Per-unit-mass reactivity changes the comparison
Surface water photochemistry has long treated DOM as the main photosensitizer, the component that absorbs light and generates reactive intermediates. This study asks whether mineral nanomaterials, present at parts-per-billion levels, can matter alongside DOM at milligram-per-litre carbon concentrations. The answer depends on which quantity is compared: concentration, mass-normalized reactivity, or total formation rate in a given water.
The authors separate these quantities explicitly. They measured hydroxyl radical (HO•) formation rates for TiO2 nanomaterials at 1, 10, 50 and 100 μg L−1 and for DOM isolates at 1, 2, 4 and 5 mgC L−1, under 222, 265 and 365 nm light. Formation rates scaled linearly with concentration in both cases, which allowed a mass-normalized rate to be fitted for each material. That normalized rate, not the raw concentration, is what makes the two classes comparable.
Probe compounds turn radical production into rates
HO• was tracked indirectly: benzene in phosphate buffer at pH 7.3 was converted to phenol by HO•, and phenol production rates were used to calculate HO• formation rates. Singlet oxygen and triplet excited DOM were tracked with separate probes, furfuryl alcohol and 2,3,5-trimethylphenol respectively, with tert-butanol added to the singlet oxygen experiments to remove HO• interference.
Light sources were LEDs peaking near 265 and 365 nm and a far-UV microplasma lamp near 222 nm, with photon flux measured by actinometry. Experiments ran in triplicate in a stirred Petri dish at 21 ± 1 °C, and nanoparticle samples were centrifuged and filtered before analysis. These choices define the scope: clean buffered solutions with one photosensitizer class at a time, not whole surface water.
TiO2 rivals DOM for hydroxyl radical formation
Across all three wavelengths, TiO2 produced HO• 10× to 100× faster per unit mass than DOM. The gap was 10× to 15× under far-UV and UV-A, and about 100× under UV-C. The consequence is that small nanomaterial masses can offset large DOM carbon masses: under far-UV, 50–100 μg L−1 of P25 or anatase generated HO• at rates comparable to 1–2 mgC L−1 DOM, while 4 mgC L−1 DOM was 3–4 times higher.
At 365 nm the balance shifts toward nanomaterials. There, 100 μg L−1 P25 exceeded 2 mgC L−1 of all three DOM isolates tested, and 50 μg L−1 P25 matched 1 mgC L−1 DOM. Even at 4–5 mgC L−1 DOM, 50–100 μg L−1 TiO2 could still account for 21–42% of total HO• formation. The authors attribute the wavelength dependence to differences in light absorption and in electron–hole separation between the anatase and mixed anatase–rutile P25 phases.
Mixtures exceed the sum of their parts
When DOM and nanomaterials were irradiated together, HO• formation was 16–21% above what the two components produced separately. This is a synergy result, not a simple addition of two independent sources. The paper's explanation is that DOM can scavenge HO• and screen light, while nanomaterials occur at much lower concentrations and do not show those effects at the tested levels.
The synergy matters for interpretation because it means measurements of filtered field samples may misattribute radical production. If photoactive mineral colloids pass through common filters, as the authors argue, then some HO• attributed to DOM in filtered water could come from particles that were never removed.
A HO•-selective pollutant reveals the nanomaterial contribution
The study tested 6PPD-quinone, a tire-derived transformation product that the authors describe as selective toward HO• and not significantly degraded by indirect photolysis with DOM. Under 365 nm light, DOM had limited effect on its degradation, whereas 100 μg L−1 P25 raised the pseudo-first-order degradation rate above direct photolysis alone, consistent with the higher steady-state HO• concentration measured in the nanomaterial case.
This result is specific to a pollutant whose transformation is dominated by HO•. The paper extends the logic with a broader set of organics using literature bimolecular rate constants and assumed steady-state concentrations, showing that DOM remains the dominant photosensitizer for contaminants transformed mainly through singlet oxygen or triplet excited DOM. The nanomaterial advantage is therefore pathway-dependent, not universal.
The advantage depends on wavelength and DOM level
The practical reading is a set of conditions rather than a single verdict. Nanomaterials contribute most where UV-C or far-UV light is present, where DOM carbon is low, and where the pollutant reacts primarily with HO•. In DOM-rich water under UV-A or far-UV, TiO2 can still contribute a minority share of HO• formation but is unlikely to dominate.
The experiments used isolated DOM and two commercial TiO2 powders in phosphate buffer, so they do not capture the full ionic, carbonate and colloidal complexity of real surface water. The authors propose spiking and unfiltered field sampling as the next tests, and they note that the mineralogy of TiO2 in surface waters is not well characterized. These are boundaries on transferability, not on the internal comparison.
Watch the denominator before ranking photosensitizers
The paper's most useful correction is methodological: a photosensitizer's importance cannot be read from its concentration alone. TiO2 at 100 μg L−1 and DOM at 4 mgC L−1 differ by more than an order of magnitude in mass, yet their HO• formation can be within a factor of a few, and mixtures can exceed both. Any ranking of DOM versus nanomaterials therefore has to state whether it is comparing per unit mass, per unit photon flux, or total rate in a specified water.
For UV-based treatment and for HO•-sensitive micropollutants such as 6PPD-quinone, the paper's evidence supports treating nanomaterials as a co-driver rather than a trace curiosity. For pollutants transformed mainly by singlet oxygen or triplet excited DOM, the same evidence leaves DOM in the dominant role.
Sources used in this answer
Environmentally Relevant Concentrations of Nanomaterials Drive Photochemical Production of Reactive Species and Pollutant Degradation in Surface Waters
TiO2 nanomaterials at 1–100 μg L−1 produce HO• 10× to 100× faster per unit mass than DOM, yielding comparable total HO• formation, 16–21% synergy in mixtures, and enhanced 6PPD-quinone degradation under 365 nm light.