Antioxidation is in demand in living systems, as the excessive reactive oxygen species (ROS) in organisms lead to a variety of diseases. The conventional antioxidation strategies are mostly based on the introduction of exogenous antioxidants. However, antioxidants usually have shortcomings of poor stability, non-sustainability, and potential toxicity. Here, we proposed a novel antioxidation strategy based on ultra-small nanobubbles (NBs), in which the gas–liquid interface was employed to enrich and scavenge ROS. It was found that the ultra-small NBs (~ 10 nm) exhibited a strong inhibition on oxidization of extensive substrates by hydroxyl radicals, while the normal NBs (~100 nm) worked only for some substrates. Since the gas–water interface of the ultra-small NBs is non-expendable, its antioxidation would be sustainable and its efect be cumulative, which is diferent to that using reactive nanobubbles to eliminate free radicals as the gases are consumptive and the reaction is unsustainable. Therefore, our antioxidation strategy based on ultra-small NB would provide a new solution for antioxidation in bioscience as well as other felds such as materials, chemical industry, food industry, etc.

In living systems, antioxidation is one of the most concerned issues since reactive oxygen species (ROS) are usually produced persistently along with normal cellular metabolism1,2. However, excessive ROS ofen causes oxidative damage to a variety of important cellular components, including lipids, proteins, and DNA molecules3–6. Currently, various antioxidants have been suggested as dietary supplements to reduce ROS-associated diseases7. Te efectiveness of those antioxidants has been proven in the treatments of many oxidative damage-caused acute diseases8,9. However, in recent decades, most clinical trials in the treatments of oxidative damage-caused chronic diseases by the supplements of antioxidants have not provided convincing evidence for the clinical benefts10.
Badly, some antioxidants even have toxic side efects11–15, and most of them are non-sustainable in use and become unstable due to their sensitivity to normal environments16–21. Terefore, novel antioxidation strategies with high stability, sustainability, and biologically-safety are demanded. Gas–liquid interface has long been recognized to have unique physical, chemical, and biochemical properties. Recently, it has been employed to regulate many oxidation/reduction reactions. Some simulations and experimental evidence have shown that gas–liquid interfaces could enrich ROS and regulate the processes of their generation and quenching22–25, resulting in enhancing/inhibiting the substrate oxidation reaction by ROS. For example, Heath and Valsaraj26 studied the process of the enrichment of ROS and the reactants at the gas–liquid interface and found that the reaction rate was largely promoted by several orders as compared with that in bulk solutions. Nam and Richard27–29 found that oxidation or reduction would occur at the gas–liquid interfaces of small water droplets for diferent kinds of substrates. In these studies, the gas–liquid interface takes efect through the adsorption of ROS and/or substrates. Tus, if the surface area of a gas–liquid interface is so small that it prefers to enrich ROS but has insufcient space for larger substances, it may exhibit a certain antioxidant activity for a series of substrates. So far, the size efects of the gas–liquid interface on the reactivity have not been investigated like that of nanodroplets30,31.

Nanobubbles (NBs), typically as a nanoscale gas-phase suspended in the water phase32,33, can provide a large number of gas–liquid interfaces that may be employed for the enrichment of ROS. Te size of the NBs varies from ~10 nm (ultra-small NBs) to hundreds of nanometers (normal NBs); therefore, it is a suitable model to study the antioxidation or oxidation of a gas–liquid interface. Previously, it has been reported that oxygen NBs promoted the formation of ROS by producing hydroxyl radicals through the collapse of the microbubbles34, while the reductive hydrogen NBs helped the quenching of ROS35,36. However, in these studies, the chemical properties of the gas phases rather than the size of NBs were focused on, in which the gases in the nanobubbles are consumptive and would run out so that the redox reaction is unsustainable.

In this study, an antioxidation strategy based on ultra-small NBs without exogenous antioxidants was provided. A signifcant size dependence was observed when the NBs were employed to determine their ability to block the oxidization of substances by the hydroxyl radicals. It was found that the ultra-small NBs exhibited a strong antioxidant efect for extensive substrates, while normal NBs worked only for some substrates. Since the gas–water interface of the ultra-small NBs is non-expendable, its antioxidation would be sustainable and its efect be cumulative. We believe that this research would help develop new solutions for removing excess free radicals in a system without reductants supply.

Results

Antioxidation of ultra‑small N2 NBs. Te experiment was frst conducted by determining the antioxidant efect of the ultra-small nitrogen (N2) NBs by detecting their ability to block the oxidization of 3, 3′, 5, 5′-tetramethylbenzidine (TMB) caused by the hydroxyl radicals (Figures S1 and S2) generated from H2O2 with the catalysis of Cu2+. Ultra-small N2 NBs were generated in cold pure water (0 °C) during a compressiondecompression process37 and then were introduced to the oxidation-reaction system under room temperature and atmospheric pressure. Te oxidation curves were obtained by monitoring the absorbance at 652 nm of the oxidized product of TMB38. It’s worth noting the N2 NB itself has no detectable absorbance at 652 nm (Figures S3), and the redox potentials of N2 NB-containing water were similar to that of pure water (Table S1). Te results showed that the oxidation rates of TMB in water containing ultra-small N2 NBs were greatly reduced in comparison to that in pure water along with the increase of reaction time, and the absorbance values at plateau were much lower than that in pure water (Fig. 1a), which suggests a strong antioxidant efect of the ultra-small N2 NBs. In addition, a comparative study indicated that the antioxidant ability of the ultra-small N2 NBs was equivalent to a common antioxidant, sodium ascorbate, in a concentration between 100 and 200 μM (Fig. 1b).

In our experiments, nanoparticle tracking analysis (NTA) and dynamic light scattering (DLS) were employed as complementary means to determine the size distribution and concentration of nanobubbles in water39. By monitoring the Brownian Motion of a relatively small number of individual objects, NTA is able to accurately measure the concentration (106 –109 particles/mL) and size (10–2000 nm) of polydisperse populations40. Due to the low light scattering of NBs in water, NTA can test their size distribution in the range of 50–2000 nm, meanwhile determining their concentration. In the case of DLS, the collective difusion of a larger number of objects is monitored and their average size is calculated. However, DLS only provides a rough size distribution of samples ranging from 0.3 nm to 15 μm without concentration information41,42. Figure 1c (upper) showed a typical size distribution of the as-generated N2 NBs as measured by NTA, with the peaks mostly between 50 and 270 nm. NTA analysis also indicated a NB concentration of 5.42× 107±5.78× 106 particle/ml and an averaged NB size of 152.7±14.1 nm. Figure 1c (bottom) showed two peaks with very strong scatter intensity in the DLS curves, indicating that the sizes mostly centered at 3.62 nm and 255 nm, respectively. Te only peak observed in the DLS number percent curve (Fig. 1c, bottom) centered at 3.62 nm, suggesting that these ultra-small NBs made up the overwhelming majority in numbers in the solution.

A degassing experiment was carried out to rule out the possibility that the introduction of impurities during NB generation might have also caused the observed antioxidant efect. By removing most of the N2 NBs in water afer degassing for 24 h under a vacuum of 0.01 atm (Fig. S4), TMB oxidization curves (Fig. 1d) showed that the antioxidation ability of the N2 NBs water was signifcantly reduced, clearly confrming that the observed antioxidant efect was originated from the N2 NBs rather than from impurities.

Size dependence of the N2 NB’s antioxidant capability. Since the size of the N2 NBs generated was widely distributed in the range of 0–300  nm (Fig. 1c), it was plausible to explore if there would have a size dependence for their antioxidant capability. We found that the normal N2 NBs generated in fresh ultrapure water at room temperature did not inhibit but slightly enhance the oxidation of TMB (Fig. 2a). NTA study showed a typical size distribution of the normal N2 NBs between 70 and 220 nm (Fig. 2b, upper), a NB concentration of 6.41× 107±1.72× 107 particle/ml, and an averaged NB size of 116.9±14.7 nm. DLS study revealed two strong scattering intensity peaks centered at 142 and 396 nm, respectively (Fig. 2b, bottom). Both NTA and DLS results of normal N2 NBs showed no detectable NBs with sizes smaller than 50 nm, implying that the antioxidant efect was only caused by the ultra-small NBs (typically<50 nm). Besides, we found that the ultra-small N2 NBs transformed from normal N2 NBs through a freeze-thawing operation also exhibit an antioxidant efect (Fig. S5). In addition to the ultra-small N2 NBs, ultra-small oxygen (O2) NBs also have a strong antioxidant efect in the TMB oxidation reaction (Fig. S6).

The antioxidation mechanism for the ultra‑small NBs. Te above results clearly showed that there was a size dependence on the NB’s antioxidant capability. Ultra-small N2 NBs inhibited the oxidization of TMB by hydroxyl radicals, while their clusters or normal N2 NBs (typically>50 nm) slightly enhance the oxidation of TMB. Te contrasting efects of the small and large NBs on the TMB oxidation seemed difcult to be understood. Currently, our knowledge about the chemical properties of the interfaces of NBs is much poor, it is wise to interpret our observations based on the existing realizations regarding the regulation of oxidation and reduction by gas–water interfaces. Since the electrical surface potential diference of NBs is normally − 20 mV, far smaller than the 3 V at the gas–liquid interface of small water droplets28,43. Tus, it is not appropriate to explain our results from the electrical surface feld mechanism proposed by Nam and Richard. Previous studies have shown
that, when free radicals and substrates were both enriched at the gas–liquid interfaces, the oxidizing reaction
could be accelerated26,44. Terefore, we believed that the selective enrichment of ROS at the gas–liquid interface
of the NBs might play an important role in our reaction systems. A plausible explanation may be that the surface areas of the ultra-small NBs were so small and had insufcient space for larger substrate molecules to be
easily adsorbed, which resulted in the fact that it preferred to enrich more ROS but fewer substrate molecules.
Te short-lifetime hydroxyl radicals would be enriched at the interface and quenched by themselves (Fig. 3). In
contrast, the big surface area of the large NBs (or NB clusters) would enrich both the TMB and the hydroxyl
radicals at their gas–liquid interfaces, and enhance the reaction between TMB and hydroxyl radicals as usual.
Tis mechanism also works for another classic hydroxyl radical probe, 2,2’-Azinobis-(3-ethylbenzthiazoline6-sulphonate) (ABTS) (Fig. S7). In addition to the hydroxyl radicals, the ultra-small NBs were also found to
scavenge superoxide anion radicals (Fig. S8).

Antioxidation of N2 NBs for hydrophilic substrates. According to our proposed mechanism (Fig. 3), normal NBs enhance oxidation due to that they simultaneously adsorb ROS and hydrophobic TMB at the gas–liquid interface, which increases their reaction probability. If this is the case, normal NBs should also exhibit antioxidant efects when substrate molecules that tend to remain in the water phase rather than at the gas–liquid interface are used. To test this hypothesis, dimethyl pyridine N-oxide (DMPO), a commonly-used electron spinresonance (ESR) spin trap, was employed for capturing hydroxyl radicals. DMPO is hydrophilic so that it should present in the water phase. In this experiment, ESR was used to measure the intensity of the oxidized DMPO (DMPO-OH). Results (Fig. 4) showed that DMPO-OH signals in reaction systems containing normal N2 NBs or ultra-small N2 NBs were much lower than that of the control group, indicating an antioxidation effect. The results further support our mechanism.