Zero tolerance for ESD
09/01/2006
Ionizers, remote sensors, monitoring, and material choices all contribute to control
By Sarah Fister Gale
For a lot of cleanroom operators, any static is too much static. Disk drive manufacturers are leading the downward spiral toward zero tolerance for electrostatic-related events in the cleanroom, with semiconductor, flat panel, medical electronics, and pharmaceutical manufacturers close on their heals.
As with all things manufactured in a cleanroom, shrinking geometries are leading to greater sensitivities, and in the case of electrostatic charge, that means even a few volts of static electricity can result in serious impacts on yield.
Most disk drive makers have tolerances of two volts or less in the cleanroom-down from five volts just a year ago, says Steve Heymann, CEO of Novx Corporation (San Jose, Calif.), supplier of instrumentation and software for universal monitoring and control solutions in cleanrooms, manufacturing, and ESD-sensitive work areas. “It takes a lot of experience to create a cleanroom environment with control to that level.”
To help manage this static-free environment, the disk drive industry follows ESD standards, produced by the International Disk Drive Equipment & Materials Association (IDEMA; Santa Clara, Calif.), that standardize procedures for testing and set the correlation and benchmarking of products. The IDEMA documents include a packaging standard for disk drives and components that includes materials requirements for the various stages of disk drive manufacture, and a human body model (HBM) document that addresses the issue of testing magnetoresistive (MR) head products for qualification prior to shipping.
AC/DC ionization
Meeting a one- or two-volt sensitivity limit requires a complex ionization program that begins with a room system ionizer. Ionizers in the cleanroom are the first defense against static and, as sensitivity levels drop, ionization equipment is getting more attention.
Room system ionizers output clusters of airborne molecules that are bound by polarization forces to a charged (typically single) nitrogen or oxygen molecule, increasing the conductivity of the air with the charged gas molecules. When ionized air comes in contact with a charged surface, the surface attracts ions of the opposite polarity. As a result, the static electricity is neutralized.
There are three common ionization methods used in commercial air ionizers: corona, alpha and photoelectric. Each method generates air ions, but corona ionization is the most commonly used method in cleanrooms.
AC and DC high voltage is used to generate corona ionization. High voltage is applied to a sharp emitter point (see Fig. 1) or a small-diameter emitter wire, resulting in an electric field around the emitter. This high-voltage field interacts with the electrons in the nearby gas molecules, resulting in positive or negative ions, depending on the type of high voltage that is applied.
Unlike AC corona ionization, which emits both positive and negative ions from the same emitter point, DC corona ionizers emit ions from separate positive and negative emitter points, creating less recombination of ions.
“Since ions are emitted independently for each polarity, it’s possible to monitor and control the amount and the equality of ions emitting from the positive and negative emitter points,” says Arnold Steinman, a member of the board of directors for the ESD Association and chief applied technologist at MKS, Ion Systems (Alameda, Calif.), provider of process control solutions for advanced manufacturing processes. “Various types of monitoring and control systems are available for DC corona ionizers. As a result, ionizers using DC corona offer great degrees of control and fine-tuning. The high level of system sophistication along with the low ion recombination rate also means that DC corona ionizers are appropriate for ESD-sensitive and contamination-critical technology applications.”
There are two types of DC corona ionization: pulsed DC and steady-state DC. Steady-state DC continuously applies positive high voltage to half of the emitter points and negative high voltage to the other half. The ionizer may contain a single pair of emitters (as in blow-off guns or nozzles, and in ceiling emitters), an array of pairs of emitters (DC blowers), or a straight line of emitter pairs (DC bars). Steady-state DC ionization may be employed with low or high airflow, depending on how far apart the emitter points are spaced. Steady-state DC ionization is commonly used in cleanroom systems, laminar flow hoods, blowers, and blow-off guns.
Pulsed DC ionizers allow positive and negative emitter points to be turned on and off alternately, creating clouds of positive and negative ions. Ionizers using pulsed DC ionization may be finely tuned to allow timing cycles and polarities to operate precisely for a specific application. Positive and negative emitters may be set to alternate in time periods of seconds. In certain areas, a greater proportion of one polarity may be needed over the other, or the time that the voltages are on may be stretched to prevent any recombination.
Local ionization requires balance
For many cleanrooms, a generic room ionizer is just the beginning of a complete ionization system. “Cleanroom operators used to assume that if they had an ionizer, everything was okay with regard to ESD, but that’s not true anymore,” Heymann says. “The problem with ESD technology is that simply installing it doesn’t mean it’s still working a day, a week, or a month later.”
Now, lower sensitivity thresholds are forcing cleanroom operators to add to their general room ionization systems, which can knock the atmosphere from 1,000 to 100 volts. Incorporating additional ionizers inside process tools and around critical areas of the environment can get voltages down under 10 volts. Workstation ionizers, including bar ionizers, blowers, and compressed-gas blow-off devices, are used in these defined work areas (see Fig. 2). Whether the ionizer is mounted above or directly on the work surface, the distance separating it from the ESD-sensitive product is much smaller than with room systems-typically less than 1 meter.
For extremely critical environments, ionizing blowers are used in place of pulsed DC systems, which normally have a voltage swing that can’t be adjusted to zero (see Fig. 3). The problem, however, is that although the blower eliminates static, it interferes with other elements of the clean environment, Steinman explains. “Blowers affect air turbulence and can become a source of contamination,” he says. “A collaboration has to occur between contamination control and ESD management to be sure you are meeting all required levels.”
To prevent the negative effects of blowers, Ion Systems uses emitter points made from its patented single-crystal silicon material that is “on the edge of meeting ISO Class 1 standards,” Steinman says. “Other materials for emitter points are just not as clean.”
Early warning systems
Adding to the challenge of controlling ESD is the fact that, in many cases, manufacturers don’t realize an event has caused damage until late in the production process. The failure may not show up until a disk drive is assembled, or it could be a latent failure in which ESD weakens or wounds the component but not enough to cause a malfunction during testing. Over time, however, the wounded component will cause poor system performance and eventually complete system failure. Because latent failures occur after final inspection or in the hands of customers, the cost for repair is very high.
To combat this problem, cleanroom operators are using remote sensors to continually monitor the atmosphere and determine the number of positive or negative ions required to neutralize the air, says Carl Newberg, general chairman of the ESD Association Symposium and president of Rivers Edge Technical Services, a materials testing firm in Rochester, Minnesota. “This is important because without sensors an ionizer can create a balance offset that makes matters worse.”
There has also been a growing trend toward independent monitoring in the process areas around key workstations to make sure the ionization process is working. These independent monitors gather and collect ESD data, giving operators ongoing reports about the cleanroom environment that detail any spikes or trends in ESD events during critical points in the manufacturing process.
This allows them to address problems in the cleanroom before they cause serious impacts to yield. For example, the resulting reports may show spikes during certain times of the day correlating to changes in personnel, placement or movement of tools or materials, or faults in grounding of equipment. By tracking the events and their locations, operators can identify where the problems are and what’s causing them.
“When sensitivity levels get that critical, you need to put these sensors directly in front of your ionizers in key target areas and use the feedback to maintain balance,” says Jim Curtis, business unit manager for the electronics manufacturing group of ITW/Simco (Hatfield, Pa.), a manufacturer of cleanroom ionization systems. “They are constantly striving for zero.”
But documenting events and trends isn’t enough to protect yields from ESD damage. Some monitors may only have a red or green light to show they are operating. Although data about events may be sent to a central database, if the sensors don’t actively alert operators when an issue arises, days or weeks could go by before it’s noticed, Heymann says. “If you’re making three drives per minute, how many will be made before you realize there’s a problem?”
Such questions have resulted in the implementation of proactive sensors that respond with an alarm the instant an event occurs, such as an ionizer needing maintenance or equipment becoming ungrounded. In more advanced systems, the monitors have a direct tool interface enabling them to shut down a tool or the entire production line when static levels rise out of specification, giving operators a chance to solve the problem and pull products off the line for evaluation before continuing. “Now, if the product is dead you can toss it before you put it through fifty more process steps,” Heymann says. “With this kind of proactive control, you don’t produce a single bad drive, panel or IC.”
Although it may seem extreme and costly, the payoff can be huge, says Curtis. He’s seen several clients realize dramatic increases in yield as the result of improving their ionization programs, including one RFID manufacturer that went from 60 to 90 percent yield after installing and tweaking a new ionization system. “The first guy in charge of the operation didn’t see the value of ionization,” he says. “They had ionization, but it wasn’t installed properly and it wasn’t in the right places.”
Another client, who debated investing in a $100,000 ionization system for two years because of the price, ultimately agreed the system paid for itself in one day, Curtis says. “Ionization systems are underappreciated, but you can’t argue with real-world results.”
S20.20 guides industry
For semiconductor manufacturers, the ESD crisis hasn’t yet reached critical urgency. Depending on the operation, they can tolerate 30 to 200 volts without damage; however, semiconductor manufacturers are already facing many of the same problems disk drive makers faced just a few years ago.
“The semiconductor industry invented ESD,” Newberg says. “When things started blowing up in the 70s, they quickly figured that ESD was causing it and learned how to put ESD protection on the chips.”
However, in the last few years semi manufacturers have whittled that protection away, giving up safety for speed and increased IC functionality. “It’s dropped their sensitivity levels dramatically.”
Novx’s Heymann agrees. “Semiconductors are trending in the same direction as the disk drive manufacturers. They dropped from 3,000 or 4,000 volts to 300 volts or less almost overnight because they gave up the oxide thickness that protected wafers against ESD.”
Heymann also points out that concern about ESD is disparate between the front and back ends of the semiconductor manufacturing process. “On the back end, where they cut the chips, they know they need to watch out for ESD, but there’s still a misconception that wafers are not ESD-sensitive at the front end of the process.”
Like the disk drive industry, semiconductor manufacturers follow standards for managing ESD. The ANSI/ESD S20.20-1999 standard, Development of an Electrostatic Discharge Control Program, covers the requirements necessary to design, establish, implement and maintain an ESD-control program to protect electrical or electronic parts, assemblies and equipment susceptible to ESD damage from HBM discharges greater than or equal to 100 volts. The areas covered by this standard will expand and overlap as industries progress along their technology roadmaps, and an update is expected next year.
ANSI/ESD S20.20 is a guidance document created using the industry’s experience with the ISO 9000 and ISO 14000 standard, says Steinman. The guidance is gaining worldwide acceptance as an international standard and is currently going through translation and draft status at the International Electrotechnical Commission (IEC).
It’s useful as a global standard in part because it was designed to work within ISO parameters. “If you have an ISO quality-control program, you already have the infrastructure in place for S20.20. You just have to add the technical elements,” Steinman says. “ISO certifiers will be able to certify to the S20.20 standard, as well.”
The S20.20 document is already available in English, Chinese and Spanish, and is being rapidly adopted by organizations across the globe. “When things are standardized, people don’t have to make decisions about whether they should do it or not,” Steinman says. “It becomes something that everyone has to do and they understand that. It’s a way for everyone in the electronics food chain to do the right thing to make sure they all use the right level of ESD protection.”
The ESD Association currently offers an ESD program documentation review service, through which the ESD Association’s Facility Certification committee will review a manufacturer’s ESD program documentation and compare it to the requirements listed in ANSI/ESD S20.20-1999. A report is provided that describes the areas that need to be improved in order for documentation to be compliant with ANSI/ESD S20.20-1999.
“This service should be considered a must for any company that is preparing for facility certification based on ANSI/ESD S20.20-1999,” Newberg says. “Getting this certification enables companies to prove to their clients that they are meeting industry standards.” For some groups, such as the military and NASA, S20.20 certification is already a requirement of doing business.
Material choices
Along with ionization and traditional grounding tools for equipment and personnel, manufacturers are also taking a hard look at the materials being used in the cleanroom and how they relate to ESD.
Electrostatic charges on materials are the result of a transfer of electrons caused by the sliding, rubbing, or separating of a material, which is a prime generator of electrostatic voltages. Plastics, fiberglass, rubber and textiles can all collect these charges. When this happens to an insulating material, which does not allow for the flow of electrons across or through its bulk, the built-up charge tends to remain in the localized area of contact. This electrostatic charge on the insulator can induce a charge on nearby conductors, such as a person or a microcircuit, which can then discharge via an arc or spark when the conductor comes in contact with a body at a sufficiently different potential.
“Disk drive manufacturers are reducing the potential for ESD events with the use of advanced polymers that can replace metal carriers with static-dissipative thermoformed or CNC ceramic/ICP parts,” says Bob Vermillion, president of RMV Technology Group (Clayton, Calif.), an ESD solution provider. “These polymers reduce rapid discharge by exhibiting surface resistance, volume resistance or two-point resistance readings.”
When evaluating ESD protective materials, the most common property to test is surface resistivity or surface resistance of a material, since that provides a measure of how well the material dissipates electrical charge in contact with its surface.
Cleanroom operators and tool manufacturers are implementing more ESD-resistant materials to meet client needs. Conductive materials, such as shielded bags, foils, and metal, are popular because they have low electrical resistance-generally less than 1 x 105 ohms/sq (surface resistivity) and 1 x 104 ohm-cm (volume resistivity)-and allow electrons to flow easily across their surfaces or through their volumes. When a conductive material becomes charged, the charge (i.e., the deficiency or excess of electrons) is uniformly distributed across the surface of the material. If the charged conductive material makes contact with another conductive material, the electrons will easily transfer between the materials. If the second conductor is grounded, the electrons will flow to ground and the excess charge on the conductor will be neutralized. Conductive materials are usually carbon-particle- or carbon-fiber-filled throughout.
Static dissipative materials used to prevent electrostatic discharge to and from humans generally have a resistivity between 106 and 109 ohms/sq. They have an electrical resistance between insulative materials and conductive materials. There can be electron flow across or through the dissipative material, but it is controlled by the surface resistance or volume resistance of the material.
For static-dissipative materials, like all materials, charge can be generated triboelectrically. However, like the conductive material, the static-dissipative material will allow the transfer of charge to ground or other conductive objects. The transfer of charge from a static-dissipative material will generally take longer than from a conductive material of equivalent size. Slowing the charge transfer is one way to prevent ESD damage. Charge transfers from static-dissipative materials are significantly faster than from insulators and slower than from conductors.
There are many kinds of antistatic or static-dissipative materials, but not all can be used in a cleanroom to reduce the risk of ESD because there may be contamination issues. For example, plastics that are surface-coated with quaternary ammonium salts, amidoamines, or salts of octanoic acid to impart nonpermanent ESD properties cannot be used in the cleanroom. ESD-control materials for the cleanroom must pass the same tests for particle generation, outgassing, and the presence of chemical residues that are applied to all materials used in the cleanroom.
The future
Looking ahead over the next couple of years, cleanroom operators should expect more of the same. The good news is that, since the disk drive industry has figured out how to control static electricity down almost to zero levels, there is a roadmap to follow.
And while semiconductor manufacturers are heading in the same direction, they are fortunate to have already automated many of their processes, removing humans-the biggest ESD risk factor in the cleanroom-from the equation. But there’s still much to be done, says Steinman. “For the semiconductor industry, there’s a wafer-level ESD problem coming that people need to think about,” he says. “The important thing to remember is that you don’t want to solve an ESD problem. That’s too expensive. It’s much more cost-efficient to prevent one from happening in the first place.”
ESA-A problem for bio/pharmaceuticals and medical devices as well
“ESD control in a cleanroom environment goes hand-in-hand with controlling electrostatic attraction (ESA) and the build-up of charge on wafers, disk drives, MR heads, microprocessor-driven devices or components,” says RMV’s Vermillion. ESA is the accelerated deposition of particles onto a surface due to the presence of an electric field created by excess electrical charge on a surface.
If surfaces are charged, ESA attracts and holds particles that would otherwise remain airborne in the cleanroom laminar airflow. Submicron-sized particles cause defects in semiconductor production in much the same way that dust on a photographic negative or print paper causes a visual defect.
“As technology changes lead to smaller feature sizes in semiconductor devices, the size of the killer particle also decreases,” says MKS’s Steinman. “Smaller particles are more easily attracted-and more difficult to remove-because of static charge on surfaces, and once they’re bonded to a charged surface, it’s very difficult to remove the contamination.”
ESA has also become a problem for pharmaceutical and medical device manufacturers who face biological contaminants in the cleanroom. Even though the allowable particles in a pharmaceutical environment are much bigger than in a semiconductor manufacturing facility, the stakes are much higher when a particle becomes charged and attracts to critical materials that could affect anything from the sterility of a tool to the viability of a pacemaker or stent, explains River’s Edge’s Newberg. “You can’t afford any failures there.”
An ionization system releases the charge on these particles so they remain in the airflow and can be swept back into the air filter systems.
Newberg has seen more and more pharmaceutical and medical device manufacturers looking into ionization systems as they come to understand the damage that ESA can cause.