As of June 2026, the best anyone could do was treat carbonate-radical formation like a bulk-water reaction wearing an interface costume. This paper changes that.

OK, so I have read this paper five times and I think what Liu and colleagues are saying is: the place where water meets air is not just a boundary. It is more like the VIP entrance to a very tiny chemical nightclub, and the hydroxyl radical is somehow on the list.

The paper, "Redefining ·CO3- Formation Chemistry: Zundel-like Switches Drive Carbonate-·OH Interfacial Reactivity," studies how carbonate radicals form when carbonate or bicarbonate meets hydroxyl radicals. That sounds like the sort of sentence that makes a room immediately check its phones, but stay with me. Carbonate radicals matter in atmospheric chemistry, water treatment, pollutant breakdown, oxidative biology, and probably several other places where chemistry is doing paperwork we forgot to file.

The old picture was fairly straightforward: hydroxyl radicals, those hyperactive little oxidizers, bump into carbonate species in the bulk liquid, and carbonate radicals appear. Simple. Tidy. Possibly too tidy, which is always when chemistry starts laughing quietly in the corner.

The Interface Is Not Just the Edge of the Pool

The authors used ab initio molecular dynamics and machine learning molecular dynamics to watch these reactions at a molecular level. Translation: they combined expensive quantum calculations with ML-accelerated simulations, because if you ask pure quantum mechanics to follow every water molecule for long enough, your computer starts aging in dog years.

Machine learning molecular dynamics is not ChatGPT writing lab notes. It is more like training a very fast assistant to imitate high-accuracy quantum calculations of atomic forces. The overworked GPU interns do the math, the model learns the energy landscape, and suddenly you can simulate more chemistry than your budget would otherwise tolerate.

What they found is the spicy bit: the gas-liquid interface dominates carbonate radical generation. Not the comfortable middle of the water. The surface.

According to the authors, hydroxyl radicals and bicarbonate ions both become enriched at the interface: 85.2% of ·OH and 92.2% of HCO3- hang out there in their simulations. If I am reading this right, the surface is basically running a singles mixer for reactive species. Bulk water, meanwhile, is still filling out the event registration form.

Tiny Proton Bridges, Big Chemical Consequences

The star mechanism involves Zundel or Zundel-like hydrogen-bonded configurations. A Zundel-type structure is where a proton gets shared between two molecules through a strong hydrogen bond. Imagine two water molecules fighting politely over the same hydrogen, like roommates pretending the last slice of pizza is "communal."

In this paper, those Zundel-like arrangements act as molecular switches. When the geometry lines up just right, proton and electron transfer can happen quickly. The authors describe two pathways: a concerted proton-electron transfer, where both moves happen together, and a stepwise route, where proton transfer happens first and electron transfer follows. Correct me if I am wrong, but the punchline seems to be: the interface pre-organizes the molecules so the reaction does not have to wander around looking for its keys.

And the numbers are not shy. The interfacial pathway reaches about 90 ± 6.13% yield, compared with 80 ± 8.94% in bulk. Even louder: the reported interfacial rate is about 1.15 × 10^11 M^-1 s^-1, roughly 100 times faster than the homogeneous bulk reaction at 9.63 × 10^8 M^-1 s^-1.

That is not "slightly better." That is the interface showing up in running shoes while bulk water is still tying one lace.

Why Carbonate Radicals Keep Sneaking Back Into the Story

Carbonate radicals used to get treated like supporting actors. Hydroxyl radicals had the marquee role: powerful, reactive, dramatic, absolutely the kind of molecule that would interrupt a meeting. But recent work keeps showing carbonate radicals doing serious environmental chemistry.

A 2025 review by Sharma, Herrmann, and Meyerstein argues that bicarbonate-rich water treatment systems can form carbonate radicals through iron and oxidant chemistry, not just the usual hydroxyl or sulfate radical pathways. Another 2025 atmospheric study found carbonate radical ions may contribute substantially to sulfate formation during dust and haze episodes. There is also 2023 work on micropollutant degradation showing that explicitly modeling hydration can change predicted carbonate-radical reaction rates by an order of magnitude. I love when water, the molecule we all pretend is simple because it is in every children’s diagram, quietly ruins everyone’s assumptions.

The new JACS paper adds a missing mechanistic layer: not only can carbonate radicals matter, but the surface microenvironment may be where much of their formation gets turbocharged.

What This Could Change, Assuming It Holds Up

If these results reproduce and extend to messier real systems, they could change how people design advanced oxidation processes for water remediation. Instead of only asking "How do we make more ·OH?", engineers might ask "How do we create better interfaces where ·OH, bicarbonate, and carbonate meet in the right geometry?"

That could matter for pollutant degradation, atmospheric aerosol chemistry, catalyst design, and maybe even how we model oxidative cycles in natural waters. It also makes the ML angle feel genuinely useful rather than decorative. The machine learning here helps simulate fleeting molecular arrangements that experiments struggle to catch directly, because hydroxyl radicals have the patience and lifespan of a fruit fly with espresso.

Still, I would not run into the street yelling that all water chemistry has been solved. The paper uses simulations, and simulations are maps, not the territory. The next questions are experimental validation, different pH conditions, salts, organics, aerosols, mineral surfaces, and all the chaotic nonsense real water brings to the party.

But I think the central lesson is sturdy: chemistry at interfaces can differ wildly from chemistry in the bulk. The edge of water is not empty space. It is a reaction zone with opinions.

References

1. Jiarong Liu et al., "Redefining ·CO3- Formation Chemistry: Zundel-like Switches Drive Carbonate-·OH Interfacial Reactivity," Journal of the American Chemical Society (2026). DOI: 10.1021/jacs.6c01510. PubMed: PMID 42308371

2. ACS Figshare supporting record for Liu et al. (2026): 10.1021/jacs.6c01510.s001

3. Virender K. Sharma, Hartmut Herrmann, and Dan Meyerstein, "Carbonate Radical Anion (CO3•-) in Carbonated Water: Generation and Importance in Environmental Processes," Environmental Science & Technology 59, 25518-25526 (2025). DOI: 10.1021/acs.est.5c05578

4. Yangyang Liu et al., "Carbonate radical ion as a key driver of rapid atmospheric sulfate formation," npj Climate and Atmospheric Science 8, 45 (2025). DOI: 10.1038/s41612-025-00905-4

5. Wei Zeng and Wei Hu et al., "The overlooked carbonate radical in micropollutant degradation: An insight into hydration interaction," Chemical Engineering Journal 474, 145245 (2023). DOI: 10.1016/j.cej.2023.145245

6. Ruijuan Zhao et al., "Radical-mediated proton transfer enables hydroxyl radical formation in charge-delocalized water," Chemical Science 16, 11954-11960 (2025). DOI: 10.1039/D5SC02206A

Disclaimer: This blog post is a simplified summary of published research for educational purposes. The accompanying illustration is artistic and does not depict actual model architectures, data, or experimental results. Always refer to the original paper for technical details.