Study points to ammonia contamination in water microdroplets saga

New mass spectrometry experiments challenge water microdroplet’s purported ability to spontaneously generate hydroxyl radicals. Led by Ryan Julian at the University of California, Riverside, US, the experiments point to trace ammonia contamination mimicking the signals previously attributed to these reactive species.

Water microdroplets are typically about 10μm in size. They’re generated naturally in the atmosphere, or using tools like ultrasonic humidifiers or gas nebulisers. The ions that do, or do not, form spontaneously within them have been a subject of significant debate since 2019, when Richard Zare of Stanford University in the US announced that sprays of microscopic water droplets spontaneously produce hydrogen peroxide.

Spraying pure water into a mass spectrometer, Zare observed an ion at m/z 36, which he interpreted as a hydroxyl radical as either [H₃O + ^•OH]⁺ or [H₂O^• + H₂O]⁺. He also exposed two radical scavengers, melatonin and caffeine, to water microdroplets. The resulting spectra showed peaks 17Da above the m/z of the protonated drugs, which he took to be their oxidised forms, and proof that the microdroplets contained hydroxyl radicals. 

‘That was really the first report of something really unusual in a droplet. Since that time, people have debated whether that is a correct result,’ explains Evan Williams who researches mass spectrometry at the University of California, Berkeley, US, and was not involved in either study. 

Julian had not followed the water microdroplet debate prior to this research, first encountering it at a conference. ‘Definitely it didn’t seem to be making a lot of sense to me,’ he says. While previous work presented alternative explanations for Zare’s results, until now, no one has faithfully reproduced his original data. ‘If you can reproduce the original results on which everything was based and show that it’s all not what it was represented to be, then it’s pretty difficult to argue against that,’ says Julian.

Working with graduate student Aidan Purcell, Julian used gas flow nebulisation on LC/MS grade water to produce microdroplets in the absence of any voltage. The resulting spectra showed the expected large m/z 36 peak, but when this ion was subjected to collision-induced dissociation (CID), the researchers saw m/z 18, the expected mass for NH₄⁺, not m/z 19, which would correspond to H₃O⁺. 

Notably, the gas flow nebulisation data did not show a peak at m/z 37, where [H₃O + H₂O]⁺ would be. ‘That’s the dominant ion that you expect to get from a protonated water droplet,’ explains Williams. Its absence is further proof that m/z 36 corresponds to protonated ammonia, not water, argues Julian. 

When the team instead electrosprayed water into the mass spectrometer, a technique known to produce highly protonated water droplets through voltage input, a peak at m/z 37 appeared alongside the one at m/z 36. CID of the m/z 37 peak gave m/z 19, corresponding to H₃O⁺, whereas CID of m/z 36 again yielded m/z 18, corresponding to NH₄⁺. 

To be certain that the observed m/z 18 CID products do correspond to NH₄⁺, the researchers repeated the nebulisation and electrospray experiments using ¹⁸O water. Again, they saw a peak at m/z 36, as well as one at m/z 38 that they interpreted as [NH₄ + H₂¹⁸O]⁺, but nothing at m/z 40, which would correspond to [¹⁸OH₃ + •¹⁸OH]⁺. CID of m/z 36 and 38 peaks yielded exclusively m/z 18 ions, consistent with NH₄⁺ but not [¹⁸OH₂]^+•, which would have given a m/z 20 peak. 

Williams describes this result as compelling, noting that ¹⁸O labels water but not ammonia, making it a definitive test. And to him, ammonia contamination makes a lot of sense. As well as being a common solvent contaminant, ‘your breath contains a few hundred parts per billion of ammonia. You also exude more ammonia out of your skin than you actually breathe, so even if you hold your breath, ammonia is going to be in the air around the instrument,’ he explains. 

Himanshu Mishra, an environmental science and engineering professor at King Abdullah University of Science and Technology in Saudi Arabia, has long questioned Zare’s experimental results. He describes Julian’s findings as beautifully logical, authoritative, and ‘an important salvo in the field’. Williams echoes these thoughts: ‘Ryan Julian does a very elegant job breaking this ion up and showing that the results are consistent with protonated ammonia with a water molecule and that they are not consistent with protonated water with a hydroxyl radical.’ 

Both Mishra and Williams note that several experiments disproving Zare’s hypothesis have already been published, but that Julian’s contribution was necessary for the water microdroplet debate to advance. However, no one Chemistry World spoke to believes this work will put an end to the debate, and Julian is prepared for pushback on it. ‘I know other people who’ve been involved with this controversy have had people question, “why did you bother to do this?” or say “you’re just being a jerk”… But at the end of the day science needs to be right for us to move on and progress, and so there’s not really another way to do it, other than to publish diverging opinions where they exist.’ 

Zare declined to comment.