ULTRAPURE MATERIALS—GASES

Using a nondestructive test to qualify corrosive specialty gas cylinders

Alan Zdunek, Tracey Jacksier, John Borzio, and Denis Rufin, Air Liquide

A reliable supply of ultrapure electronic specialty gases is critical to maintaining the high tool uptimes necessary to maximize productivity and manufacturing yield in semiconductor fabs. The delivery of such gases poses special challenges, however, because of their highly corrosive and reactive nature. Halogen gases such as boron trichloride, for example, can hydrolyze in the presence of moisture and react with metal container surfaces, forming particles that compromise gas purity.¹ And if such particles reach the point of use, the wafer surface can become contaminated.

The first challenge suppliers of corrosive specialty gases face in improving the reliability of their products is to develop a delivery cylinder that remains nonreactive to the gases over many product fills, yet minimizes cost. Although the steel material used to form gas cylinders is itself corrosion resistant, that characteristic can be enhanced by various techniques. An electroless-nickel surface treatment has been found to be a good choice for corrosive gas delivery because it can be applied consistently to large numbers of cylinders. The resulting coating does not react readily in corrosive environments, and the procedure is economically attractive compared with other alternatives.

Figure 1: Typical measurement results showing thickness uniformity of electroless-nickel cylinder coating.

Figure 2: Micrographs illustrating the properties of electroless-nickel cylinder coatings: (a) cross section and (b) surface morphology at 200x magnification.

Described in detail elsewhere, the electroless-nickel coating is a metallic, nickel-phosphide glass formed on the interior surface of the steel cylinder by an autocatalytic chemical reaction.^2—4 The electroless nickel inherently passivates, forming a strongly bonded, low-porosity surface layer that resists undercutting and has a consistent thickness from top to bottom, as seen in Figure 1. The two scanning electron microscope (SEM) micrographs in Figure 2 also illustrate the coating's beneficial properties. The cross section in Figure 2a shows the excellent adhesion of the electroless nickel to the steel cylinder substrate, while the surface view in Figure 2b shows the coating's smooth morphology, which mimics the underlying polished-steel substrate. Qualitative evidence of electroless nickel's thermodynamic stability in corrosive environments is presented in Figure 3, which shows a galvanic series of metals and alloys in 5% sodium chloride.⁵ In this series, electroless nickel has a relatively anodic corrosion potential compared to nonpassivated 316 and 304 stainless steel, indicating it is more nonreactive, or noble. It is approximately as noble as the Hastelloy B and C alloys in aqueous halide environments. In contrast, aluminum alloys and low-carbon steel have more cathodic potentials and are relatively active, or easily oxidizable, in halide environments.

Figure 3: Galvanic series of metals and alloys in 5% NaCl.⁵ A chrome-oxide layer has been grown on the passivated stainless steels.

Figure 4: Corrosion rates of various metals, alloys, and metallic coating in 35% hydrochloric acid (in mils per year) determined during accelerated corrosion-rate testing.¹⁰

Our research with treated gas cylinders at Air Liquide has confirmed electroless nickel's corrosion resistance. The results of accelerated corrosion-rate testing, presented in Figure 4, indicate that electroless-nickel-coated steel cylinders are less reactive in aggressive halide (35% hydrochloric acid) environments than either stainless steel or carbon steel and, as was seen in Figure 3, are comparable to Hastelloy C- and B-series alloys. Figure 5 shows additional tests in which various materials were exposed to dry HC1 gas revealed that the amount of leached metals from an electroless-nickel-coated steel cylinder surface is comparable to that for 316L electropolished stainless steel and is less than the concentration for 4130 (chrome-molybdenum) steel, which is the base material usually used for cylinder construction. The remainder of this article focuses on a quality control (QC) method that has been developed to qualify surface-treated cylinders for use with corrosive gases. The results of tests of in-service cylinders are also provided.

Figure 5: Soluble-metal concentrations of various materials after 14-day exposure to HCl gas.¹⁰

Cylinder Quality Control

A consistent cylinder package is a critical parameter for reliable delivery of ultrapure specialty gases. It is common practice in the industry to inspect the inside of cylinders for surface smoothness prior to filling, and several methods have been developed to quantify particle shedding.^6—8 However, visual observation is not sensitive to coating thickness or such microdefects as porosity and cracks, which can adversely affect gas quality. Our studies have shown that coating parameters such as thickness, porosity, and adhesion are dependent on the coating process and on the steel-surface preparation done prior to applying the coating. The cross-section micrograph in Figure 6, for example, shows an electroless-nickel-coated cylinder surface that had been poorly prepared by an outside plating supplier. A coating defect is apparent and adhesion of the coating to the substrate is uneven, compared to that seen in Figure 2a. Although such defects can be minimized with the use of good coating practices, quality control techniques also should be employed to ensure that coating specifications have been met by the outside plating supplier on all cylinders. Indeed, such techniques are routinely used in the plating industry to ensure that the tolerances of plated parts are maintained over large production runs.

Figure 6: Micrograph of an electroless-nickel-coated cylinder cross section with a coating defect, which resulted from poor surface preparation.¹⁰

Figure 7: Schematic drawings showing the nondestructive QC method for measuring cylinder coating quality.¹⁰

To meet this need for strict quality control, we have developed a nondestructive test specifically for qualifying electroless-nickel-coated cylinders destined for use with electronic specialty gases.^9,10 Based on two key properties of electroless nickel—coating thickness and coating porosity—the test can be used to verify the acceptability of the cylinder coating before gas filling is performed. The coating thickness is measured by inserting a probe into the small opening of a gas cylinder using a specially designed tool so that a hard contact can be made with the coated cylinder surface, as shown in Figure 7. The measurement method is based on magnetic principles and uses the substrate's magnetic properties to determine the coating thickness. Figure 8 provides typical results for the coating-thickness component of the test. In that instance, visual observation had indicated that all of the cylinders in the lot being tested were smooth and had no visible defects. Although the majority of the cylinders did pass the test, several did not meet the minimum thickness specification (nos. 17—20) and were rejected.

Figure 8: Typical results of the coating-thickness component of the QC test for electroless-nickel-coated gas cylinders. Results shown represent the average thickness over the length of each cylinder.¹⁰

The coating-porosity component of the QC test is based on electrochemical measurements and is illustrated in Figure 7. The test yields a quantitative number for porosity that is normalized to an arbitrary scale from 1 to 100. A cylinder with a coating porosity value of less than 3 is considered to have negligible porosity and would be accepted, but any cylinder with a porosity value greater than 3 would be rejected and returned for reprocessing. Figure 9 shows coating-porosity values for 20 electroless-nickel-coated cylinders. As was the case for the coating-thickness test results shown in Figure 8, several of the cylinders had to be rejected.

Figure 9: Typical results of the coating-porosity component of the QC test for electroless-nickel-coated gas cylinders.

It is standard industry practice to verify product integrity after cylinder filling by analyzing metals in the liquid-phase gas, which is accepted as a worst-case scenario. Figures 10a—c provide such data for cylinders of Cl^₂, BCl^₃, and HBr, respectively. The consistently low impurity levels that have been achieved provide further evidence of the corrosion resistance of the treated and qualified cylinders.¹¹

Figure 10: Results of liquid-phase metal analyses of cylinders containing (a) chlorine (Cl^₂), (b) boron trichloride (BCl^₃), and (c) hydrogen bromide (HBr), respectively.

Cylinder Preservation

The QC method described above is also useful for periodically requalifying cylinders. The quality of such cylinders must be preserved by controlled processing and maintenance throughout their life cycle. For example, gas suppliers should specify the amount of product that should be left in the cylinder by customers. For electronic specialty gases, it is especially important that the surface of the cylinders remain unaffected by corrosion.

Figure 11: Micrographs of an electroless-nickel-coated cylinder after 27 months of exposure to hydrochloric acid: (a) surface morphology at 200x magnification and (b) cross section.

Analyses that we have performed indicate that electroless-nickel surface properties can be maintained for more than 2 years with controlled returned-cylinder processing. Figure 11 shows two micrographs of an electroless-nickel-coated cylinder after 27 months in Cl^₂ service. The surface morphology (Figure 11a) is similar to that seen in Figure 2b (which shows a cylinder prior to gas exposure), with no pitting, cracks, or defects. The excellent adhesion of the coating after Cl^₂ exposure (illustrated in the cross section in Figure 11b) was verified by ASTM coating-adhesion tests.⁹ Additional testing has indicated that cylinders of HBr and HCl that have been filled, returned, and reprocessed exhibit metal levels that are comparable to or lower than the initial fill. Extended shelf-life studies are also being conducted to determine the life span of the coated cylinders.

Conclusion

A nondestructive quality-control method has been developed to ensure that electroless-nickel-coated cylinders destined for high-purity corrosive gas delivery meet strict specifications. Specifically, compliance with coating-thickness and coating-porosity specifications is used as the basis to accept or reject such cylinders. Metals analyses performed after the initial fill has confirmed the consistent quality of the gases contained in cylinders qualified via this QC technique and SEM analysis of qualified cylinders after 27 months of Cl₂ gas service showed no degradation of the coating.

References

1. T Ohmi et al., "Formation of Chromium Oxide on 316L Austenitic Stainless Steel," Journal of Vacuum Science and Technology A 14, no. 4 (1996): 2505—2590.

2. "Surface Engineering," in ASM Handbook, Vol. 5 (Materials Park, OH: ASM International, 1994), 290—310.

3. GO Mallory and JB Hajdu, Electroless Nickel Plating: Fundamentals and Applications (Orlando, FL: American Electroplaters and Surface Finishers Society, 1991).

4. YV Matilis et al., "The Corrosion Protection of Chlorine Compressors and Accessories by Nickel Plating," Khimicheskoe i Neftyanoe Mashinostroenie 11 (November 1971): 972—973.

5. MG Fontana, Corrosion Engineering (New York: McGraw-Hill, 1986).

6. G Kasper, HY Wen, and HC Wang, "Developing Particle Standards for Cylinder Gases," Microcontamination 7, no. 1 (1989): 18—26, 68—71.

7. HW Wang, HY Wen, and G Kasper, "Mechanical Shocks on Gas Cylinders: Characterization and Effects on Particle Content," in Proceedings of the Annual Meeting of the Institute of Environmental Sciences (Mount Prospect, IL: IES, 1989).

8. HC Wang, HY Wen, and G Kasper, "Factors Affecting Particle Content in High-Pressure Cylinder Gases," Solid State Technology 32, no. 5 (1989): 155—158.

9. A Zdunek, P Vanecek, and E Kernerman, "Method and Apparatus for Measuring Coating Quality," U.S. patent pending, 1997.

10. A Zdunek and T Kimura, "A Corrosion Engineering Perspective to Specialty Gas Distribution Systems," in Proceedings of the Workshop on Gas Distribution Systems at Semicon West 98 (Mountain View, CA: SEMI, July 1998), K1—K21.

11. T Jacksier and J Borzio, "Field Implementation of Sampling Techniques for Metals Analysis in Electronic Specialty Gases," in Proceedings of the Institute of Environmental Sciences and Technology Conference, (Mount Prospect, IL: IEST, 1997), 255—261.

Alan Zdunek, PhD, is a scientist at Air Liquide's Chicago Research Center, where his group performs research on corrosive gas handling and cylinder technology. Zdunek has a BA in chemistry from Knox College and an MS and PhD in chemical engineering from the Illinois Institute of Technology. He serves as task force leader for the SEMI International Corrosion testing task force of the contamination in gas distribution systems subcommittee. (Zdunek can be reached at 708/579-7864 or alan.zdunek@airliquide.com.)

Tracey Jacksier, PhD, is a scientist at Air Liquide's Chicago Research Center, where she is a group leader responsible for inorganic analysis of electronic gases. Jacksier has a BS from Purdue University and a PhD in physical chemistry from the University of Massachusetts. She has authored more than 35 articles and technical presentations and is serving as secretary of the SEMI international gases committee.

John Borzio is the plant manager of Air Liquide's electronic specialty gas production facility in Morrisville, PA. He has a BS from Rutgers University and has over 25 years of experience in electronic gas production, handling, and analysis.

Denis Rufin, PhD, is the business manager of Air Liquide's electronic specialty gas production facility in Chalon-sur-Saône, France. He has spent the last 14 years in marketing and project management of all aspects of bulk and process gas supply for the semiconductor manufacturing industry. He participated in the creation of Air Liquide Laboratories in Japan and organized a major U.S. semiconductor manufacturer's bulk silane supply. Rufin has an MS in physical engineering from the Polytechnic Institute of Grenoble and a PhD in low-temperature physics from the University of Grenoble. He has authored more than 12 technical papers and has many patents in the field of gas application technology.

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