Radical Discovery Prompts A Rethink Of Cloud Chemistry

How long does a cloud live? As a nation, Britain is famous for its preoccupation with the weather. So as a scientist growing up here, it seemed only natural to look to the skies and make this into a professional concern as well as an opportunity for small talk.

I may be biased as a result of this obsession, but cloud formation is an exciting process. You won’t be surprised to hear that the key ingredient is water vapor. However, in most locations in the atmosphere, clouds don’t form spontaneously from water alone – and this is where aerosols enter the picture.

Aerosols are tiny particles which provide a surface for water vapor to condense onto and form droplets. They play a key role in climate by enabling this cloud condensation, as well as directly scattering and absorbing sunlight. This is to say nothing of their huge role in human health at the planet’s surface, as the most important factor in air pollution-related mortality.

Aerosols and clouds also provide fascinating and important spaces for atmospheric chemistry. In a recent study, we looked at one such process, the formation of hydroxyl radicals (OH) in cloud droplets. Hydroxyl radicals are known as “nature’s detergent” due to their ability to react with a multitude of atmospheric species in air and water. These reactions can transform pollutants and greenhouse gases and change the abundance of aerosols. OH chemistry, therefore, has a key role in the atmosphere.

Where does the OH come from in cloud droplets? Previous studies have suggested that transfer from the surrounding gas phase (via the breakdown of ozone in sunlight) was probably the most important known source [1]. Other work has collected water from real clouds and found that species dissolved in the droplets such as hydrogen peroxide and nitrate also decompose in sunlight to form moderate amounts of OH [2]. Often this production is assisted by trace amounts of transition metal ions such as iron (Fe2+).

Our recent study [3] found a new and potentially dominant source of OH in cloud droplets. The work, led by Prof Suzanne Paulson at UCLA, took a different approach to previous studies and looked for OH production during the conversion of aerosol particles into cloud droplets. And it turns out that during cloud drop formation, a substantial light-driven “burst” of OH is observed. This burst lasts at most a few minutes (the time resolution of our measurements) before settling down to the more sedate OH production rates seen by groups who collected the cloud water directly.

The samples in our study were collected from three urban sites in California (Claremont, Fresno, and West Los Angeles) with distinct characteristics in terms of aerosol sources. Aerosol samples were collected onto filters and transported to the UCLA laboratory for analysis. Cloud formation was simulated by adding water to the aerosol filter samples (at dilutions relevant to atmospheric cloud droplets) in the presence of UV light. The water contained a small quantity of a probe molecule which fluoresces after reaction with OH, and this allowed us to monitor OH production from the solution over time.

Remarkably, all the aerosol samples we tested exhibited some form of OH burst, regardless of location, season (summer/winter), or time day (morning/afternoon/evening). It seems from our results so far that this is potentially a widespread phenomenon during cloud droplet formation.

A couple follow up questions arose when we saw these results: what is producing the OH this quickly, and why didn’t it happen in the aerosols before we collected them?

Puzzlingly, the known OH sources I mentioned above, such as hydrogen peroxide and nitrate, decompose too slowly to produce rapid bursts of OH we saw. We investigated alternatives and found that in the laboratory we could recreate the burst in simulated cloud water containing organic peroxides (analogous to hydrogen peroxide, but often more reactive) and iron ions (Fe2+). This may sound exotic, but organic peroxides are produced in the atmosphere from reactions of plant and vehicle emissions with (you guessed it) OH radicals. From our tests, we expect them to be abundant enough and decompose quickly enough to be a plausible source of OH in our ambient samples. We think that the role of light is to regenerate the Fe2+ which catalyzes peroxide breakdown.

The second question (why didn’t the OH form in the aerosol?) returns to the distinction between cloud droplets and aerosols. While cloud droplets are highly dilute, aerosols contain relatively little water and are often observed to be highly viscous or even solid in the atmosphere. So, while aerosols may contain all the ingredients for the burst, we hypothesize that they are locked in an unreactive state until dissolved by water during cloud formation. Our previous work [4] speculated that there may also be similar “pent-up” reactive molecules in aerosols inhaled by humans, and the highly humid lung environment could act as a “quick release,” allowing damaging reactions to occur.

At the beginning of the article, I posed the question: “How long does a cloud live?” I haven’t answered it yet, but now, almost at the end, it’s key to putting the “burst” we measured into context. It’s impossible to give a single answer, but cloud droplets may seem surprisingly short-lived. Typically, they persist for around 15 minutes before evaporating again, back to aerosols and water vapor. And in their 15-minute lifetime, the size of bursts we measured would often translate to being the largest source of OH in the droplets.

Observation of this large new source has the potential to improve our understanding of cloud and aerosol chemistry and its consequent impacts on climate, visibility, and health. Perhaps mention it next time you’re chatting to your neighbor about the latest weather forecast!

These findings are described in the article entitled A light-driven burst of hydroxyl radicals dominates oxidation chemistry in newly activated cloud droplets, recently published in the journal Science Advances.


  1. Herrmann et al., Atmos. Environ., 39, 4169-4183, 2005.
  2. Bianco et al., Atmos. Chem. Phys., 15, 9191-9202, 2015.
  3. Paulson et al., Sci. Adv., 5, eeav7689, 2019.
  4. Gallimore et al., Atmos. Chem. Phys., 17, 9853-9868, 2017.