Registration Dossier

Ecotoxicological information

Short-term toxicity to aquatic invertebrates

Currently viewing:

Administrative data

Link to relevant study record(s)

Description of key information

The acute toxicity to aquatic invertebrates was tested using titanium dioxide and titanium. Hence, for titanium carbide this endpoint is derived by read-across from titanium dioxide and titanium.
Reliable, publically available relevant data for TiO2 short-term toxicity to aquatic invertebrates do not report mortality in any of the species tested. These data are used in a weight-of-evidence approach. Due to similar or lower transformation/dissolution results for titanium carbide (the target substance) than titanium dioxide and titanium (the source substances), the resulting toxicity potential would also be expected to be similar or lower, so read-across is appropriate. Therefore, the dose descriptors are expected to be sufficiently similar or higher for the target substance, and read-across to the source chemical is adequately protective. For more details refer to the attached description of the read-across approach.

Key value for chemical safety assessment

Additional information

Griffitt et al. (2008) examined the effect of TiO2 nanoparticles (primary particle size distribution: 20.5 +/- 6.7 nm) to adult Daphnia pulex as well as to neonate Cerodaphnia dubia (< 24 h). Test organisms were exposed for 48 h to nominal concentrations of 0 (control) and 10 mg/L. Major particle diameters observed in suspension were 687.5 nm. 10 mg TiO2/L did not result in immobilisation in both species.

In a limit study according to OECD guideline 202 (Johnson et al., 1986) Daphnia magna was exposed to a single nominal concentration of 1000 mg TiO2/L for 48 h. Microscopic examinations of daphnids following exposure to TiO2 suspensions showed that the daphnids filtered the TiO2 particles, passing them to the gut. Thus combined oral exposure and exposure via the aqueous phase did not result in 50 % mortality. The 48 hour EC50 was > 1000 mg/L.


Apart from Johnson et al. (1986) also Wahrheit et al. (2007) investigated the effects of TiO2 on Daphnia magna after 48 h. While the particle size of TiO2 used in the test was not stated by Johnson et al. (1986), Wahrheit et al. (2007) used fine (defined as ~ 100 nm) and ultrafine (defined as < 100 nm) test material at nominal test concentrations of 0 (control), 0.1, 1.0, 10, 100 mg TiO2/L. Actual particle size of ultrafine and fine TiO2 was 140 nm and 380 nm, respectively. No toxic effects could be observed after exposure for 48 h.


This result is in agreement with the result of Lovern & Klaper (2006) who studied the short-term toxicity of TiO2 to Daphnia magna (U.S. EPA standard operating procedure 2024) with different TiO2 samples in two tests. TiO2 mixtures were prepared either by sonication or placement in tetrahydrofuran (THF) and filtration. Sonicated TiO2 samples were used at nominal test concentrations of 50, 200, 250, 300, 400 and 500 ppm, whereas lower concentrations of 0.2, 1, 2, 5, 6, 8 and 10 ppm were used for the assessment of filtered (THF) samples. The way the particles were prepared, however, influenced their toxicity: while unfiltered, sonicated TiO2 samples did not cause mortality or only 9 % effects at 500 ppm, filtered (THF) samples had toxic effects. Transmission-electron micrograph (TEM) images of the solutions show that the particles in the filtered solutions (filter size 200 nm) had a mean particle diameter of 30 nm, while particle size of sonicated TiO2 solution samples ranged from 100 to 500 nm. Thus, the study emphasises that the type of dispersion and size of the TiO2 particles may influence toxicity. Lovern & Klaper (2006) suggested that particle aggregation likely is the reason for the absence of effects in unfiltered, sonicated TiO2 samples.

The results obtained with the filtered (THF) TiO2 samples are not considered relevant for real-world environmental exposure due to the dispersion treatment of TiO2 before the toxicity test, and only the results for sonicated, unfiltered TiO2 samples (48-h EC50 > 100 ppm) are further considered in the assessment.


Short-term toxicity experiments with TiO2 (particle size: 100 nm and 200 nm) conducted by Dabrunz et al. (2011) did not result in mortality of Daphnia magna after the standard exposure duration of 48 h at the maximum concentration tested (8 mg/L) in any of the experiments. However, prolonged exposure up to 96 h led to an increase in toxicity. In addition, in experiments with prolonged exposure duration smaller particle size resulted in higher toxicity. Besides, in a prolonged experiment up to 96 h the authors observed “biological surface coating” of neonate daphnids with TiO2 particles. This biological surface coat completely disappeared with the first molting (shedding of shell) but reoccurred within 1 h after the first molting and continued steadily during the 96-h exposure period, causing a delay in molting and significantly lower molting success of only 10% compared to the control (p = 0.0065).

However, as the standard duration of short-term toxicity experiments with daphnids is 48 h, and substance assessment as well as classification and labelling are based on standard experiments with an exposure duration of 48 h, and toxic effects may mainly be attributed to physical effects, results obtained in the prolonged experiments are not carried forward to the effects assessment of titanium carbide. In addition, (eco-)toxicologically relevant release of Ti ions from titanium carbide is not expected as the concentration of soluble Ti ions was below the method detection limit (< 0.4 µg/L) in the T/D test. Thus, TiC is considered to be practically insoluble and the formation of insoluble Ti compounds due to Ti ion release form TiC with subsequent biological surface coating and physical effects is considered to be irrelevant.


Besides the publications considered as reliable (see above), additional publically available information of lower reliability is available on TiO2/Ti short-term toxicity to aquatic invertebrates. These data are considered supporting information only and are not decisive for the substance assessment:


Heinlaan (2008) investigated the 48-h toxicity of nano TiO2 (25–70 nm) to T. platyurus and D. magna according to Standard Operational Procedures “Thamnotoxkit F TM magna” (1995) and “Daphtoxkit F TM magna” (1996), respectively. Nanosize TiO2 did not induce mortality in T. platyurus at 20,000 mg/L whereas the same concentration induced 60 % mortality in D. magna. No toxicity could be observed at lower levels (incubation in the dark).

Effects (48-h) of TiO2 nanoparticles (50–150 nm) at 100 mg/L on Chydorus sphaericus were tested in a Chydotox test (Velzeboer et al., 2008). Under the conditions of the Chydotox test no mortality was observed.

Pérez-Legaspi and Rico-Martínez (2001) investigated the influence of Ti (speciation as well as test concentrations not specified) to neonate females of three rotifer species (Lecane hamata,Lecane luna, and Lecane quadridentata). Test organisms where exposed to the test substance for 48 h. The 48-h LC50 (nominal concentrations) for Lecane hamata, Lecane luna, and Lecane quadridentata are 15.6 mg Ti/L, 11.9 mg Ti/L and 8.5 mg Ti/L, respectively, demonstrating different susceptibility to the test chemical.

Borgmann et al. (2005) conducted a one-week toxicity tests using the freshwater amphipod Hyalella azteca. The 7-d LC50 of Ti (element) to Hyalella azteca was determined to be 0.979 mg Ti/L (nominal) or <0.272 mg Ti/L (measured).


The absence of ecotoxic effects in most of the experiments referenced in this section may at least partly be explained by low concentrations of small TiO2 particles in suspension due to aggregation and agglomeration (Lovern & Klaper, 2006; Velzeboer, 2008; Wahrheit et al., 2007; Griffitt et al., 2008). It is assumed that the actual concentrations of nano-sized TiO2 material in the tests were probably far lower than the nominal test concentrations applied. The results of Dabrunz et al. (2011) suggest that toxic effects that could be partly observed at very high concentrations and/or prolonged exposure duration is caused be physical effects.

Based on lower transformation/dissolution results for titanium carbide (the target substance) than titanium dioxide (the source substance) the resulting toxicity potential is also be expected to be lower. Therefore, the dose descriptors are expected to be sufficiently high for the target substance, and read-across to the source chemical is adequately protective. In fact, (eco-)toxicologically relevant release of Ti ions from titanium carbide is not expected as the concentration of soluble Ti ions was below the method detection limit (< 0.4 µg/L) in the T/D test. Thus, TiC in considered to be practically insoluble. Release of Ti ions to any ecotoxicologically relevant extent (and potential subsequent formation of soluble and/or insoluble Ti compounds) is not expected. Therefore, any toxic effects to aquatic invertebrates are not expected to arise from TiC.