The Toxic Substances Control Act (TSCA) of 1976 provides the U. S. Environmental Protection Agency (EPA) with authority to regulate certain industrial chemical substances.1 Under the TSCA, chemicals used in industrial applications are reviewed for their risks to human health and the environment while other substances (such as foods, drugs, cosmetics, and pesticides) are excluded. The scope of the TSCA reviews includes an evaluation of the potential risks to workers and other populations, and the environment. The TSCA definition of a chemical substance is based on molecular identity, not on physical properties such as particle size.2 Given this definition and the types of uses covered, many nanomaterials are chemical substances subject to TSCA. Nanomaterials based on chemical substances already on the TSCA Inventory are considered existing chemicals. Examples of nanomaterials based on existing chemicals are metals like iron and gold, and some metal oxides such as titanium dioxide and silicon dioxide. Nanomaterials that are not on the TSCA Inventory are considered new chemicals. Examples of nanomaterials that are new chemicals are carbon nanotubes and fullerenes. Because carbon nanotubes and fullerenes are different allotropes of carbon, the EPA considers them to have different molecular identities from other forms of carbon and to be new chemical substances.3 New chemical substances are subject to reporting and review prior to commercialization. Nanomaterials based on existing chemical substances are not subject to reporting before commercialization.
New chemicals are examined by the EPA before manufacturing as part of the Premanufacture Notice (PMN) process under section 5 of the TSCA.4 Under section 5 of the TSCA, the EPA has the authority to evaluate the potential risks in a PMN and to take actions to prevent any unreasonable risk including banning production of the chemical substance. This includes potential unreasonable risks to workers. The Occupational Safety and Health Administration (OSHA) does not have rulemakings, regulations or Permissible Exposure Limits (PEL) that would specifically apply to the new chemical nanomaterials described in this article. However, OSHA does have a framework of existing requirements that covers nanomaterials. These include section 5(a)1 of the OSH Act at 29 USC 654 known as the General Duty Clause, the Hazard Communications Standard at 29 CFR 1910.1200, the Personal Protective Equipment Standard at 29 CFR 1910.132, the Respiratory Protection Standard at 29 CFR 1910.134, the Hazardous Chemicals in Laboratories at 29 CFR 1910.1450, and several substance-specific standards at 29 CFR 1910 Subpart Z.
Basic data required in a PMN submission include chemical identity, use information, anticipated production volume, byproducts, exposure and release information, disposal practices, and existing available health and environmental effects test data. Exposure monitoring data are not required since the submission is completed before the substance is manufactured or imported. The Agency evaluates these data in the open literature, and data submitted with earlier analogous PMN materials in making its determination of potential risks. This process involves a team of reviewers, including chemists, engineers, toxicologists, ecotoxicologists, exposure assessors, and risk assessors. Where there are sufficient concerns for risks to human health or the environment, the EPA will require additional information on substance identification, effects information, and/or exposure information. Since 2005, the Agency has reviewed over 100 nanomaterials as part of the PMN process, and other nanomaterials as part of its Nanoscale Materials Stewardship Program.5
The purpose of this article is to discuss some of those experiences, recognizing that many of the specific data and materials reviewed are regarded as confidential business information by the companies who submitted the PMN information. Early publication of PMN information will enable stakeholders to have improved understandings of the directions of emerging data and its implications for risk assessment. These TSCA approaches and experiences related to worker health are relevant to gauging the scope of materials in commerce and extent to which workers are exposed, the approaches to gathering exposure data and estimating worker exposures for exposure registries, and routes of exposure most likely relevant to any potential adverse effects.6 While the focus of this article is on worker exposure, the EPA has also established interim technical guidance for assessing environmental fate and transport, general population, environmental, and consumer exposures.7
In this article, aspects of the characterization of materials proposed for commercialization, use and exposure information for these materials, and the EPA's methods for estimating worker and general population exposures based on this information are reviewed. Next, toxicity information relevant to worker exposures is addressed, and general approaches to estimating concern levels for workers are examined. Finally, risk management methods to address potential concerns for worker health will be discussed.
A good understanding of the physicochemical characteristics of the nanomaterials under evaluation is critical to an accurate assessment of potential risks to workers, consumers, the general population, and environmental receptors such as aquatic species. Furthermore, such factors are important if exposure registries are developed, because the associations between the physicochemical characteristics of specific nanomaterials and workers who are exposed to them should be understood. Many of the questions posed by the EPA to commercial submitters as part of the reviews of nanomaterials under the TSCA new chemicals process are aimed at better understanding the type, and size distributions, of nanomaterials to which humans and environmental receptors may be exposed. Such questions are often pivotal to the overall risk assessment conclusions made by the Agency on these materials. The importance of understanding the physicochemical characteristics of nanomaterials, before drawing toxicological conclusions based on the testing of such materials, is illustrated in the literature: for example, an inhalation study in mice tested the effects of what was described as multiwalled carbon nanotubes (MWCNTs), and found that while the concentrations tested did not result in significant lung inflammation or tissue damage, they were correlated with systemic immune function alterations.8 This was a novel finding at the time, given that the most sensitive endpoints in similar studies were associated with adverse lung effects rather than effects on the immune system; such a finding could have altered the overall risk assessment approaches for MWCNTs in terms of the endpoints considered in the risk assessment and the associated toxicity values on which an occupational exposure limit would be based. However, later examinations of the materials used indicated that the material tested was more likely a mixture of carbon nanofibers and carbon nanotubes.9,10 Beyond the chemical characterization of the nanomaterial itself, it is also critical to the analysis of potential worker risks to understand whether the nanomaterial is associated with other materials such as resins which are used to encapsulate nanomaterials such as CNTs for certain commercial applications. If nanomaterials are associated with resins and are respirable, then different occupational exposure limit considerations come into play, as opposed to those applicable to the nanomaterial in isolation. For example, if the nanomaterial is encapsulated in an insoluble resin then the resin characteristics and size distribution would be important to consider in the assessment of the overall risks to workers.
In the case of workers, it is also important to understand the particle size distribution to which workers may be exposed, and what fractions of that distribution would deposit in different portions of the respiratory tract. For example, risks to workers due to effects of particles that lodge in the alveolar region of the lung are greatly lessened if nanomaterials are present in size ranges above that which would be inhaled into the deep lung. The EPA11 and others12 note an increased concern for respirable poorly soluble particulates with sizes less than 10-μm aerodynamic diameter; this same size fraction is noted as having high deposition rates in the human alveolar region. The PMN submissions for nanomaterials can contain limited information on the nanomaterial product size distribution or airborne particle size distribution, and the methodology used can impact the usefulness of the data in characterizing exposures. For example, the EPA prefers condensation and optical particle counting methods involving direct air sampling to provide estimates of respirable airborne particle fractions in the workplace, rather than a dynamic light scattering method that measures particle sizes in solution since measurements done in solution may not be representative of size distributions in exposure situations. The EPA also prefers data on size distributions in the workplace air to better characterize potential risks rather than relying upon assumptions as described in the exposure screening methods explanation below to assess potential exposures to nanomaterials.
The EPA, as part of the PMN review process, evaluates worker exposures and potential risks to workers. When appropriate, the EPA specifies personal protective equipment, engineering controls, and/or modifications to nanomaterials’ manufacture to address these potential risks. Some understanding of potential exposures to workers, and others such as the general population can be gained by examining the types of industrial chemical uses seen to date in the TSCA new and existing chemicals’ programs. Table 1 may be useful in identifying study populations for epidemiological studies of workers since current industries that produce and use nanomaterials, and market trends for future study populations, can be identified. A large proportion of the PMN nanomaterial applications received to date indicate that certain nanomaterial classes are represented in commercial products more than others. For example, at this time carbon-based nanomaterials are seen more than other material classes, followed closely by metal oxides/metals/other metal-containing nanomaterials. In the case of the metal-associated materials, approximately half of them are represented by modified silicas (Table 1).
Another level of understanding is gained by examining the production volumes, number of downstream users, and numbers of exposed workers for various nanomaterials reviewed to date. Looking across all PMN submissions, some have had very targeted uses and customers (eg, one customer and one use), while others have had very broad and diversified uses (eg, up to five uses) and numbers of downstream users (e.g., up to eight downstream users per use). Generally, the number of potentially exposed workers per site has not exceeded 10, although the EPA has estimated higher numbers per site in a case using literature data from the plastics industry. While production volumes and workers involved for specific materials cannot be released because of the confidentiality provisions under TSCA, general information about a class of materials, such as carbon nanotubes, can be helpful. For example, the EPA has received over 30 PMN submissions for carbon nanotubes, and production volumes are generally in the low tens of thousands of kilograms per year. The number of workers projected by industry to be involved in the manufacture of these CNTs is typically less than six per site, but can be as high as 100 workers for a single site where CNTs are incorporated into a final product.
Key information and data gathered for PMNs by the EPA includes locations of facilities, chemical identity information, environmental releases, manufacturing diagrams, throughput volumes of materials in kg/day, operating days in days/yr, physical states and concentrations of the nanomaterial of interest at key stages of handling, worker activities with exposure potential, personal protective equipment, engineering controls that limit exposures/releases, on-site waste treatment processes, and number of workers for each of these activities. This information is requested in the PMN Notice software for electronic submissions.13 The information in the PMN Form could be useful in establishing standardized fields for exposure registries. To date, the EPA has estimated worker exposures to nanomaterials using the same methods that are applied to substances that are not nanosized. This is due at least in part to the metrics currently used for reporting toxicity, exposure, and risk data; and due to the limitations of current nanomaterial measurement techniques. Current EPA risk evaluations use mass-based concentrations for inhalation estimates, where the units used are micrograms or milligrams per cubic meter (μg/m3 or mg/m3) as a time-weighted average (TWA) for the worker's shift, assumed normally to be 8 hours. The EPA recognizes that mass based metrics may not be the best approach for estimating exposures to nanomaterials and will continue to work with its researchers, other federal agencies, and other sources to identify and develop more appropriate methods.
The worker exposure evaluations of nanomaterials described in this article have been solely for Toxic Substances Control Act (TSCA) section 5 purposes, in which the EPA employs screening level approaches for estimating worker exposures to new chemical substances for which exposure monitoring data are unavailable. Information on these approaches, and the primary worker exposure estimation tool the Chemical Screening Tool for Exposures and Environmental Releases, which is a PC-based software program, are available from the EPA's public exposure Web site.14 The EPA uses a hierarchy of preferred methods for estimating worker exposures to chemical substances, and the following three main tiers of this hierarchy apply equally to nanomaterials:
I. Personal monitoring data for the chemical of interest in the workplace of interest;
II. Personal monitoring data for the chemical of interest in a workplace situation that is similar to the workplace of interest (surrogate workplace situation) OR personal monitoring data for a chemical that is similar to the chemical of interest in the workplace of interest (surrogate chemical);
III. Modeled estimates or concentration assumptions based on regulatory limits.
Literature searches have found several studies that document personal monitoring data for specific workplace settings. Maynard et al15 carried out a laboratory based study to evaluate the physical nature of the aerosol formed from single-walled carbon nanotube material during mechanical agitation, complemented with airborne and dermal exposure while handling unrefined material. Handling resulted in very low airborne concentrations (from 0.7–53μg/m3), consistent with the tendency to aggregate into larger masses. The EPA has used these concentrations in several new chemical cases as tier II surrogate data where carbon nanotubes that were not identical to the single-walled carbon nanotube materials in the Maynard study were used, and workplace activities have seemed to match well to those documented. No other studies with mass concentration monitoring data found in the literature have yet matched new chemical case situations well enough to apply the data in these studies in any new chemicals cases involving nanomaterials. For example, for data from a particular study to be applicable to a given case, the nanomaterial's chemical structure and the workplace scenario and handling (eg, physical state during handling, specific worker activities that can result in dermal and inhalation exposures, and amounts of materials handled) must be adequately similar in both the study and the case.
The NIOSH (the National Institute for Occupational Safety and Health) has conducted site studies at several sites where carbon-based nanomaterials of several types (nanotubes and nanofibers) are manufactured. These studies have used several different methods for generating particle number concentrations. The EPA has used this information to indicate the potential presence of nanomaterials and nanoparticles in workplace air but has not changed its quantitative exposure assessments. Some of these studies have also documented mass concentrations using the NIOSH method 5040 for Diesel Particulate Matter (as elemental carbon).16 In several new chemicals cases, submissions have included very limited amounts of data from this method, but the EPA has not found these data robust enough to consider them to be representative of worker exposures to the nanomaterials for these cases. Furthermore, the presence of more than one species of nanomaterial (eg, the manufactured nanomaterial, diesel exhaust, and others) can present additional challenges toward characterizing the exposure concentrations of the nanomaterial of interest.
In most new chemicals cases, the limited amount of applicable literature data leads EPA to employ standard screening methods for estimating particulate exposures. Several of the primary screening methods for estimating dust exposures include the tier II “Small Volume Solids Handling Inhalation Model” and the tier III “OSHA PEL for Particulate, Not Otherwise Regulated, total and respirable particulate” models. Also, several primary screening methods for estimating aerosol exposures in “end-use” scenarios (eg, liquid spraying or roll coating mist generation) include the tier II “UV Roll Coating Inhalation Model (non-volatiles)” and the tier II “Automobile Spray Coating Inhalation Exposure Model (nonvolatile non-polyisocyanates)” models. When airborne particle size distribution data are unavailable or potentially not representative, the EPA assumes 100% of particles may be respirable in exposure concentration estimates at or below the OSHA respirable Particulate, Not Otherwise Regulated PEL of 5 mg/m3, with any remaining exposure concentration above 5 mg/m3 assumed not to be respirable. The EPA also uses a suite of standard dermal exposure models to estimate dermal exposures (in mg/day) to nanomaterials. These inhalation and dermal models are documented in the Chemical Screening Tool for Exposures and Environmental Release help system.13
As part of EPA's analysis of potential human and environmental exposures to nanomaterials, under section 5, the EPA often found that, where there are exposures, they frequently involve workers who could be exposed to airborne nanoparticles during their manufacture. Other exposures later in the life cycle of the nanomaterial's production and use are also possible. For example, life cycle analyses of the use of CNTs in batteries, textiles, and epoxy resins have been examined.17 In these cases, exposures are possible with shattering/manual recycling of batteries containing CNTs, shredding/recycling of fabrics with CNT external coatings, and sanding of epoxy resins containing CNTs. In the future, increased exposures to consumers are possible from spray-applied nanomaterials such as metal oxides used to treat hard surfaces.18
UNDERSTANDING POTENTIAL HAZARD AND SETTING OELs
Chronic or subchronic studies in animals are often used to estimate worker inhalation concentrations of concern, but only a few chronic studies are available at this time for traditional nanomaterials such as titania and carbon black.19 To target nanomaterials, which are most likely to cause concerns for workers for future surveillance, exposure registries, and epidemiologic research, better understandings of their toxicity/carcinogenicity potential is necessary. For newer manufactured nanomaterials such as carbon nanotubes and fullerenes, no chronic (and few subchronic) studies are currently available. While the OECD's Working Party on Manufactured Nanomaterials is engaged in testing nanomaterials in subchronic inhalation studies, other Federal or independent industry testing have yielded data at this time using subchronic protocols accepted by regulatory bodies. A set of subchronic inhalation tests on 0.05 um and 1 um C60 fullerenes in rats has been completed by the National Toxicology Program, and results indicate that there were no biologically significant effects at the highest concentrations tested: 2.5 mg/m3 for the nanosized fullerenes, and 30 mg/m3 for the 1 um fullerenes.20 These data indicate higher concern concentrations for some fullerenes via the inhalation route in subchronic studies, relative to another class of newer carbon-based nanomaterials MWCNTs. Data from subchronic inhalation studies on two different MWCNTs have indicated lower concentrations of concern, based on adverse lung effects in rats for MWCNTs which tend to agglomerate in air: Ma-Hock et al21 identified a low observed adverse effect level of 0.1 mg/m3; Paulhun22 found a no observed adverse effect level of 0.1 mg/m3. While data on MWCNTs are becoming available, it is unclear at this time how these data on two types of MWCNTs can be applied in a quantitative manner to estimate the adverse lung effects of MWCNTs with different physicochemical and agglomerate properties, and to other CNTs such as single-walled CNTs. This lack of data for different nanomaterials within a class of nanomaterials (such as the MWCNT class), as compared to the vast number of commercial CNTs coming into the marketplace, make other qualitative approaches to estimating adverse pulmonary effects for these materials necessary to complete premanufacture regulatory reviews in a timely manner under TSCA. Beyond adverse pulmonary effects, there are indications that other endpoints such as those associated with cardiovascular effects may need to be evaluated.23
The EPA has pursued a category approach for setting OELs, and identifying testing recommendations for certain categories of chemicals, when coupled with chemical-specific exposure information for 53 specific chemical categories applicable to the TSCA New Chemicals Program.3 One of these categories is a health category that addresses certain respirable, poorly soluble particulates (RPSPs), including nanomaterials. For certain analogs of RPSPs (including crystalline silica, talc, titanium dioxide, lithium manganese oxide, and carbon black), if the particle size is less than 10 μm in diameter and their respective NIOSH Recommended Exposure Limit (REL) or OSHA PEL is exceeded then a subchronic inhalation toxicity test in rats is recommended. These tests may be required of the PMN submitter to better characterize potential risks. This approach can be expanded by considering new OELs such as the draft NIOSH RELs for titania: 0.1 mg/m3 TWA for up to 10 hr/day during a 40 hour work week for ultrafine titania, and 1.5 mg/m3 TWA for up to 10 hr/day during a 40-hour work week for fine size titania.24 Thus, a range of OELs can be identified for particles of differing potencies, and analog nanoparticles with similar physicochemical properties from PMN submissions can be aligned with applicable OELs. For other unique nanomaterials which do not align with the particles that have established OELs, subchronic test data can be used to derive an occupational exposure limit for that material by the EPA, or the EPA can use OELs derived from the literature such as the 50 μg/m3 TWA for a certain MWCNT as noted in Pauluhn.25 Such a MWCNT-specific occupational exposure limit could be adjusted for other MWCNTs by taking factors such as the degree of agglomeration and/or differing catalyst effects into consideration.
While there is uncertainty associated with the risk assessment of nanomaterials, the EPA has identified potential hazards and exposures. To address potential environmental and health risks, the EPA has used its new chemical authority to prevent or limit human and environmental exposures. For example, the EPA limits use of the nanomaterial to the specific uses in the notice, does not allow spray applications of the nanomaterials, and controls any potential exposures to workers with protective equipment such as impervious gloves and the NIOSH-approved respirators when workers are reasonably likely to be exposed. For further details see the Federal Register 65751, 57430, and 5546.26–28 The EPA requires that the nanomaterial be embedded in a polymer or metal matrix or other article before any consumer uses. In certain cases, the EPA has limited the amount of nanoparticles less than 100 nm for the as-manufactured nanomaterial. The EPA generally does not allow environmental releases directly to surface waters but in some cases has allowed limited releases resulting in stream concentrations of the nanomaterial less than one part per billion. Disposal of new chemical nanomaterials are usually via incineration or landfill. While some nanoparticles are released, most of the nanomaterials are disposed of after they have been embedded in a polymer or metal matrix or other article.
For nanomaterials subject to PMN review there are requirements in administrative orders agreed upon between the EPA and the PMN submitter to establish these restrictions limiting exposures and environmental releases while also requiring development of data such as subchronic inhalation toxicity studies (for examples, see Federal Register 65751 and 5743026,27), material characterization, particle size distribution and other physical chemical property. The EPA also issues significant new use rules to establish these same requirements for all other manufacturers and processors of the same nanomaterial.
To better understand nanomaterials that are existing chemicals already in commerce,
The EPA is also developing a proposed rule under its TSCA section 8(a) information gathering authority to require the submission of additional information. This rule would propose that persons who manufacture these nanomaterials notify the EPA of certain information including production volume, methods of manufacture and processing, exposure and release information, and available health and safety data. The EPA also intends to propose a section 4 test rule for certain nanomaterials that are already in commerce. The proposed rule would require testing for health effects, ecological effects, and environmental fate as well as provide material characterization data. Finally, the EPA is developing a proposed significant new use rule to require reporting of new nanomaterials based on existing chemical substances. The significant new use rule would require persons who intend to manufacture, import, or process these new nanomaterials to submit a significant new use notice to the EPA before manufacturing the new nanomaterial. The EPA would review and manage any potential risks using the same process described for new chemicals. See http://www.epa.gov/oppt/nano/index.html for additional information.
The EPA will continue to gather data obtained through its new and existing chemical's authorities and other valid sources such as the peer-reviewed literature to improve its approaches to nanomaterial risk assessment and risk management. Such information may lead to new and/or improved approaches for nanomaterial worker surveillance, registries, and related epidemiological research.
The authors wish to thank Scott Sherlock for his review and comment on this manuscript.
3. U.S. EPA. 2008.73: Federal Register. 64946. October 31, 2008.
6. Trout DB, Schulte PA. Medical surveillance, exposure registries, and epidemiologic research for workers exposed to nanomaterials. Toxicology. 2010;269:128–135.
7. U.S. EPA 2010. Interim Technical Guidance for Assessing Screening Level Environmental Fate and Transport of, and General Population, Consumer, and Environmental Exposure to Nanomaterials. Available at: http://www.epa.gov/oppt/exposure/pubs/guidance.htm
. Accessed May 19, 2011.
8. Mitchell L, Gao J, Vander Wal R, Gigliotti A, Burchiel A, McDonald J. Pulmonary and systemic immune response to inhaled multiwalled carbon nanotubes, Tox Sci. 2007;100:203–214.
9. Lison D, Muller J. Letter to the Editor. Tox Sci. 2008;101:179–180.
10. McDonald J, Mitchell L. Letter to the Editor. Tox Sci. 2008;101:181–182.
12. Miller F. Dosimetry of particles in laboratory animals and humans in relationship to issues surrounding lung overload and human health risk assessment. A critical review. Inhal Tox. 2000;12:19–57.
15. Maynard A, Baron P, Foley M, Shvedova A, Kisin E, Castranova V. Exposure to carbon nanotube material: aerosol release during the handling of unrefined single-walled carbon. Toxicol Environ Health A. 2004;67:87–107.
16. National Institute for Occupational Safety and Health. NIOSH Manual of Analytical Methods. (NMAM), 4th ed. DHHS (NIOSH) Publication No. 94-113. Cincinnati, OH: National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention, US Department of Health and Human Services, 1994.
17. UK Department for Environment, Food, and Rural Affairs. Research into the likelihood and possible pathways of human exposure via inhalation arising throughout the life cycle of a selection of commercially available articles containing carbon nanotubes—CB0423, 2010. Available at: http://www.defra.gov.uk/
Accessed May 19, 2011.
18. Chen B, Afshari A, Stone S, et al. Nanoparticles-containing spray can aerosol: characterization, exposure assessment, and generator design. Inhal Toxicol. 2010;22:1072–1082.
19. Heinrich U, Fuhst R, Rittinghausen S, et al. Chronic inhalation exposure of Wistar rats and two different strains of mich to diesel-engine exhaust, carbon-black, and titanium dioxide. Inhalation Toxicol. 1995;7:466–533.
20. Walker N, Baker J, Gregory L, Dill J, Germolec D, White K. Evaluation of the effect of particle size on the toxicity and toxicokinetics of fullerene C60 in rats and mice following nose-only inhalation exposure. Abstract Soc Toxicol Ann Meeting; 2009.
21. Ma-Hock I, Treumann S, Strauss V, Brill S, Luizi I, Martiee L. Inhalation of multiwall carbon nanotubes in rates exposed for 3 months. Tox Sci. 2009;112:468–481.
22. Pauluhn J. Subchronic 13-week inhalation exposure to rats to multiwalled carbon nanotubes: toxic effects are determined by density of agglomerate structures, not fibrillar structures. Toxicol Sci. 2010a;113:226–242.
23. Nurkiewicz T, Porter D, Hubbs A, et al. Nanoparticle inhalation augments particle-dependent systemic microvascular dysfunction.Particle Fibre Toxicol. 2008;5:1–12.
25. Pauluhn J. Multi-walled carbon nanotubes (Baytubes): approach for derivation of occupational exposure limit. Reg Toxicol Pharmacol. 2010b;57:78–89.
26. U.S. EPA 2008c. 73 Federal Register 65751. November 6, 2008.
27. U.S. EPA. 2009. 74 Federal Register 57430. November 6, 2009.
28. U.S. EPA. 2010e. 75 Federal Register 5546. February 3, 2010.