As mentioned in chapter 5 documents referring specifically to the application of nanoparticles for ecotoxicological tests are not available. For tests with daphnids and chironomids the method described by Hund-Rinke et al. (2010) was used, and supplementing studies con-cerning filtration and the use of stabilisers were performed.
6.1 Basic procedure
The method described by Hund-Rinke et al. (2010) was applied. For insoluble nanoparticles in powder form the required amounts were weighed in brown glass vessels (600 mL) using a suitable balance. Test medium was added, the mixture was stirred (magnetic stirrer, 900 rpm) and ultrasonified (3 min, 500 W) in a bath sonicator (Bandelin Sonorex RK 514 BH; 35 kHz; 215/860 W) filled with water to one third of the dispersion height in the bottles. For con-centrations in the range of 5 - 100 mg/L every concentration was prepared individually. For concentrations below 5 mg/L concentrated stock suspensions were prepared in most cases.
For the test with daphnids, a 20 mg/L stock dispersion was used.
For silver (NM-300K) stabilised in an aqueous medium suitable stock dispersions were pre-pared in the test medium. A homogenous distribution was achieved via stirring.
6.2 Filtration
The method to be applied for the exposure of daphnia was elaborated in pre-tests. According to the guideline, the test substance is dissolved in the exposure medium, the daphnia are added and the incubation is performed without any movement of the vessels or the exposure medium. This is a suitable procedure for soluble test substances. Nanoparticles, however, will sediment in tests without movement and the exposure concentration in the test media will decrease. Stirring or shaking during the incubation period is not advisable as daphnids and their reproduction rate are sensitive towards such procedures. Therefore, it was investigated whether a stabile dispersion can be achieved by elimination of the large agglomerates via filtration.
The following procedure was applied:
Two different filter types and two different filter devices were tested.
Filters: Mixed cellulose esters combined with filters of the disposal type Polycarbonate membrane filter combined with a filtration device using vacuum
In a first step the dispersion of TiO2 nanoparticles (P25; 10 mg/L) was filtered using a filter with a pore size of 0.45 µm (mixed cellulose esters). In a second step the filtrate was filtered again using a filter with a pore size of 0.2 (polycarbonate membrane filter) or 0.22 µm (mixed cellulose filter). In every fraction the Ti concentration was analysed.
In contrast, with mixed cellulose filters, almost no filtrate was obtained using polycarbonate membrane filters even when the filters were changed several times during the filtration proc-ess.
In Table 19, representative results from several filtration experiments with both filter types and filter devices are presented. In the first experiment different pore sizes were used for the two filtration processes in accordance with the scheme above. Although the primary particle diameter of P25 was 21 nm, almost no Ti was detected in the filtrate. Particles and agglom-erates mainly remained above the filter. The maximum recovery in the filtrate of step 1 was 0.01%. Using the filtration device with vacuum, in step 2 a slightly higher Ti concentration was measured than in the filtrate of step 1. It is assumed that some agglomerates of the first filtration process remained in the filter device although the device was carefully cleaned be-fore the second filtration process started. During the second filtration process the remaining agglomerates were rinsed in the new filtrate resulting in the increased Ti concentrations. For justification a second experiment was performed. In this experiment the dispersion was fil-tered twice with a filter of the same pore size. Again, a higher concentration of Ti was meas-ured in the second filtrate when using vacuum filtration.
Between replicate samples the results can differ by a factor of 2 - 4. For ecotoxicological ex-periments with concentration-effect curves, concentrations differing by a factor of 2 - 3 have to be investigated. As the concentrations of the replicate samples differed by a factor of 2 - 4 no clear dose-response curves were expected by investigating the filtrates. Furthermore, filtration is very time consuming and does not seem to be suitable for routine testing.
There-0.45 µm 0.22 µm
Analysis of Ti concentration in every fraction TiO2 dispersion
fore, the testing of filtrates was cancelled for the test with daphnids and original dispersions were tested.
Table 19: Ti concentration in the filtrate of a TiO2 dispersion after diverse filtration proc-esses.
TiO2 nanoparticles: 10 mg/L; Ti: 6 mg/L; filter material: mixed cellulose ester Filtration - Step 1 Filtration - Step 2 First experiment: step 1 and step 2: different pore sizes
Filter 0.45 µm cellulose mixed ester 0.22 µm cellulose mixed ester Filtration device using vacuum Sample 1: 0.62 µg/L
Sample 2: 0.33 µg/L
Sample 1: 1.27 µg/L Sample 2: 1.43 µg/L
Filter 0.45 µm cellulose mixed ester 0.2 µm polycarbonate membrane
filter Filters of the disposal type Sample 1: 0.17 µg/L
Sample 2: 0.17 µg/L
Sample 1: 0.26 µg/L Sample 2: 0.13 µg/L Second experiment: step 1 and step 2: same pore sizes
Filter 0.22 µm cellulose mixed ester 0.22 µm cellulose mixed ester Filtration device using vacuum Sample 1: 0.34 µg/L
Sample 2: 1.39 µg/L
Sample 1: 2.87 µg/L Sample 2: 2.24 µg/L
Filter 0.2 µm polycarbonate membrane
filter
0.2 µm polycarbonate membrane filter
Filters of the disposal type Sample 1: 0.55 µg/L Sample 2: 0.22 µg/L
Sample 1: 0.17 µg/L Sample 2: 0.26 µg/L
6.3 Stabilisers
The investigation of stabilisers was not within the scope of this project as stabilisers are as-sumed toxic. However, due to many discussions in the scientific community, and especially the recommendations of scientists dealing with the preparation of homogenous test sub-stances, the effect of the stabiliser sodium hexametaphosphate (0.01 %) was investigated.
To give a complete overview on the experiments dealing with application methods for nanoparticles performed at Fraunhofer IME, the results are presented in the following. Ex-periments focusing on the preparation of homogenous test suspensions used concentrations of about 1 % (10 g/L). In the test guidelines for aquatic ecotoxicological tests a maximum concentration of 0.01 % is fixed for the use of solubilisers for organic chemicals insoluble in water. In our project, the same limit was applied for the stabilisers used for nanoparticles.
The test was performed in Erlenmeyer flasks with TiO2 nanoparticles and algae as a growth test according to the OECD test guideline 201 (Hund-Rinke et al., 2010). The addition of the stabiliser resulted in stable suspensions. No sedimentation of the test substance occurred.
Although the stabilizer was applied in the maximum tolerable concentration mentioned for stabilizers in the test guidelines a toxic effect was still observed and growth of algae was significantly reduced . However, in the presence of the stabiliser, TiO2 nanoparticles (P25) affected the growth of algae less than without stabiliser (Figure 6). Bioavailability of P25 to algae is reduced in the presence of a stabiliser although the test suspension has a better homogeneity.
In the first step of hazard assessment, ecotoxicological tests must simulate a worst case scenario. The use of stabilisers at this stage should guarantee that the level of ecotoxicity does not decrease. As shown in Figure 6, the use of a stabiliser can reduce bioavailability.
Therefore, the application and the benefit need to be justified.
0 2000 4000 6000 8000 10000 12000 14000 16000
d0 d1 d2 d4
Incubation time [d]
Fluoreszence
Control Control + Stabilizer P25 P25 + Stabilizer
Figure 6: Effect of stabilisers (sodium hexametaphosphate, 0.01%) and P25 in the growth test with algae.