Material Corrosion Study Belies Conventional Wisdom

Maria A. Lester, Associate Editor -- Semiconductor International, 11/1/1999

Conventional wisdom in the semiconductor industry has asserted that low manganese (Mn) in electropolished (EP) 316L stainless steel improves corrosion resistance, particularly in welded applications such as gas delivery lines. This is based on research that showed stainless steel tubing had better corrosion resistance when manganese was removed from base material. The standard accepted level of manganese in the industry is 0% to 2%. But researchers at Swagelok Co. (Solon, Ohio) wanted to validate the standards before switching to lower-manganese stainless steel. Their findings refuted the industry s common assumptions.

Testing was performed on 21 heats of electropolished tubing, including 6 low-Mn heats, in diameters ranging from 1/4 in. to 2 in. The heats were selected from among 300 commercially available heats representing a variety of mills, melt techniques and chemistries. Following more than 2100 orbital welds, standard ASTM (American Society for Testing and Materials) corrosion resistance tests (ASTM G31 for general corrosion resistance and ASTM G150 for critical pitting temperature [CPT]) were performed on the inside surfaces of as-received and as-welded tubing.

Results indicate manganese has little or no effect on the corrosion resistance of EP 316L tubing. However, there was a significant correlation between the tubing s

2. This graph shows pitting occurs on the weld and in the heat-affected zone on each side of the weld with no enhanced corrosion downstream of the weld due to redeposited manganese from welding. (Source: Swagelok)

base chemistry and corrosion resistance before and after welding: The results indicated base chemistry of the tubing determines corrosion resistance. Chromium and nickel for as-welded tubing were shown to improve pitting corrosion resistance, while sulfur had a detrimental effect (Fig. 1). General corrosion resistance was enhanced by molybdenum and copper, with manganese having a slight negative role.

More than 300 orbitally welded test specimens that had undergone pitting corrosion tests were examined for pit location with respect to the weld and purge gas direction. Results were contrary to current industry suppositions: Instead of increased pitting downstream of the weld, where the re-deposited manganese layer would occur, more pitting occurred upstream of the weld (Fig. 2).

Although the industry has accepted surface science figures of merit as predictors of actual corrosion resistance,  our test results showed no correlation, said Dr. Sunniva Collins, research metallurgist at Swagelok. Other researchers have studied different passivation methods for welded parts; however, many of these methods are based on the manganese/corrosion assumption or the use of surface science numbers to indicate enhanced surface passivity. For example, a high chrome-to-iron ratio, high chrome oxide-to-iron oxide ratio, or a very thick oxide layer traditionally have been assumed to show improved corrosion resistance.

1. The graph shows corrosive response of the materials is due not to Mn but to other alloying elements and that pitting corrosion is a strong function of base chemistry. The low Mn heats (Mn l 0.3%) are shown with open triangles. Each data point is the average of at least six valid repeats. (Source: Swagelok)

To test the bounds of these approaches, Swagelok also did surface science analysis on the heats to evaluate the numbers and determine what, if any, bearing they have on the material s corrosive response. Again, no correlation was found. Said Collins:  The implication has always been that you need these high numbers in order to get better corrosion resistance. The truth of the matter is that corrosion response is governed not by surface chemistry but by base chemistry. These findings call into question the necessity of many passivation techniques and other surface finish requirements that increase the values of these figures.

Further efforts to improve corrosion resistance are in progress at Swagelok, such as studying different post-weld passivation programs. Current research summaries will be posted on the company s Web site, www.swagelok.com. More detailed work on the basic metallurgy of the tubing will evaluate correlations between other metallurgical parameters, such as mechanical properties and welding conditions. " .