Proposed Chemical Classification Scheme
In 2009, prior to the establishment of the RCC Nanotechnology Initiative, the US EPA worked independently on the development of a nanomaterial classification scheme based on similarities in chemical composition, drawing primarily on scientific literature and submissions to their program. This classification scheme was further developed in concert with their Canadian counterparts at Environment Canada and Health Canada under the RCC Nanotechnology Initiative’s Work Element 2, and resulted in the chemical classification scheme proposed in this document.
This classification scheme will be used: (1) to focus on nanomaterials which are expected to typically behave differently on the nanometer scale; and, (2) to select appropriate analogue/read-across information within a class of nanomaterials. As the science develops, approaches for selecting analogue/read-across information across different classes will be considered.
Experts and stakeholders were asked to provide input into the proposed classification scheme during the March 20, 2013 RCC Nanotechnology Initiative Workshop. The refined classification scheme, reflecting that input, is presented below in Figure 2. Stakeholders and experts at the workshop agreed that this proposed classification scheme is an appropriate starting point7.
By identifying these classes of nanomaterials, the Canada/US Programs are indicating which nanomaterials they believe behave differently from their non-nanoscale forms. For example, gold nanoparticles display different properties as compared to bulk gold. Substances such as organic polymers and pigments, however, have not typically been found to exhibit unique nanoscale properties/phenomena. (Those substances have been on the nanometer scale due to their synthetic route, and, as such, have undergone traditional chemical risk assessments.) The classes in Figure 2 are not exhaustive and will be modified as new nanomaterials are notified to the Canada/US Programs, and as the scientific knowledge increases.
Hybrid nanomaterials (for example, a carbon nanotube with a metal oxide surface modification which displays unique behaviour) are not part of the proposed classification scheme as they fall into multiple chemical composition classes. For that reason, all hybrid nanomaterials will continue to be assessed on a case-by-case basis without the inclusion of analogue/read-across information.
The class for organics (section 3.6) was added to the classification scheme based on stakeholder feedback received at the March 20, 2013 RCC Nanotechnology Initiative Workshop ; it represents an emerging area of nanotechnology, one which the Canada/US Programs know relatively little about at this point. That is expected to change as scientific knowledge increases, especially with the emergence of nano-cellulosic materials into the marketplace.
Figure 2: Proposed classification scheme based on similarities in chemical composition.
The blue boxes in Figure 2 represent the classes the Canada/US Programs and their stakeholders believe that nanomaterials can fall into, based on similarities in chemical composition. The physiochemical parameters listed (in the white boxes) represent the intrinsic physicochemical parameters which must be similar between two nanomaterials for them to be considered for analogue/read-across information (e.g., if two nanomaterials within a class have the same physiochemical parameters, it is likely that they will have similar properties and behavior in, for instance, wastewater treatment plants).
In this document, the terms “solubility”, “degradation” of the nanomaterial surface, and “dissolution” are used interchangeably and are meant to broadly encompass the release of ions from the nanoparticle in solvent media over time. It is important to note that measuring solubility is still complex (e.g., measuring dispersions vs. solubility vs. dissolution). “Size” refers to primary particle size or mean size in the absence of primary particle size information. “Surface modification” refers to chemical changes to the surface of the nanoparticle (including surface oxidation, chemical functionalization, etc.). “Surface chemistry” refers to the properties of the surface (e.g., surface charge). Both surface chemistry and surface modification are integrally linked and inter-dependent.
In addition to supporting analogue/read-across information, many of these physiochemical parameters are important markers for understanding nanomaterial fate and behavior during risk assessments. As such, it is likely that the Canada/US Programs will request all relevant information (in the case of the identification of analogues/read-across, all physiochemical parameters listed are required) during the regulatory process. Other information, including extrinsic parameters - such as aggregation, agglomeration, and de-agglomeration - which are used in the risk assessment process but are not part of this classification scheme may, nonetheless, be requested during the regulatory process to help assess hazard, fate, and effects.
The classification scheme presented in this document should not be used to infer toxicological modes of action for nanomaterials as the science for this is still emerging (e.g., Nel et al. (2012)). The work by Nel and others on toxicological modes of action still needs to be evaluated for reproducibility both within the proposed classes and across different classes to determine their applicability in classification processes. However, available information on two nanomaterials which have been found to have similar chemical composition and parameters, as outlined in Figure 2, could be used to increase the weight-of-evidence to support toxicology assessments. (For example, if two submissions for multi-walled carbon nanotubes (MWCNTs) had sufficiently similar* physiochemical parameters (as listed in Figure 2), data from one could potentially be used as read-across for the other to increase the weight-of-evidence in the assessment). In addition, while still very early, it may also be of benefit to consider extrapolations between nanoparticles of different compositions if their physicochemical parameters are sufficiently similar* within a class. (For example, if titanium dioxide and silicon dioxide display the same physicochemical parameters, could they also potentially display the same environmental fate?)
Sections 3.1-3.7 contain information on each of the nanomaterial classes listed in Figure 2, including how differences in their physiochemical parameters can lead to differences in fate and effects.
Carbon nanotubes (CNTs) are typically described as seamlessly rolled sheets of graphite. These rolls can be of single sheets (single-walled carbon nanotubes, or SWCNTs), or multiple sheets (double and MWCNTs). Both the Canadian New Substance Program and US New Chemicals Program consider CNTs as new substances that do not have any non-nano counterparts (this includes graphite and graphene) and, as such, have assessed and will continue to assess each CNT (SWCNT and MWCNT) individually.
There exists scientific information showing links between the physicochemical parameters outlined in Figure 2 for CNTs and their fate and effects.
- Length and diameter (aspect ratio) have been demonstrated to be physical features of CNTs that are considered determinants for their pulmonary toxicity. Bussy et al. (2012) showed linkages between the changes in surface modifications and surface chemistry as a result of changing aspect ratios of CNTs, and corresponding CNT-induced inflammation.
- Surface modification and surface chemistry: Pasquini et al. (2012) have demonstrated the differences in cell viability (invitro toxicological endpoint) as a function of surface modification and surface chemistry on SWCNTs, suggesting that surface chemistry and surface modifications may be an important parameter in understanding CNTs behavior.
- Liu et al. (2013) have reviewed the physicochemical parameters important in understanding the toxicity of CNTs. The number of walls and reactivity, driven in part by the CNT end-caps (capped/uncapped) and chirality, were found to be important factors in understanding effects.
The Canada/US Programs have concluded that the physicochemical parameters listed in Figure 2 are important to distinguish CNTs within the same class. The examples cited clearly demonstrate that differences in these parameters can lead to differences in behaviors. If these parameters are the same, or sufficiently similar (see footnote), it is expected that analogue/read-across information can be used.
This approach was recently used on a CNT assessment in Canada. Through the selection of an appropriate analogue using the criteria above, differences in CNT environmental behavior and effects due to the dispersability in environmental media were identified.
The Canada/US Programs have limited datasets on the inorganic carbon class; past evidence indicates that this class includes graphenes (2D sp2 bonded carbon sheets), fullerenes (soccer ball-shaped carbon macrostructures), and nano-carbon black (carbon-based filler). Although inorganic carbons are similar to CNTs, CNTs were excluded from this category because there is sufficient information indicating that their behavior is dictated by physical attributes unique to CNTs’ tubular structures. There is a significant amount of literature suggesting that inorganic carbons exhibit differences in their behavior and effects based on the physiochemical parameters identified in Figure 2. Many uncertainties remain for this class, however, including which other materials could fall into the inorganic carbon class, and whether information from one type of material can be used to increase the weight-of-evidence for another type of material within the class.
- Jachak et al. (2012) found that the biological effects of graphenes are driven by the number of layers, surface area, size and shape (lateral dimensions), stiffness, surface modifications, and surface chemistry.
- Similar findings were found for fullerenes in a review by Sergio et al. (2012). They note that size, chemical modifications (such as the introduction of zinc inside the fullerenes) and surface chemistry, among other properties, affect reactivity.
This class of inorganic carbon is also consistent with the work done by Stone et al. (2010) on the development of classes (see Section 2.1).
Metal Oxides and Metalloid Oxides
According to a global marketplace report, metal oxide and metalloid oxide nanoparticles represent one of the largest classes of nanomaterials in terms of volumes, uses, and applications. This class does not represent a specific chemical composition, but rather generic compositional information: MOx, MaMbOx, in which M is a metal/metalloid and O is oxygen. There is a wealth of information on the fate and effects of metal oxides and metalloid oxides being driven by size, shape, composition, crystal structure (e.g., titanium dioxide), surface chemistry and surface modifications. Horie and Fujita (2011), for example, demonstrated the importance of those physiochemical parameters on the effects of metal oxide and metalloid oxide nanoparticles.
In addition to those parameters, where the metal oxides or metalloid oxides are soluble (see previous discussion on solubility/dissolution), the solubility will also need to be measured before analogue/read-across information can be shared between two substances. The concept of dissolution/solubility of nanomaterials is currently being discussed internationally within the OECD WPMN, as well as within certain European projects, in order to increase knowledge on how the dissolution of a nanomaterial into its ionic forms would impact its consideration from a risk-assessment perspective.
With this and the following class (3.4 Metal, Metal Salts, and Metalloid Nanoparticles), only nanoparticles of the same chemical composition will be considered for use of analogue/read-across information; for example, two nanoparticles of titanium dioxide with similar physiochemical parameters can be considered for analogue/read-across information. As scientific knowledge increases, considerations will be given to using analogue/read-across information for compounds of varying compositions when their physiochemical parameters are sufficiently similar*.
Metal, Metal Salts, and Metalloid Nanoparticles
Metals, metal salts, and metalloids (M0+) behave similarly to the metal oxides and metalloid oxides (section 3.3) in terms of key physicochemical parameters (see common physicochemical parameters in Figure 2). In addition, solubility is of particular importance for this class, a fact reflected by the inclusion of both “solubility” and “oxidation states” in its physiochemical parameters (Figure 2). The role of solubility on fate and effects of metal, metal salts, and metalloid nanoparticles is well documented in literature (See Casals et.al  for the fate and effects of solubility on nanoparticles). In biological or environmental systems, nanoparticles will likely be driven to higher or even complete dissolution. As such, metal, metal salts and metalloid nanoparticles may possess associated toxicity and environmental risks because they will act as a source of potentially toxic cations (e.g., silver nanoparticles have a bactericidal effect that has been correlated with the number of released Ag+ ions). In addition, consideration should also be given to the creation of nanoparticles different from the parent particle due to the dissolution of surface ions (e.g., different sizes, shapes, surface chemistry, etc.).
Semiconductor Quantum Dots
Quantum dots are semiconductor nanoparticles with composition and size-dependent electronic properties. In addition to the importance of the physiochemical parameters outlined in the preceding classes (size, shape, composition, crystal structure, surface chemistry, and surface modifications), liberation of the ions through degradation and core-shell composition are key parameters in understanding the fate (such as releases) and effect of quantum dots. The comprehensive review by Hardman (2006)  and the study by Liu et. al. (2012) investigating releases of quantum dots from nanocomposite lighting demonstrate the importance of the physicochemical parameters identified for this class within the proposed classification scheme.
The Canada/US Programs acknowledge that many organic chemical substances may be on the nanoscale, but are not engineered on this size scale to exploit any nano-specific property. These typically include organic dyes, polymers, and organic pigments. However, there are situations where an organic substance, such as nanocrystalline cellulose (NCC), takes advantage of a nanoscale property. NCC has unique nanoscale properties including high specific strength and modulus, optical properties, and high surface area. It is these types of substances that are considered part of this class. Further discussion is required to understand what specific nanoscale properties or behaviors of engineered organic substances in this class would be of interest for regulatory oversight.
This category includes emerging nanomaterials and/or nanomaterials with which the Canada/US Programs have had very limited experience, or for which there is insufficient science to classify based on similarities in chemical composition. To date, these have included metal alloys (e.g., tungsten carbide),nanoclays, and tubular structures of metals/metal salts/metalloids. It is believed that for tubular structures of different metals/metal salts/metalloids, the requirements to consider two nanomaterials similar are likely similar to those of carbon nanotubes.
Bionanomaterials, or substances which combine biotechnology and nanotechnology to produce advanced functional materials, were identified by stakeholders at the March 20, 2013 RCC Nanotechnology Initiative Workshop as an emerging area that should be considered for regulatory purposes because of their potential commercial impact. Bionanomaterials - in this context - does not refer to nanomaterials interacting with an organism. At this point, the Canada/US Programs have not received any notifications for bionanomaterials. The ISO (the International Organization for Standardization) is carrying out work regarding the vocabulary around the interface between nanomaterials and biology: http://www.iso.org/iso/home/store/catalogue_tc/catalogue_detail.htm?csnumber=51767
It is up to the discretion of the Canada/US Programs to make classes and use analogue/read-across information where appropriate for the nanomaterials which fall into the “other” category. Further discussions on this category will be required between Canada and the US to ensure their New Substance/Chemicals Programs remain aligned.
 Nel, A.; Xia, T.; Meng, H.; Wang, X.; Lin, S.; Ji, Z.; Zhang, H. Acc. Chem. Res., 2013, 46, 607-621.
* The term “sufficiently similar” is undefined and will be discussed and agreed to once this classification scheme is implemented in the regulatory process. The authors welcome any ideas on what constitutes two parameters to be similar – e.g., differences of 10 %? Or similarity based on a minimum number of identified parameters?
 Baughman, R.; Zakhidov, A.; de Heer, W. Science, 2002, 297, 787-792.
 Poland, C.A.; Duffin R.; Kinloch I.; Maynard A.; Wallace WA.; Seaton A.; Stone, V.; Brown, S.; MacNee, W.; and Donaldson, K. Nature Nanotechnol., 2008, 3, 423-428.
 Bussy, C.; Pinault, M.; Cambedouzou, J.; Landry, M.; Jegou, P.; Mayne-L’hermite, M.; Launois, P.; Boczkowski, J.; Lanone, S. Particle and Fibre Tox., 2012, 9, 46.
 Pasquini, L.; Hashmi, S.; Sommer, T.; Elimelech, M.; Zimmerman, J. ES&T, 2012, 46, 6297-6305.
 Liu, Y.; Zhao, Y.; Sun, B.; Chen, C. Acc. Chem. Res., 2013, 46, 702-713.
 Skandani, A.; Zeineldin, R.; Al-Haik. Langmuir, 2012, 28, 7872-7879.
 Steurer, P.; Wissert, R.; Thomann, R.; Mulhaupt, R.Macro. Rapid Comm., 2009, 30, 316-327.
 Tegos, G.; Demidova, T.; Arcila-Lopez, D.; Lee, H.; Wharton, T.; Gali, H.; Hamblin, M. Chem & Bio., 2005, 12, 1127-1135.
Praveen, S.; Chattopadhyay, P.; Albert, P.; Dalvi, V.; Chakraborty, B.; Chattopadhyay, S. Comp. App Sci. & Manuf., 2009, 40, 309-316.
 Jachak, A.; Creighton, M.; Qiu, Y.; Kane, A.; Hurt, R.MRS Bull., 2012, 37, 1307-1313.
 Sergio, M.;Behzadi, H.; Otto, A.; Spoel, D. Env. Chem. Let., 2012, DOI: 10.1007/s10311-012-0387-x.
 “The Global Nanotechnology and Nanomaterials Industry” Technology Report No. 68 by Future Markets, Inc.
 Horie, M., and Fujita, K. (2011) Toxicity of metal oxide nanoparticles, p: 145-178, in Advances in Molecular Toxicology, volume 5. Oxford, United Kingdom, 251 p.
 Casals, E.; Gonzalez, E.; Puntes, V.F. J. Phys. D.: Appl Phys., 2012, 45, 443001.
 Michalet, X.; Pinaud, F.; Bentolila, L.; Tsay, J.; Doose, S.; Li, J.; Sundaresan, G.; Wu, A.; Gambhir, S.; Weiss, S. Science, 2005, 307, 538-544.
 Hardman, R. Env. Health. Persp. 2006, 114, 165-172.
 Liu, J.; Katahara, J.; Li, G.; Coe-Sullivan, S.; Hurt, R. ES&T, 2012, 46, 3220-3227.
 Cranston, E.; Gray, D. BioMacromolecules, 2006, 7, 2522-2530.
 Peng, B.L., Dhar, N., Liu, H.L., and Tam, K.C. (2011) Can. J. Chem Eng. 9999: 1-16.
Kühnel, D.; Bush, W.; Meissner, T.; Springer, A.; Potthoff, A.; Richter, V.; Gelinsky, M.; Scholz, S.; Schirmer, K. Aq. Tox., 2009, 93, 91-99.
 Lordan, S.; Kennedy, J.; Higginbotham, C. Jo. App. Tox., 2011, 31, 27-35.
 Kar, A.; Smith, Y.; Subramanian, V.ES&T, 2009, 43, 3260-3265.
 Whitesides, G. Nature Biotechnology, 2003, 21, 1161-1165.
- Date Modified: