Report on E171 (titanium dioxide) and E172 (iron oxide) - analytical perspectives from research

Submitted on behalf of TDMA & TC E172 1./2. April 2019 Parma Report on E171 (titanium dioxide) and E172 (iron oxide) - analytical perspectives from research Five E171 grades were analysed by 25 E172 grades were analysed by SEM, BET, Brookhaven XDC and SEM, TEM, BET, Brookhaven XDC CPS DC and LD Sample B Sample A Yellow 3 Red 11

Sample C Sample D Red 2 Black 1

Sample E Sample Anatase (%) Product Name Colour Crystal Phase A 99,9 Yellow 3 Yellow Goethite FeO(OH)

B 99,7 Red 2 Red Hematite Fe2O3 C 99,6 Red 11 Red Hematite Fe2O3 D 99,1 Black 1 Black Magnetite Fe3O4 E 99,4 2 SEM results E171 E171 is not a nano product • SD low for E171 Manufacturer (M1, M2, M3) • Inexperienced laboratories could find very different results

60 0.250 50 0.200 40 0.150 30 20 0.100 10 0.050 Q0<100nm (%) Q0<100nm 0

median x50 (µm) x50 median 0.000 A B C D E A B C D E M-1 SEM 18 46 36 10 44 M-1 SEM 0.137 0.105 0.115 0.163 0.106 M-2 SEM 18 50 37 10 47 M-2 SEM 0.140 0.101 0.110 0.173 0.102 M-3 SEM 19 41 36 14 44 M-3 SEM 0.138 0.108 0.113 0.162 0.105 60 0.250 50 0.200 40 0.150 30 0.100 20 0.050

Q0<100nm (%) Q0<100nm 10 0 (µm) x50 median 0.000 A B C D E A B C D E LAB A SEM 4 5 3 0 22 LAB A SEM 0.179 0.162 0.165 0.222 0.142 LAB A STEM 12 12 19 7 42 LAB A STEM 0.145 0.145 0.126 0.183 0.12 LAB B TEM 29 55 41 15 59 LAB B TEM 0.127 0.095 0.109 0.164 0.095

3 TEM vs BET Correlation x50 with dminVSSA high

Yellow 1 • VENATOR is the sole manufacturer

800 100% of E172 700 80% 600

500 60% • Primary particle size analysis 400 300 40% possible for red and yellow but not 200 20% 100 for black Iron Oxides 0 0% • Rubout of crystals in a monolayer on Frequency Cumulative % the surface is important for accurate Red 11 results 100 100% 80 80% • Will be difficult to find an external lab 60 60%

40 40% which is able to measure Iron Oxides 20 20% • High correlation with dminVSSA 0 0%

1 x50 feret.min vs dminVSSA Frequency Cumulative % x50 ECD vs dminVSSA Linear (x50 feret.min vs dminVSSA) Red 2 Linear (x50 ECD vs dminVSSA)

300 100% 250 80% 0.1 200 60%

150 x50(µm) 40% 100 50 20% y = 0.945x - 0.0049 y = 1.0082x + 0.0047 0 0% R² = 0.9975 R² = 0.9865 0.01 0.010 0.100 1.000 Frequency Cumulative % dminVSSA (µm) 4 Particle Size for Risk Assessment? Constituent Particles

 We support the EFSA approach to analyse the particle size of food additives as pristine material, in the food matrix and in biological media  However, the guidance document is not consistent on page 21 in the evaluation of agglomeration or aggregation state by two independent methods  Standard EM can only measure the constituent particles size and not the agglomeration or aggregation state  The second method must corelate with EM to be a suitable screening method and the sample must be dispersed to a plateau  Nevertheless, for many products (Black Iron Oxides) EM doesn’t work and alternatives must be possible  The Nano Define decision-flow scheme for dispersion criteria on page 79 should be adjusted by the dispersion technique and mandatory correlation with EM

5 Mobile Particle Size for Risk Assessment Defined Dispersion Energy

 “Once a material is classified as nano according the constituent particle size of the EU definition, the information on the mobile particle size under realistic conditions (food matrix, biological media, etc.) is needed.  Dispersability has been reported as a founding base for the grouping of the nano materials” (EChA, “Appendix R.6-1 for nano materials applicable to the Guidance on QSARs and Grouping)  The OECD TG318 for aqua toxicity could be a good base for food application, a defined dispersion energy according to NIST 1200 is recommended and indicates the particle size in aqueous media  The same issues apply for food, what is the right dispersion energy, what is the dispersion medium, degradation, dissolution…….  An initial approach for the determination of the E171 and E172 smallest mobile particle size of the pristine material or in the food matrix is presented on the next slides

6 Volume weighted Particle Size Standardised dispersion energy

Dispersion with low energy Particle size control of E171 and E172 by a standardised dispersion procedure:

. Comparing production dispersion energy density with laboratory scale using NIST 1200 TG Aggregate size is determined by the . The energy density of 240 J/ml is production process (page 17 line 14) three times higher than normally applied in production wet milling of

TiO2 DispersionEnergy . Full redispersion of agglomerates without breaking up aggregates or Not possible to disperse completely to constituent particles unbounded constituent particles . M. Stintz propose 270 J/ml for synthetic amorphous silica (SAS), Powder Technology 318 (2017) 451- 458 and Nanomaterials 8 (2018) 454

7 Comparison E171/ E172 Volume based particle sizes for risk assessment . “Dispersibility has been reported as a founding base for the grouping of the nano materials” (EChA, “Appendix R.6-1 for nano materials applicable to the Guidance on QSARs and Grouping) Brookhaven XDC E172 E172 10.000 Dispersion Energy 240 J/ml 10.000

3.373 5 (µm)

50 2

1 1.000 1.000

Volume X Volume 0.328

0.261 0.253 Volume Nano (%) Nano Volume 0.100 0.100 Yellow1 Red2 Red11 Black1 Yellow1 Red2 Red11 Black1 E171 E171

10.000 10.000

(µm) 1.3

50 1.1 1.000 1.000 0.8 0.8 0.369

0.280 0.264 0.267 0.304 Volume X Volume

Volume Nano (%) Nano Volume 0.1

0.100 0.100 A B C D E A B C D E

8 E171/ E172 Conclusion on Particle Size Analysis

 E171 is not a nano material according the EU definition  E172-Black Iron Oxides are not nano products  The dispersion criteria of the Nano-Define decision-flow scheme shall be corrected for “constituent particles”  Realistic volume based particle sizes for toxicological testing can be achieved by dispersion energies <300 J/ml  Using the NIST 1200 TG volume based particle sizes of different products can be easily compared by different laboratories  Implementation of a dispersibility criterium in the risk assessment  CPS DC and Brookhaven XDC are suitable instruments to measure mobile nano particles

9 Back Up

10 E172 Particle Size by Brookhaven XDC & LD Absolute particle size depends on the method

. Dispersion Energy 240 J/ml . Same trend for both methods 1.000 Average XDC X50 Volume (µm) Average MS2000 X50 Volume (µm) 0.900 0.900 0.870 0.800

0.770 0.747 0.700 0.717 0.690 0.690

0.600 0.640 0.610 0.623 0.590 0.600 0.580 0.570 0.500 0.527 [µm] 0.500 0.493

0.400 0.425

0.300 0.319 0.328

0.261 0.253 0.253 0.236 0.234 0.200 0.220 0.192

0.100 0.145

0.000 Yellow1 Yellow2 Yellow3 Red1 Red2 Red3 Red4 Red5 Red6 Red7 Red8 Red9 Red10 Red11 Geothite Hematite

11 E172 Nano Content by XDC & LD Good sensitivity for high nano contents

• Brookhaven XDC is more sensitive to fine particles than LD

14 Average XDC Q3 (%) <100nm Averagen MS2000 Q3 (%) <100nm

12 12.0

10

8 8.4 [%]

6 6.0 5.7 5.0 5.0 5.0 4

3.0 2 2.0 1.7 1.2 0.1 0.7 0.0 1.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 0 Yellow1 Yellow2 Yellow3 Red1 Red2 Red3 Red4 Red5 Red6 Red7 Red8 Red9 Red10 Red11 Geothite Hematite

12 Long Dispersion experiment No complete dispersion of aggregates after 3h

3600 J/ml 7200 J/ml 10800 J/ml

Anatase 1h Horn Sonication Anatase 2h Horn Sonication Anatase 3h Horn Sonication SomeAnatase Crystal Debris 1h probe sonication MoreAnatase Crystal Debris 2h probe sonication IncreasedAnatase Crystal 3h probeDebris sonication some tip debris more tip debris increased tip debris aggregates still survived 13 E171 Long Dispersion experiment Particle size analysed by CPS DC

Number Median x50 (µm) Q0.<0,1µm (%) Volume Median x50 (µm) Q3<0,1µm (%) 100

28.3 28.5 30.1 28.9 28.4 31.9 30.2 23.2 22.9 27.2

10

3.48 3.28 2.59 2.82 2.88 3.02 1.98 2.26 1.31 1.57 1

0.29 0.28 0.27 0.26 0.24 0.24 0.24 0.23 0.23 0.23 0.15 0.14 0.14 0.13 0.13 0.13 0.13 0.13 0.13 0.1 0.12

SEM number feret.min: SEM number ECD: Median x50 = 0,100 µm Median x50 = 0,120 µm Q0<100nm = 47% Q0<100nm = 33% 0.01 120 300 600 900 1800 3600 5400 7200 9000 10800 2 5 10 15 30 60 90 120 150 180 E171/E171 OECD P25 food TG318 nano Dispersion Energy (J/ml) ) SCCS Dispersion Time (min) UV-Filter

14 E171 Brookhaven XDC & CPS DC results High agreement between different manufacturer

. Small interlaboratory SD on volume particle sizes and nano contents . High reproducibility of results 2.0 0.400 1.5 0.300

0.200 1.0

0.5

0.100 (%) Q0<100nm x50 Volume (µm) Volume x50 0.000 0.0 A B C D E A B C D E M 1 XDC 0.281 0.260 0.264 0.360 0.301 M 1 XDC 0.5 0.8 0.3 0.0 1.0 M 2 CPS DC 0.276 0.264 0.268 0.363 0.310 M 2 CPS DC 0.9 1.3 1.1 0.2 1.5 M 1 CPS DC 0.283 0.269 0.270 0.383 0.301 M 1 CPS DC 1.1 1.2 1.0 0.2 1.5

2.0 0.400 D, 0.369

E, 0.304 A, 0.280 1.5 E, 1.3 0.300 B, 0.264 C, 0.267 B, 1.1 0.200 1.0 A, 0.8 C, 0.8

0.100 0.5

D, 0.1

0.000 0.0 x50 Volume (µm) Q3 Volume <100nm (%)

15 VSSA E172 Screening of particle sizes and nano contents

100.0 푚2 퐷 2퐷 VSSA cut off = 60 ∗ dmin (D*) = 91.0 푐푚3 3 VSSA 푉푆푆퐴 90.0 Smallest dimension D= 3 spheres, 2 needles, 1 flakes

80.0 77.0 76.4 74.3 71.9 72.6 69.2 70.0 65.4 61.6 59.8

) 60.0

3 Cubes&Spheres = 60 m2/cm3 /cm 2 49.3 49.6 50.0 47.5 46.8 Needles & Spheres = 50 m2/cm3 43.6 42.9 39.0 VSSA VSSA (m 40.0 34.2 Needles = 40 m2/cm3 33.3 29.8 29.4 30.0 27.3 26.9

22.1 21.0 20.0

10.0

0.0

Red1 Red6 Red2 Red3 Red4 Red5 Red7 Red8 Red9

Red10 Red11

Black1 Black2 Black3 Black4 Black5 Black6 Black7 Black8

Brown1 Brown2 Brown3

Yellow1 Yellow2 Yellow3 Yellow Red Brown Black BET

According to DIN ISO 9277, samples are heated up to 180°C in 30 min and left at this temperature for one hour before determination of the SSA under Nitrogen at 5 different pressure ratios: 0,1 – 0,15 – 0,2 – 0,25 – 0,3.

The Volume Specific Surface Area (VSSA) is calculated by multiplication of the SSA with the gravity ()

VSSA = SSA x  Equation 1

The Monodisperse Diameter of the Volume Specific Surface Area (dminVSSA) is calculated for different particle shapes by the following equation

dminVSSA (D*) = 2퐷 Equation 2 푉푆푆퐴

The VSSA cut off is calculated by the equation below.

푚2 퐷 VSSA cut off = 60 ∗ Equation 3 푐푚3 3

*Smallest dimension D= 3 spheres, 2 needles, 1 flakes

17 Disc Centrifuge and Dispersion

Instrumentation, parameters and formulation Company Instrument Detector Disc Speed Formulation UV-Light M-2 CPS DC 18000 rpm 0,1g E171, 100 ml water, 0,06 g Calgon N 405nm M-1 Brookhaven XDC X-Ray 3000 rpm 2g E171 in 50 ml water, 0,02 g calgon N UV-Light 0,1g E171, 38 ml water, 12g PDO*, 0,05 g M-1 CPS DC 18000 rpm 470nm Calgon N

Dispersion Equipment

Company Sonic Probe Tip Diameter Tip Length Eglible volume Nominal Power

M-2 Sonoplus 2200 13 mm 250 mm 25-500 ml 750 W

M-1 Sonics VCX750 13 mm 136 mm 25-500 ml 750 W

18 NIST 1200 Technichal Guideline Power Calibration

19