Extra precautions sometimes backfire

Harold D. Fitch

The phone rang just as the manufacturing project manager was about to head home for the weekend.

“Hello,” he said. He heard a sigh on the other end of the line. “We found some products with signs of black aluminum,” said the production manager. “I`ll be right there,” said the manufacturing product manager.

As he was walking to production, he remembered that black aluminum meant that the copper aluminum alloy used for the circuit wiring layer had a corrosion problem. It had always been caused by some kind of halogen contamination in one of the aluminum processing steps. The culprits had usually been chlorine or fluorine related.

He immediately started calling personnel to determine who could start sampling materials on the second shift that night.

Fortunately, several key manufacturing engineering people with past experience with this type of problem were still on site. They began collecting samples of materials such as photoresist, developer, rinse, and cleaning solutions.

The chemical analysis department lined up enough personnel to start on the analysis early Saturday morning. At 7 a.m., the group met to discuss material samples to be tested.

No halogen

The product in the aluminum section had been put on hold at the first sign of the problem. Tests revealed that only about 10 lots of product showed any of the black aluminum. These lots were set aside. The various chemical solutions had been collected and turned over to the analysis group. It was now time to do a general cleaning of the aluminum processing area and get ready to process product as soon as the culprit was identified and eliminated.

Unfortunately, as the analysis was completed, no halogen contamination was found anywhere. Later, a couple of tests still exhibited the aluminum corrosion.

The next week was spent continuing to look for halogen contamination, but none was found. Each test showed the corrosion problem. By the end of the week, no visible progress had occurred. The manufacturing/engineering team was at a loss to find the cause of the problem.

Yellow acetone?

On Monday of the second week, the project manager and an engineer were touring the affected process area when one of the process operators asked if she could talk to them about her process.

“I`ve noticed that the acetone is yellow lately,” the process operator said. “I don`t know why.”

The engineer and project manager went to the process station with the operator, who was responsible for the resist strip process. They all observed the yellow acetone. The next step: determining if the incoming acetone was yellow. It was. The team then exposed non-corroded aluminum to the yellow acetone. Exposure to the yellow acetone did indeed cause the good aluminum to corrode!

In addition, exposure to clean acetone from the chemical lab did not produce any corrosion. The acetone feeding the strip tank was changed to another batch of clear acetone and the problem disappeared.

A sample of the yellow acetone was sent to the lab for examination. Acetic acid and a trace of iron were found. The acetic acid could certainly cause corrosion of the aluminum copper alloy.

How did it happen?

A review of the acetone handling procedures showed that acetone was received in 55 gal. drums, stored in the chemical storage area, and delivered to the manufacturing line in the drums as needed. Receiving inspection was performed on the incoming acetone, and all of the acetone in question had passed the incoming inspection. What had happened?

The quality group`s inspection the acetone drums showed two things: one, several drums with yellow acetone all tested positive for acetic acid and iron; two, there were about three times the usual number of drums of acetone in the chemical storage area. Now the picture was becoming clearer. There had been concern about the storage of acetone and several extra lots had been bought in advance, therefore, a large amount was in storage. Further investigation showed that the incoming material had passed inspection but quality control had decided that additional inspection during storage would be a good idea because of the lengthened storage time. The drums were stored in an outdoor chemical storage area near the river, and a QC inspector opened selected drums and took samples for spot checking. All the spot checks passed inspection, but the problem with yellow acetone was traced to drums that had been opened for this spot check.

When the drums were opened for inspection, moisture contaminated the drums because of the high humidity from the river. In turn, this caused the formation of acetic acid that attached to the steel drums causing them to rust.

Harold Fitch is president of Future Resource Development, a consulting firm in Burlington, VT, specializing in cleanroom education and problem-solving.