated with species within different molecular weight ranges, as shown in Table I. The category of very-high-volatility species with molecular weights of <50 amu consists mainly of water with a limited presence of carbon monoxide and carbon dioxide. Even though hydrogen is a common outgassing species in ultra-high-vacuum environments, it cannot be detected by this test method because of its very low boiling point. The majority of surface contaminants, such as hydrocarbons, usually fall into the high-, medium-, and low-volatility categories, depending on the species' chemical structure and boiling point. Thus, an additional parameter, sub-sum, which is the sum of the mass losses of high-, medium-, and low-volatility species, was used in this study as an indicator of outgassing hydrocarbons detected on the metallic specimens.

Ion Chromatography. To quantify ionic contamination from the material samples and the metallic parts each specimen was immersed in a static DI-water bath at temperatures between 68° and 73°F for 24 hours. The DI water, which has a very high resistivity of approximately 18 M‡, acted as a leaching or extraction solution. Following the immersion period the water solution containing trace amounts of ionic species was analyzed using a DX-300 IC system (Dionex, Sunnyvale, CA) in accordance with the general guidelines established in U.S. EPA Methods 300.0 and 300.7 for anions and cations, respectively. The surface concentration, expressed in terms of 10¹² molecules/cm², was derived by normalizing the total amount of ions in the solution to the total wetting surface of the test specimen. This technique allowed a quantitative comparison of the surface ionic species on the polymeric materials and the metallic parts.

Results and Discussion

Outgassing Data. The results from the first phase of evaluation, during which one specimen each from various types of materials including eight polymeric films was tested, are displayed in Figure 2. As seen in this figure, FEP and the polychlorofluoroethylene-based materials Aclar 22C and 33C exhibited a relatively low outgassing rate (TML = 0.1 wt %) with very insignificant or no detectable CVCM (0.01 wt %). Natural PE had a moderate outgassing rate (0.12 wt %) and a moderate CVCM (0.03 wt %), while antistatic PE, natural polypropylene, and the cleanroom wipe were found to exhibit moderate outgassing rates (0.15—0.45 wt %) but relatively higher CVCMs (0.06—0.15 wt %). Finally Capran 980, nylon 6, and the lint-free packaging cloth (commonly used to protect delicate surfaces) showed much higher outgassing rates (>=2.0 wt %) and moderate CVCMs (0.01—0.03 wt %).

Figure 2: Comparative outgassing test results for 10 types of material specimens, including eight polymeric films. (Note that results for CVCM are on a different scale than those for TML and WVR.)

To further understand these variations in outgassing levels, additional tests were performed on natural PE, antistatic PE, and antistatic nylon. Among these three types of polymeric films, antistatic nylon had displayed the lowest level of CVCM, although the material's hygroscopic property was evident from the very high level of WVR (>6.0 wt %). The additional test results for antistatic nylon showed a very high TML level of >2.0 wt % and a moderate CVCM of ~ 0.03 wt %, which were very similar to the previous results for that material, shown in Figure 2. Test results for natural and antistatic PE are summarized in Figure 3. The typical TML range for natural PE (0.12—0.20 wt %) was similar to that for antistatic PE (0.18—0.21 wt %), but the typical CVCM range for natural PE (0.03—0.06 wt %) was approximately 50% lower than that for antistatic PE (0.08—0.10 wt %). Thus, on a quantitative basis, antistatic PE appears to be a greater potential source of hydrocarbon contaminants than natural PE, which suggests that replacing antistatic PE with natural PE should reduce the risk of such contaminants being introduced on process tools by packaging materials. However, it was surprising to observe that in one of the outgassing tests (Series 2), the antistatic PE specimen yielded no detectable CVCM.

Figure 3: Comparative outgassing test results for natural and antistatic PE films. Two test series were run for natural PE, four for antistatic PE.

Ionic Contamination. The surface concentrations of the various ionic species that had been leached from the material specimens and which were quantified by IC are expressed as 10¹² molecules/cm² in Figure 4. These data clearly indicate that both natural PE and FEP had relatively low levels of leachable ions. Aclar 22C and Aclar 33C exhibited some leaching of chloride (Cl^—) and fluoride (F^—), while sodium (Na⁺) was clearly the primary ionic species leached from the nylon-based materials Capran 980 and nylon 6. Indeed, Capran 980 showed a particularly high level of Na⁺. Alone among the materials tested, the lint-free fabric exhibited relatively high levels of such ionic contaminants as sulfate (SO^₄—2), magnesium (Mg⁺²), and calcium (Ca⁺²); it also had high levels of Na⁺.

Figure 4: Comparative IC results for 10 types of material specimens, including eight polymeric films.

Figure 5: Comparative IC results from additional testing of natural PE, antistatic PE, and antistatic nylon films.

Additional IC tests of natural PE, antistatic PE, and antistatic nylon were conducted to understand the variation in ionic contamination. The significant ionic species and typical surface concentrations detected in this test series are compared in Figure 5. Three ionic species—nitrite (NO^₂—), phosphate (PO^₄—3), and lithium (Li⁺)—were rarely detected and are not displayed in this figure. With the exception of Na⁺, the levels of ionic species that were detected on the natural and antistatic PE were within the same order of magnitude. In contrast, the antistatic nylon yielded higher ionic contamination levels than did either type of PE. Specifically, the levels of Cl^—, ammonium (NH^₄+), potassium (K⁺), and Ca⁺² detected on the antistatic nylon were substantially higher. In other words, the amounts of ionic contaminants detected on natural and antistatic PE were lower and showed less variation when compared with the ionic contamination levels detected on antistatic nylon. These results can be further used to derive acceptance criteria for setting clean packaging specifications.

	Outgassing Classification, Mass Loss (µg/cm2)
Metallic Specimen Type	Total	Very High Volatility	High Volatility	Medium Volatility	Low Volatility	Sub- Sum
Chemically cleaned aluminum	0.2— 0.45	0.2— 0.45	0.005— 0.02	0.02— 0.05	0.005— 0.015	0.03— 0.085
Electroless nickel	0.035— 0.12	0.025— 0.75	0.005— 0.02	0.002— 0.03	0.02— 0.03	0.03— 0.08
Electro- polished 316 stainless steel	0.08— 0.15	0.05— 0.12	0.006— 0.02	0.02— 0.02	0.02— 0.04	0.045— 0.08

Table II: Typical ranges of outgassing test results for three types of metallic specimens.

Cleanliness Comparison. The typical ranges of the outgassing test results for the three types of metallic specimens are shown in Table II. Because of the differences in collection methods and collection duration used for the outgassing tests of metallic and polymeric material specimens, it is difficult to compare the results for the two types of samples. Additionally, presenting the measurement results in terms of weight percent makes it difficult to make quantitative comparisons. However, a comparison is possible by converting CVCM to mass loss per unit area using the known density and thickness of the polymeric films. For natural PE and antistatic PE, it was estimated that the CVCM levels were in the 3—6 mg/cm² range based on the film density. For nylon, the CVCM level was in the range of 1—2 mg/cm². Compared with the sub-sums shown in Table II for the various metallic surfaces, these levels are at least two to three orders of magnitude higher.

What was commonly detected on the metallic specimens during outgassing tests were significant amounts of various plasticizers used in processing the films in which the specimens had been packaged. Recontamination caused by the hydrocarbon outgassing of polymeric film appears to be the rule rather than the exception for direct packaging of metallic components in plastic films. Further testing revealed that this high level of hydrocarbon contaminants introduced by packaging materials was eliminated when the polymeric film was replaced with ultra-high-vacuum-compatible aluminum foil.

	Anion Surface Concentration (1012 molecules/cm2)
Metallic Specimen Type	F—	Cl—	NO2—	Br—	NO3—	SO4—2	PO4—3
Chemically cleaned aluminum	~4	~60	~25	<2	~20	~3	<2
Electroless nickel	~10	~125	~50	~2	~100	~35	~750
Electropolished 316 stainless steel	<3	~10	~3	<2	~5	~45	~25

Table III: Typical anion surface concentrations detected on three types of metallic specimens.

	Cation, Surface Concentration (1012 molecules/cm2)
Metallic Specimen Type	Li+	Na+	NH4+	K+	Mg+2	Ca+2
Chemically cleaned aluminum	~8	~10	~5	~3	~125	~60
Electroless nickel	~8	~55	~85	~20	~5	~220
Electropolished 316 stainless steel	~8	~85	~10	~5	~10	~10

Table IV: Typical cation surface concentrations detected on three types of metallic specimens.

The IC testing of the metallic specimens revealed that, in general, the levels of detected ionic species on the metallic surfaces were higher than those on the polymeric materials. The typical surface concentrations of ionic species detected on the three types of metallic parts are summarized in Tables III and IV for anions and cations, respectively. Based on the rule of thumb that the cleanliness level of packaging films should be at least three to five times better than the cleanliness specifications of the parts to be packaged in order to minimize the risk of cross-contamination, a comparison of the typical ionic surface concentrations of the three common polymeric films (shown in Figure 5) with the data for metallic specimens reveals that nylon is not an appropriate choice for packaging metallic components because of its high levels of various ionic species. It also appears that both natural and antistatic PE exhibit acceptable levels of ionic cleanliness. However, it should be noted that IC is only effective for the detection of soluble species, and some antiblocks used in antistatic PE are not soluble and thus cannot be detected by this technique.

Conclusion

Based on the results of this study, it appears that, because the levels of outgassing from natural PE in terms of CVCM are relatively lower, the replacement of antistatic PE with natural PE can reduce the potential for cross-contamination of packaged components. However, it is also clear that most currently available films, including natural PE, are not adequate for the direct packaging of metallic components that require very low levels of hydrocarbon contaminants. Even though there are a number of high-end products such as FEP and Aclar 33C on the market, the use of such films for general-purpose packaging is unlikely because of their very high costs (15 times more expensive than PE). There are also problems with their use related to the limits of various physical properties. This situation means that there is an urgent need to develop and commercialize affordable polymeric films with lower hydrocarbon outgassing. It is estimated that potential polymeric films for ultraclean packaging should have a CVCM of no greater than 0.003 to 0.006 wt %. As an interim solution, Applied Materials has documented best-known practices using natural PE for many clean packaging applications.

References

1. Havens MR, "Understanding Pink Poly," in EOS/ESD Symposium Proceedings, Rome, NY, EOS/ESD Association, pp 95—101, 1989.

2. Losev YP, "High-Temperature Stabilization of Polyolefins," in Handbook of Engineering Polymeric Materials, Cheremisinoff NC (ed), New York, Marcel Dekker, pp 81—92, 1997.

3. Anderson J, Denton R, and Smith M, "Contaminated Antistatic Polyethylene," in EOS/ESD Symposium Proceedings, Rome, NY, EOS/ESD Association, pp 36—70, 1987.

4. Goodman J, and Mikkelsen K, "Materials for Microenvironment Construction," in Proceedings of the 39th Annual Technical Meeting of the Institute of Environmental Sciences, Mount Prospect, IL, IES, pp 538—543, 1993.

5. Wilson G, and Van Wees H, "Technological Advancements for ESD Problems," CleanRooms, 10(7): 46—48, 1996.

6. Garrett J, private communication.

Steve Lin, PhD, a member of the technical staff of supplier management operations at Applied Materials in Austin, TX, is responsible for the development of specifications and the establishment of practices that are critical to the continuous acquisition of clean equipment components. He has held a number of technical positions at the company, including senior manufacturing engineer, senior mechanical engineer, and project manager. He has also worked extensively in various areas of the semiconductor process equipment industry, including product design, process development, and manufacturing operations. Lin has a BS in chemical engineering from National Tsing Hua University, Taiwan, ROC. He received MS and PhD degrees in materials science and engineering from the University of Texas, Austin. (Lin can be reached at 512/272-2164 or steve_lin@amat.com.)

Sarah Graves is an engineering manager for central engineering and technology at Applied Materials in Austin, where she runs the microcontamination reduction programs for the Austin and Santa Clara manufacturing cleanroom operations. During her 10 years with the company, she has also been responsible for microcontamination reduction efforts in both oxide and metal etch and for particle analysis and data interpretation relating to semiconductor equipment manufacturing in the corporate microcontamination group. Graves received her BS in metallurgical engineering from the University of Wisconsin and an MS in metallurgical engineering and materials science from Carnegie Mellon University. Before joining Applied, she held positions with Data General, Surface ScienceLaboratories, and Kevex. (Graves can be reached at sarah_graves@amat.com.)

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