Cleanroom Technologies

Evaluating the ESD performance of gloves under realistic cleanroom conditions

Roger W. Welker, R.W. Welker Associates; Arleigh Hartkopf and Peter G. Lehman, Ansell Occupational Healthcare; Steven B. Heymann, NOVX; and Carl Newberg, River's Edge Technical Services

The last installment in a three-part series presents the results of a multivariable study that compares the discharge times of nitrile, PVC, and natural latex gloves.

As the functional structures of electronic and electromechanical devices continue to shrink, their ability to withstand the effects of electrostatic overstress and electrostatic discharge (ESD) also diminishes. Among the devices most sensitive to ESD are magnetoresistive and giant magnetoresistive heads, solid-state lasers, and semiconductor devices with gate widths smaller than 0.35 µm. These products have design and performance characteristics that necessitate the establishment of an aggressive and comprehensive ESD control program in the manufacturing environment. Cleanroom gloves are a critical part of such a control program.

A review of the literature reveals that the need for comprehensive ESD control programs in high-technology manufacturing environments was recognized as early as 1983.¹ A year later, a researcher pointed out the importance of involving operating personnel in any ESD solution.² Other studies have addressed such issues as the need for ground- and charge-monitoring systems, the selection of materials for ESD applications, the choice and management of air ionizers, the appropriate fabrics for cleanroom garments, and the potential for ESD from pressure-sensitive adhesives.^3­10 A comprehensive ESD control program designed specifically for magnetoresistive heads has also been described.¹¹

However, none of those articles focuses specifically on the issue of the ESD performance of cleanroom gloves. This lack of published work is somewhat surprising, given that gloves are an integral element in the generation of electrostatic charge and the transfer of that charge to and from products during manufacturing operations. One purpose of this series of articles is to fill that information gap. The first article (MICRO, May 1999) discussed the contamination and ESD tests used to certify cleanroom gloves.¹² The second article (MICRO, September 1999) examined the contamination performance of both natural latex and nitrile gloves and described the beneficial effects of washing on glove contamination levels.¹³ To conclude the series, this article focuses on the factors affecting the ESD performance of gloves and glove liners under realistic use conditions.

Experimental Variables and Procedures

When gloves are used in a cleanroom, the use of wrist straps and glove liners, among other variables, has to be included in the evaluation of the gloves' ESD performance. All of these components are part of a comprehensive ESD control system and must work together to ensure satisfactory performance. The study described in this article considered various glove materials and types of glove liners, and examined glove cleanliness and relative humidity (RH). The goal was to determine the effects of these variables on glove performance in real-world conditions. Discharge tests were chosen as the means to compare performance levels. As discussed in the first article in this series, discharge time tests are useful because they reflect the expected performance of materials in their intended application and are based on generally accepted test methods, including Method 404 in Federal Standard 101C.¹⁴

Gloves. Five different gloves considered suitable for use in cleanrooms were included in the study. To determine whether glove chlorination affects discharge performance, three different types of nitrile gloves were evaluated: unchlorinated ones, those chlorinated on the inside only, and those chlorinated inside and out. The other two gloves tested were made of PVC and natural rubber latex. The three sets of nitrile gloves were provided by the Ansell Occupational Healthcare critical environment development group (Coshocton, OH). The PVC gloves were from Oak Technical Products (Stow, OH), and the natural latex glove was the CR100 from VWR Scientific (Westchester, PA). The gloves were evaluated in three different conditions: directly out of the original package, after a deionized (DI) water wash, and after contamination using magnesium silicate as a model contaminant. Because these last gloves were dusted off prior to testing, they were not visibly contaminated.

A charged-plate monitoring system was used to measure discharge times from applied voltages to target voltages.

Glove Liners. Four different glove/glove liner combinations representative of those commonly used in high-technology manufacturing were tested. These combinations were: gloves without glove liners; gloves with a full-finger, insulative glove liner from Berkshire (Great Barrington, MA); gloves with a half-finger, insulative glove liner (also from Berkshire); and gloves with a full-finger, X-static brand static-dissipative glove liner from Ansell Golden Needles.

A technician performs discharge time tests in accordance with Method 404 in Federal Standard 101C.

Test Protocol. All of the gloves were tested at 23° ± 2°C (72° ± 3°F) and 50 ± 5% or 12 ± 3% RH after being properly conditioned for a minimum of 48 hours. Tests were run alternately with the tester wearing a wrist strap and not wearing a wrist strap. The tester stood on a Teflon insulative sheet to ensure that discharge occurred only through the charge monitor or the wrist strap. When no wrist strap was worn, discharge was solely through the charge monitor. A 20-pF charged-plate and Series 5000 monitoring system from NOVX (San Jose) was used to measure discharge times from applied voltages of ±1000 V to targets of ±100, 50, 20, and 10 V. These times were recorded using NOVX data acquisition software. Because the monitor has a 100-G resistance in the path to ground through its electrometer, even an insulative material such as natural latex should discharge to some voltage relatively rapidly.

A minimum of three samples of each type of glove were tested twice under each test condition. The procedure was as follows:

• The tester put on the appropriate glove or glove and liner combination, either with or without a wrist strap.
• Standing on an insulative sheet, the tester placed a gloved hand onto the charged plate, applying normal pressure.
• The 20-pF charged plate and the tester were charged to >1200 V.
• As the charge on the tester's body was discharged via the wrist strap or, when no wrist strap was worn, through the monitoring system, the times required to discharge from ±1000 to ±100, 50, 20, and 10 V were recorded.

Statistical Analysis Methods. The primary objective of the study was to determine which variables had a dominant influence on discharge time. Thus, it was possible to combine such factors as type of glove material and similar glove conditions (straight out of the package versus washed versus subsequently recontaminated) into various groups for statistical analysis. Combining variables permitted a clear interpretation of the test data.

Results

Wearing vs. Not Wearing a Wrist Strap. One statistical grouping included results for all five types of gloves tested fresh out of the original packaging at 50% RH with the tester wearing a groundable wrist strap and not wearing a wrist strap. These discharge time data, which are averaged over all glove-liner conditions, are summarized in Table I.

In all cases, not wearing a wrist strap interfered with the discharge time performance of the glove being tested. In addition, the natural latex gloves were not capable of discharging to 50 V within 10,000 milliseconds (10 seconds), even when the tester was wearing a wrist strap. Because these data indicate that wearing a wrist strap is mandatory when manufacturing products with extreme ESD sensitivity—such as magnetoresistive heads, which require a rapid discharge to <50 V—the remainder of this article only discusses data from tests in which a wrist strap was worn.

Relative Humidity. When the data for gloves that were tested fresh out of the original packaging at RH levels of 12 and 50% were analyzed, it was found that the relative humidity in the test environment had no effect on discharge times for the static-dissipative PVC glove. However, its effect could be measured for the three types of nitrile gloves. These data are shown in Table II.

For the nitrile gloves, discharge times were generally higher at 12% RH, especially for discharging to 20 V. The effect was most pronounced for the unchlorinated nitrile gloves and least noticeable for the gloves that were chlorinated inside and out. The PVC gloves also exhibited consistently higher discharge times at 12% RH, but the differences were very small. Most importantly, all four glove types met even the most demanding discharge time requirement of 1000 to <10 V in under 500 milliseconds.

Glove Liners. Analysis of the test results when the four static-dissipative gloves were worn in combination with the four glove-liner conditions yielded the discharge time data, averaged over all glove types and conditions, shown in Table III. For all of the liners tested the discharge time increased as the target discharge voltage decreased. However, the average discharge time for each glove/glove-liner combination was never more than 300 milliseconds, so all of the combinations met the most demanding disk-drive manufacturer's requirement of 1000 to <10 V in under 500 milliseconds.

It is not surprising that the fastest discharge times were achieved when the tester did not wear a liner. Human skin resistance is typically modeled as a 1500- resistor, which is well below the 10⁶- lower limit for dissipative materials. The discharge times with the X-static liner were nearly as fast as those without liners except for the discharges to ±10 V in a low-RH environment. In testing at 50% RH, the half-finger, insulative glove liner performed well for discharges to ±20 V, but the amount of time it required to reach ±10 V was significantly higher. At the low RH level, this liner's discharge times to ±10 V were particularly adversely affected, although not as much as those of the X-static glove liner. The full-finger, insulative glove liner performed the worst of all at the 50% RH level, but, interestingly, showed little sensitivity to the relative humidity of the test environment.

Hydration Time. During the 50% RH glove-liner tests, it was observed that the performance of the full-finger, insulative liner improved rapidly with time. The timing of the effect was not measured, but it was assumed to be the result of the hydration of the liner material by perspiration, since tests were conducted by a wearer who perspires relatively heavily. Discharge time improvements for persons who do not sweat heavily may be greater than those reported here.

To test the hypothesis that the time to achieve acceptable discharge performance may be significantly affected by the relative humidity in the test environment and by the degree that a person wearing gloves and liners perspires from the palms, more-careful observations were made during the testing at 12% RH, where the effect of glove liner hydration should be amplified over testing at 50% RH. In these tests, the time for the glove and liner combination to reach a stable discharge time was measured and found to be about 5 minutes. Table IV, which includes data for two types of nitrile gloves and two different glove liners, presents the comparable results of tests performed immediately after the tester put on a glove and liner, and again after 5 minutes. (In contrast, all data reported in Table III are for fully equilibrated liners—i.e., stable readings.)

The discharge times of both the X-static and the insulative full-finger glove liner were affected by wear time at 12% RH, especially for discharges to ±10 V. However, in no case was the discharge time slower than that required by the most demanding specification in the industry.

Nitrile Chlorination. When data for all liner and glove condition combinations were averaged for the three nitrile gloves to determine the effect of chlorination, it was found that the more a glove is chlorinated, the better its discharge time performance is. The effect, which is small but measurable, is most prominent for discharges to ±10 V. These results are summarized in Table V.

Data for all glove conditions were also averaged for the various combinations of gloves and liners, yielding the results summarized in Table VI. These data show that while chlorinating nitrile gloves affects discharge time, the effect is inconsistent. Compared to unchlorinated gloves, the nitrile gloves that were chlorinated on both sides exhibited an improved discharge time when worn with any of the glove liners. These results also suggest that wearing a full-finger glove liner increases discharge times. However, in no case was the discharge time greater than 500 milliseconds.

Glove Condition. All four static-dissipative gloves were tested under three conditions: directly out of the original package, after washing in DI water (followed by towel drying), and after light contamination with magnesium silicate. These results were averaged for all glove types, yielding the data listed in Table VII. These results indicate that glove washing provides a slight improvement in discharge time, particularly for discharging to ±10 V. Also, contamination on the surface of the glove by an insulative contaminant, even though not visible, adversely affects discharge times.

Conclusion

Measuring the discharge times of nitrile, PVC, and natural latex cleanroom gloves under various conditions provided a wealth of information that should prove helpful to high-technology manufacturers. Statistical analyses of the data demonstrate that not wearing a wrist strap causes all gloves to fail even the most generous discharge time requirement. Low relative humidity increases the discharge time for nitrile static-dissipative gloves but has relatively little effect on PVC gloves. The use of a glove liner also tends to increase discharge time. With a full-finger glove liner, this effect is time dependent; the increase in discharge time decreases with time, probably as a result of glove-liner hydration from hand perspiration.

A nitrile glove chlorinated on both sides discharges more rapidly than a nitrile glove that is chlorinated only on the inside, which in turn discharges more rapidly than an unchlorinated nitrile glove. Glove liner performance does not seem to be significantly affected by choice of glove material, since discharge times for all the gloves tested met even the most demanding manufacturing requirements. Finally, washing improves the discharge time performance of gloves slightly, while surface contamination degrades their discharge time performance significantly.

References

1. RC Walker, "Implementing an ESD Control Program," Microcontamination 1, no. 2 (1983): 20­24.
2. GE Hansel, "The Role of the Production Operator in Preventing Damage," Microcontamination 2, no. 4 (1984): 43­46.
3. SC Heymann et al., "Voltage-Detection Systems Help Battle ESD," EE—Evaluation Engineering (November 1997): S-6­S-12.
4. JC Hoigaard, "ESD Test Equipment and Workstation Monitors," EE—Evaluation Engineering (July 1998): 58­61.
5. M Banks, "Watch Those Electrons, ESD Battle Heats Up," Data Storage (July 1998): 61­62.
6. SL Thompson, "All About ESD Plastics," EE—Evaluation Engineering (July 1998): 62­65.
7. E Greig et al., "Controlling Static Charge in Photolithography Areas," MICRO 13, no. 5 (1995): 33­38.
8. A Steinman, "How to Select Ionization Systems," EE—Evaluation Engineering (June 1998): 62­69.
9. BI Rupe, "Electrical Properties of Synthetic Garments with Interwoven Networks of Conductive Filaments," Microcontamination 3, no. 5 (1985): 24­28.
10. RJ Pierce and J Shah, "Potential ESD Hazards from Using Adhesive Tapes," EE—Evaluation Engineering (November 1996): S-30­S-31.
11. RW Welker, "A Comprehensive ESD Control Program for MR Heads" (paper presented at the Asia Pacific Magnetic Recording Conference, Singapore, July 29­31, 1998).
12. RW Welker and PG Lehman, "Using Contamination and ESD Tests to Qualify and Certify Cleanroom Gloves," MICRO 17, no. 5 (1999): 47­51.
13. RW Welker, "Controlling Particle Transfer Caused by Cleanroom Gloves," MICRO 17, no. 8 (1999): 61­65.
14. Federal Standard 101C, Method 404 (Washington, DC: Government Services Administration).

Roger W. Welker is a senior ESD control engineer at Jet Propulsion Lab, Pasadena, CA. He is founder and principal scientist of R.W. Welker Associates, a consulting firm specializing in contamination and electrostatic discharge control. He has 17 years experience in high-technology development and manufacturing at IBM, Seagate, and Micropolis. He also spent 11 years in applied R&D, focusing mainly on applications of fine particles. Welker has authored or coauthored more than 60 papers and is a member of the IEST, the American Association for Aerosol Reasearch, the ESD Association, and the Data Storage Institute. He received his BS in chemistry from the University of Maryland in College Park. (Welker can be reached at 818/368-0557 or rwwlws@aol.com.)

Arleigh Hartkopf, PhD, is director of analytical and technical services for Ansell Occupational Healthcare (Coshocton, OH). Before joining Ansell in 1996, he worked for 18 years as an analytical chemist for Mobil Chemical in its R&D department. While at Mobil he was involved at various times in most of their original broad line of products. Hartkopf earned a BS in chemistry at Harvey Mudd College (Claremont, CA) and an MS in inorganic chemistry and a PhD in analytical chemistry from Northeastern University, Boston. (Hartkopf can be reached at 800/800-0444 or ahartkopf@ansell.com.au.)

Peter G. Lehman, PhD, recently retired as vice president, R&D, of Ansell Occupational Healthcare. Before joining Ansell in 1997, he worked for 10 years in the pharmaceutical industry in Melbourne, Australia, where he managed the manufacture of pharmaceutical active raw materials and designed and built cleanrooms. He has a PhD from the University of East Anglia (Norwich, UK) and did postdoctoral work at the University of Groningen, the Netherlands, and the research school of chemistry at the Australian National University in Canberra.

Steven B. Heymann is the CEO of Novx (San Jose). Before cofounding the company, he held several senior management positions in both start-up and established high-technology companies. He has more than 20 years of experience in various segments of the electronics industry and successfully cofounded two companies that manufactured or sold products for that industry. Heymann is coauthor of Novx's first patent on its ESD monitoring technology as well as subsequent patents. He earned a BS in information technology from California Polytechnic University (San Luis Obispo, CA).

Carl Newberg is president and owner of River's Edge Technical Service (Rochester, MN), an independent testing laboratory and consulting service to the ESD and contamination control industries. He was ESD program manager for Western Digital and was involved in the corporate ESD program at Seagate and IBM. He is the cochairman of the ESD subcommittee of IDEMA and a member of the ESD Association. He has a BS in metallurgical engineering and an MS in materials science and engineering from the University of Arizona (Tucson), as well as a professional engineer's license in the state of Arizona.

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