They thought they might solve the world’s energy problems. Then they realized they were wrong.

Researchers are retracting a 2016 PNAS paper that described a way to create gasoline-like fuels directly from carbon dioxide in the atmosphere.

Senior author Frederick MacDonnell, a professor at the University of Texas at Arlington (UTA), told us he originally thought his team had made a preliminary breakthrough that might “solve the world’s energy problems.” Instead, he said:

It was an elaborate trap we fell into.

In a retraction notice that contains more information than we usually see, MacDonnell and his co-authors wrote:

We have now demonstrated that our results … are largely due to artifacts and that the underlying thesis of this work has not been shown.

They explained how their results were actually due to impurities, likely graphite, in the catalyst driving the reaction. MacDonnell told us that the impurities — and a field-wide failure to account for them — have likely tainted the results of other published studies that used the same catalyst, titanium dioxide.

The paper, “Solar photothermochemical alkane reverse combustion,” was originally published Feb. 22, 2016, in the Proceedings of the National Academy of Sciences, and will be retracted Jan. 8. It described a chemical reaction — so-called “reverse combustion” — that created liquid hydrocarbons out of carbon dioxide and water. Other researchers have also achieved reverse combustion, but have mostly produced methane, MacDonnell said.

In a press release issued by UTA at the time, MacDonnell explained the potential impact of the reaction as the basis for producing more sustainable fuels:

many of the hydrocarbon products from our reaction are exactly what we use in cars, trucks and planes, so there would be no need to change the current fuel distribution system.

The paper received media coverage in a variety of publications, including news outlets such as the Fort Worth Star-Telegram and Daily Mail India, and blogs covering green technology such as Sustainable Skies. The 2016 research has been cited 14 times, according to Clarivate Analytics’ Web of Science.

MacDonnell and first author Wilaiwan Chanmanee, a postdoc at UTA, took responsibility for the paper’s problems. In the notice, they said they had not conducted a “sufficiently rigorous analysis,” which led them to overlook impurities in the catalyst as a source of the carbon that showed up in the products.

“Let’s question everything”

In the notice, the authors thanked University of Toronto professor Geoffrey Ozin for raising questions about their results. MacDonnell told us Ozin contacted him Feb. 24, 2016, just days after publication, expressing excitement, but suggested the authors should have used carbon dioxide containing only carbon isotopes. This would allow them to trace the atoms through the reaction and ensure the carbon dioxide gas was involved. MacDonnell said he initially dismissed the suggestion:

I was kind of full of myself at the time. I was 100 percent convinced we did this.

But months later, when his team tried to scale up the reaction to produce more than trace amounts of hydrocarbons, they couldn’t get it to work as expected. MacDonnell said they realized there was a big problem in May 2017; it took about five months of work to convince themselves that the paper was unsalvageable.

PNAS told us the authors first contacted the journal about the problems on Nov. 13, 2017, and submitted a draft retraction notice later that month, on Nov. 30. Since the beginning of May 2017, when the team realized there was a problem, the paper has been cited five times.

MacDonnell told us he chose not to alert the journal until the very end, because he didn’t want to compound the mistake:

We didn’t rush into publishing that paper and we weren’t going to rush into retracting it until we knew what we were talking about… I’m only going to get one chance to correct the record. If I come back and correct my corrections I’m going to look like a fool.

So:

We tried different things and finally said, “Let’s question everything and go back to basics.

He said part of his reluctance to use 100 percent isotopic carbon, as Ozin suggested, was cost — a single reaction would cost $4,000 to run:

When we finally bit the bullet and spent that money, we didn’t see anything in the product and we were like, “Holy shit, what’s going on?”

The team then ran the reaction using helium gas without any carbon dioxide at all. According to the retraction notice, this experiment also produced hydrocarbons, indicating they were made from the impurities in the catalyst.

MacDonnell said his field needs to learn from the mistake:

There’s a vast body of literature where no such labelling experiments have been done and no controls such as just using an inert gas in lieu of carbon dioxide.

I think that’s going to have to become the standard of the field.

MacDonnell told us:

It’s a horrific experience to go through as a scientist.

But he added that the people he’s dealt with during the retraction process have been “understanding and gracious. They’ve been very kind.”

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here. If you have comments or feedback, you can reach us at [email protected].