lable at ScienceDirect

Regulatory Toxicology and Pharmacology 73 (2015) 137e150

Contents lists avai

Regulatory Toxicology and Pharmacology

journal homepage: www.elsevier .com/locate/yrtph

Comparative assessment of nanomaterial definitions and safetyevaluation considerations

Darrell R. Boverhof a, Christina M. Bramante b, John H. Butala c, Shaun F. Clancy d,Mark Lafranconi e, Jay West f, *, Steve C. Gordon g

a The Dow Chemical Company, 200 Larkin, E-157, Midland, MI 48674, USAb Cabot Corporation, 157 Concord Road, Billerica, MA 01821, USAc Toxicology Consultants, Inc., 7 Glasgow Road, Gibsonia, 15044 PA, USAd Evonik Corporation, 299 Jefferson Road, Parsippany, NJ 07054, USAe Tox Horizons, LLC, 7569 Kings Mills Rd, Maineville, OH 45039, USAf American Chemistry Council, 700 2nd Street NE, Washington, DC 20002, USAg 3M Company, 3M Center, Building 220-6E-03, St. Paul, MN 55144, USA

a r t i c l e i n f o

Article history:Received 9 January 2015Received in revised form21 May 2015Accepted 2 June 2015Available online 23 June 2015

Keywords:NanomaterialDefinitionsRisk evaluation

* Corresponding author. American Chemistry CoWashington, DC 20002, USA.

E-mail addresses: rboverhof@dow.com (D.R. Bovcabotcorp.com (C.M. Bramante), butala@jhbutala.comevonik.com (S.F. Clancy), toxhorizons@gmail.comamericanchemistry.com (J. West), scgordon@mmm.co

http://dx.doi.org/10.1016/j.yrtph.2015.06.0010273-2300/© 2015 The Authors. Published by Elsevier

a b s t r a c t

Nanomaterials continue to bring promising advances to science and technology. In concert have comecalls for increased regulatory oversight to ensure their appropriate identification and evaluation, whichhas led to extensive discussions about nanomaterial definitions. Numerous nanomaterial definitions havebeen proposed by government, industry, and standards organizations. We conducted a comprehensivecomparative assessment of existing nanomaterial definitions put forward by governments to highlighttheir similarities and differences. We found that the size limits used in different definitions wereinconsistent, as were considerations of other elements, including agglomerates and aggregates, distri-butional thresholds, novel properties, and solubility. Other important differences included considerationof number size distributions versus weight distributions and natural versus intentionally-manufacturedmaterials. Overall, the definitions we compared were not in alignment, which may lead to inconsistentidentification and evaluation of nanomaterials and could have adverse impacts on commerce and publicperceptions of nanotechnology. We recommend a set of considerations that future discussions ofnanomaterial definitions should consider for describing materials and assessing their potential for healthand environmental impacts using risk-based approaches within existing assessment frameworks. Ourintent is to initiate a dialogue aimed at achieving greater clarity in identifying those nanomaterials thatmay require additional evaluation, not to propose a formal definition.© 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND

license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Definitions of nanomaterials and their use in regulatory evalu-ations have been, and continue to be, an area of active scientific andpolicy debate (ICCA, 2010; Maynard, 2011; Stamm, 2011; Bleekeret al., 2013). Nanomaterials may exhibit properties different fromtheir non-nano forms, and these different properties have raised

uncil, 700 2nd Street, NE,

erhof), christina.bramante@(J.H. Butala), shaun.clancy@(M. Lafranconi), jay_west@m (S.C. Gordon).

Inc. This is an open access article u

questions about potential human health and environmental risks.The International Organization for Standardization (ISO) hasdefined “nanomaterial” as a “material with any external dimensionin the nanoscale or having internal structure or surface structure inthe nanoscale” (ISO, 2010) and “nanoparticle” as a “nano-objectwith all three external dimensions in the nanoscale” where nano-scale is defined as the size range from approximately 1e100 nm(ISO, 2008). These technical definitions, based on size only, may beinsufficient from a risk evaluation standpoint because they do notinclude other important elements that should be considered whendetermining whether a nanomaterial may need additional review.

Discussions about developing a definition for nanomaterialshave been challenging because of the need to satisfy two divergingconsiderations. A definition should be broad enough to define

nder the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

D.R. Boverhof et al. / Regulatory Toxicology and Pharmacology 73 (2015) 137e150138

materials that may warrant additional evaluation, yet it should notbe so broad as to include those materials for which additional ex-amination or evaluation would not be meaningful in terms ofprotecting human health or the environment. A balance is neces-sary to ensure appropriate allocation of resources to most effec-tively protect the health and safety of humans and theenvironment. Definitions proposed to date have taken a variety ofapproaches in attempting to strike a suitable balance, but whilethey may satisfy specific jurisdictional mandates, their frequentlycontradictory inclusions and exclusions present a complex regu-latory maze for producers of nanomaterials and products contain-ing them. Difficulties associated with attempting to comply withcontradictory nanomaterial definitions and potential regulationscan impede international trade and, more fundamentally, reducepublic confidence in the adequacy of regulatory protections.

Nanomaterials are neither inherently hazardous nor inherentlysafe (Auffan et al., 2009; Donaldson and Poland, 2013), and it hasbeen broadly recognized that they should not be treated as such inevaluation programs (SCENIHR, 2007; Holdren, 2011; Hamburg,2012). Likewise, the informational elements of a nanomaterialdefinition presented in this paper are not intended to identifyinherently hazardous or non-hazardous materials. Rather, they areintended to be used in conjunction with available hazard andexposure information to identify nanoscale materials whichmay beof interest for potential priority setting, risk assessment, and riskmanagement activities. While the elements we identify in thispaper can help strengthen or inform developing definitions, theyshould not be viewed as a comprehensive list of factors to beconsidered in a safety assessment of nanomaterials. Regardless ofthe definition that is applied, there is an obligation of both regu-lators and the regulated community to ensure that a material isevaluated appropriately to determine whether it poses a risk tohuman health or the environment. This evaluation should be basednot only on the intrinsic hazard potential of the material but onconsideration of exposure potential (e.g. during manufacturing,use, and disposal) as well.

The purpose of this paper is twofold: (1) to compare andcontrast existing nanomaterial definitions and (2) to present a set ofinformational elements to be considered as discussions on the needfor definitions and nanomaterial regulatory frameworks continue.The intent is not to propose a formal definition for nanomaterials,but to initiate a dialogue aimed at achieving greater clarity inidentifying those nanomaterials that may require additional eval-uation. That process should account for each of the elements pre-sented in this paper at some point in the evaluation, while seekingto eliminate differences that create further ambiguity. The ideasexpressed in this paper apply to the commercial manufacturing anduse of nanomaterials.

2. Comparison/contrast of current definitions

Numerous definitions for nanomaterials have been proposed byvarious government, industry, and standards organizations. Thesedefinitions are often inconsistent in their elements and scope,which can lead to confusion in determiningwhether amaterial is oris not considered to be a “nanomaterial.” To better understand thesimilarities and differences among definitions, we performed acomprehensive comparative assessment of 14 definitions fromvarious regulatory authorities (Table 1). The assessment includedformal regulatory definitions as well as definitions stated in guid-ance or policy documents (“advisory definitions” in Table 1). Thedefinitions are generally intended to address safety impacts topeople and the environment, though there may appear to be aparticular emphasis on human health impacts. The elements

considered in this analysis are applicable for human and environ-mental safety, as well as toxicological and ecotoxicological effects.

Size (or external dimensions) was the only common elementacross all of the definitions, though the upper size limits weresometimes variable among the definitions. Several important coreelements that were not consistently mentioned included: consid-eration of agglomerates and aggregates, distributional thresholds,novel properties, and solubility. Other important differencesincluded number size distributions versus weight distributions andthe inclusion of natural and incidental nanomaterials along withthose manufactured intentionally. A summary of these core ele-ments and the similarities and differences among the 14 definitionsis presented in Table 2. This comparative assessment highlightssome of the key differences that can lead to a lack of clarity andconsistency with respect to the term “nanomaterial” and whatmaterials may be subject to existing or developing regulations.What follows is a discussion on some of the factors that should beconsidered when developing regulatory definitions or identifyingnanomaterials that may be of interest, with a focus on those ele-ments identified through the comparative assessment in Table 2.

3. Core elements for describing nanomaterials

3.1. Size

Size is the fundamental defining characteristic of all nano-materials. While size is an easy concept to understand, it is moredifficult to apply because there are no natural physical or chemicalboundaries that delineate the “nanoscale.” By convention,1e100 nm is the size range most commonly used in reference tonanomaterials, but there is no bright line that clearly demarks thenanoscale from a chemical or biological perspective. At the lowerend of the range, 1 nm is intended to distinguish between indi-vidual molecules and nanomaterials, although some molecules(e.g., some proteins and biomolecules) may have at least onedimension larger than 1 nm. Many nanomaterial definitionsexplicitly include materials that may have dimensions below 1 nm(e.g., fullerenes and graphene; Table 2). At this lower end of“nanoscale,” the characteristics and properties of the material arelargely defined by the chemistry of the molecule and not by thephysical nature of the formed nanoscale materials.

The upper end of the “nanoscale” at 100 nm is an arbitrary cut-off since the size-dependent behavior of materials does not stop orbegin abruptly at 100 nm. Many properties characteristic of thenanoscale, such as solubility, light scattering, and surface area ef-fects, are predictable and continuous characteristics of the bulkmaterials (Donaldson and Poland, 2013). In an attempt to be in-clusive of all the characteristics that may be important for regula-tory oversight, some authorities have expanded the upper range ofthe nanoscale well beyond 100 nm (Table 2) (Health Canada, 2011;US FDA, 2011; Taiwan, 2012).

While size limits are somewhat arbitrary, there is generalagreement that any unique nano-specific phenomena of particu-lates are most likely to occur between 1 and 100 nm. For instance,the properties of inorganic particles were evaluated by Auffan et al.(2009) who found that novel, size-dependent properties of nano-scale materials, such as catalytic properties of gold, the photo-catalytic activity of TiO2, and the tunable fluorescent behavior ofquantum dots, occur below 30 nm (see the Novel Properties sectionof this paper for further discussion). Most regulatory authoritieshave used 1e100 nm to define the nanoscale, which is consistentwith the ISO standard (ISO, 2008). We agree that this provides areasonable range, provided there is recognition that particle sizealone is not sufficient for the evaluation of a nanomaterial and that

Table 1Current regulatory and advisory definitions of “nanomaterial”.

Organization Definition Product category Status

European CommissionCosmetics Directive(EC, 2009)

“ ‘[N]anomaterial’ means an insoluble or biopersistant and intentionallymanufactured material with one or more external dimensions, or aninternal structure, on the scale from 1 to 100 nm.”

Cosmetics Regulatory Definition

Australian GovernmentDepartment of Healthand Ageing (NICNAS, 2010)

“[I]ndustrial materials intentionally produced, manufactured or engineeredto have unique properties or specific composition at the nanoscale, that is asize range typically between 1 nm and 100 nm, and is either a nano-object(i.e. that is confined in one, two, or three dimensions at the nanoscale) or isnanostructured (i.e. having an internal or surface structure at the nanoscale).Aggregates and agglomerates are included and apply to materials where 10%or more of the particles by number count meet the above definition.”

All non-food Advisory Definition

Health Canada(Health-Canada, 2011)

Health Canada considers any manufactured substance or product and anycomponent material, ingredient, device, or structure to be nanomaterial if itis at or within the nanoscale in at least one external dimension, or has internalor surface structure at the nanoscale, or if it is smaller or larger than thenanoscale in all dimensions and exhibits one or more nanoscaleproperties/phenomena. Nanoscale properties/phenomena refer to propertiesthat are attributable to the size of the substance and size effects.

All Advisory Definition

United States Food and DrugAdministration (US FDA, 2011)

There is no formal agency definition. However, “when considering whether anFDA-regulated product contains nanomaterials, or otherwise involves theapplication of nanotechnology, FDA will ask: Whether an engineered materialor end product has at least one dimension in the nanoscale range (approximately1 nm to 100 nm); or whether an engineered material or end product exhibitsproperties or phenomena, including physical or chemical properties or biologicaleffects, that are attributable to its dimension(s), even if these dimensions falloutside the nanoscale range, up to one micrometer.”

Cosmetics,pharmaceuticals,food, and foodcontact

Advisory Definition

United States EnvironmentalProtection Agency (US EPA,2007, 2011b, 2014)

There is no formal agency definition. However, the agency has outlined key criteriaacross several documents, including: solid at 25 �C and atmospheric pressure,particle size between 1 and 100 nm in at least 1 dimension; the material exhibitsunique and novel properties because of its size; the material is engineered at thenanoscale; inclusion of primary particles, aggregates and agglomerates; and adistribution of particles with greater than 10% by weight less than 100 nm.

All except:cosmetics,pharmaceuticals,food and foodcontact

Advisory Definition

European Commission,Recommendation on theDefinition of Nanomaterials(EC, 2011a)

“ ‘Nanomaterial’ means a natural, incidental or manufactured materialcontaining particles, in an unbound state or as an aggregate or as an agglomerateand where, for 50% or more of the particles in the number size distribution, oneor more external dimensions is in the size range 1 nme100 nm.”“In specific cases and where warranted by concern for environment, health, safetyor competitiveness the number size distribution threshold of 50% may be replacedby a threshold between 1 and 50%.”

“[F]ullerenes, graphene flakes and single wall carbon nanotubes with one or moreexternal dimensions below 1 nm should be considered as nanomaterials.”

All Advisory Definition

European Parliament and theCouncil of the EuropeanUnion on the provisionof food information toconsumers (EC, 2011b)

“’[E]ngineered nanomaterial’ means any intentionally produced material that hasone or more dimensions of the order of 100 nm or less or that is composed of discretefunctional parts, either internally or at the surface, many of which have one or moredimensions of the order of 100 nm or less, including structures, agglomerates oraggregates, which may have a size above the order of 100 nm but retain propertiesthat are characteristic of the nanoscale.“Properties that are characteristic of the nanoscale include: (i) those related to thelarge specific surface area of the materials considered; and/or (ii) specificphysico-chemical properties that are different from those of the non-nanoform ofthe same material.”

Food Regulatory Definition

(continued on next page)

D.R.Boverhof

etal./

RegulatoryToxicology

andPharm

acology73

(2015)137

e150

139

Table 1 (continued )

Organization Definition Product category Status

European Commission,Biocides Directive (EC, 2012)

“ ‘[N]anomaterial’ means a natural or manufactured active substance or non-activesubstance containing particles, in an unbound state or as an aggregate or as anagglomerate and where, for 50% or more of the particles in the number sizedistribution, one or more external dimensions is in the size range 1e100 nm.“Fullerenes, graphene flakes, and single wall carbon nanotubes with one or moreexternal dimension below 1 nm shall be considered as nanomaterials.”

Biocides Regulatory Definition

French Ministry of Ecology,Sustainable Developmentand Energy (ANSES, 2012)

“Substance at nanoscale” is defined as a substance “intentionally producedat nanometric scale, containing particles, in an unbound state or as an aggregateor as an agglomerate and where, for a minimum of 50% of particles in the numbersize distribution, one or more external dimensions is in the size range 1 nme100 nm.By derogation from this definition, fullerenes, graphene flakes and single-wall carbonnanotubes with one or more external dimensions below 1 nm are consideredas ‘substances at nanoscale’.”

All Regulatory Definition

Taiwan Council of Labor Affairs

(Taiwan, 2012)

A nanomaterial is one which is intentionally manufactured or designed and meetsany of the following conditions:A) Material with one or more external dimensions or an internal or surface structureon the scale from 1 to 100 nm;

B) It is smaller or larger than the nanoscale above in all spatial dimensions and exhibitsone or more nanoscale phenomena/property (for example, increased intensity andchemical reactivity).

C) 13 engineered nanomaterials by OECD in 2010 which can be used as referenceare provided in the definition.

All Advisory Definition

Swiss Federal Office of PublicHealth and Federal Officefor the Environment(H€ock et al. 2013)

Utilizes European Commission, Recommendation on the Definition of Nanomaterials(EC, 2011), but also includes particles up to 500 nm and respirable particles of up to10 microns with nanoscale side chains.

All Advisory Definition

Norwegian EnvironmentAgency (2014)

Follows the European Commission, Recommendation on the Definition ofNanomaterials (EC, 2011)

All sectors, butonly for chemicalsclassified as hazardousaccording to statedcriteria.

Regulatory Definition

Belgian Federal Public ServiceHealth, Food Chain Safetyand Environment(Belgium, 2014)

“Substance produced in nanoparticular state” is defined as a substance containingunbound particles, or aggregate or agglomerate of those particles, where 50% or moreof the particles in the number size distribution have one or more external dimensionis in the size range of 1e100 nm. The definition excludes natural, non-chemically modifiedsubstances and those for which the fraction in the 1e100 nm range is “a byproductof human activity,” which is further defined.

Scope is defined bya complex set ofexemptions.

Regulatory Definition

Danish Ministry of theEnvironment (2014)

Follows the European Commission, Recommendation on the Definition of Nanomaterials(EC, 2011)

Scope is defined bya complex set ofexemptions.

Regulatory Definition

D.R.Boverhof

etal./

RegulatoryToxicology

andPharm

acology73

(2015)137

e150

140

Table 2Comparative assessment of elements considered in current regulatory and advisory definitions of “nanomaterial”.

Organization and/or regulation Size Solubility Aggregates andagglomerates

Distribution threshold Intentionallymanufactured/engineered

Novelproperties

European Commission Cosmetics Directive 1e100 nm Yes Yes No specific mention Yes NoAustralian Government Department of Health and Ageing 1e100 nm No Yes 10% by number Yes YesHealth Canada 1e100 nm and largera No Yes No specific mention Yes YesUnited States Food and Drug Administration 1e100 nm and largerb No No specific

mentionNo specific mention Yes Yes

United States Environmental Protection Agency 1e100 nm No Yes 10% by weight Yes YesEuropean Commission Recommendation for a Definition 1e100 nm No Yes 50% by number No NoEuropean Parliament and the Council of the European

Union on the Provision of Food Information to Consumers1e100 nm and largera No Yes No specific mention Yes Yes

European Commission Biocides Directive 1e100 nm No Yes 50% by number No NoFrench Ministry of Ecology, Sustainable Development,

Transport and Housing1e100 nm No Yes 50% by number Yes No

Taiwan Council of Labor Affairs 1e100 nm and largera No No specificmention

No specific mention Yes Yes

Swiss Federal Office of Public Health and Federal Officefor the Environment

1e100 nm and largerc No Yes 50% by number No No

Norwegian Environment Agency 1e100 nm and larger No Yes 50% by number No NoBelgian Federal Public Service Health, Food Chain Safety

and Environment (Belgium, 2014)1e100 nm No Yes 50% by number Yes No

Danish Ministry of the Environment (2014) 1e100 nm No Yes 50% by number No No

a Health Canada, the European Commission (for food and food contact materials) and the Taiwan Council of Labor Affairs have indicated the inclusion of materials largerthan the nanoscale in all dimensions if they exhibit one or more nanoscale properties/phenomena.

b The US FDA includes materials up to one micron if the material exhibits properties or phenomena that are attributable to its dimensions.c The Swiss definition includes substances with primary particles, aggregates and agglomerates up to 500 nm, as well as respirable materials of up to 10 microns that may

have nanoscale side branches.

D.R. Boverhof et al. / Regulatory Toxicology and Pharmacology 73 (2015) 137e150 141

there should be adequate consideration given to the inclusion ofthe additional criteria discussed below.

3.2. Distributional thresholds and mass vs. particle number

An additional practical aspect of understanding how size definesa nano-object or nanomaterial is the size distribution around amedian or mean. Few particulate substances are “monodisperse”(i.e., with a geometric standard deviation <2). Instead, most natu-rally occurring and manufactured nanomaterials have size distri-butions that may vary widely. Fig. 1A shows the hypotheticaldistribution of two particle populations with the samemedian size.For these two examples, the distribution varies greatly, but most ofthe particles are still in the nanoscale (less than 100 nm). However,if only a portion of the particles in the distribution is within thenanoscale, should that substance be considered a nanomaterial?Fig. 1B illustrates a particle size distribution in which only 10% of

A B

Fig. 1. Particle size distribution considerations. (A) Hypothetical size distribution of two sidistributions. (B) A hypothetical particle size distribution highlighting that a substance wpopulation less than 100 nm.

the particles fall within the nanoscale. Inclusion of a distributionalsize cut-off as a component of a definition is needed to provideincreased clarity as to what should be considered a nanomaterial(Bleeker et al., 2013). Many definitions include particle size distri-bution as an important consideration, though inconsistently. Somedefinitions define the particle size distribution based on the weightpercentage, while others define it based on the number countdistribution. For example, the European Commission has suggestedparticle number count of �50%, whereas US EPA has suggested>10% by weight (Table 2).

While the number count approach may seem intuitive whenspeaking of particulate substances (Bleeker et al., 2013), the weightpercent approach is more consistent with current detection capa-bilities and sensitivity of relevant analytical instrumentation, aswell as current practices for toxicological evaluations and estab-lishment of occupational exposure limits. Current nanoparticlemetrology standards have been largely focused on mass and

mple particle populations with the same mean particle size but different overall sizeith a mean particle size greater than 100 nm may still have a portion of the particle

D.R. Boverhof et al. / Regulatory Toxicology and Pharmacology 73 (2015) 137e150142

volume interpretations, and there have been limited efforts tovalidate particle count distributions that go below 100 nm (Brownet al., 2013). Despite this limitation, it has been noted that a focuson mass or volume as part of a distributional threshold may resultin larger particles skewing the distribution and an overall under-estimate of the true size distribution (Bleeker et al., 2013). Such asituation would be most applicable for non-homogenous or multi-modal populations of particulate substances and could beaddressed by specific analysis guidance. It is also true in that adistribution based on particle number will skew the distributiontoward the smallest particles. While number count may be moreinclusive in terms of defining substances as nanomaterials, the lackof analytical approaches and examples of how it can be appliedeffectively does not allow for an evaluation of the true utility ofsuch an approach. Using a number count criterion may result in theinclusion of materials that have mass-based distributions muchgreater than 1000 nm and would never be intended for inclusion aspart of a discussion of nanomaterials. For example, under the Eu-ropean Commission's recommendation for a definition (EC, 2011a),which includes particulates having �50% of particles by numberbelow 100 nm, millimeter sized ball bearings containing a traceamount of sub-100 nm wear debris could be classified as a nano-material, whereas a population of relatively monodisperse particleswith a median diameter of 110 nm may not meet the definition(Brown et al., 2013). Issues with contamination have also beenhighlighted with air, water, and even laboratory vessels where sub-100 nm particles may be present in low mass but high numbers(Brown et al., 2013). These examples illustrate how a definitionbased on a number count distribution cut-off may result in theinappropriate identification of materials, which could impact thepractical implementation of such an approach.

Arguments for using particle number-based cut-offs in nano-material definitions are based on the concept, supported by someexperimental findings over a number of years, that biological ef-fects caused by particulates tend to correlate more closely with theadministered dose expressed as particle number or total surfacearea rather than as mass (Tsuji et al., 2006; Wittmaack, 2007).However, examination of the more recent literature revealsdiffering opinions on the appropriate dose metric for describing thebiological or toxicological effects of nanomaterials. Data have beenpresented to show that doseeresponse relationships with nano-materials can be appropriately defined by particle mass or volume(Pauluhn, 2009, 2010). Other studies have suggested that totalparticle surface area may be an appropriate metric (Stoeger et al.,2006), while others have suggested that mass or surface area canbe effective descriptors depending on the material (Ho et al., 2011).

Another consideration for selection of the dose-metric is theintended use of the dose characterization. For comparison of dosewithin a given particle type and size, any of the dose-metric optionswould likely be sufficient. However, there is still uncertainty aboutthe toxicologically relevant dose-metric when comparing doseacross classes of particle types and size range. In these cases, it hasbeen suggested that particle numbers or surface area may be therelevant choice (OECD, 2012).

Interestingly, in the area of occupational exposure limits, dis-cussions on particle number versus mass measurements havepreviously been addressed, and it was recognized that the use ofmass would be more relevant to defining the health hazard whileproviding a simpler, less expensive, and more reproducible methodthan a particle count approaches (Tomb and Haney, 1988). Inaddition, a recent workshop on occupational exposure limits (OELs)for nanomaterials indicated that mass-based sampling and relatedanalytical methods are viewed as the most practical means forroutinely monitoring airborne particulates in the workplace(Gordon et al., 2014). These factors are expected to drive OEL

development for nanomaterials towards mass-based measurementvalues.

The analytical challenges of determining whether a materialmeets the particle number-based EC definition have been recentlyreviewed, the primary obstacle being the lack of standardized,validated methods (JRC, 2014). Specific technical challenges ofnumber-based particle analyses identified in the JRC report include:sample preparation techniques which can change the size distri-bution; obtaining representative samples; counting primary par-ticles within aggregates; determining the adequacy of dispersionmethods for powder; choosing the most appropriate size metric fornon-spherical particles (e.g., minimum external dimension, equiv-alent sphere, etc.); and suitability of standard reference materials.Other obstacles include the lack of interlaboratory proficiencytesting to identify laboratories which can reliably assess whether amaterial is a nanomaterial according to the EC definition and thehigh costs of performing these determinations, especially whenelectron microscopy is required.

Until this issue is resolved, some have recommended thattoxicological studies of nanomaterials include the characterizationof particles in a manner that allows for conversions betweendifferent metrics. While this may seem reasonable, it is importantto recognize that laboratory instruments used to measure particlemass or volume often employ mathematical conversions to deriveparticle count information and that, for these instruments, a 1%error in the ability of a method to accurately describe a mass orvolume distribution at the nanoscale could translate to a greaterthan 50% error in a number distribution (Linsinger et al., 2012;Brown et al., 2013). Opinions from the Organization for EconomicCooperation and Development (OECD) also recognize that it isappropriate and likely that dose results will continue to be reportedin mass-based units and that any deviation from this might havemajor consequences for the international mutual acceptance ofdata and would impact classification and labeling where hazards ofsubstances are related to mass concentration (OECD, 2012).

Based on these considerations, a mass-based distributionalthreshold is considered more appropriate than a number-basedthreshold. As with the 100 nm cut-off used for defining nano-objects in general, the choice of a weight percent cut-off such as10% is considered a reasonable distributional cut-off that willeffectively capture materials of the size range that have been thesubject of nanotechnology discussions within various agencies andorganizations, while at the same time providing increased clarityfor manufacturers and users as to which materials should beincluded and brought forward for discussion. A 10% weight cut-offis also consistent with the resolution of many of the analyticalmethodologies that currently exist for particle characterization(Linsinger et al., 2012; Brown et al., 2013), which will help with theapplication of definitions. As analytical resolution increases and ourcollective understanding of the biological and toxicological impli-cations of nanomaterials improves, cut-offs such as 100 nm and 10%by weight may change. In the meantime, the use of these limitsallows for a consistent and conservative approach for the assess-ment of nanomaterials.

3.3. Aggregates and agglomerates

Manufacturing processes can have an important impact on theform of particulates and how particles may associate with oneanother. Nanomaterials are produced by either “top down” or“bottom up” manufacturing approaches (Luther, 2004). Most con-ventional nanomaterial manufacturing processes are “top down” inwhich a larger sized material is ground or milled to smaller particlesizes. Depending on the amount of energy applied in the process,the final material may range in size from micronscale to nanoscale.

Fig. 2. Considerations for primary particles, aggregates and agglomerates. (A) A bottom-up synthesis approach for synthetic amorphous silica through a pyrogenic process. Theprocess involves hydrolysis of volatile silanes, such as silicon tetrachloride (SiCl4), in the flame of an oxygenehydrogen burner. This results in the growth (nucleation, condensation,coagulation) and aggregation of pyrogenic silica particles. Tightly bound aggregates can then form agglomerates of aggregates. Reprinted from Synthetic Amorphous Silica (CAS No.7631-86-9)” JACC No. 51, September 2006, p. 45.©2006, with permission from European Centre for Ecotoxicology and Toxicology of Chemicals. (B) Relationship of primary particle,aggregates of primary particles, agglomerates of primary particles, and agglomerates of aggregated primary particles.

D.R. Boverhof et al. / Regulatory Toxicology and Pharmacology 73 (2015) 137e150 143

Some graphenes are examples of nanomaterials produced by a “topdown”method, inwhich graphite is mechanically reduced in size to“sheets” of graphene having nanoscale dimensions (Knieke et al.,2010). Pigments are another example of nanomaterials that arecommonly produced by a top down approach.

By contrast, nanomaterials produced by “bottom up” processesare synthesized from atomic or molecular species by chemical re-action(s), allowing the precursor particles to grow in size. Solegelchemistry is an example of this approach in which solid nano-particles dispersed in a liquid (a solution or sol) agglomerate toform a continuous three-dimensional network extendingthroughout the liquid (a gel) (Young, 2002). Thin films and xerogelsare examples of nanomaterials produced by the solegel process(Brinker et al., 1992; Dur~aes et al., 2012). Another type of “bottomup” nanomaterial manufacturing involves high temperature com-bustion of chemical feedstock to produce airbornemolecules whichimmediately collide to form solid nuclei. During this process,coagulation rates are very rapid, with nuclei sintering or coalescinginto spherical primary particles (also called “nodules”). Withinmilliseconds, the primary particles coalesce (sinter) into largerparticles, called aggregates. As the manufacturing process con-tinues, the aggregates collide, resulting in secondary structures,known as agglomerates (Donnet et al., 1993). Industrial aciniformaggregates (e.g., titanium dioxide, carbon black and some forms ofsynthetic amorphous silica) are examples of nanostructured ma-terials produced by thermal combustion processes. Fig. 2A illus-trates the production of aggregated and agglomerated substancesmanufactured by a bottom-up high temperature combustionprocess.

The manufacturing processes described above can producethree different forms of particulates which may exist at the nano-scale: discrete primary particles, aggregates, and agglomerates. Keyproperties of each form are discussed below and depicted in Fig. 2B.ISO defines a “particle” as a minute piece of matter with definedphysical boundaries (ISO, 2008). Primary particles are consideredthe smallest discreet entity. Theymay exist at the nanoscale and are

considered to be an indivisible entity. ISO defines an “aggregate” asa particle comprised of strongly bonded or fused particles wherethe resulting external surface areamay be significantly smaller thanthe sum of calculated surface areas of the individual components.Aggregates may be comprised of constituent primary particles thatrange in size from nanoscale to greater than 100 nm and,depending on the size and number of the primary particles, ag-gregates themselves can range in size from the nanoscale tomicronscale. It is the size and shape of the aggregates that have afundamental influence on the properties of the nanomaterial(Donnet et al., 1993). Importantly, aggregates are robust, essentiallyindivisible structures, similar to primary particles, in that theycannot be broken down by external forces typically encounteredduring subsequent handling and processing (Gray and Muranko,2006). In contrast, an “agglomerate” is a collection of weaklybound particles, or aggregates ormixtures of the two, and for whichthe resulting external surface area is similar to the sum of thesurface areas of the individual components (ISO, 2008). Agglom-erates are held together by relatively weak forces, such as van derWaals forces or by simple physical entanglement. Therefore, incontrast to aggregates, agglomerates may break down into theirconstituent entity(s) with sufficient external energy (Donnet, 1993;Gray and Muranko, 2006). Agglomerates themselves can vary insize but are typically greater than 100 nm.

Most definitions specifically include aggregates and agglomer-ates (Tables 1 and 2). Although aggregates and agglomerates can bemuch larger than 100 nm, many definitions include these entitiesbased on consideration of the potential for these structures to breakdown into smaller nanoscale entities if sufficient force is applied.The EU Commission's recommendation on the definition of ananomaterial notes “agglomerated or aggregated particles mayexhibit the same properties as the unbound particles (EC, 2011a).Moreover, there can be cases during the life-cycle of a nanomaterialwhere the particles are released from the agglomerates or aggre-gates.” This concern merits scientific examination. Gray andMuranko (2006) investigated the robustness of two different

Fig. 3. Pulmonary clearance of inhaled particulates. Pulmonary clearance of particlescan be achieved through dissolution or mechanical clearance which can impact theoverall toxicity of the material. Mechanical clearance includes particle transport bymucocilliary activity (primarily upper respiratory), phagocytosis by macrophages andmigration via the mononuclear phagocyte system, or direct translocation of particlesthough respiratory epithelium.

D.R. Boverhof et al. / Regulatory Toxicology and Pharmacology 73 (2015) 137e150144

industrial aggregates (synthetic amorphous silica and carbon black)by evaluating the mechanical forces needed to break down theaggregate structure. They concluded that despite the extremeprocessing of aggregate materials, there was no significant libera-tion of nodules (primary particles), although there was somereduction in size of aggregates by internal fracture. The fracturingwas reported to occur at weak spots in the necks between nodules.Continued application of mechanical energy did further reduceaggregate dimension once the weak necks had been fractured.Thus, in the aggregate materials studied, the fusing together ofprimary particles within aggregate structures is tight and strongand therefore the aggregates represent robust discreet particulates.The authors further concluded that aggregate structures that areable towithstand these intensemechanical forces are unlikely to bebroken down in a biological system in which forces are very weakby comparison.

This conclusion is supported by the work of Maier et al. (2006)who investigated whether lung surfactant may promote disaggre-gation of titanium dioxide particles. They evaluated whether theenergy of interaction between the main component of lung sur-factant and titanium dioxide is sufficient to split the bonds betweenprimary particles of titanium dioxide aggregates and/or the bondsbetween aggregates of titanium dioxide agglomerates. Usingmathematical modeling, they determined that the interaction en-ergy is insufficient to split the bonds. To validate the mathematicalmodel, the authors measured the particle size distribution of tita-nium dioxide suspensions in simulated lung fluid over time andfound no appreciable change in the particle size distribution,indicating that the titanium dioxide did not break down in thesimulated lung fluid. The authors conclude that lung surfactantdoes not promote the disaggregation of titanium dioxide aggre-gates or agglomerates.

Using various exposure modes (i.e. intratracheal instillation andthe inhalation of aerosols with nano-sized or microsized particles),Creutzenberg et al. (2012) investigated the fate of nanoparticleagglomerates after uptake in the lung and found a tendency ofnanoscale particles to form larger size agglomerates followingdeposition and interaction with cells of the respiratory tract. Theauthors concluded that an increase in particle number due to adisintegration of agglomerates does not seem to be of high bio-logical relevance.

Less work has been done in environmental systems, but theresults found in mammalian studies are generally applicable toenvironmental media. The energy found in environmental settingsis not sufficient to break apart aggregates, though it has beendemonstrated that agitation in combination with surfactancy andcontributions to ionic strength provided by natural organic mattercan cause some agglomerates to break apart (Brant et al. 2005; Seealso Taurozi et al., 2011).

Overall, given the current scientific literature, the inclusion ofaggregates and agglomerates in a regulatory definition may notalways be warranted. Specifically, aggregates that have been shownnot to break down into smaller forms under normal industrialprocessing or anticipated use conditions should not be treated asdiscreet nanomaterials for assessment purposes. These aggregatesmay have biological properties that are effectively represented bylarger-scale or bulk forms of the same material. These conceptsshould be considered in the overall identification and assessment ofnanomaterials and will be important in read-across assessments.

3.4. Solubility

Solubility in biological and ecological media is another impor-tant consideration in the evaluation of nanomaterials because ithelps distinguish between characteristics of the material that are

based on molecular identity alone and characteristics that areinfluenced by particle behavior, such as surface activity. Most cur-rent definitions do not consider solubility although the importanceof solubility is recognized in the global guidance document fromthe International Council on Cosmetic Regulations (ICCR) (Rauscheret al., 2012) and is incorporated into the European CommissionCosmetics Directive definition (EC, 2009).

Solubility has not been explicitly defined for nanomaterials.Without a fundamental definition of solubility to identify whichnanomaterials would be soluble and which would not, a practicalapproach may be to use existing guidelines of classifying solubilitysuch as used by the EC's Scientific Committee on Consumer Safety(SCCS) in their Guidance on the Safety Assessment of Nano-materials in Cosmetics (SCCS, 2012). Using this approach, poorlysoluble materials are classified as having an aqueous solubility ofless than 1 mg/L. The corollary of that would be substances equal toor greater than 1 mg/L would be considered soluble.

Rapid dissociation of particles to ions or molecules wouldindicate that the toxicology of the particle (and the assessment ofrisk) should be based primarily on the molecular identity(SCENIHR, 2007) rather than properties of the particle (surfacereactivity or crystallinity, for example). The properties of materialsthat quickly dissociate in biological or ecological fluids are likely tobe dominated by the properties of the molecular and/or ionic form.They will be absorbed, distributed, and eliminated as the molecularform, while those that dissociate slowly will primarily behave asparticulates. For example, zinc oxide and cadmium selenide(quantum dots) particles have been shown to dissociate into ions(Hardman, 2006; Xia et al., 2008; Gulson et al., 2010; Galeone et al.,2012), and the dissociated ions thereby contribute to the biologicalproperties of the particles. When considering solubility for regu-latory purposes, the kinetics of dissolutionwill be important ratherthan total solubility at equilibrium. The rate of solubility willdetermine the amount and form of substance available for possiblebiological activity from the time thematerial is inhaled towhen it iscleared. While a less common phenomenon, it should be noted thatfor some materials, nanoparticles can form from dissolved speciesunder ecological conditions (Slowey, 2010; Akaighe, 2011).

The relationship between dissolution and mechanical clearancemechanisms has been well studied and shown to be important forpredicting the biological effects of inhaled particulate and man-made fibers (Fig. 3) (Oberd€orster, 1989; Davis, 1994; Oberd€orster

D.R. Boverhof et al. / Regulatory Toxicology and Pharmacology 73 (2015) 137e150 145

et al., 1994; Borm et al., 2006). In this context, mechanical clearanceincludes particle transport bymucocilliary activity (primarily upperrespiratory), phagocytosis by macrophages and migration via themononuclear phagocyte system, or direct translocation of particlesthough respiratory epithelium. As an example, consider the dif-ference in potential biological effects between three nanoscalematerials: sea salt (NaCl), zinc oxide (ZnO) and titanium dioxide(TiO2). The salt will dissociate almost instantly upon inhalation andcontact with the mucosa. The pattern of dissociation of ZnO par-ticles is more complex but eventually leads to some proportion ofsolubilization (Lopes et al., 2014), while the nanoscale TiO2 isessentially insoluble with a half-life of weeks in the lung(Oberd€orster et al. 1994, Bermudez et al. 2004). The clearance ofinhaled particulate NaCl begins with rapid dissolution followed byabsorption and elimination of the solubilized ions. Clearance of ZnOis initially by a combination of mechanical means and throughdissolution, while TiO2 clearance is almost exclusively by me-chanical means due to its limited solubility in biological fluids(Bermudez et al., 2004). The physical form of the sea salt exposureis largely unimportant to the biological effects since it is immedi-ately dissociated into its ionic components while the effects of theTiO2 will be determined primarily by its physical form. The po-tential effects of inhaled ZnO will be a mix of particle effects(initially) and the effects from the ionized form of Zn produced bythe dissolution. Thus, the consideration of solubility and dissolutionrate is important in distinguishing between materials that have thepotential for displaying size-dependent particulate effects andthose materials whose effects are due to the molecular formregardless of their initial physical particulate form.

Factors that affect dissolution include composition of the me-dium, ionic strength, temperature, pH, and the presence of ligands,such as proteins, lipids, or formed elements in the blood, that canpreferentially bind to the nanomaterial and shift the kinetics ofclearance (Rauscher et al., 2012). Surface architecture, coatings, andagglomeration state also influence dissolution (Borm et al., 2006;Rauscher et al., 2012). Thus, evaluating the extent and rate ofdissolution is not trivial. Unfortunately, there are no standards fordefining solubility relative to the rate of dissolution. Categoricalstatements such as “insoluble” have little meaning for solid parti-cles. Indeed, there is little agreement about what soluble/insolubleand slow/rapid clearance mean with reference to inhaled particles.For example, amorphous silica has a half-life in the lungs of rats ofaround 50 days yet it is said to be cleared “rapidly” mainly by“dissolution” in lung fluid (Fruijtier-Polloth, 2012). Instead, the rateof dissolution may be more valuable in characterizing whether aparticle remains intact or dissociates into components that maydrive further assessment.

Because the biological effects of materials that rapidly dissolvein biological fluids are primarily related to their chemical propertiesand not their physical form, they would not appear to warrantadditional regulatory interest beyond that given to the non-nanoform of the same materials. Presumably, larger-scale solubleinhaled particles will pass through the nanoscale size range as theydissolve and would exhibit the biological effects relevant to thenanoscale form. Thus, the toxicological information from the largerparticles may also be relevant to the toxicology of the nanoscaleform. Ultimately, solubility and dissolution are important factorsthat need to be considered in the assessment of nanomaterials.

3.5. Intentionally manufactured nanomaterials

Nanoscale particles are not new to nature. They have beenubiquitous from the formation of the earth to the present, and weare constantly surrounded by nanoparticles. For example, the airwe breathe contains tens of thousands of nanoparticles per cubic

centimeter (Buzea et al., 2007; Hochella et al. 2008; Slezakova et al.2013). Some of these nanoparticles are incidental particulates ofman-made origin such as combustion products from the burning ofcoal and petroleum products, residues from abrasion, and evenbaking. Many others are of natural origin, coming from weather,volcanic dust, sea spray, and natural combustion sources such asforest fires. The volume of naturally occurring nanomaterials farexceeds the volume of those produced from man-made sources(Buzea et al., 2007). While it is recognized that some very smallparticles can cause health concerns due their composition and/orconcentration in air, more than the simple presence of very smallparticles is needed to focus attention on these materials.

We recommend that a core element of any definition fornanomaterials in commerce include only those nanomaterials thatare intentionally manufactured (i.e., engineered). Inclusion of thiselement would avoid inappropriate focus on nanomaterials that areproduced incidentally and those from natural sources, allowingstakeholders to focus on materials that are intentionally manu-factured at the nanoscale to have specific properties. ISO (2010)provides useful definitions for two types of nanomaterials, manu-factured and engineered. A manufactured nanomaterial is inten-tionally produced for commercial purposes to have specificproperties or specific composition (e.g., transparency). An engi-neered nanomaterial is designed for specific purpose or function(e.g., quantum dots).

In each case, the intent of manufacturers is to take advantage ofproperties the nanomaterial provides based on its size. Hence it isappropriate to consider intentional manufacture as a core elementin the assessment of the nanomaterial. Most current definitionsinclude this important element (Table 2), but we believe it is onethat should be adopted uniformly.

3.6. Size-dependent properties of nanomaterials

Nanomaterials are typically developed and used to takeadvantage of size-dependent properties that do not exist for com-parable materials of larger sizes and are therefore considered un-usual or novel. For some nanomaterials there are properties thatare size-dependent and are not easily predictable based on theproperties of larger forms of the same material. From Auffan et al.(2009), examples include:

� The melting point of tin is 232 �C when its particle size is>100 nm but at a particle size of 6 nm, its melting point is only14 �C.

� Cadmium selenide is normally a semiconductor, but whenformed into a 5e10 nm quantum dot the movement of electronsis much more limited (quantum confinement) resulting in thegeneration of intense size-dependent visible light emissions(fluorescence) of a narrow band of wavelengths (color) whenexcited by less intense white or blue light.

� Gold is generally considered to be chemically inert, making it anideal material in many applications such as electronic connec-tions and jewelry. Interestingly, gold becomes catalytic when itsparticle size is reduced to the nanoscale.

In contrast, many nanomaterials do not exhibit unusual or novelproperties when compared to their larger scale forms. An exampleof a size-dependent property that is not unusual is a change insurface area compared to mass. This property is easily determinedboth analytically andmathematically, and, for a givenmass, smallerparticles have more surface area than larger particles. An applica-tion of this property is in absorption, where high surface area solidmaterials can be used to absorb more liquid than a comparableamount of low surface area material. Another example is in

D.R. Boverhof et al. / Regulatory Toxicology and Pharmacology 73 (2015) 137e150146

reinforcing agents where solid materials can be added asstrengthening agents in plastics, and smaller sized materials pro-vide more strength for a given mass than larger particles. As noted,this property is size-dependent but is not novel or unusual (Fuet al., 2008).

From an assessment standpoint, the important issue is whetherthe nanomaterial has biologically relevant size-dependent prop-erties that differ from those of the same material at a larger scaleand which make read-across of toxicological data from the largermaterial questionable. Some size-dependent properties may havelittle or no relevance for hazard identification (e.g., melting point,color), while others may be highly relevant (e.g., surface area ac-tivity, solubility). Therefore, we recommended that nanomaterialevaluation frameworks focus on nanomaterials that have unusual(i.e., not readily predictable) size-dependent properties that maymerit additional review.

Many definitions already include the concept of “novel” or“unique” properties or “nanoscale phenomena” (Tables 1 and 2).However, most current definitions provide little or no guidance asto which properties should be considered novel and unique forhazard assessment purposes and fail to adequately consider thedifficulties entailed in accurately measuring these properties. Forexample, surface reactivity is frequently mentioned as an impor-tant contributing factor to the (eco)toxicity of nanomaterials.However, there is no consensus on the most appropriate methodsfor measuring surface reactivity, and available methods may beprone to artifacts and misinterpretations (Horst et al., 2013;Petersen et al., 2014). Some definitions include extension of theseunusual properties to materials greater than 100 nm in size, and itis recommended that in these cases the definition guidance shouldclearly indicate the basis for the inclusion of larger materials.Overall, additional discussion and clarity in this area is critical toensure consistency and transparency in the identification andassessment of nanomaterials.

4. Discussion

Taken together, the elements discussed abovedparticle size, adistributional threshold, intentional manufacture, the state ofagglomeration/deagglomeriation and aggregation/disaggregation,solubility/dissolution rate and precipitation/reformation, andrelevant size-dependent propertiesdrepresent important compo-nents that should be considered when discussing nanomaterialdefinitions or establishing evaluation frameworks. Purely technicalnanomaterial definitions based only on particle size, such as ISO'sdefinition (ISO, 2008), are not useful from a safety evaluationstandpoint because they can include virtually every substance thatmay exist in particulate form. Nanomaterial evaluation frameworksshould be established to help to narrow the scope to those nano-scalematerials that maymerit additional evaluation for purposes ofprotecting human health and the environment while at the sametime having appropriate clarity of scope and being practical toimplement.

Each of the elements discussed in this paper is included in oneor more of the definitions we reviewed; however, they are not usedconsistently and often differ in the scope specified by a particularelement. As a result, a substance considered a nanomaterial underone regulatory scheme may not be considered a nanomaterial inanother. Such inconsistency creates considerable difficulty fornanomaterial manufacturers and users in complying with multipleregulatory requirements and may lead to the perception amongcustomers and the public that nanomaterials are not beingadequately evaluated or regulated. As governmental authoritiesconsider moving forward with discussions on nanomaterial defi-nitions and risk evaluation frameworks, they should strive for

greater alignment and consistency in what is considered to be ananomaterial. Consistency will be instrumental for ensuring thetransparent identification of materials that warrant additionalconsideration, stakeholder confidence in the current regulatoryframeworks, and continued development of nanotechnology.

It is also important to give adequate consideration to ongoingdevelopments in research that informs the human health andenvironmental assessment of nanomaterials. Many of the initialconcerns regarding nanotechnology revolved around un-certainties concerning the potential impact of nanomaterials onhuman health and the environment. From a regulatory perspec-tive, there were two key areas of concern: (1) whether hazard,exposure, and risk assessment methods used for other materialscan be applied to nanomaterials, and (2) whether existing regu-latory frameworks, which do not distinguish materials on thebasis of size, can ensure proper identification, review, and regu-lation of nanomaterials.

The first concern about the applicability of hazard assessmentapproaches for other materials was largely based on concerns thatnanomaterials might harbor unknown modes of toxicity or exhibitnano-specific biological effects that would not be detected throughthe use of current hazard evaluation approaches. Today there is anextensive and growing body of research on nanomaterial toxicitythat addresses this concern. One of the most comprehensive effortshas been implemented through the OECD, which launched a pro-gram of work to ensure that the approaches for hazard, exposure,and risk assessment are appropriate for the evaluation of nano-materials (OECD, 2006). The methods and approaches consideredare applicable to human, animal, and environmental situations.Preliminary conclusions from this work have indicated that currentapproaches for the testing and assessment of traditional chemicalsare appropriate for assessing the safety of nanomaterials (OECD,2009; OECD, 2012). It is acknowledged that certain test methodsmay need to be adapted to allow for appropriate compatibility withnanomaterials. This is in alignment with a recent perspectivewhichindicated that there is no evidence that nanomaterials exhibit novelmechanisms of toxicity and that conventional particle toxicologydata are useful for the determination of nanoparticulate hazardpotential (Donaldson and Poland, 2013). Such conclusions repre-sent important advancements in our understanding of nano-materials and should be considered as discussions on safetyevaluation and the need for definitions continue to evolve.

With respect to the second concern, the ability of current reg-ulatory frameworks to ensure proper identification and review ofnanomaterials, the OECD has recommended that its membercountries apply existing international and national frameworks tomanage any risks associated with the manufacture and use ofnanomaterials (OECD, 2013b). As part of its recommendation, OECDnotes that current frameworks may need to be adapted; however,completely new frameworks are not needed for nanomaterials.Efforts in the OECD are now focused on nanomaterial evaluationapproaches including those related to identification, characteriza-tion and read-across assessment approaches (OECD, 2013a).

While the discussion on elements of nanomaterials that may beimportant for safety evaluation is important, we believe manygroups have focused on the development of definitions for identi-fication of nanomaterials without consideration of how theywill beused. If definitions are developed, they should be developed inconjunction with plans for implementation. In this sense, demon-stration of how definitions would be applied through the use ofcase studies may be informative.Without such a parallel strategy, itis difficult to understand the true scope of the definition or howspecific elements of the definition can best be addressed, especiallyconsidering the complex analytical characterization that may berequired.

D.R. Boverhof et al. / Regulatory Toxicology and Pharmacology 73 (2015) 137e150 147

This concept was introduced, in part, within the EuropeanCommission recommendation (EC, 2011a), which states that therecommendation should be used as a reference for determiningwhether a material should be considered as a “nanomaterial” forlegislative and policy purposes. The Commission appears torecognize that its recommendation should not simply be imple-mented as written, but that individual regulationsmay have uniquerequirements that a single, rigid definition may not satisfy. TheCommission also states that the identification of a nanomaterialshould not prejudge or reflect the scope of legislation or provisionsfor potentially establishing additional requirements for nano-materials, including those relating to risk management. The Com-mission goes on to state that in some cases it may be necessary toexclude materials from the scope of any legislation even if they fallwithin the definition (EC, 2011). These statements clearly indicatethat implementation of the definition as part of an evaluationframework will require additional guidance and scope refinement.

A recent example of the complexity associated with imple-menting a regulatory definition can be seen with the French Reg-istry for nanomaterials. The registry requires reporting ofnanomaterials marketed in France and leverages the EuropeanCommission's recommended definition (ANSES, 2012). Imple-menting the registry required extensive additional guidancebeyond the definition to determine what was within scope andtherefore reportable. Initial guidance was supplemented byextensive question and answer documentation during the actualreporting process. For example, one point of clarity focused on thedefinition term “intentionally manufactured”whichwas to indicate

Fig. 4. Example approach for implementation of a nanomaterial definition into a regulatoauthority and moves beyond identification towards a framework for assessment. Importantlyevaluated to ensure that they are safe for their intended uses. While definitions may not inpresented in this paper (listed on the left of the figure) should be considered at some point dreview.

the deliberate manufacturing of a substance at the nanoscale andexcluded nanomaterials that may be unintentional by-products ofthe synthesis process. Another point of clarity was provided on thelack of a need to declare a substance if it was bound in a mixturefrom which it was not expected to be released or extracted undernormal or foreseeable conditions of use. This exclusion highlightsthe consideration of release and/or exposure potential as part of theframework for identification of materials, which is an importantcomponent in the evaluation of potential downstream risk. Thisexample represents the activity of a single country within the Eu-ropean Union, but additional country-specific registries have beenimplemented (Tables 1 and 2). Clearly, any variation on the use ofdefinitions and how they are implemented within and amongdifferent regions can lead to inefficiency, inconsistency, andmiscommunication.

An additional example illustrating the challenge of a regulatorydefinition comes from a proposed rule issued by the US Environ-mental Protection Agency (US EPA, 2011a). The Agency proposed arule that sought to obtain information pertaining to the use ofnanomaterials in pesticide products. However, it was noted in oneof the public comments that the lack of clarity in how the Agencydefined a nanomaterial at that time would make it difficult forthose who develop pesticide products to know if their materialswere considered to be nanomaterials (ACC, 2011). The authors ofthe response recommended that the Agency consider developing amore clear definition using many of the concepts explained above.

These examples highlight the difficulty of implementing adefinition, which in turn provides support for considering the

ry decision framework. The framework starts with the definition from the regulatory, regardless of the definition used, all substances and materials should be appropriatelycorporate all of the elements that are described above, we believe that the elementsuring the evaluation process to determine if the material requires additional regulatory

D.R. Boverhof et al. / Regulatory Toxicology and Pharmacology 73 (2015) 137e150148

regulatory implementation of a definition as part of the develop-ment process. Implementation of a definition without integrationinto a regulatory framework results in the identification of a list ofmaterials without purpose or action, which could lead to unnec-essary stigmatization of nanomaterials as a whole. The potential forformation of such perspectives is well-recognized and specificallyhighlighted as an area of concern within a White House memo-randum on regulatory frameworks for nanomaterials (Holdren,2011). The memorandum notes that agencies should avoid mak-ing unfounded generalizations that categorically judge nano-materials as intrinsically benign or harmful. Importantly,nanomaterials are not inherently hazardous, and therefore defini-tions and frameworks that intend to identify nanomaterials shouldbe established to avoid such generalizations and avoid establishingnanomaterial inventories without purpose. Even for those nano-materials for which health or environmental hazards have beenidentified, it is important to recognize that the actual health orenvironmental risk will be determined by the potential for expo-sure to the material. The concept of considering both hazard andexposure as a determination of risk is inherent in the regulatoryprocess and should be maintained for the assessment ofnanomaterials.

Ultimately, if a definition or definition system for identifyingnanomaterials is needed, its application within the currentassessment framework should be addressed. An example of how ananomaterial definition can be incorporated into a risk-basedassessment framework is presented in Fig. 4. The first step is todetermine whether the substance falls into the scope of interest fornanomaterials. If a material meets the criteria for a potentialnanomaterial (which may be specified by certain definition ele-ments), the next step should be to determine whether it is a new orexisting chemical substance according to the applicable regula-tions. If the nanomaterial is a new substance, the manufacturer orimporter must follow the standard notification and registrationprocess to ensure that the nanomaterial undergoes regulatory re-view. Registrations of new nanomaterials should include availableinformation that may be relevant to the evaluation of potentialhazards and risks.

If the nanomaterial is considered an existing chemical substanceunder the applicable regulations, the next level of assessmentshould be to determine whether existing toxicology and/or safetydata for the nanomaterial (or the larger form of the same material)are available and adequate for assessing potential human health orenvironmental risks under anticipated or foreseeable exposurescenarios. Examples of existing nanomaterials for which robusttoxicology datasets are available include carbon black, syntheticamorphous silica, and zinc oxide, all of which have been used fordecades and evaluated in numerous toxicological and epidemio-logical studies. The potential risks and appropriate safe handlingpractices for these so-called “historical” nanomaterials are wellunderstood. Except in unusual cases (e.g., a previously untestedroute of exposure), additional regulatory review of these nano-materials would seem to offer little added value and divert scarceresources. This concept is consistent with existing perspectives thatsome materials should be excluded from any developing nano-material legislation (EC, 2011).

If the nanomaterial does not have a specific toxicological dataset sufficient for evaluating potential risks, the next consideration iswhether the nanoscale material possesses novel size-dependentproperties relative to larger forms of the same material and, if so,whether such properties could plausibly influence the hazard/exposure/risk profile of the nanomaterial relative to its largerforms. As discussed previously, some nanomaterials do not possessnovel size-dependent properties (i.e., ones not readily predictablebased on simple size scaling), and some novel properties (e.g., color

differences observed at the nano-scale) have no known or plausiblecorrelation to hazard or risk. Additional elements could also beconsidered at this stage (if not previously addressed as part of thescoping stage), including solubility/dissolution and impact of ag-gregation/agglomeration. Doing so provides an opportunity toevaluate the material with consideration of the ability to use read-across approaches to predict the hazards/risks of the nanomaterialbased on available information from the larger-scale material. Asnoted previously, the size boundaries for defining nanomaterialsare arbitrary, and many material properties are predictable andscalable as their size decreases from >100 nm to smaller sizes. Onthe other hand, if a nanomaterial possesses novel size-dependentproperties that could plausibly influence hazard or risk, read-across from the larger-scale material may be inappropriate, andadditional information on the nanomaterial should be gathered orgenerated.

Throughout this proposed approach and regardless of the defi-nition used, it is critical at each stage of the evaluation process toensure confidence in the hazard and risk assessment of the sub-stance, consistent with good product stewardship and the princi-ples of Responsible Care®, to ensure that all substances are safe fortheir intended use (ACC, 2014; ICCA, 2014). Furthermore, it isacknowledged that more detailed guidance than that presentedhere will be necessary as part of a nanomaterial evaluationframework, regulatory or otherwise (e.g., Environmental DefenseFund and DuPont, 2007; Oomen et al., 2014). While emergingdefinitions may not incorporate all of the elements that aredescribed above, we believe that the elements presented in thispaper should be considered at some point during the evaluationprocess to determine if the material requires additional review.

Transparency document

Transparency document related to this article can be foundonline at http://dx.doi.org/10.1016/j.yrtph.2015.06.001.

References

ACC, 2011. Comments Submitted in Response to: Pesticides; Policies ConcerningProducts Containing Nanomaterials; OpportunA-HQ-OPP-20ity for PublicComment (EP10-0197). American Chemistry Council. Available at: http://www.regulations.gov/#!documentDetail;D¼EPA-HQ-OPP-2010-0197-0179. Accessed14.01.14.

ACC, 2014. Responsible Care. American Chemistry Council. Available at: http://responsiblecare.americanchemistry.com. Accessed 14.01.14.

Akaighe, N., MacCuspie, R., Navarro, D., Aga, D., Banerjee, S., Sohn, M., Shar,a, V.,2011. Humic acid-induced silver nanoparticle formation under environmentallyrelevant conditions. Environ. Sci. Technol. 45, 3895e3901.

ANSES, 19 February 2012. Ministry of Ecology, Sustainable Development, Trans-portation, and Housing. Decree no. 2012-232 of 17 February 2012 on the annualdeclaration on substances at nanoscale in application of article R. 523-4 of theEnvironment code. Off. J. Fr. Repub. Text 4/44. Available at: https://www.r-nano.fr/?locale¼en Accessed 14.11.13.

Auffan, M., Rose, J., Bottero, J.Y., Lowry, G.V., Jolivet, J.P., Wiesner, M.R., 2009. To-wards a definition of inorganic nanoparticles from an environmental, healthand safety perspective. Nat. Nanotechnol. 4, 634e641.

Belgium, 2014. Royal Decree Concerning the Placing on the Market of SubstancesProduced in Nanoparticular State. Federal Public Service Health, Food ChainSafety and Environment, Kingdom of Belgium.

Bermudez, E., Mangum, J.B., Wong, B.A., Asgharian, B., Hext, P.M., Warheit, D.B.,Everitt, J.I., 2004. Pulmonary responses of mice, rats, and hamsters to sub-chronic inhalation of ultrafine titanium dioxide particles. Toxicol. Sci. 77,347e357.

Bleeker, E.A., de Jong, W.H., Geertsma, R.E., Groenewold, M., Heugens, E.H., Koers-Jacquemijns, M., van de Meent, D., Popma, J.R., Rietveld, A.G., Wijnhoven, S.W.,Cassee, F.R., Oomen, A.G., 2013. Considerations on the EU definition of ananomaterial: science to support policy making. Regul. Toxicol. Pharmacol. 65,119e125.

Borm, P., Klaessig, F.C., Landry, T.D., Moudgil, B., Pauluhn, J., Thomas, K., Trottier, R.,Wood, S., 2006. Research strategies for safety evaluation of nanomaterials, partV: role of dissolution in biological fate and effects of nanoscale particles. Tox-icol. Sci. 90, 23e32.

D.R. Boverhof et al. / Regulatory Toxicology and Pharmacology 73 (2015) 137e150 149

Brinker, C., Hurd, A., Schunk, P., Frye, G., Ashley, C., 1992. Review of sol-gel thin filmformation. J. Non-Cryst. Solids 147, 424e436.

Brant, J., Lecoanet, H., Hotze, Wiesner, M., 2005. Comparison of electrokineticproperties of colloidal fullerenes (n-C60) formed using two procedures. Environ.Sci. Technol. 39, 6343e6351.

Brown, S.C., Boyko, V., Meyers, G., Voetz, M., Wohlleben, W., 2013. Towardadvancing nano-object count metrology: a best practice framework. Environ.Health Perspect. 121, 1282e1291.

Buzea, C., Pacheco, I.I., Robbie, K., 2007. Nanomaterials and nanoparticles: sourcesand toxicity. Biointerphases 2, MR17e172.

Creutzenberg, O., Bellmann, B., Korolewitz, R., Koch, W., Mangelsdorf, I., Tillmann, T.,Schaudien, D., 2012. Change in agglomeration status and toxicokinetic fate ofvarious nanoparticles in vivo following lung exposure in rats. Inhal. Toxicol. 24,821e830.

Davis, J.M., 1994. The role of clearance and dissolution in determining the durabilityor biopersistence of mineral fibers. Environ. Health Perspect. 102 (Suppl. 5),113e117.

Denmark, 2014. Order on a Register of Mixtures and Articles that Contain Nano-materials as Well as the Requirement for Producers and Importers to Report tothe Register. Danish Environmental Protection Agency Statutory. Order no. 644of 13 June 2014.

Donaldson, K., Poland, C.A., 2013. Nanotoxicity: challenging the myth of nano-specific toxicity. Curr. Opin. Biotechnol. 24, 724e734.

Donnet, J., Bansal, R.C., Wang, M.J., 1993. Carbon Black. Marcel Dekker, Inc., NewYork.

Dur~aes, L., Benedini, L., Costa, B., Portugal, A., 2012. Sol-gel synthesis and washing ofamorphous g-FeO (OH) xerogels. Mater. Werkst. 43, 427e434.

EC (2009). European Commission, 12 December 2009. Regulation (EC) No 1223/2009 of the European Parliament and of the Council of 30 November 2009 oncosmetic products. Off. J. Eur. Union (EN) 52, 59e209. L342.

EC, 2011a. European Commission. Commission Recommendation of 18 October2011 on the definition of nanomaterials. Off. J. Eur. Union (EN) 54, 38e40, 20October 2011. L275.

EC, 2011b. European Commission. Regulation (EU) No 1169/2011 of the EuropeanParliament and of the Council of 25 Ocboer 2011 on the provision of food in-formation to consumers, amending Regulations (EC) No 1924/2006 and (EC) no1925/2006 of the European Aprliament and of the Council, and repealingCommission Directive 87/250/EEC, Council Directive 90/496/EEC, CommissionDirective 1999/10/EC, Directive 2000/13/EC of the European Parliament of thethe Council, Commission Directives 2002/67/EC and 2008/5/EC and Commis-sion Regulation (EC) No 608/2004. Official J. Eur. Union (EN) 18e63, 22November 2011. L304.

EC (2012). European Commission, 27 June 2012. Regulation (EU) No 528/2012 of theEuropean Parliament and of the Council of 22 May 2012 concerning the makingavailable on the market and use of biocidal products. Off. J. Eur. Union (EN)1e123. L167.

Environmantal Defense Fund and Dupont, 2007. NANO Risk Framework. Availableat: http://www.nanoriskframework.com. Accessed 19.05.15.

Fruijtier-Polloth, C., 2012. The toxicological mode of action and the safety of syn-thetic amorphous silica-a nanostructured material. Toxicology 294, 61e79.

Fu, S.Y., Feng, X.Q., Lauke, B., Mai, Y.W., 2008. Effects of particle size, particle/matrixinterface adhesion and particle loading on mechanical properties of partic-ulateepolymer composites. Compos. Part B: Eng. 39, 933e961.

Galeone, A., Vecchio, G., Malvindi, M.A., Brunetti, V., Cingolani, R., Pompa, P.P., 2012.In vivo assessment of CdSe-ZnS quantum dots: coating dependent bio-accumulation and genotoxicity. Nanoscale 4, 6401e6407.

Gordon, S.C., Butala, J.H., Carter, J.M., Elder, A., Gordon, T., Gray, G., Sayre, P.G.,Schulte, P.A., Tsai, C.S., West, J., 2014. Workshop report: strategies for settingoccupational exposure limits for engineered nanomaterials. Reg. Tox. Pharm.68, 305e311.

Gray, C.A., Muranko, H., 2006. Studies of robustness of industrial aciniform aggre-gates and agglomerates: carbon black and amorphous silicas: a review ampli-fied by new data. J. Occup. Environ. Med. 48, 1279e1290.

Gulson, B., McCall, M., Korsch, M., Gomez, L., Casey, P., Oytam, Y., Taylor, A.,McCulloch, M., Trotter, J., Kinsley, L., Greenoak, G., 2010. Small amounts of zincfrom zinc oxide particles in sunscreens applied outdoors are absorbed throughhuman skin. Toxicol. Sci. 118, 140e149.

Hamburg, M.A., 2012. FDA's approach to regulation of products of nanotechnology.Science 336, 299e300.

Hardman, R., 2006. A toxicologic review of quantum dots: toxicity depends onphysicochemical and environmental factors. Environ. Health Perspect. 114,165e172.

Health Canada, 2011. Policy Statement on Health Canada's Working Definition forNanomaterials. Avaialble at. http://www.hc-sc.gc.ca/sr-sr/pubs/nano/pol-eng.php. Accessed 14.11.13.

Ho, M., Wu, K.Y., Chein, H.M., Chen, L.C., Cheng, T.J., 2011. Pulmonary toxicity ofinhaled nanoscale and fine zinc oxide particles: mass and surface area as anexposure metric. Inhal. Toxicol. 23, 947e956.

Hochella Jr., M.F., Lower, S.K., Maurice, P.A., Penn, R.L., Sahai, N., Sparks, D.L.,Twining, B.S., 2008. Nanominerals, mineral nanoparticles, and earth systems.Science 319, 1631e1635.

H€ock, J., Epprecht, T., Furrer, E., Gautschi, M., Hofmann, H., H€ohener, K., Knauer, K.,Krug, H., Limbach, L., Gehr, P., Nowack, B., Riediker, M., Schirmer, K., Schmid, K.,Som, C., Stark, W., Studer, C., Ulrich, A., von G€otz, N., Weber, A., Wengert, S.,Wick, P., 2013. Guidelines on the Precautionary Matrix for Synthetic

Nanomaterials. Federal Office of Public Health and Federal Office for the Envi-ronment, Berne. Version 3.0. Available at: http://www.bag.admin.ch/nanotechnologie/12171/12174/index.html?lang¼en. Accessed 18 May 2015.

Holdren, J.P., June 9, 2011. Memorandum for the Heads of Executive Departmentsand Agencies: Policy Principles for the U.S. Decision-making ConcerningRegulation and Oversight of Application of Nanotechnology and Nanomaterials.Available at: http://www.whitehouse.gov/sites/default/files/omb/inforeg/for-agencies/nanotechnology-regulation-and-oversight-principles.pdf. Accessed10.01.14.

Horst, A.M., Vukanti, R., Priester, J.H., Holden, P.A., 2013. An assessment of fluo-rescence- and absorbance-based assays to study metal-oxide nanoparticle ROSproduction and effects on bacterial membranes. Small 9, 1753e1764.

ICCA, 2010. Core Elements of a Regulatory Definition of Manufactured Nano-materials. International Council of Chemical Associations. Available at: http://www.icca-chem.org/ICCADocs/Oct-2010_ICCA-Core-Elements-of-a-Regulatory-Definition-of-Manufactured-Nanomaterials.pdf?epslanguage¼en. Accessed10.01.14.

ICCA, 2014. Responsible Care. International Council of Chemical Associations. http://www.icca-chem.org/en/Home/Responsible-care/. Accessed 14.01.14.

ISO, 2008. International Organization for Standardization. Technical Specification:Nanotechnologiesdterminology and Definitions for Nano-objectsdNanoparti-cle, Nanofibre and Nanoplate. ISO/TS 80004-2:2008.

ISO, 2010. International Organization for Standardization. Nano-technologiesdvocabularydpart 1: Core Terms. ISO/TS 80004-1:2010.

JRC, 2014. Joint Research Center. Towards a Review of the EC Recommendation for aDefinition of the Term “nanomaterial” Part 2: Assessment of Collected Infor-mation Concerning the Experience with the Definition. EUR 26744 EN. Avaialbleat: https://ec.europa.eu/jrc/sites/default/files/jrc_nm-def_report2_eur26744.pdf. Accessed 19.05.15.

Knieke, C., Berger, A., Voigt, M., Taylor, R.N.K., R€ohrl, J., Peukert, W., 2010. Scalableproduction of graphene sheets by mechanical delamination. Carbon 48,3196e3204.

Linsinger, T.P.J., Roebben, G., Gilliland, D., Calzolai, L., Rossi, F., Gibson, N., Klein, C.,2012. Requirements on Measurements for the Implementation of the EuropeanCommission of the Term “nanomaterial”. European Commission Joint ResearchCentre. Reference Report. Report EUR 25404 EN.

Lopes, S., Ribeiro, F., Wojnarowicz, J., Lojkowski, W., Jurkschat, K., Crossley, A.,Soares, A.M., Loureiro, S., 2014. Zinc oxide nanoparticles toxicity to Daphniamagna: size-dependent effects and dissolution. Environ. Toxicol. Chem. 33,190e198.

Luther, W., 2004. Industrial Application of Nanomaterials e Chances and Risks.Technology Analysis, Future Technologies Division of VDI TechnologiezentrumGmbH, No. 54, Dusseldorf.

Maier, M., Hannebauer, B., Holldorff, H., Albers, P., 2006. Does lung surfactantpromote disaggregation of nanostructured titanium dioxide? J. Occup. Environ.Med. 48, 1314e1320.

Maynard, A.D., 2011. Don't define nanomaterials. Nature 475, 31.NICNAS, 2010. Australian Government, Department of Health, National Industrial

Chemicals Notification and Assessment Scheme. Chemical Gazette of 5 October2010. Available at: http://www.nicnas.gov.au/communications/publications/chemical-gazette/chemical-gazette-archive. Accessed 14.11.13.

Norway Environment Agency, 2014. What Chemicals Must Be Reported?. Availableat: http://www.miljodirektoratet.no/no/Tema/Kjemikalier/Produktregisteret/The_Product_Register/What-Chemicals-Must-Be-Reported/. Accessed 18.05.15.

Oberd€orster, G., 1989. Dosimetric principles for extrapolating results of rat inhala-tion studies to humans, using an inhaled Ni compound as an example. HealthPhys. 57 (Suppl. 1), 213e220.

Oberd€orster, G., Ferin, J., Lehnert, B.E., 1994. Correlation between particle size,in vivo particle persistence, and lung injury. Environ. Health Perspect. 102(Suppl. 5), 173e179.

OECD, 2006. Report of the OECD Workshop on the Safety of Manufactured Nano-materials: Building Co-operation, Co-ordination and Communication. Organi-sation for Economic Co-operation and Development. ENV/JM/MONO(2006)19.28 April 2006.

OECD, 2009. Preliminary Review of OECD Test Guidelines for Their Applicability toManufactured Nanomaterials. Organisation for Economic Co-operation andDevelopment. ENV/JM/MONO(2009)21. 10 July 2009.

OECD, 2012. Important Issues on Risk Assessment of Manufactured Nanomaterials.Organisation for Economic Co-operation and Development. ENV/JM/MONO(2012)8. 28 March 2012.

OECD, 2013a. Co-Operation on Risk Assessment: Prioritisation of Important Issueson Risk Assessment of Manufactured Nanomaterials. Organisation for EconomicCo-operation and Development. Final Report. ENV/JM/MONO(2013)18. 21August 2013.

OECD, 2013b. Recommendation of the Council on the Safety Testing and Assess-ment of Manufactured Nanomaterials. Organisation for Economic Co-operationand Development. C(2013)107. 19 September 2013.

Oomen, A.G., Bos, P.M., Fernandes, T.F., Hund-Rinke, K., Boraschi, D., Byrne, H.J.,Aschberger, K., Gottardo, S., von der Kammer, F., Kuhnel, D., Hristozov, D.,Marcomini, A., Migliore, L., Scott-Fordsmand, J., Wick, P., Landsiedel, R., 2014.Concern-driven integrated approaches to nanomaterial testing and assessment:report of the NanoSafety Cluster Working Group 10. Nanotoxicology 8,334e348.

Pauluhn, J., 2009. Retrospective analysis of 4-week inhalation studies in rats withfocus on fate and pulmonary toxicity of two nanosized aluminum

D.R. Boverhof et al. / Regulatory Toxicology and Pharmacology 73 (2015) 137e150150

oxyhydroxides (boehmite) and pigment-grade iron oxide (magnetite): the keymetric of dose is particle mass and not particle surface area. Toxicology 259,140e148.

Pauluhn, J., 2010. Subchronic 13-week inhalation exposure of rats to multiwalledcarbon nanotubes: toxic effects are determined by density of agglomeratestructures, not fibrillar structures. Toxicol. Sci. 113, 226e242.

Petersen, E.J., Henry, T.B., Zhoa, J., MacCuspie, R.I., Kirschling, T.L.,Dobrovolskaia, M.A., Hackley, V., Xing, B., White, J.C., 2014. Identification andavoidance of potential artifacts in nanomaterial ecotoxicity measurements.Environ. Sci. Technol. 48, 4226e4246.

Rauscher, H., Ansell, J., Araki, D., Bronaugh, R., Ikarashi, Y., Lefebvre, M.-A.,Ogasawara, H., Suwa, T., Tayabali, A.F., Tooley, I., Yasuda, N., 2012. ICCR WorkingGroup: Characterization of Nanomaterials II e Insolubility, Biopersistence andSize Measurement in Complex Meida. International Cooperation on CosmeticRegulations. ICCR/NANOCHAR1/R/FINAL/. Available at: http://iccrnet.org/topics/. Accessed 01.05.15.

SCCS, 2012. Guidance on the Safety Assessment of Nanomaterials in Cosmetics.Scientific Committee on Consumer Safety. Available at: http://ec.europa.eu/health/scientific_committees/consumer_safety/docs/sccs_s_005.pdf. Accessed08.04.15.

SCENIHR, 2007. Opinion on the Appropriateness of the Risk Assessment Method-ology in Accordance with the Technical Guidance Documents for New andExisting Substances for Assessing the Risks of Nanomaterials. 21e22 June 2007.Scientific Committee on Emerging and Newly-Identified Health Risks.

Slezakova, K., Morais, S., Pereira, M.C., 2013. Atmospheric nanoparticles and theirimpacts on public health. In: Rodriguez-Morales, A. (Ed.), Current Topics inPublic Health. InTech, ISBN 978-953-51-1121-4. Chapter 23, Available from:http://www.intechopen.com/books/current-topics-in-public-health/atmospheric-nanoparticles-and-their-impacts-on-public-health.

Slowey, A., 2010. Rate of formation and dissolution of mercury sulfide nano-particles: the dual role of natural organic matter. Geochim. Cosmochim. Acta 74,4693e4708.

Stamm, H., 2011. Risk factors: nanomaterials should be defined. Nature 476, 399.Stoeger, T., Reinhard, C., Takenaka, S., Schroeppel, A., Karg, E., Ritter, B., Heyder, J.,

Schulz, H., 2006. Instillation of six different ultrafine carbon particles indicates asurface area threshold dose for acute lung inflammation in mice. Environ.Health Perspect. 114, 328e333.

Taiwan, 2012. Guidance for Supplementary Existing Chemical Substance Nomina-tion (SECN), first ed. Taiwan Council of Labor Affairs. (既有化學物質增補提報作業

指引).Taurozi, J., Hackley, V., Wiesner, M., 2011. Ultrasonic dispersion of nanoparticles for

environmental, health and safety assessment: issues and recommendations.Nanotoxicology 5, 711e729.

Comparison of number and respirable mass concentration determinations. In:Tomb, T., Haney, R. (Eds.), 1988. Advances in Air Sampling. American Conferenceof Governmental Industrial Hygienists. Lewis Publishers, Chelsea, MI.

Tsuji, J.S., Maynard, A.D., Howard, P.C., James, J.T., Lam, C.W., Warheit, D.B.,Santamaria, A.B., 2006. Research strategies for safety evaluation of nano-materials, part IV: risk assessment of nanoparticles. Toxicol. Sci. 89, 42e50.

US EPA, 2007. Nanotechnology White Paper. U.S. Environmental Protection Agency.EPA 100/B-07/001. Accessed 14.11.13 at. http://www.epa.gov/osa/pdfs/nanotech/epa-nanotechnology-whitepaper-0207.pdf.

US EPA, 17 June 2011. Pesticides; policies concerning products containing nanoscalematerials; opportunity for public comment (2011a). Fed. Regist. 76 (14),35383e35395.

US EPA, 28 December 2011. U.S. Environmental Protection Agency. Proposed sig-nificant new use rules on certain chemical substances (2011b). Fed. Regist. 76(249), 81441e81447.

US EPA, 6 April 2014. U.S. Environmental Protetion Agency. Chemical substanceswhen manufactured or processed as nanoscale materials; TSCA reporting andrecordkeeping requirements (2014). Fed. Regist. 80 (65), 18330e18342.

US FDA, 2011. Considering Whether an FDA-regulated Product Involves the Appli-cation of Nanotechnology: Guidance for Industry: Draft Guidance. U.S. Food andDrug Administration. Available at: http://www.fda.gov/RegulatoryInformation/Guidances/ucm257698.htm. Accessed 14.11.13.

Wittmaack, K., 2007. In search of the most relevant parameter for quantifying lunginflammatory response to nanoparticle exposure: particle number, surface area,or what? Environ. Health Perspect. 115, 187e194.

Xia, T., Kovochich, M., Liong, M., Madler, L., Gilbert, B., Shi, H., Yeh, J.I., Zink, J.I.,Nel, A.E., 2008. Comparison of the mechanism of toxicity of zinc oxide andcerium oxide nanoparticles based on dissolution and oxidative stress proper-ties. ACS Nano 2, 2121e2134.

Young, S., January 2002. Overview of Sol-Gel Science and Technology (2002). ArmyResearch Laboratory, Maryland. ARL-TR-2650.