RequestLink
MICRO
Advertiser and
Product
Information

Buyer's Guide
Buyers Guide

tom
Chip Shots blog

Greatest Hits of 2005
Greatest Hits of 2005

Featured Series
Featured Series


Web Sightings

Media Kit

Comments? Suggestions? Send us your feedback.

 

MicroMagazine.com

Critical Materials

Understanding the new SEMI standards for stainless-steel welding applications

Sunniva Collins and Ed Wolf, Swagelok; and Kevin Nguyen, SEMI

The SEMI International Standards Program, a key service offered by SEMI, provides the framework for the production of timely and technically accurate specifications and other standards documents of importance to the semiconductor industry. The past two years have seen several changes in those standards pertaining to the composition, surface finish and quality, corrosion resistance, and welding of stainless steel and other corrosion-resistant alloys for gas- and liquid-distribution components and systems. Each of the following standards has undergone significant modification or has been newly created to replace older standards:

  • SEMI F20-0704, Specification for 316L Stainless Steel Bar, Forgings, Extruded Shapes, Plate, and Tubing for Components Used in General Purpose, High Purity and Ultra-High Purity Semiconductor Manufacturing Applications. This standard outlines metallurgical cleanliness and material composition requirements.
  • SEMI F19-0304, Specification for the Surface Condition of the Wetted Surfaces of Stainless Steel Components. This standard covers surface quality requirements and relevant test methods, including tests to determine the physical morphology and chemical qualities of a wetted surface.
  • SEMI F77-0703, Test Method for Electrochemical Critical Pitting Temperature Testing of Alloy Surfaces Used in Corrosive Gas Systems. This standard provides a new technique for ascertaining the relative resistance of wetted surfaces to pitting.
  • SEMI F78-0304, Practice for Gas Tungsten Arc (GTA) Welding of Fluid Distribution Systems in Semiconductor Manufacturing Applications. This standard outlines accepted procedures for welding stainless steel and other corrosion-resistant metals and alloys.
  • SEMI F81-1103, Specification for the Visual Inspection and Acceptance of Gas Tungsten Arc (GTA) Welds in Fluid Distribution Systems in Semiconductor Manufacturing Applications. This standard provides visual inspection and acceptance criteria for welds of stainless steel and other corrosion-resistant metals and alloys.

This article focuses on three of the above standards: SEMI F78 and F81, which describe welding best practices and acceptance criteria, and SEMI F20, which contains basic information about material composition requirements for successful welds. Each of these standards was approved within the last year. SEMI F20 is a revised standard, while SEMI F78 and F81 replace SEMI F3, which was cursory and contained only the briefest instruction and guidelines on welding best practices and testing. Although all SEMI standards are voluntary, understanding and implementing these three will help prevent the corrosion and contamination problems associated with process fluid-distribution systems, an area of great concern to the semiconductor industry as linewidths continue to shrink and new processes are brought on-line.

Revised Standard for Distribution-System Materials

The recently approved revisions to SEMI F20 define the metallurgical cleanliness material composition and requirements for 316L stainless steel used in the manufacture of components for general-purpose, high-purity (HP), and ultra-high-purity (UHP) fluid (gas or liquid) distribution systems. The standard defines three grades of material, based on its potential use. General-purpose-grade materials must have the chemical composition specified in ASTM standards A269 and A632, with the exception of sulfur, which is limited to a maximum of 0.012% by weight. Additional composition requirements for the HP and UHP grades of 316L are presented in Table I. SEMI F20 also contains two important appendices describing the effects of trace and residual sulfur and copper on weldability.

Effects of Sulfur Content on Welding. Type 316L austenitic stainless steel is the preferred material for the components of systems that distribute the process fluids used in semiconductor fabrication. However, several properties affecting the manufacture and applications of 316L can vary significantly with sulfur content. Thus, although the ASTM standards permit sulfur content of up to 0.030%, SEMI F20 sets lower maximums of 0.012% for general-purpose 316L and 0.010% for HP and UHP grades of the material.

Table I: Chemical composition requirements for high-purity and ultra-high-purity grades of 316L stainless steel.

Sulfur has a very low solubility in austenitic stainless steels, where it exists as discrete inclusions of manganese sulfide. These inclusions, which increase in number and size as the sulfur content increases, initiate pits and other defects that decrease the corrosion resistance of the finished part. On the other hand, inclusions improve the material’s machinability. Typically, stainless steels that will be machined have sulfur compositions near the 0.030% maximum, since those with lower sulfur levels require lower feeds and speeds and may reduce tool life.

Figure 1: Convection currents in the weld pool during welding of stainless steel with sulfur levels of >0.005%.
Figure 2: Convection currents in the weld pool during welding of stainless steel with sulfur levels of <0.005%.
Figure 3: Asymmetric weld pool during welding of two stainless-steel pieces with significantly different sulfur levels.

Sulfur also strongly affects the welding of 316L stainless steel. A variation of sulfur content from very low levels to the maximum permitted can increase weld bead penetration by approximately a factor of two for similar weld parameters. When sulfur content is below approximately 0.005%, there are well-documented changes in weld pool dynamics, called the Marangoni effect. Convection currents flow outward, resulting in a wide, shallow weld pool (see Figure 1). At such low levels of sulfur, the heat input required for a full-penetration weld increases. Conversely, when sulfur contents are significantly above 0.005%, convection currents flow downward from the welding arc, resulting in a narrow, deep weld pool (see Figure 2).

In addition, serious problems can occur when welding two pieces of stainless steel with substantially different sulfur contents. The weld pool can become asymmetric, favoring the low-sulfur side; the root of the weld may shift away from the joint, as shown in Figure 3. This effect can be minimized by selecting parts with sulfur contents that vary within ±0.010 percentage points. Greater differences in sulfur contents require adjustments to weld setup to achieve full penetration to the root of the joint.

Effects of Copper Content on Welding. Copper is not an intentional addition to 316L steel, but is present as a result of the composition of the furnace charge materials used during melt, which are primarily scrap with elemental and master alloy additions. Copper content can only be reduced by dilution, and, over time, as more scrap material is recycled, the level of residual copper is likely to increase.

According to the Aerospace Materials Specification AMS 2248E, if a material’s copper content is not specified, its maximum level is 0.50%. However, SEMI F20 restricts copper content further, to 0.30%, because of its effects on weldability. In welding, copper is a surface-active element (like sulfur, oxygen, and selenium) and can contribute significantly to weld bead meander. It volatizes and redeposits downstream of the weld, providing a site for the initiation of corrosion under some conditions.
Although copper is an austenite phase stabilizer and has been shown to improve resistance to stress corrosion cracking in chloride solutions, other studies have shown that there is a correlation between a high copper content and weld bead meander, slag, and discoloration. Adding hydrogen to the argon shield gas also affects weldability adversely. The following research findings on copper content informed the revision to SEMI F20:

  • Because of copper-sulfur interaction, the effect of a material’s copper content on welding is more pronounced at sulfur levels <0.010%.
  • Oxidation of steels containing 0.31% copper and 0.010% sulfur is known to occur after welding in argon, but acceptable welds are possible with materials containing 0.30% copper and 0.011% sulfur.
  • Cosmetically unacceptable welds, with bead meander and copper redeposited in the heat-affected zone, have been observed in 316L steels containing 0.41% and 0.47% copper.
  • Steels with more than 0.30% copper may be welded successfully in argon but may exhibit bead meander and electrode attack when hydrogen is added to the argon. Significant electrode damage has been known to occur when the shield gas was Ar/H2.

New Standards for Welding Procedures and Inspections

The new standards SEMI F78 and F81 are the result of a long, collaborative process of surveying and comparing welding and weld-testing procedures from many companies. After the GTA Welding Task Force had evaluated these written documents, the best practices and test methods became the SEMI standards. Thus, the two standards represent an industry-wide consensus about the specifications that define an acceptable weld and the procedures necessary to produce consistent, optimum welds that will prevent contamination and ensure weld strength and integrity.

At its first meeting, the task force decided not to revise SEMI F3 (Guide for Welding Stainless Steel Tubing for Semiconductor Manufacturing Applications) but to replace it with the two new standards that would address two distinct audiences: welders and welding equipment operators, and quality assurance and quality control inspectors. The task force recognized that these two audiences must share a common vocabulary in order to communicate effectively with each other. Each also must understand the other’s expectations. The two standards, therefore, were organized in a similar fashion with correlating sections relating to purpose, scope, limitations, and terminology. SEMI F81 specifically references SEMI F78 as the baseline for evaluating all welds by stating that “Welds shall be produced in conformance with the procedures and requirements outlined in SEMI F78.”

SEMI F78 and F81 contain significantly more information and direction for welders and weld inspectors than the previous standard. While SEMI F3 was only about three pages long, the two new standards taken together are about 27 pages. In addition, SEMI F3 was a “guide,” which, according to SEMI, is “an option or instruction with options as choices,” but SEMI F78 and F81 carry greater authority. SEMI F78 is a “practice” (also known as a “definitive procedure”), and SEMI F81 is a “specification,” which “defines the requirements for a product or service.” In other words, the new standards do not just provide options, they also prescribe specific methods.

Purpose, Scope, Limitations, and Terminology. Each of the two new standards describes its purpose as relating to the welding of stainless steels and other corrosion-resistant metals and alloys for fluid (liquid or gas) distribution systems in semiconductor manufacturing applications. Each standard’s scope is confined to the GTA welding of autogenous butt joints in fluid-distribution-system components, including tubing, pipe, fittings, subassemblies, and other parts that contact process fluids. Safety and health procedures are not part of the two standards, although SEMI F78 does contain an appendix on welding fumes and other potential hazards that welders and weld equipment operators may face.

The two standards are concerned only with austenitic and superaustenitic grades of stainless steel and other corrosion-resistant metals and alloys with solid-solution grades of nickel or titanium alloys. The welding applications covered are limited to autogenous circumferential butt-joint GTA welds performed on components 6 in. or less in diameter, automatic GTA welding processes, and welds performed without fillers and fluxes.

SEMI F78 and F81 reference several external documents that served as the points of departure for many of the prescribed practices and procedures. Of particular note are documents from the American National Standards Institute (ANSI); the American Society of Mechanical Engineers (ASME), specifically its bioprocessing equipment standard; and the American Welding Society (AWS). Finally, the standards define 63 terms used by welders, weld equipment operators, and weld inspectors. The definitions cover both standard terms, such as those referenced in ANSI/
AWS A3.0 and the ASME bioprocessing equipment standard, and nonstandard terms commonly used in the field.

The standard and nonstandard terms are cross-referenced for clarity. For example, halo, haze, and heat tint/color are each defined as nonstandard terms for “discoloration resulting from a welding procedure,” while discoloration is a standard term for any change in color from that of the base metal. According to the standards, “[discoloration] is usually associated with oxidation occurring on the weld and heat-affected zone on the outside and inside of the weld joint as a result of heating the metal during welding. Colors may range from pale bluish-gray to deep blue and from pale straw color to a black crusty coating.”

Welding Requirements. SEMI F78 describes all apparatus, equipment, and purge-gas requirements for welding fluid- distribution-system components, including those relating to fixtures, weld heads, and electrodes. According to the standard, all seamless austenitic stainless-steel tubing is to conform with SEMI F20 or customer specifications.

SEMI F78 also delineates a process for making and evaluating test specimens, or coupons. Before a weld of a new size or wall thickness or one containing a new alloy is used in production welding, a primary standard sample weld must be made, sectioned, and analyzed on the job site. Once the sample is found to be acceptable, all the essential variables of the welding process must be documented in the procedure qualification record. That sample then becomes the standard against which other welds of the same size, wall thickness, and alloy will be judged. Any significant deviation from the sample should result in a rejection of the noncompliant weld.

Following the procedure qualification, coupons should be made at the start and end of all shifts, whenever a discrepancy is discovered by the operator, or when any of the following events occur:

  • A change in weld parameters.
  • A change in material (heat number).
  • A change in tube size or wall thickness.
  • A change in equipment, such as the weld head or weld-head extensions.
  • A change in ambient temperature of ±20°F.
  • A change in power source or the addition or removal of extension cords.
  • A significant change in the purge gas source or flow rate.

SEMI F78 also details procedures for joint preparation and tube cleaning and purging (including suggested purge settings based on the tube size and wall thickness). Specific protocols are identified for cleanroom welding and field installation, and postweld inspection criteria and procedures are described. Such procedures are to be performed on all welds by the welder or weld equipment operator. In addition, the standard mandates daily logs containing weld results. Figure 4 provides a summary overview of the process that should be followed to ensure consistent, acceptable welds.
Weld Inspection Requirements. SEMI F81 is concerned with the tools and methods of visual weld inspection. A key point made in the standard is that a weld that appears acceptable under one magnification may be deemed unacceptable under a higher one. Thus, acceptable tools, magnification levels, and illumination techniques should be specified and agreed upon by the welded component supplier and purchaser. The standard cites sight pipes, rigid borescopes, calipers, v-blocks, dial indicators, and comparators as examples of inspection tools. Magnifying glasses and optical microscopes (2x to 40x), flashlights, bright fluorescent lights, natural (ambient) light, and white paper (for background illumination) are named as magnification or illumination methods.

Figure 5: Examples of rejectable defects in cross-sectioned butt welds, depicted with the inside curvature of the tube facing the viewer: (a) an undercut, (b) center-line shrinkage, and (c) an ID concavity. In all three cases, the weld itself provides a wall thickness that is less than the nominal thickness of the tube.

Section 7 of the standard provides criteria for the evaluation of welds. Specifically, all welds should exhibit complete penetration around the entire internal weld surface, with uniform penetration and bead width. There should be no visible surface cracks, porosity, or inclusions when the weld is viewed under magnification. Examples of rejectable defects identified in the standard include:

• Minimum wall thickness Tmin below that of the parent metal at any point in the weld.
• An undercut or center-line shrinkage (shown in Figures 5a and 5b, respectively).
• ID concavity (shown in Figure 5c).
• OD concavity that exceeds 10% of the nominal tube wall thickness T on ≥1-in. (25-mm) tubing.
• OD concavity of any size on <1-in. (25-mm) tubing.
• OD convexity that exceeds 10% of the nominal wall thickness T.
• ID convexity that exceeds 10% of the nominal wall thickness T.
• Minimum ID weld-bead width that is less than the nominal wall thickness T. (The maximum ID bead width should be 2.5 times the nominal wall thickness T, but in any individual weld, the maximum ID bead width should not be ≤1.25 times the minimum bead width.)
• ID and OD weld bead meander that exceeds 35% of the nominal wall thickness T.
• Porosity, inclusions, or slag on the ID weld root surface when viewed without magnification. (A small slag inclusion may be acceptable if it is at the end of the downslope, has a diameter that is less than 10% of the nominal wall thickness T, and does not affect weld integrity.)
• OD weld face width that is not a minimum of two times the nominal wall thickness T.
• Weld bead overlap that is less than 80% on the OD and 70% on the ID along the complete length, except the downslope.
• Tack welds that are not totally consumed and undetectable in or around the OD or ID weld bead.
• Visible discoloration on the tube or weld ID when viewed under a bright fluorescent light without magnification (HP and UHP systems only); for welds greater than 2 in. in diameter, a slight blue color may be acceptable. Oxidation on the OD weld face (except a light straw color) is permissible. It should be removed with a stainless-steel wire brush immediately after welding, unless prohibited by the end-user.
• Axial misalignment that exceeds 10% of the nominal wall thickness T.
• Angular misalignment that exceeds ±1⁄2° (1⁄8 in./ft).
• Weld downslope not present or not of sufficient length to prevent a crater at the end of the weld. The distance between the ID downslope and the OD downslope should be a minimum of three times the nominal wall thickness T.

Conclusion

SEMI F78 and F81, together with the revisions to F20, represent important steps in the evolution of the semiconductor industry. In the past, different welding practices appeared in various SEMI and other industry standards and in company procedures. It was even not unusual for a company to define welding practices and criteria differently at different facilities. Now, SEMI F78 and F81 bring together all required welding and weld-testing procedures for fluid-distribution-system applications, including those in the open lab or cleanroom and in component manufacturing. Anyone performing a weld for the semiconductor industry, whether they are building a component or subassembly or working on equipment or an equipment hookup, need only refer to these documents.

GTA welding of stainless steel and other corrosion-resistant alloys involves a number of variables. The systematic, standardized approach outlined in the standards ensures that human error is held to a minimum and that welders and weld equipment operators understand that unsuccessful welds are the result of specific, identifiable variables. Implementing the standards will help prevent process fluid contamination, which in turn will enable semiconductor manufacturers to maximize device yields.


Sunniva Collins, PhD, is a manager of metallurgy and surface science at Swagelok (Solon, OH). In this role, she is responsible for assessing technical issues concerning materials, with special emphasis on critical systems and high-purity applications. Collins serves on SEMI’s North American task forces on surface analysis and stainless steel and cochaired the gas tungsten arc welding task force. She received a bachelor’s degree from the University of Michigan in Ann Arbor and a PhD and MSE in materials science and engineering from Case Western Reserve University in Cleveland. (Collins can be reached at 440/248-4600 or unniva.collins@swagelok.com.)

Ed Wolf is a training specialist with the Swagelok distributor training services group. He has more than 35 years of experience in the semiconductor industry and is a member of the SEMI standards and ASME A-12.20.1 committees. In addition, he cochaired SEMI’s gas tungsten arc welding task force. He received a technical degree in electronic science from Texas A&M University in College Station. (Wolf can be reached at 440/248-4600 or edwin.wolf@swagelok.com.)

Kevin Nguyen has been a standards engineer with SEMI (Mountain View, CA) since 2000. He works with industry volunteers to develop standards related to semiconductor manufacturing facilities, process gases, and silicon wafers. Nguyen holds a bachelor’s degree in chemical engineering from the University of California, Davis. (Nguyen can be reached at 408/943-7997 or knguyen@semi.org.)


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.