Brilliant Idea #1: Helium arc welding and hydrogen isotopes

Marilyn and Tom, 1980

Not many women can say they’ve owned a welding rig. I purchased a helium arc welder (i.e., heliarc welder) as a new staff scientist to try out a possible new method for measuring hydrogen isotopes in animal tissues.

“A welder?” you ask.

“What’s she doing with a welder?”

It wasn’t just any welding machine. It was a fancy, brand new welder designed for detail work—sealing small bits of metal together, rather than the rigs used to build bridges. We had it installed down in the basement of the old Geophysical Lab, where the custodial staff made their lunches and changed into their uniforms.

We’d been having problems with our combustion vacuum line when I was analyzing protein-rich animal tissues. Tom Hoering had read a paper that reported sealing organic matter into a pure iron tube, heating it to over 1500°C at which point the iron metal became pervious to hydrogen gas which seeped out. It wasn’t easy to find pure iron tubing, but our instrument shop had a small stock that they weren’t using. They donated it to our new, bright idea.

Here was our plan. Once we learned how to weld, we’d seal our samples in the iron tubes, place them in a specially-made glass vacuum line and heat them up. Because of the extreme temperature that we needed to achieve, we couldn’t use a regular furnace—not even a platinum one—because the glass, even quartz, would melt. Instead, we needed to use an induction heater [think the fancy cooktops people have these days] that was the size of a modern isotope ratio mass spectrometer, about a cubic meter and weighing over 200 kilograms. With induction heating, only the iron metal tube would get hot, allowing the hydrogen gas out. We’d collect it and measure its isotope patterns.

The day the welding machine was installed, Tom and I joined “the Guys from the Shop” and were instructed by the welding salesperson how to use the welder. Tom took a stab at it first. “Damn!” he said, as he burnt through the metal. He tried a second time—it was better, but not pretty.  I was next.

I donned the welding shield, settled in, watched by Andy, Steve, and Herb from the Shop, Tom and the welding rep. No pressure! I sparked it up—and wouldn’t you know it, burned a hole just like Tom. He and I laughed—a loud chortle from Tom and a small snicker from me. I did three more tries, and by the last one, it wasn’t half bad. Tom and I were satisfied and proud.

The Shop guys inspected our work. They were clearly not impressed. They were accustomed to precision work. Our work was “good enough for government work”, but it didn’t look too hot. Tom and I carried the ‘sealed’ tubes to our vacuum line and checked them out. Only two out of five or six were actually sealed. With a bit of practice, we finally got it figured out and were ready to start the measurements.

I weighed out 5 milligrams of dried mouse liver and tapped it into an iron tube that had been welded on the bottom. We took five of these down to the basement, fired up the welder, taking turns welding them shut, and brought them back upstairs to Tom’s lab for the analyses.

Picture the older scientist with his necktie tucked into his shirt along with me, bushy long hair held back by a rubber band, a highly unlikely pair trying out a wild idea.

The induction heater was fired up. It was noisy, required 240 Volt electrical power and had huge vacuum electric tubes that needed to be warmed up for 30 minutes before it was ready to go. It took about an hour to process each of the five samples, followed up by running them on our hydrogen isotope mass spectrometer that I had on and ready. One of the samples was not sealed properly. Two of the others had low yields. The remaining two had measureable amounts of hydrogen with isotope values similar to each other, close to measurement we’d made by more “traditional” methods.

The old hydrogen mass spec, 1979

I wouldn’t call this a success in any sense of the word. Today, a similar analysis takes about 5 minutes to weigh a sample, and another 5 for the analysis. No welding, no monster furnaces. Tom and I did another two sets of samples with similar results. 

We had to admit to ourselves that this wasn’t ever going to be a robust method. We spent our time re-designing and building a better combustion vacuum line that ultimately worked well. The welder was given to the Shop, but none of the Guys wanted to learn how to use it. Eventually we gave it to our sister lab, DTM, who had folks who knew heliarc welding.

Sometimes you get “good” ideas. They make sense. They’ll solve a problem or open up a new scientific door. And other times, well, more work is necessary.