Titanium dioxide

Bioaccumulation: aquatic / sediment

Description of key information

Several laboratory studies of the bioaccumulation of nanosized TiO2 by arthropods, annelids, molluscs and fish exposed via food, sediment or the water point to a low bioaccumulation potential. However, the accumulation kinetics were not addressed in several studies and often it remains unknown if steady-state was reached. In sum, available data on bioaccumulation of nanosized-TiO2 in invertebrates and fish indicate that nanosized TiO2 does not appear to bioaccumulate or biomagnify. A similar low bioaccumulation potential may be assumed for microsized material based on common physico-chemical properties including the poor solubility. 

Additional information

Microsized TiO2:

Based on ECHA’s Guidance on the Application of the CLP Criteria (Version 4.1, June 2015), hazards associated with metals and inorganic metal compounds may arise “when the substance is dissolved in the water column, exposure from this source is limited by the solubility of the substance in water and bioavailability of the substance to organisms in the aquatic environment. Thus, the hazard classification schemes for metals and metal compounds are limited to the acute and long-term hazards posed by metals and metal compounds when they are available (i.e. exist as dissolved metal ions, for example, as M+ when present as M-NO₃), and do not take into account exposures to metals and metal compounds that are not dissolved in the water column.”

Transformation/dissolution data (generated according to the standard protocol in Annex 10 to UN GHS, i.e. OECD Series on Testing and Assessment No. 29) of different microsized TiO2 materials indicate a low solubility in environmental media as dissolved (opperationally defined as the Ti fraction after centrifugal filtration (~2.1 nm)) Ti concentrations during and after 28 d were below the respective LOD/LOQ (< 0.11 / < 0.34 µg Ti/L). Thus, microsized TiO2 materials are poorly soluble and dissolution does not increase with time.

Data on bioaccumulation of microsized TiO2 in aquatic organisms do not exist, and the performance of such tests may not be feasible:

According to the CLP guidance Annex III.3.2 Bioaccumulation (version 4.1 – June 2015), bioaccumulation studies in which the test substance concentration is below the LOD of the analytical method are considered to be invalid, since problems will arise in interpreting the bioconcentration potential. Therefore, the performance of bioconcentration/bioaccumulation studies with microsized TiO2 is not feasible.

Given the low bioaccumulation potential of nanosized TiO2 a similarly low potential may be assumed for microsized material based on common physico-chemical properties including the poor solubility. 

Nanosized TiO2:

According to the recent ECHA guidance on aquatic bioaccumulation (Appendix R 7-2 Recommendations for NM applicable to Chapter R7c) it is not possible to predict the bioaccumulation of nanomaterials from the log Kow or solubility. Hence, BCF/BAF values have to be experimentally determined. Further, for nanomaterials that undergo dissolution it is recommended to obtain information on the form of the substance as present in the animal tissue. Since nano-TiO2 does not to dissolve under environmentally relevant conditions, a potential uptake is unlikely to be associated with the uptake of dissolved Ti forms. Thus, nanosized TiO2 (if any) could potentially only be taken up of via adsorption to biological surfaces, direct penetration, or endocytic processes such as via phagocytosis, pinocytosis, and caveolae-dependent or clathrin-mediated endocytosis.

Furthermore, Handy et al. (2012) critically mentions that nanomaterials form colloidal dispersions, which are dynamic systems, therefore a “steady-state” cannot be reached in bioaccumulation studies. According to the fish bioaccumulation OECD 305 (2012) guideline it is recommended to use the dietary approach, if stable concentrations of the test substance cannot be demonstrated. Therefore, the use of dietary exposure in bioaccumulation studies is recommended for nanomaterials such as nanosized TiO2: 

 

Two studies examined the bioaccumulation potential of nano-TiO2 in fish and in snails exposed via spiked food, consisting of previously exposed daphnids and algae:

 

In a reliable study, zebrafish were fed with daphnia previously exposed to 0.1 mg/L n-TiO2, and the resulting BMF value amounts to 0.024. Thus, TiO2 nanomaterials are not biomagnified in zebrafish since BMFs are well below 1. Zebrafish (Danio rerio) were further exposed to 0.1 mg n-TiO2/L (nominal) via the water phase for 14 days resulting in a bioaccumulation factor (BAF) of 25.38 (Zhu et al. 2013). However, the adsorption of n-TiO2 to the fish cannot be excluded. The BAF and BMF values derived from experiments with exposure concentrations of 1 mg/L n-TiO2 (Zhu et al. 2013) are not reliable since steady state according to OECD 305 was not reached.

 

In a supporting study, the uptake of nanosized titanium dioxide by algae biofilms consisting of Synedra ulna, Cenedesmus quadricauda and/or Stigeoclonium tenue as mono- or polyculture was measured by Kulacki et al (2012). Biofilms were exposed to 0, 0.1 and 1.0 mg/L n-TiO2 in the medium for 32 days. Furthermore, the uptake of nanosized TiO2 by Physa acuta grazing for 3 days on biofilms previously exposed for 32 days to n-TiO2 was examined. The Ti concentrations of algae (around 0.3 to 1.0 µg/mg dry mass) and snails (around 0.0030 -0 .0060 µg/mg dry mass) were different from concentrations of unexposed control biofilms and organisms feeding on unexposed control biofilms. Considering, that Ti tissue concentrations of snails after the depuration were lower than corresponding Ti concentrations of the biofilm, these results indicate that nanosized TiO2 does not biomagnify. Steady state was, however, not confirmed in the snail bioaccumulation study. Therefore, results are considered as supporting data.

 

In conclusion, these two reliable food-spiked bioaccumulation studies indicate a low bioaccumulation potential at different trophic levels and that nanosized TiO2 does not biomagnify in aquatic organisms, such as freshwater fish and snails.

 

These results are supported by a bioaccumulation study, demonstrating that nanosized TiO2 does not bioaccumulate in water-exposed sediment worms. The titanium concentrations of sediment worms (Lumbriculus variegatus) exposed to 100 mg/L n-TiO2 via the overlaying water for 28 days was not different than the concentrations of control worms (Schaefers & Weil, 2013).

Additionally, in a sediment-water Lumbriculus toxicity test using spiked sediment according to OECD test guideline 225, Lumbriculus variegatus were exposed to TiO2 nanoparticles (TiO2-uf-2; primary particle size: 19 nm) at 62.5 to 1000 mg TiO2/kg sediment dw for 28 d in artificial sediment (Simon, 2019). Accumulation of total Ti by Lumbriculus variegatus was evaluated at test end after a subsequent 6-h depuration phase. A reproducible ratio of 0.4 for total Ti concentrations of worm tissue at test end and mean measured concentrations of the sediment, based on dry mass, indicate that the potential for bioaccumulation of TiO2-uf-2 by Lumbriculus variegatus is very low. Considering that the worms were depurated for a brief period only, the Ti measured in worms may be due to ingested sediment.

Further supporting evidence is available from a study performed by Kuehr et al. (2021), who examined bioaccumulation and biomagnification of nanoparticles in benthic aquatic invertebrates, i.e., Hyalella azteca in a flow-through system using two different concentrations of TiO2 NPs (1.49 and 3.3 µg TiO2/L medium, time-weighted average). Regarding bioaccumulation, a significant increase in tissue concentrations was not observed in organisms exposed to high and low TiO2 concentrations. Measured total Ti concentrations of acid-digested animals, as equivalent to total TiO2 body burdens, were mostly below the limit of quantification (1.390 µg/L) and the limit of detection (0.463 µg Ti/L).

Regarding biomagnification, a slight increase in body burden of TiO2 was observed in animals treated with low dietary exposure to TiO2 NPs, leading to a biomagnification factor of 0.12. At the higher concentration, the body burden increased until 72 hours of exposure and remained stable afterwards (initial: 0.59 +/- 0.07 mg TiO2/kg; stable: 1.07 +/- 0.03 mg TiO2/kg), yielding a biomagnification factor of 0.02. However, due to the missing depuration phase for the lower test concentration and methodological shortcomings (no negative control) and insufficient documentation (not reported if mortality was below 10% and if the temperature validity criteria of ± 2°C was met), derived biomagnification factors are of low reliability.

Furthermore, several supporting water- or sediment-spiked bioaccumulation studies of TiO2 nanomaterials exist. However, in these studies the accumulation kinetics were not addressed, it is not clarified if steady state was reached, organisms were not depurated before body burdens were measured, or the validity of the test was not confirmed as described further below:

Hyalella azteca (7 - 8 d old) were exposed to the nanosized TiO2 material P25 (25 nm) via natural sediment (20 and 100 mg/L sediment, nominal) in a chronic sediment toxicity test according to EPA 600/R-99/064 guideline under laboratory light and simulated solar radiation. The body burden of organisms exposed for 21 d and subsequently left to purge their guts for further five days in untreated natural lake water was with 14 µg/mg not different from control burdens (Wallis et al. 2014).

Exposure of water fleas (Daphnia magna) to 0.1 and 1.0 mg/L n-TiO2 for 24 h resulted in bioaccumulation factors (BAFs) of 56563 L/kg and 118063 L/kg, respectively, for non-depurated organisms. Thus, BAFs include TiO2-NP in the digestive tract (Zhu et al. 2010). When daphnia were fed with uncontaminated algae in a more realistic scenario, body concentrations were significantly lower resulting in a BAF of 1232 L/kg at 1.0 mg/L n-TiO2. Further, after the 72 h depuration, body concentrations decreased significantly. Unfortunately, adsorption of TiO2-NP to the Daphnia cannot be excluded and BAFs are only reported for non-depurated organisms.

In a bioaccumulation study by Kuehr et al. (2019), freshwater bivalves (Corbicula fluminea) were exposed to dispersions of titanium dioxide nanoparticles (NM-105, anatase/rutile, primary particle size: 21 nm) at two concentrations in a newly developed flow-through system for 120 h with a subsequent depuration period of 144 h. The initial total TiO2 concentration in the bivalves' soft tissue increased during the uptake phases from 0.22 mg TiO2 per kg to a plateau level of around 0.6 mg TiO2 per kg after 24 h of the uptake phase (lower concentration treatment) and to around 5.4 mg TiO2 per kg after 96 h of the uptake (higher concentration treatment), respectively. Plateau concentrations remained stable until the end of the uptake phase (120 h), resulting in BAFss (steady state) values of 6150 (lower concentration) and 9022 (higher concentration). During the following depuration phase, the TiO2 body burden of the bivalves decreased rapidly to stable concentrations of 0.22 mg TiO2/kg after 16 h of depuration (lower exposure concentration) and 0.16-0.17 mg TiO2/kg after 24 h depuration (higher exposure concentration). Derived BAF values should be considered carefully, since a negative control was not performed and background Ti concentrations were not determined. Actual total exposure concentrations were 0.099 and 0.589 µg TiO2/L (TWA) corresponding to 0.059 and 0.353 µg Ti/L (TWA), respectively, and were thus around lower background levels of Ti in European freshwaters (5th and 95th percentile: 0.1 and 4.2 µg Ti/L in fraction < 0.45 µg/L (Salminen, 2005)). Furthermore, even though it cannot be definitively differentiated between incorporated and ingested particles, localisation of particles in digestive tracts and viscera as well as the fast and almost complete depuration indicate that TiO2 NPs were only ingested.

In a non-guideline, environmental monitoring study, specimens of oysters (Crassostrea gigas), mussels (Mytilus edulis), scallops (Chlamys farreri), clams (Ruditapes philippinarum) and ark shells (Scapharca subcrenata) were obtained from an aquaculture facility in Hongdao, Jiazhou Bay (Qingdao, China) and analyzed for content of metals (Ti, Zn, Cu & Ag). Mass-based total titanium concentrations ranging from 1.11 to 1.69 µg/g fresh weight are reported in mollusc tissues. However, mass-based concentrations in surrounding seawater and/or sediment are not reported, therefore the bioaccumulation potential cannot be assessed. The authors further report number-based concentrations of 0.65 ± 0.13 *10^7 Ti-bearing nanoparticles (mean size < 100 nm) per mL in seawater and 2.1 ± 0.31 to 2.76 ± 0.2 *10^7 Ti-bearing nanoparticles per g in the molluscs. The authors conclude that molluscs concentrate Ti-bearing nanoparticles in their bodies and respective number-based concentration factors are presented in the publication. However, from the presented results it is not clear if the nanoparticles are truly incorporated into the molluscs’ tissues or are attached to the surface. Hence, it is not clear if the specimens really accumulate Ti-bearing nanoparticles in their tissues or if the nanoparticles aggregate within the shells of the animals without being incorporated and accumulated in the tissues. Especially for digestive organs and gills it can be assumed that particles are attached to the surface since these organs are in direct contact with the surrounding seawater.

Schirmer et al. (2014) measured the uptake and depuration of the radioactively labelled nano-TiO2 material P25 ([48V]TiO2, 28 nm) in the aquatic nematode Plectus aquatilis during a 24 h lasting uptake (1.3 x 10-4 M TiO2) and a 24 h lasting depuration period (0.0 M TiO2) in a non-standardized bioaccumulation test. Unpurged, equilibrium TiO2 tissue concentrations amounted to around 0.012 kg TiO2/kg nematode (graphical estimate) after the uptake period of 24 h and around 0.002-0.004 mg TiO2/kg nematode remained in the nematodes after the depuration period of 24 h. Reported BAF values cannot be related to the P25 material and TiO2 water concentrations were not reported. Hence, no valid BAF values can be derived from this study. Nematode TiO2 tissue concentrations in the uptake phase represent TiO2 tissue concentrations of unpurged organisms, since organisms were not purged before analysis.

Uptake and depuration of the radioactively labelled nano-TiO2 material P25 ([48V]TiO2, 28 nm) in the aquatic nematode Plectus aquatilis was measured during a 24 h uptake (1.3 x 10-4 M TiO2) and a 24 h lasting depuration period (0.0 M TiO2) in a non-standardized bioaccumulation test by Isaacson et al. (2017). An uptake rate constant of 11 L/kg/h as well as an elimination rate constant of 0.45 /h for the first 2 h of the elimination phase and an elimination rate constant of 0.0065 /h after 2 h of elimination. The reported BCF(kinetic), which has not been growth-corrected, was calculated to be 24 L/kg and the BCF (steady state) was calculated to be 9.7 L/kg. However, TiO2 concentrations in water and temperature regime during the experiment were not reported. The derived BCF values can thus not be considered reliable.

Isotherms of the adsorption of the nanosized TiO2 material P25 (20-30 nm) to eggs of Danio rerio were determined at equilibrium after 8 h of exposure (0, and 2.5-120 mg/L, nominal), either by directly digesting exposed eggs (treatment 1: rigidly attached and loosely attached TiO2) or by rinsing eggs before digesting them (treatment 2: rigidly attached TiO2). Adsorption of TiO2 to eggs was characterized by a three-step process of rapid initial layer formation, followed by break-up of aggregates and finally rearrangement of floc structures. The adsorption isotherms of treatment 1 and 2 were described by Dual-Langmuir and Langmuir equations, respectively, resulting in maximum adsorption capacities of inner rigid layers of 0.81-0.84 µg TiO2/egg and outer softly flocculated layers of 1.01 µg TiO2/egg. The Langmuir adsorption constants (Kl) for the inner layers were 7.72 L/mg (treatment 1) and 7.61 L/mg (treatment 2), and for the outer layer it was 1.41 L/mg (treatment 1) (Shih et al. 2016).

Whole body concentrations of > 20 < 40 ng/embryo (zebrafish, Danio rerio, 8dpf) were measured after exposure in water to different TiO2 NPs at 1 mg/L for 8 days followed by 24 h of depuration. Concentrations after a similar exposure to microsized TiO2 resulted in < 20 ng Ti/ embryo indicating that accumulation of TiO2 in nano- and microsize are somewhat comparable (Faria et al, 2014).

Titanium concentrations in various tissues of Oncorhynchus mykiss (i.e., liver, muscle, brain and gills) remained constant over the tested concentration range of TiO2-NP in water (0 - 1 mg TiO2/L), resulting in decreasing BAFs with increasing TiO2 concentrations (Federici et al. 2007). Significant differences in Ti concentrations of stomach, intestine and carcass of juvenile rainbow trout exposed to wastewater-borne, effluent-supplemented or water-dispersed TiO2 NP(up to 50 µg Ti/L) were also not observed in a 28-d juvenile growth test when compared to the respective dilution water control (Zeumer et al. 2019). Similar trends were observed by Ates et al. (2013) for gills and intestine whereas titanium could not be detected in muscle and brain of goldfish (Carassius auratus). Furthermore, Ramsden et al. (2009) exposed juvenile rainbow trouts (Oncorhynchus mykiss) to titanium dioxide nanoparticles (Aeroxide P25, 24.1 ± 2.8 nm) via spiked fish food (5.4 and 53.6 mg Ti/kg food, measured) for 8 weeks with a subsequent depuration phase of 2 weeks in a recirculating system with dechlorinated tap water, without following a specific test guideline. Significantly different but relatively low levels of titanium in the gill, intestine, liver, brain and spleen of Oncorhynchus mykiss were measured after dietary exposure of fish to both nano-TiO2 food concentrations in comparison to the control fish during the 8 weeks uptake phase and after the depuration period in gills, intestine, liver and brain. Steady-state was not confirmed in the test after 8 weeks of exposure and it was further not possible to derive kinetic biomagnification parameters as only one time point was sampled during the depuration period. Nevertheless, uptake in fish occurs only in the lower nmol/g dw range (max. 20 nmol/g dw= 0.95 mg/kg based on graphed data and tissue-specific BMF appear to be < 1.

In sum, available reliable data on bioaccumulation of nanosized-TiO2 in invertebrates and fish indicate that nanosized TiO2 does not bioaccumulate or biomagnify.

 

References:

Handy R, van den Brink N, Chappell M, Mühling M, Behra R, Dušinská M, Simpson P, Ahtiainen J, Jha AN, Seiter J, Bednar A, Kennedy A, Fernandes TF, Riediker M. 2012. Practical considerations for conducting ecotoxicity test methods with manufactured nanomaterials: what have we learnt so far? Ecotoxicology. 21:933-972.