newest developments in modeling nanomaterial flows · technology & society laboratory ......

Willkommen

Welcome

Bienvenue

Newest Developments in Modeling Nanomaterial Flows

Bernd Nowack

Empa – Swiss Federal Laboratories for Materials Science and Technology

Technology & Society Laboratory

9014 St. Gallen

Switzerland

Overview

Material flow analysis and environmental fate modeling

Short history

Newest developments

Extension of covered materials and processes

Dynamic MFA

Use in risk modeling

Validation

Why nanomaterial flows?

Flows from production and manufacturing to use, followed by release and end-of-life

Knowledge about flows forms the basis for exposure assessment

The size of the flows determines the relevance of different nanomaterials

A life-cycle approach forms the basis for the flow assessment

The two kinds of models

Material flow models (MFA) predict releases from products, fate in technical systems and final release to the environment

Environmental fate models (EFM) describe the further fate in the environment and distribution within environmental compartments

PEC-values

Simplified

PEC-values,

Worst-case

(no further

degradation

)

MFA models

Model name Reference Model type Uncertaintyconsideration

ENM

n/a (Boxall et al., 2007) Simple algorithms 3 scenarios 11 ENM

MFA (Mueller and Nowack, 2008)

Excel-based MFA 2 scenarios TiO2, CNT, Ag

PMFA (Gottschalk et al., 2010)

Probabilistic model in R Probabilistic TiO2, ZnO, Ag.CNT, fullerenes,SiO2, Au,

O’Brien (O'Brien andCummins, 2010a)

Excel with add onpackages

10’000 iterations TiO2, Ag, CeO2

PFA (Arvidsson et al.,2012)

scenarios Ag, TiO2

LEARNano (Keller et al., 2013;Liu et al., 2015)

Integrated in RedNano,web-based

scenarios 10 ENM

DPMFA (Bornhöft et al.,2016)

Dynamic probabilisticmodel in Python

Probabilistic TiO2, CNT, Ag, ZnO

PMFA Version1.0.0

(SUN, 2016) Probabilistic model in Rwith GUI (graphical userinterface)

Probabilistic

Input data required for MFA

Parameter Comment UncertaintyProduction volume within the system boundary

Directly available or scaled up/down from other regions

Depending on the material the uncertainty is medium to very high

Distribution of mass to product categories

The most critical parameter in MFA

Very high: quantitative data are largely absent

Release from products/ applications

Transfer factors needs to be estimated based on release studies or expert knowledge

Real-world studies using real products are mostly missing, therefore quite high uncertainty. Often worst-case assumptions are used.

Transfer factors for technical compartments

Data for WWTP are abundant, for WIP only few available, almost nothing for landfills

Low uncertainty for WWTP and WIP, high for landfills

Transfer factors during recycling

Only considered in some models, no quantitative data available

Uncertainty medium

Transfer factors for environmental compartments

Although transfer between environmental compartments is part of EFM, some MFA models include limited transfers

Often worst-case scenarios

History of MFA Models

Generation 1: Deterministic MFA

High emission scenarioSwitzerlandFlows in tons/yearSTP: sewage treatment plantWIP: waste incineration plant

Müller and Nowack (2008) Environ. Sci. Technol. 42: 4447–4453

A and B are deterministic values

AB

Flow=A x B

A: ENM amount in compartment AB: transfer coefficient

c

Deterministic MFA: Global Flows

Keller et al. (2013) J. Nanopart. Res. 15: 1692

TiO2

History of MFA Models

Gottschalk et al. (2015) Int. J. Environ. Res. Public Health 12: 5581

Generation 2: Probabilistic (stochastic) MFA

Modeling by Empa

Nano-TiO2, ZnO, Ag, CNT, fullerenes

Pigment TiO2

Region: Europe, Switzerland, USA

Latest update by Sun et al. (2014), Environmental Pollution 185: 69-76

Nano-gold: Prospective Flows

Mahapatra et al. (2015) J. Nanobiotechnol. 13: 93

Nano-Au used in medical applications

Prospective scenario

Annual flow of nano-Au in the UK in kg/year

Nano-silica

Landfill

Sewage treatment

Wastewater

Air

Surface water

Overflow

STP sludge

Production

Manufacture

Consumption

PMC

Sediment

2,600

13,000

7,500

27,000

11,000

31,000

1,700

340

6,800

310

4,400

9.2

52

0.0029

290

Nano-SiO2 (EU, 2014)2,600

Wet scrubber

Burning(WIP)

Filter

Recycling

Export

2,100

340

2,100

2,600

4,500

1,900

2.8

2.9

4,600

3,300

1,700

35,000

6,800

SoilST soil4,600

Soil4,700

Production 91,000

Cement plantCement plant3,700Paints 62.7%

Polymers 7.1%

Coating 7.1%

Ceramics 6.5%

Car tires 5.7%

Cleaning agent 5.1%

Catalyst 2.8%

Food products 1.4% Cosmetics, personal care products 0.8%

Others 0.6%

Wang et al. (2016) Sci Total Environ. 545-546: 67-76.

Nano-SiO2 flows in the EU in tons/y

Nano iron oxides

LandfillLandfill

Sewage treatment

Sewage treatment

WastewaterWastewater

SoilSoil

AirAir

Surface water

Surface water

OverflowOverflow

STP sludge

Production

Manufacturing

Consumption

PMC

SedimentSediment

200

1000

1100

1700

Production7500

800

3100

67

25

410

46

69

1.4

2.7

0.00039

45

nano iron oxides (EU, 2014)

560

Acid washing 2Acid washing 2

Burning(WIP)

Burning(WIP)

FilterFilter

Recycling

ExportExport

180

25

103

410

330

0.39

8204700

410

140

N&U

soil

ST soil390

110

0.37

390

CementCement

1700

440

210

103

150

unconsidered500

Downcycling

Use of bottom ash

590

Wang et al. (2016) Nanotoxicol. in press

Nano iron oxide flows in the EU in tons/y

Nine nanomaterials in Denmark

Gottschalk et al. (2015) Int. J. Environ. Res. Public Health 12: 5581

Int. J. Environ. Res. Public Health 2015, 12 5590

Photostable TiO2 Photocatalytic TiO2 ZnO

Ag CNT CuCO3

CeO2 QD CB

Figure 2. Mass flow diagrams for nine nanomaterials in Denmark. Rounded modal values

are shown in tons/year (except CB where the unit is kt/y). Boxes show accumulation or

transformation. The modes, combining all the Monte Carlo simulations show what

has to be most likely expected at each place of the figure without necessarily reflecting

in detail and holistically mass balance of the flow system. CNT: Carbon nanotubes;

QD: Quantum dots; CB: Carbon Black.

New: TiO2: photocatalytic and photostable, CeO2, QD, CuCO3, Carbon black

Flows out of recycling

Sun et al. (2014) Environ. Pollut. 185: 69-76.

Flows out of recycling

Caballero-Guzman et al. (2015) Waste Management 36: 33–43.

Dynamic probabilistic MFA

Bornhöft et al. (2016) Environ. Modeling Software 76: 69-80

Sun et al. (2016) Environ. Sci. Technol. 50: 4701-4711



Product

category i

Use

release

EoL

release

Compartment

Compartment

Use release schedule

time

am

ou

nt o

f re

lea

se

EoL release schedule

am

ou

nt o

f re

lea

se

Compartment

Compartment

time

Allocation to technical

and environmental

compartments after release

Release over time;

schedule of EoL release

depends on product life

distribution

❷

Release over time;

Schedule of release varies

by products

❹

❸

ENM

production

Product

category

Product

categoryManufacture of ENM

enabled products

Allocation of ENM to

different products

❶

Dynamic MFA

Release Module: Time dependent ENM release from products

Pro

bab

ilit

y5 year 10 year 15 year

Life span of product category

EN

M r

ele

ase

1 2 3 4 5 6 7 8 9 10 11 12 13 year

ENM release kinetic of each product category

Dynamic MFA

Release Module: Development of ENM in stocksa. nano-TiO2 in in-use

stocksb. nano-TiO2 in accumulative stocks




Mean Mode Median Q0.15 Q0.85

STP Effluent 44.4 13.7 16.3 2.77 76.1 µg/L

STP sludge 1.60 0.47 0.84 0.16 3.42 g/kg

Solid waste to Landfill 12.9 7.67 10.3 5.37 21.3 mg/kg

Solid waste to WIP 10.3 6.19 7.93 4.24 16.9 mg/kg

WIP bottom ash 395 161 237 93.6 729 mg/kg

WIP fly ash 543 238 327 129 979 mg/kg

Surface water 2.17 0.61 1.10 0.19 4.40 µg/L

Sediment 43.1 30.0 38.7 21.3 65.0 mg/kg

Natural and urban soil 2.94 1.86 2.57 1.44 4.53 µg/kg

Sludge treated soil 61.1 40.8 54.6 30.9 93.3 mg/kg

Air 2.05 0.86 1.24 0.43 3.98 ng/m3

STP Effluent 0 0 0 0 0 µg/L

STP sludge 0 0 0 0 0 µg/kg



WIP bottom ash 6.43 3.13 4.10 1.49 10.2 mg/kg

WIP fly ash 12.2 4.85 7.32 2.16 20.8 mg/kg

Surface water 0.38 0.16 0.23 0.05 0.64 µg/L

Sediment 6.97 4.95 6.26 3.64 10.5 mg/kg


Sludge treated soil 1.82 1.01 1.52 0.74 2.96 µg/kg

Air 0.94 0.39 0.48 0.08 1.67 ng/m3

STP Effluent 2.65 0.71 1.04 0.16 4.57 ng/L

STP sludge 61.3 20.2 32.9 7.41 113 µg/kg

Solid waste to Landfill 79.0 35.4 51.0 18.7 139 µg/kg

Solid waste to WIP 19.7 10.6 14.7 6.87 33.5 µg/kg

WIP bottom ash 170 111 141 75.4 267 µg/kg

WIP fly ash 340 169 259 114 582 µg/kg

Surface water 1.51 0.63 1.01 0.40 2.78 ng/L

Sediment 30.1 23.8 27.8 18.3 43.3 µg/kg



Air 0.01 0.00 0.01 0.00 0.02 ng/m3

STP Effluent 2.60 0.59 1.06 0.18 4.52 ng/L

STP sludge 63.3 23.7 37.4 7.88 117 µg/kg

Solid waste to Landfill 82.3 42.6 54.9 27.7 146 µg/kg

Solid waste to WIP 24.8 14.4 19.6 11.6 40.0 µg/kg

WIP bottom ash 287 205 260 149 437 µg/kg

WIP fly ash 571 335 474 223 934 µg/kg


Sediment 164 163 162 127 201 µg/kg



Air 0.01 0.00 0.01 0.00 0.02 ng/m3

STP Effluent 8.58 6.50 7.00 0.92 15.6 ng/L

STP sludge 326 273 277 33.9 593 µg/kg



WIP bottom ash 0.43 0.17 0.30 0.09 0.79 mg/kg

WIP fly ash 0.92 0.30 0.54 0.14 1.74 mg/kg


Sediment 6.74 6.34 6.46 4.32 9.24 µg/kg

Natural and urban soil 35.0 33.8 33.9 23.6 46.2 ng/kg


Air 0.02 0.01 0.02 0.00 0.03 ng/m3

CNT

Nano-Ag (1900-2020)

EU (2014)

Nano-TiO2

Nano-ZnO

Nano-Ag (1990-2020)

Use of MFA in risk modeling

For risk assessment, both exposure and hazard data are needed

The MFA modeling can provide PEC values

As hazard data PNEC values are needed

Obtained from literature by species sensitivity distributions

Risk= PEC/PNEC

ModeledEnvironmental Concentrations

ExposureAssessment

HazardCharacterization

RiskCharacterization

Material flowmodeling

Evaluation of ecotox data

Hazard characterization (freshwater)

• Species sensitivity distributions (SSD)• State: 2014• Available for water, soil, (sediment)• Used to derive PNEC from HC5

Coll et al., Nanotoxicology (2016) 10: 436-444

Nano-Ag CNT Nano-TiO2

Nano-ZnO

Risk characterization (freshwater)

• Predicted environmental concentration (PEC) from Sun et al. (2014)

• Predicted no effect concentrations (PNEC) from SSD

Nano-Ag CNT Nano-TiO2

Nano-ZnO


Log µg/l Log µg/l Log µg/l

Log µg/l

Risk characterization

Risk Characterization Ratio (RCR) = PEC/PNEC

Freshwater Soil


Validation of models

Nowack et al. (2015) Environmental Science: Nano 2: 421 – 428

Issues with validation

Currently, PEC values cannot be validated,

therefore the model as a whole cannot be

validated

Applies mainly to environmental fate modeling

Processes within the models can be validated

Modeling and analytical studies are able to

provide an orthogonal view

Conclusions

MFA methods are central to determine flows of ENM to the environment

Methods have improved over the last years, with including more and more processes and compartments

The covered ENM have also increased, with models available for almost all relevant ENM

The dynamic modeling of the flows constitutes a major improvement of the model

The simple PEC values provided by MFA provide first assessments of environmental exposure

Release flows from MFA are basic inputs to fate modeling of ENM

Thank you for your attention!

contact: [email protected]

www.empa.ch/nanoimpact

Abstract deadline November 15, 2016

Aim: Integrate scientific knowledge on about nanomaterial exposure, fate, effects and risks in the environmentKey question: Have nanomaterials have novel fate routes and effects resulting in new and unknown threats under environmentally relevant conditions ?

OrganizersBernd Nowack, Empa, SwitzerlandRalf Kägi, Eawag, SwitzerlandMark Wiesner, Duke University, USAPeter Gehr, University of Bern,Switzerland

http://www.empa.ch/nanoimpact

newest developments in modeling nanomaterial flows · technology & society laboratory ......

Documents