Titanium dioxide

Endpoint summary

Description of key information

Micro-and nanosized TiO2 is not acutely and chronically toxic to aquatic organisms. Thus, nano- and microsized TiO2 are not a classified or non-classified acute and chronic hazard to aquatic organisms.

Additional information

Acute toxicity

 

Short-term toxicity of microsized TiO2:

Acute toxicity data are available for several aquatic fresh- and saltwater species covering three trophic levels (primary producers, primary and secondary consumers). The tables below provide an overview of the effect values (all unbounded) for acute intrinsic toxicity. For microsized TiO2 in freshwater and marine water, the trophic level most sensitive cannot be identified as only unbounded EC/LC50 values are available for fish, daphnids and algae.

 

Table 1: Overview of acute aquatic (intrinsic) toxicity data for titanium dioxide bulk material in freshwater:

Species	Parameter	Endpoint	Value (mg TiO2/L)	Reference
Pimephales promelas, Oncorhynchus mykiss, Danio rerio	mortality	96 h LC50	> 100 to > 1,000 (nominal)	Hutton et al. 1992 a-c, Turner et al. 2006, Mahjoubian et al. 2021
Oncorhynchus mykiss, Danio rerio	mortality	14 d LC50	> 0.87 to > 1.1 (measured)	Boyle et al. 2013, Ramsden et al. 2013
Daphnia magna	immobility	48 h EC50	> 100 to > 1,000 (nominal)	four studies*
Pseudokirchneriella subcapitata	growth rate (filtrate)	72 h EC50	> 100 (nominal)	Schlich et al., 2015

* Haley and Kurnas et al. (1993), Johnson et al. (1986), Wiench et al. (2009), Mendoza & Cerrillo (2016)

 

Table 2: Overview of acute aquatic (intrinsic) toxicity data for titanium dioxide bulk material in marine water:

Species	Parameter	Endpoint	Value (mg TiO2/L)	Reference
Cyprinodon variegatus	mortality	96 h LC50	> 10,000	Thomson et al. 2007
Acartia tonsa	mortality	48 h LC50	> 10,000	Thomson et al. 2007
Skeletonema costatum	growth rate	72 h EC50	> 10,000(nominal)	Hudson et al. (2007)

 

Short-term toxicity of nanosized TiO2:

For nanosized TiO2 in freshwater and marine water, the trophic level toxicologically most sensitive can also not be identified as only unbounded EC/LC50 values are available for fish, daphnids, algae and aquatic plants. The tables below provide an overview of the effect values (unbounded) for acute intrinsic toxicity.

 

The only bounded EC/LC 50 values from single studies relate to effects of dispersed nanosized TiO2 on the mobility of Daphnia magna (48 h EC50: 19.3 mg/L; 48 h EC50: 103.9 mg/L; 48 h LC50: 28.4 mg/L), Ceriodaphnia dubia (48 h EC50: 2.41-2.54 mg/L) and Artemia salina (48 h EC50: 284.8 mg/L), the survival of Hyalella azteca (96 h LC50: 631 mg/L) and the growth rate of Pseudokirchneriella subcapitata (72 h EC50: 415 -1028 mg/L):

 

i) observed at test concentrations above the limit test concentration of 100 mg/L defined in OECD 201 and 202 (Artemia salina, Daphnia magna, Pseudokirchneriella subcapitata), and/or

 

ii) that are presumably linked to mechanical effects and not to the intrinsic toxic properties of nanosized TiO2 (Artemia salina, Ceriodaphnia dubia, Daphnia magna, Hyalella azteca).

 

Non-dissolved TiO2 present in test media has the potential to exert physical effects on test organisms unrelated to toxicity. Since all bounded EC/LC 50 values of dispersed nanosized TiO2 are above the acute hazard classification criteria defined in Regulation (EC) No 1272/2008 and observed effects are presumably linked to mechanical effects, it is concluded that nanosized TiO2 is not acutely toxic to aquatic freshwater organisms.

 

Table 3: Overview of acute aquatic (intrinsic) toxicity data for titanium dioxide nanomaterial in freshwater:

Species	Parameter	Endpoint	Value (mg TiO2/L)	Reference
Carassius auratus, Oncorhynchus mykiss	mortality	96 h LC50	> 100 (nominal)	Ates et al. 2013, Turner et al. 2006
Oncorhynchus mykiss, Danio rerio	mortality	14 d-LC50	> 1.0 to > 1.1 (measured)	three studies*
Danio rerio	embryo mortality, hatching rate	96 h LC50	> 100 (nominal)	Clemente et al. 2014
D. magna, D. pulex, D. similis, C. dubia	immobility	48 h EC50	> 10 to > 1,000 (nominal)	six studies**
Pseudokirchneriella subcapitata	growth rate	72 h EC50	> 50 (nominal)	Nicolas et al. 2015
Lemna minor	growth rate	7 d EC50	> 100 (nominal)	Schlich, 2019

* Federici et al. (2007), Boyle et al.(2013), Ramsden et al. (2013)

** Wiench et al. (2009), Wyrwoll et al. (2014), Zhu et al. (2010), Griffit et al.(2008), Clemente et al. (2014), Mendoza & Cerrillo (2016)

 

Table 4: Overview of acute aquatic (intrinsic) toxicity data for titanium dioxide nanomaterial in marine water:

Species	Parameter	Endpoint	Value (mg TiO2/L)	Reference
Artemia salina	mobility	48 h EC50	> 100	Ates et al. 2013
Phytoplankton (multi- species)	growth rate	96 h NOEC	≥ 1 (nominal)	Miller et al. 2010
Dunaliella tertiolecta	growth rate	72 h NOEC	≥10 (nominal)	Morelli et al. 2018

 

 

Chronic toxicity

 

Long-term toxicity of microsized TiO2:

Chronic toxicity data are available for several aquatic freshwater species covering three trophic levels and marine algae. The tables below provide an overview of the effect values for chronic intrinsic toxicity. For dispersed microsized TiO2 in freshwater, the trophic level most sensitive cannot be identified as only unbounded NOEC values are available for long-term effects on fish, crustacea and algae.

 

For microsized TiO2 in marine water, a 72 h NOErC of 5,600 mg/L is available for the alga Skeletonema costatum (Hudson et al. 2007). However, effects were observed at concentrations more than fiftyfold above the acute test limit of 100 mg/L (OECD TG 201) and above the solubility limit of microsized TiO2. Hence, microsized TiO2 is not chronically toxic to aquatic organisms up to the acute OECD test limit of 100 mg/L and up to its solubility limit. 

 

Table 5: Overview of chronic aquatic (intrinsic) toxicity data for titanium dioxide bulk material in freshwater:

Species	Parameter	Endpoint	Value (mg TiO2/L)	Reference
Danio rerio	hatching success, time to hatch, larvae length	6 d NOEC	≥ 160 (nominal)	Shaw et al. 2016
Daphnia magna	reproduction, further chronic endpoints	21 d NOEC	≥ 10 (nominal)/ ≥ 2.3 (measured)	Campos et al. 2013
Pseudokirchneriella subcapitata	growth rate (filtrate)	72 h NOEC	≥ 100 (nominal)	Schlich et al., 2015

 

Table 6: Overview of chronic aquatic (intrinsic) toxicity data for titanium dioxide bulk material in marine water:

Species	Parameter	Endpoint	Value (mg TiO2/L)	Reference
Skeletonema costatum	growth rate	72h NOErC	5,600 (nominal)	Hudson et al. (2007)

 

Long-term toxicity of nanosized TiO2:

Chronic toxicity data are available for several aquatic freshwater species covering three trophic levels and marine algae. The tables below provide an overview of the effect values (unbounded) for chronic intrinsic toxicity. The only bounded NOEC/EC10 values from single studies relate to effects of dispersed nanosized TiO2:

 

i) on the length of Danio rerio (8 d LOEC: 1000 mg/L) observed at test concentrations above the limit test concentration of 100 mg/L defined in OECD 212 (Danio rerio), and

 

ii) on the growth rate of Pseudokirchneriella subcapitata (72 h EC10: 2.1 mg/L, nominal; 72 h EC10: 5 -35 mg/L) presumably linked to a removal of phosphate as nutrient from the closed artificial test system and subsequent nutrient deficiency and not to the intrinsic toxic properties of nanosized TiO2 (Nicolas et al. 2015; Hund-Rinke et al. 2016).

 

The absence of long-term toxicity of nanosized TiO2 to freshwater algae is further confirmed by a 32 d study of Kulacki et al (2013) on biofilms (algal polycultures).

 

The limited data on the effects in saltwater, i.e. two studies on marine algae, indicate that nanosized TiO2 is not hazardous to marine algae and cyanobacteria up to 10 mg/L (96 h NOEC:≥1 mg/L, Miller et al. 2010; 72 h NOEC:≥10 mg/L, Morelli et al. 2018).

 

Since all bounded NOEC/EC 10 values of dispersed nanosized TiO2 are more than twofold above the chronic hazard classification criteria defined in Regulation (EC) No 1272/2008 and/or observed effects are presumably due to an experimental artefact, it is concluded that nanosized TiO2 is not chronically toxic to aquatic organisms.

 

Table 7: Overview of chronic aquatic toxicity data for titanium dioxide nanomaterial in freshwater:

Species	Parameter	Endpoint	Value (mg TiO2/L)	Reference
Danio rerio	mortality/morphological effects (ELS)	8 d NOEC	≥ 1000 (nominal)	Faria et al. 2014
Danio rerio	Hatching success/larvae length	6 d NOEC	≥ 160 (nominal)	Shaw et al. 2016
Oncorhynchus mykiss	growth rate	28 d NOEC	≥ 0.07 (measured)	Zeumer et al. 2020
Daphnia magna	reproduction and further chronic endpoints	21 d NOEC	≥ 0.17 (nominal) – ≥ 100 (nominal)	three studies**
Chironomus riparius*, Lumbriculus variegatus*	emergence rate/ development rate, reproduction/biomass	28 d NOEC	≥ 100 (nominal)	Hund-Rinke et al. 2013 Schaefers & Weil, 2013
Polycultures freshwater	biomass	32 d NOEC	≥ 1 (nominal)	Kulacki et al. 2012
Lemna minor	growth rate	7 d NOEC	≥100 (nominal)	Schlich, 2019

* exposed via the water phase

** Campos et al. 2013, Hartmann et al. 2019, Hund-Rinke et al. 2013

 

 

Table 8: Overview of chronic aquatic toxicity data for titanium dioxide nanomaterial in marine water:

Species	Parameter	Endpoint	Value (mg TiO2/L)	Reference
Phytoplankton	growth rate	96 h NOEC	≥ 1	Miller et al. 2010
Dunaliella tertiolecta	growth rate	72 h NOEC	≥10 (nominal)	Morelli et al. 2018

 

Potential photoactivity:

Nano- and microsized titanium dioxide appears to have a low potential for acute and chronic toxicity to aquatic organisms as described below. However, there are indications that the photoactivity of uncoated TiO2 nanomaterials may lead to an underestimation of the environmental hazard as several studies report much lower effect concentrations of TiO2 in nano- and microsize under illumination. Only uncoated TiO2 materials show photoactivity, as a direct contact between the generated electron hole pair and water, hydroxide ions or oxygen is possible. This allows the formation of reactive oxygen species under illumination with the UV light included in solar radiation. Uncoated nanosized TiO2, however, enters aquatic systems only in ecotoxicologically not relevant concentrations as the majority is removed in sewage treatment plants (STP). Johnson et al. (2011) studied the fate of TiO2-NPs in an activated sewage sludge treatment plant serving over 200,000 people. Apparently, concentrations of Ti in the fraction < 0.45 µm differ significantly between influent and effluent. Whereas the tested unfunctionalized TiO2-NPs do not undergo removal during the primary settlement stage, the greatest decline (ca. 90%) followed the biological stage of activated sludge with mean Ti concentrations in the fraction < 0.45μm in raw sewage influent, following primary settlement and in final effluent of 30.5μg/L; 26.7μg/L; and 3.2μg/L, respectively. Results from a lab-scale sequencing reactor indicate that biological wastewater treatment plants operating with suspended biomass, such as activated sludge, could remove n-TiO2 from wastewater up to 97 ± 1% (Wang et al. 2012). The removal efficiency of STPs was confirmed by Kuhlbusch et al. (2012) who studied the fate of the nanosized TiO2 material P25 (21 nm), stabilized with sodium metahexaphosphate, in a laboratory STP according to OECD 303 A. Kuhlbusch et al. (2012) measured 4% of the initial mass of the applied TiO2 in the overflow after 22 d, mainly determined by particles in the fraction < 0.6 µm. However, results by Kuhlbusch et al. (2012) have to be interpreted carefully since the stabilizing agent may have influenced the fate of P25 in the laboratory sewage treatment plant. 

 

Furthermore, a direct entry of uncoated TiO2 nanomaterials via run-off from facades does not occur in ecotoxicologically relevant concentrations. Al-Kattan et al. (2013) exposed panels covered with paint containing either only microsized TiO2 or a mixture of bulk and nanosized TiO2 to simulated weathering by sunlight and rain in climate chambers. Both paints release limited amounts of materials (including material < 100 nm) into the environment during the study period (113 cycles of 6 hours). Background corrected Ti concentrations of both paints amounted to 1.5 µg/L in the first leachates and decreased further with each leaching event afterwards. The released concentrations stabilized at about 0.7 µg/L after 10 cycles resulting in a release of about 0.007% of the total Ti content of the paints. Measured Ti in the size fraction < 100 nm released from the paint containing nanosized TiO2 was more or less constant throughout the test and amounted to about 0.5 µg/L.

 

Kaegi et al. (2008) investigated the release of TiO2 from new and aged outdoor paints containing uncharacterised pigment-grade but not nanosized TiO2 particles. Apparently, the size of 85-90% of the total Ti in runoff samples from facades range from 20 to 300 nm without further size separation. Although the presence of incidental TiO2 nanoparticles (< 100 nm) was demonstrated in the runoffs by TEM-EDX analysis, information on the concentration, number or origin of the TiO2 nanoparticles (<100 nm) was not provided. Further shortcomings of the study include that background concentrations were not measured, a background correction was not performed and information on titanium particles in the rain was not provided. Thus, quantitative assessment of the release of nanosized TiO2 particles from new and aged outdoor paints is not possible based on the information published by Kaegi et al. (2008), and this study is not considered reliable.

Coated nanomaterials, such as Al(OH)3 coated TiO2 nanomaterials in sunscreens, are applied in personal care products that enter directly aquatic systems. However, the Al(OH)3 coating prevents photoactivity of the TiO2 core and is not removed by ageing under environmental relevant conditions (Labille et al. 2010, Auffan et al. 2010). Consequently, TiO2 nanomaterials coated with Al(OH)3 are not photoactive in natural environments and consequently are not expected to elicit phototoxic effects to aquatic organisms.

 

Thus, uncoated TiO2 nanomaterials do not enter directly the aquatic system via run-off from facades at ecotoxicologically relevant concentrations.

 

Conclusion:

In sum, micro-and nanosized TiO2 is not acutely and chronically toxic to aquatic organisms. Thus, nano- and microsized TiO2 are not a classified or non-classified acute and chronic hazard to aquatic organisms.