1. Surface Modification by Coating or Capping
The rationale for the surface modification of CuO nanoparticles by coating or capping is that the coating or capping may act as a scavenger for copper ions and reactive oxygen species
[1][2][3]. It may be noted, though, that coating or capping may also increase nanoparticle hazard mechanisms
[4]. An example thereof is the substantial increase in CuO nanoparticle toxicity to the green alga
Chlamydomonas reinhardtii and the aquatic macrophyte
Lemna gibba by coating CuO nanoparticles with styrene-butylacrylate co-polymer
[4]. Perreault et al.
[4] have suggested that this increase in toxicity is linked to changing nanoparticle interactions with cells and toxicity mechanisms.
In a safe-by-design study by Cai at al.
[2] CuO nanoparticles were coated with citrate, polyvinylpyrrolidone (PVP) and aminomethylphosphonate. Testing was for cytotoxicity in the (human) cell lines BEAS-2B and THP-1 and for inflammation of mouse lung tissue. The reduction in negative impacts by CuO nanoparticles was very limited when coating was with citrate. In the case of coating with PVP, the reduction in negative impact was moderate and in the case of phosphonate coating was relatively large, though not complete.
In a safe-by-design study, Fiandra et al.
[5] studied the impact on A549 human lung epithelial cells and
Xenopus laevis embryos of capping of CuO with polyethyleneimine (PEI) and polyethyleneglycol (PEG). They found that modifying the surface of CuO nanoparticles with PEG reduced hazard, but capping with PEI did not. As to the mechanism underlying reduced hazard, Fiandra et al.
[5] found in the case of exposure to CuO nanoparticles capped with PEG, that the generation of reactive oxygen species in cells was reduced. The presence of Cu ions in lung cells was higher in the case of PEI-capped CuO nanoparticles than in the case of their PEG-coated counterparts. Hazard was not eliminated by capping with PEG. Extrapolation of the findings of Fiandra et al.
[5] to humans and the different ways of intake by humans is subject to uncertainty
[6][7][8] and the extrapolation to other organisms that can be exposed to CuO nanoparticles is beset by uncertainties
[7][9].
In a safe-by-design study, Gosens et al.
[1] studied the impact of CuO nanoparticles surface-modified with ascorbate and PEI on short-term pulmonary inflammation in rats. These surface modifications had been tested before as to their cytotoxicity in a mouse macrophage cell line
[10]. In the latter study ascorbate-modified CuO particles scored best in reducing cytotoxicity. However, Gosens et al.
[1] found no significant differences as to toxic effects and toxic potency in the lungs of rats between the two surface modifications. This underlines the uncertainty in extrapolating outcomes of tests in cell lines to organisms. Inhalation hazard to rats was not eliminated by capping.
There is a set of studies testing the hazard of CuO nanoparticles coated with polyethylene glycol, carboxylate and methylaminated compounds, if compared with the hazard of pristine CuO nanoparticles, in a variety of biological settings
[3][11][12][13][14]. These studies will be briefly presented below.
Tatsi et al.
[11] used CuO nanoparticles with polyethylene glycol, carboxylate and methylaminated compounds in 14 days toxicity tests with earthworms. In fresh soil, CuO nanoparticles with a carboxylate and methylaminated coating were more toxic than pristine CuO nanoparticles, whereas PEGylated CuO nanoparticles had the lowest toxicity. In aged soil, Cu nanoparticles that had a methylaminated organic coating were more toxic than pristine CuO nanoparticles, whereas carboxylated and PEGylated CuO nanoparticles had (similar) lower toxicities than pristine CuO nanoparticles. Using CuO nanoparticles with the same coatings as used by Tatsi et al.
[11], Gajda-Meissner et al.
[12] concluded that coated CuO nanoparticles were more toxic in acute tests with
Daphnia magna than pristine CuO nanoparticles. Kubo et al.
[13] found that coatings of CuO nanoparticles with PEG and carboxylate reduced the cytotoxicity in the human cell lines THP-1 and HACAT, and that a methylaminated organic coating increased cytotoxicity. Extrapolation of these tests to human organisms is subject to uncertainty
[6][7][8]. Ilves et al.
[14] did show that pristine, methylaminated- and carboxylate-coated CuO nanoparticles strongly exacerbated allergen-induced lung inflammation in mice, but that the exacerbation was much less in the case of PEGylated CuO nanoparticles. Conolly et al.
[3] studied the effect on mussels (
Mytilus spp.) of pristine CuO nanoparticles and CuO nanoparticles coated with the same organic substances as used by Tatsi et al.
[11]. The focus was on gill cells, lysosomes and haemocytes. Genotoxicity affecting DNA in gill cells and haemocytes was found for both pristine and coated CuO nanoparticles. Based on acute toxicity to lysosomes and haemocytes, the hazard potential of PEG-coated CuO nanoparticles was found to be larger than for the pristine CuO nanoparticles. Chronic exposures suggested lower levels of oxidative stress associated with pristine CuO nanoparticles than with CuO nanoparticles that had carboxylate coatings.
Ribeiro et al.
[15] found that in an acute toxicity test coating or capping with organic substances (citrate, ascorbate, PEI and polyvinylpyrrolidone (PVP)) of CuO nanoparticles increased the negative impact on earthworm coelomocytes compared with pristine CuO nanoparticles.
In a safer-by-design study, Mendes et al.
[16] considered surface-modification of CuO nanoparticles with citrate, ascorbate, PEI and PVP. These modifications were tested in a mesocosm with six soil invertebrate species (consumers) during relatively long periods, up to 84 days
[16]. In this test, overall hazard was reported to be reduced by PEI, but actually increased by citrate and ascorbate modifications, whereas PVP had hardly any effect. Responses differed between species, suggesting species-specific response mechanisms. Mendes et al.
[16] stressed the importance of long-term testing to assess nanoparticle hazard and stated that multispecies testing increases the relevance to ecological hazard. Multigenerational tests would seem preferable because, as pointed out by Yu et al.
[17] and Gomes et al.
[18], there is evidence for transgenerational epigenetic effects. Extrapolation of these findings to other organisms that may be exposed to surface-modified CuO nanoparticles in a variety of environments is beset by uncertainties
[7][9][19].
In proving that CuO nanoparticles with coatings or cappings are safe or safer, the robustness of coatings or cappings under conditions encountered in their use stage and beyond should be considered
[7]. None of the studies discussed here addressed the robustness of coatings or cappings.
In the studies discussed here, the negative impacts of CuO nanoparticles can be decreased, increased and remain unaffected by coating or capping with organic substances. The series of papers regarding CuO nanoparticles coated with polyethylene glycol, carboxylate and methylaminated compounds (
[3][11][12][13][14] did show that, for a specific coating, toxicities can differ substantially across species and components thereof. PEGylated CuO nanoparticles did in several cases show a lower toxicity than pristine CuO nanoparticles
[11][13][14] but not in studies with
Daphnia magna [12] and mussel lysosomes and haemocytes
[3]. Tatsi et al.
[11] also found that relative toxicities may differ in different environments (fresh and aged soil). This underlines that in extrapolations to all organisms of findings showing a decrease in the negative impacts for specific organisms, parts thereof and derived cell lines are beset by uncertainties. None of the studies discussed here showed the elimination of CuO nanoparticle hazard by coating or capping. Furthermore, none of these studies addressed the impact of safe-by-design modifications on an antioxidant effect of CuO nanoparticles, though changes in antioxidant effects can be relevant to safety
[20].
2. Doping
In the Introduction it was noted that two studies
[21][22] claimed that safety-by-design was achieved by doping.
As pointed out in the Introduction, Naatz et al.
[21] studied the hatching inhibition of zebrafish embryos by Fe-doped CuO nanoparticles and found reduced inhibition of hatching, if compared with pristine CuO nanoparticles. Naatz et al.
[21] also studied the effect of Fe-doped CuO nanoparticles on cytotoxicity in BEAS-2B and THP-1 (human) cell lines. In addition, Joshi et al.
[23] studied exposure of (human) C6 glioma cell lines to Fe-doped and non-doped CuO nanoparticles. Both Naatz et al.
[21] and Joshi et al.
[23] found CuO nanoparticles were reduced by Fe-doping. As Joshi et al.
[23] did show that the generation of reactive oxygen species (and its associated hazard) was to be unaffected by Fe-doping, whereas the release of Cu ions from CuO nanoparticles was slowed, reduced cytotoxicity was ascribed to the latter effect. Pugazhandi et al.
[24] tested the impact of CuO nanoparticles doped with 3.6% Fe against three microbial species (two bacteria and one yeast). A substantial antimicrobial activity was found. The experiments presented by Naatz et al.
[21], Pugazhandi et al.
[24] and Joshi et al.
[23] did not show that CuO nanoparticle hazard was eliminated by doping with Fe. Taking into account the uncertainties besetting extrapolation
[6][7][8][9][19], these studies did not demonstrate that Fe-doped CuO nanoparticles can be safely used in the environment.
The induced generation of ROS by Mn
3O
4 nanoparticles as tested by Feng et al.
[22] was not eliminated by Zn doping. The aim of the doping experiments performed by Feng et al.
[22] has been to reduce the generation of ROS by keeping conduction band energy out of the biological redox potential range and having the edge of the Fermi energy (which dominates charge transfer) far away from the valence band energy edge. Feng et al.
[22] were successful as to the latter but not regarding the former. As the conduction band of CuO nanoparticles has been reported in the same range as biological redox potentials
[25], it would seem that the doping strategy used by Feng et al.
[22] is also unlikely to eliminate the generation of ROS induced by doped CuO nanoparticles. Feng et al.
[22] did not address the impact of Zn-doping on the release of metal ions from Mn
3O
4 nanoparticles, as they stated that these nanoparticles are insoluble. Insolubility, however, does not apply to CuO nanoparticles. In view of research presented by Ivask et al.
[26], Naatz et al.
[21] and Joshi et al.
[23], the release of Cu ions from CuO nanoparticles in biologically relevant settings is well established. It might also be pointed out that Katsnelson et al.
[27] and Illarionova et al.
[28] have presented evidence for the release of Mn ions from Mn
3O
4 nanoparticles in organisms and cell lines. Finally, the extrapolation of the experiments performed by Feng et al.
[22], regarding the cytotoxicity Zn-doped CuO nanoparticles to cell lines, to all organisms that can be exposed to CuO nanoparticles, is beset by uncertainties
[6][7][8][9][19]. The experiments presented by Feng et al.
[22] did not show that metal oxide nanoparticle hazard was eliminated by doping with Zn. The impact of doping with Zn on an antioxidant effect of metal oxide nanoparticles was not addressed by Feng et al.
[22].
3. Are the Modified CuO Nanoparticles Discussed Previously Safe?
None of the studies previously discussed in this section provided evidence for the elimination of hazard by safe(r)-by-design doped, coated or capped CuO nanoparticles. Still, one might argue that such elimination is not necessary for the absence of risk, as there may be exposure levels that do not give rise to negative impacts (no-negative effect levels) and exposure may remain below these levels. However, there is as yet no proof that this actually applies. Such proof is also complex. No-negative effect levels for modified CuO nanoparticles have not as yet been established. Their establishment is likely to be difficult, as Mendes et al.
[16] have shown that there are substantial differences in the response to coated nanoparticles between species, and it is also known that the toxicity of Cu ions (central to the safety claim of Naatz et al.
[21], shows large differences between species, between varieties and even between individuals
[29][30][31]. Furthermore, as pointed out above, there are currently no data about the robustness of CuO nanoparticle coatings or cappings in the real world. In the real world, no-negative effect levels in practice depend on the presence of other substances. There may be co-exposure to other nanoparticles, e.g., to ZnO nanoparticles, which may give rise to strong interactions
[32]. Furthermore, background-exposure data regarding other substances that induce the generation of ROS (central to the safety claim of Feng et al.
[22]) and can release Cu ions are needed to establish no-negative effect levels for modified CuO nanoparticles in the real world. Such exposure data are currently patchy at best. Finally, realistic fate and exposure studies are needed to show that actual exposure remains below no-negative effect levels. Such studies are currently lacking. It can be concluded that the safe(r)-by-design studies regarding the modified doped, coated or capped CuO nanoparticles discussed here do not prove that these modified nanoparticles are indeed safe. Providing such proof does not seem feasible in the near future, due to its complexity and the present lack of data.