Adding Iodide to Plasma-Activated Water: Faster Kills, a Proposed Hypoiodous Acid Mechanism, and a Salmonella Exception (McClenaghan et al., 2025)

Plasma-activated water on its own often struggles against bacteria living in biofilms. A 2025 open-access paper in the journal Biofilm, by Laura A. McClenaghan, Thomas P. Thompson, Akash Shambharkar, Ross M. Duncan, Paula Bourke, Timofey Skvortsov and Brendan F. Gilmore (accepted 20 August 2025), tests whether a trace of potassium iodide (KI) changes that. Its headline result is that micromolar iodide turned a weak treatment into one that wiped out two food pathogens within minutes. Its most useful result for a sceptical reader is that a third pathogen, Salmonella enterica, was never fully killed under any condition tested. The authors' acknowledgements name the Elephant's toothpaste experiment, a well-known demonstration of iodide reacting with hydrogen peroxide, as the study's inspiration.

What is the idea?

Plasma-Activated Water (PAW) contains hydrogen peroxide, nitrate and other Reactive Oxygen and Nitrogen Species (RONS) in strongly acidic water. Chemists have long known that peroxide oxidises iodide in acid to iodine, triiodide and hypoiodous acid (HIO), all of which kill microbes. The authors' hypothesis is that the oxidants in plasma-treated water drive on-the-spot formation of reactive iodine species (RIS), a family that includes HIO and triiodide. They present this in-situ generation as a way of minimising chemical storage and reducing harmful residues. They build on a 2019 study by Ke and colleagues, who found that iodide boosted bacterial killing by a helium plasma jet.

How was the water made and tested?

Water was activated with a needle electrode held 5 mm above 10 mL of deionised water, run in open air for 10 min. Two arrangements were compared. A spark discharge (4 kV, about 25 kHz) gave water with 1.357 mM hydrogen peroxide plus nitrate. A glow discharge (2 kV) gave mainly nitrate (228.26 mg/L) and nitrite (223.80 μM) with negligible peroxide. Both brought the pH down to about 2.5. Because the proposed iodine chemistry needs peroxide, only spark water was used from then on, with KI from 10 to 100 μM present.

The test organisms were reference strains of Escherichia coli, Listeria monocytogenes and S. enterica, challenged both as free-floating cells and as 24-hour biofilms grown on polycarbonate discs. The team also used live/dead fluorescent staining under a confocal microscope, chemical scavengers meant to remove particular oxidants, water re-adjusted to higher pH, and storage at 4 °C for up to 14 days.

What did they find?

The chemistry shifted. Adding iodide did not use up peroxide, as might be expected. Instead, measured peroxide rose from about 1.2 mM to about 1.8 mM at 30 μM KI. Nitrate rose too, from 136 mg/L to 273.9 mg/L at 100 μM KI in the paper's summary table, roughly a doubling, although the abstract puts the increase at only about 17 %; and the pH stayed between 2.15 and 2.37. The authors offer two possible reasons (catalysis involving iodine, or slower peroxide breakdown) and settle on neither.

Two pathogens were killed quickly. Plasma water alone produced only small reductions. With iodide, the paper reports complete kills of free L. monocytogenes within 3 min at 10 μM KI and of E. coli at 50 μM; the table gives log reductions of 7.06 and 6.74 respectively after 5 min. Biofilms of both species were eradicated after 10 min with 100 μM KI, and E. coli biofilms also with 50 μM.

Salmonella was the exception. Free S. enterica cells were only partly inactivated; the paper calls the species partially tolerant. The summary table's planktonic rows carry three unlabelled values, which appear to correspond to plasma water alone, 10 μM KI and 50 μM KI. For S. enterica these read 1.46, 0.8 and 0.95 log, so iodide was no better than, and numerically lower than, plasma water alone. These are single table values from three replicates, and the paper does not comment on the difference. Biofilms fell by 2.67 log at best. The microscope images matched: E. coli biofilms turned largely red (dead) after 5 min while S. enterica stayed mostly green. The authors suggest Salmonella's extra peroxide-destroying enzymes and denser biofilm matrix may explain this, but they did not test either explanation.

Peroxide alone with iodide was not enough. Mixing 1.2 mM hydrogen peroxide with 50 or 100 μM KI did not add to peroxide's killing power. The authors take this as evidence that something else in plasma water is involved.

Scavengers. Sodium pyruvate, which removes peroxide, and L-histidine, which removes singlet oxygen, each abolished killing of E. coli. Uric acid (for ozone) and Tiron (for superoxide) made little difference.

Storage. Peroxide in plain plasma water had dropped significantly by day 7. In iodide-treated water it held up longer: at 50 μM KI it did not fall at any point over the 14 days. The paper reports that water with 10 or 50 μM KI retained its antibacterial activity against E. coli and L. monocytogenes after 7 days at 4 °C, but the text gives no figure for the size of the kill.

The proposed mechanism and what is unverified

The paper proposes four steps. Plasma makes acidic, peroxide-rich water. Iodide is oxidised to triiodide and HIO. Other plasma-made oxidants, such as singlet oxygen and nitrogen species, intensify that iodine chemistry. Finally, uncharged HIO crosses bacterial membranes and does oxidative damage. Support for the role of HIO comes from raising the pH, which the authors report suppressed killing, consistent with HIO losing its uncharged form. The authors concede that changing the pH also alters the rest of the water's chemistry.

They are frank that much remains unshown. In their words, "key steps remain unverified". In particular, uptake of HIO into cells and the damage it does "have not been directly demonstrated". They also state that they did not directly measure any iodine species. The mechanism is therefore inferred from peroxide and nitrate readings, scavenger results and kill curves, not observed.

How far does the evidence go?

• Iodine species never measured. The abstract says the iodine species stayed active throughout 14 days of storage, but they were never measured. Antimicrobial activity after storage was tested only at day 7.
• Imperfect peroxide control. The comparison control used 1.2 mM peroxide, the level in iodide-free plasma water. The paper's own table gives 1.70–1.74 mM for plasma water with 50–100 μM KI, so the control did not exactly match the treatment it was compared against.
• Internal inconsistencies. The methods list the adjusted pH values as 5, 7.5 and 10.5, while a figure caption gives 5, 7 or 10. The L-histidine concentration appears as 20 μM in the methods and 20 mM in the results. The text says plasma water alone did nothing detectable to S. enterica, yet the table reports 1.46 log. The abstract gives a nitrate rise of about 17 % with iodide, while the table's figures (136 to 273.9 mg/L) roughly double.
• No chemical neutralizer described. Samples were transferred into phosphate-buffered saline and diluted before plating. The methods describe no chemical neutralizer. That is the practice questioned in The Neutralization Problem: Why Plasma-Activated Water Kill Rates May Be Overstated, where leftover oxidants can keep acting after sampling.
• Narrow conditions. The experiments used one discharge setup, deionised water, three strains, a high starting load of 1 × 10^7 CFU/mL, and polycarbonate discs. The authors' suggestions of food-processing, wound-care and sanitation uses, and their low-residue claim, go beyond what was tested. They acknowledge that the short life of the iodine species was not characterised in this study.

Why it matters for this dossier

This paper is a concrete case of the enhanced plasma water that the wider field promotes. It also shows why such claims need species-by-species reading. The same treatment that eradicated E. coli and Listeria only partly inactivated Salmonella, and against free Salmonella cells the iodide-treated water did no better in the table than plasma water alone. It also supports the dossier's broader point that peroxide in spark-made water is a working ingredient, not a fleeting trace (see Non-Thermal Plasma (Cold Atmospheric Plasma) for the discharge types). The work was funded by the US National Institutes of Health and Northern Ireland public bodies, with a studentship linked to Linden Foods.