The D. magna used in this study was provided by the Korea Institute of Toxicology (Daejeon, Korea). The D. magna was cultured and treated as described in the US Environmental Protection Agency (US EPA) manual [10]. D. magna cultured and raised in a chamber were retained at 20±1℃ in 2 L beakers containing 1.5 L of hard, reconstituted water (0.008 g L^-1 KCl, 0.12 g L^-1 CaCO₃, 0.12 g L^-1 MgSO₄, and 0.192 g L^-1 NaHCO₃ in distilled water filtered through a Minipore Milli-Q apparatus). For performance of the reproduction tests the pH of the medium was adjusted to 8.2±0.2. Cultures were maintained under an 18 hours light: 6 hours dark photoperiod. The medium for the D. magna culture was changed three times per week with fresh HRW(hard reconstituted water) and fed with green algae (Selenastrum capricornutum) and YTC (a mixture of yeast, cerophyll, and Trout chow) purchased from Aquatic Biosystem Inc., (Colorado, US). D. magna culture was maintained in 2-liter glass beakers containing 1.6 L of cultured medium and 30-50 daphnids.

The acute toxicity test was performed according to the standard protocol for D. magna acute test [10]. Approximately 30 neonates aged less than 24 hours were separated in to three groups and exposed to titanium (iv) oxide (Sigma, US) with different sizes and several concentrations for 96 hours in a static test. All test concentrations were prepared using a volume of 50 mL in triplicate and maintained at 20±1℃ during a 96-hours photoperiod of 16 hours light: 8 hours dark without any feeding. Test beakers were covered with vinyl wrap.

The chronic toxicity test was performed according to the standard protocol through D. magna chronic toxicity testing [11]; neonates aged less than 24 hours at the start of the test were exposed to approximately three concentrations of each size of the nanoparticles for a period of 21 days. Neonates were fed two to three times per week, and the media and nanoparticle solution were changed, simultaneously. Survival and offspring were observed and monitored every day. The test beakers were maintained in a chamber at 20±1℃ during a 96 hours photoperiod of 16 hours light: 8 hours dark.

All of the data were obtained from three independent samples carried out simultaneously for error analysis, and the results were shown along with the standard deviation and the correlation between the mortality under several experimental conditions. The data were analyzed using Sigma Plot (SPSS Inc., Chicago, IL. USA). A p value < 0.05 was considered significant.

We performed the acute toxicity test for 96 hours with various concentrations and sizes of TiO₂ particles. The acute experiment determined the concentration range to be used in the survival experiments. Sodium phosphate buffer (pH 6, 0.1 M) was used as the control, and its mortality was shown in Figure 1. All sizes and concentrations of TiO₂ were not acutely toxic over 48 hours. After 96 hours, four neonates treated with anatase (250 nm) and rutile (500 nm) particles at 8 mM survived, but of those treated with anatase (21 nm) at 0.75 mM, four survived. From these results, we determined that the higher concentration led to a higher mortality (see all graphs). Furthermore, the smaller size led to a higher mortality (Figure 1c). Given these results, we found that the regression curves of the acute toxicity data, LC50 and LC20 were determined as the lethal concentrations of a chemical when the percentage of D. magna mortality was 50% and 20%, respectively. In this study, D. magna was exposed to two different LC values (LC20, LC50) of anatase (250 nm) 1.29 and 3.47 mM, respectively, for 96 hours to examine the effects of the anatase form (250 nm) of TiO₂ (Figure 2a). Likewise, characteristics of the rutile (500 nm) and anatase (21 nm) varieties were measured. The LC20 and 50 of the rutile (500 nm) and anatase type (21 nm) particles were 1.98, 5.94, 0.19, and 0.42 mM, respectively (Figure 2b, c). Based on the above results, the D. magna toxicity of the three types of TiO₂ particles decreased from anatase (21 nm) to anatase (250 nm) to rutile (500 nm) particles.

In addition, the chronic toxicity over 21 days using various concentrations and sizes of TiO₂ particles determined in this study. Moreover, we used and compared the concentrations (0, 1, 10 mM) in all chronic toxicity tests. We measured the concentration- and size-dependent survival of neonates over 21 days. Survival of anatase (250 nm) and rutle (500 nm) were observed at approximately 10, 40% at 1 mM (Figure 3a, b). In contrast, the survival for 1 mM of anatase (21 nm) particles was not measured because there were killed after the cells were exposed to the solution for 16 days (Figure 3c). From these results, we determined that the toxic response of anatase (250 nm) and rutile (500 nm) progressed slowly, but the response of anatase (21 nm) particles progressed rapidly. There were more than 120 live offspring of the control after 21 days of testing (Figure 4a). That of the anatase (250 nm) and rutile (500 nm) particles with 1 mM was approximately 80 and 20%, respectively. However we did not find any neonates after exposure to anatase (21 nm) particles at 1 mM (Figure 4). Further, we found that no offspring were found at the 10 mM concentration of additional sizes of TiO₂ (Figure 4c). After finishing the chronic toxicity tests, we measured the size of neonates living until last day using microscope. Averagely, the size of neonates with 0, 1, and 10 mM anatase (250 nm) materials is 4, 3, and 2 centimeters, respectively. For rutile (500 nm) nanoparticles, the sizes are 4, 2.5, and 2 centimeters, respectively (Figure 5). However, the size of anatase (21 nm) particles could not be measured because all neonates died.