Criteria for assessment of the effectiveness of protective measures

N: particle number concentration required for attainment of a mass concentration of 0.1 mg/m³ with particles of the stated size in nm.

¹ In order to illustrate the relationship between the particle number and the dimensions and density of the materials, two different densities of CNTs are used for calculation. The density of 1350 kg/m³ approximates to the density of CNT as a substance. The density of "CNT as a commercial product" approximates to the agglomerate density of the matted microscale agglomerates forming the basis of the commercial product, and is used as such in the work by Pauluhn (2011) [8]. As the agglomeration size falls, this value approaches the substance density again. The density of inhaled CNTs and how a dose is then to be determined in relation to the mass was called into question by Oberdörster in a lecture in 2011 [9].

²GBD: granular bio-resistant dusts

The substances stated in the OECD list for which model calculations have not been performed here are:

• Carbon black, the true density of which based upon its microcrystallinity is approximately 1 850 kg/m³, whereas the density of pelletedagglomerates is in the region of 100 to 500 kg/m³.
• The density of layered silicates (nanoclays) and silicon dioxide generally ranges from 2 200 kg/m³ (amorphous) to 2 650 kg/m³ (crystalline), and is thus in the magnitude of typical respirable dust.

As can be seen from the table, for 200 nm gold particles, a concentration of 1 236 of these particles per cm³ of air would result in a mass concentration of 0.1 mg/m³. Application of the value of 20 000 particles/cm³, as stated in the BSi PAS standard referred to above, to gold particles with a size of 200 nm results in a mass concentration of approximately 1.6 mg/m³. This concentration is above the existing general dust limit value for the respirable dust fraction, and is substantially higher than the threshold value currently under discussion, which is intended to prevent the inflammatory effects of the biopersistent granular dusts.

For all substances with a particle size of 200 nm and a density greater than 1, it may be assumed that a particle concentration of 20 000/cm³ corresponds to a mass concentration (or a multiple of it) of 0.1 mg/m³. Conversely, 20 000 gold particles with a size of 20 nm per cm³ of air corresponds to a mass concentration of only 0.0016 mg/m³. This would be substantially below the respirable dust limit value. At the same time, a concentration of 1 235 400 gold particles (with a size of 20 nm) per cm³, equivalent to 0.1 mg/m³, would be readily measurable and could be substantially reduced with application of the precautionary principle by engineered measures.

The table also shows that the range in both the size of the nanoparticles and their density over more than one order of magnitude results in a range in particle number concentration of over five orders of magnitude. This cannot be covered by current instruments. The size and density of the nanoparticles must therefore be employed as classification criteria for derivation of the recommended exposure limits.

In view of the prevailing uncertainty concerning the effect of nanoparticles and the need to find pragmatic solutions for company level, the IFA proposes the following recommended benchmark levels as increases over the background exposure during an entire shift (8 hrs) for monitoring the effectiveness of protective measures in the plants, based upon its experience in measurement and the detection limits of the measurement methods currently employed. Information on the measurement of this background exposure can be found in the proposal for a tiered type measurement strategy [10]. These recommended benchmark levels are geared to minimizing the exposure in accordance with the state of the art, and are not substantiated toxicologically. Even where these recommended exposure levels are observed, a health risk may still exist for the employees. Depending upon the chemical composition, occupational exposure limit values for particular substances may have to be considered.

1. For metals, metal oxides and other biopersistent granular nanomaterials with a material density of > 6 000 kg/m³, a particle number concentration of 20 000 particles/cm³ in the range of measurement between 1 and 100 nm should not be exceeded.
2. For biopersistent granular nanomaterials with a material density below 6 000 kg/m³, a particle number concentration of 40 000 particles/cm³ in the measured range between 1 and 100 nm should not be exceeded.
3. For biopersistent granular nanomaterials without known significant specific toxicity (mean agglomerate density 1 500 kg/m³), a particle number concentration of 130 000 particles/cm³ in the range of measurement between 1 and 100 nm should not be exceeded.
4. In our view, a substantial need for discussion remains with regard to evaluation of the effectiveness of protective measures against nanoscale particles or agglomerates/aggregates larger than 100 nm. For 500 nm titanium dioxide aggregates, 360 particles/cm³ corresponds to a mass concentration of 0.1 mg/m³. Measurement of this number concentration by means of the existing instruments would necessitate virtually clean-room conditions. In a typical industrial environment, this concentration would no longer be detectable, owing to the ubiquitous background level of 20 000 particles/cm³ or more. The corresponding mass concentration of 0.1 mg/m³ titanium dioxide can however be determined reliably by means of conventional analysis methods for documentation of the in-plant conditions. For 200 nm titanium dioxide aggregates, 20 000 particles/cm³ corresponds to a mass concentration of 0.35 mg/m³. This already exceeds the value of 0.3 mg/m³ proposed by NIOSH for nanoscale titanium dioxide (refer to the chapter entitled "Proposals for limit values in the USA and UK").
5. Owing to the mounting evidence that biopersistent CNTs which satisfy the WHO fibre definition or have similar dimensions may harbour effects similar to those of asbestos, we urgently recommend that only CNTs be used that have been tested for this end point (according to the manufacturer's declaration) and which do not exhibit these properties. For carbon nanotubes (CNTs) for which no such manufacturer's declaration is available, a provisional fibre concentration of 10 000 fibres/m³ is proposed for assessment, based upon the exposure risk ratio for asbestos [11]. In addition to use of state-of-the-art protective measures, the wearing of respiratory protection and protective clothing is advisable even if the recommended exposure limits are observed. The demarcation of contaminated and non-contaminated zones should be reviewed.
   
   At present, however, monitoring of the above value in plants is hampered by a lack of collection methods of verified suitability, corresponding analysis methods, and criteria for counting of the fibres and determining of the fibre count concentration. An urgent need exists here for the development of analysis methods and conventions for interpretation.
   
   For a transitional period, a particle number concentration of 20 000 particles/cm³ should not be exceeded. In a worst-case scenario, however, this would correspond to a fibre concentration of 20 billion fibres/m³, and illustrates that the existing methods for determining the particle number concentration of CNTs at the workplace are unsatisfactory. Some companies have employed internal guide values as mass concentration values based upon the residual content of metallic catalysts in the CNTs.
6. For ultrafine liquid particles (such as fats, hydrocarbons, siloxanes), the applicable maximum workplace limit (MAK) or workplace limit (AGW) values should be employed owing to the absence of effects of solid particles.
7. The recommended benchmark levels stated above should not be applied to ultrafine particles (see definition). For some processes and technologies in which ultrafine particles are produced, proven protective measures and binding provisions exist for the handling of them. Welding fumes and diesel-engine emissions are examples. The body of rules and regulations which exists in this area and which has been drawn up with reference to the current state of knowledge should be applied until new findings become available.

Reception and discussion of the proposed assessment metrics

Since the first proposal for assessment metrics was made in mid-2009, a number of limit values have been published at national and international level. In 2009 for example, the MAK Commission published a health-based value of 0.1 mg/m³ for zinc oxide (respirable fraction) in consideration of the effect of zinc oxide smoke [12]. The value proposed by the IFA for zinc oxide nanoparticles of 40 000 particles/cm³ approximates to this mass concentration; for particles smaller than 100 nm, the value remains substantially below this mass concentration.

As already reported, NIOSH proposed a value of 0.3 mg/m³ in 2011 for ultrafine/nanoscale titanium dioxide, based upon toxicological findings for the avoidance of lung cancer. For titanium oxide particles 100 nm in size, this would correspond to a particle number concentration of approximately 135 000 particles/cm³. This value can be monitored easily using current measurement technology, and further reduction appears possible. Conversely, the particle number concentration of 40 000 particles/cm³ for particles 50 nm in size proposed by the IFA corresponds to a mass concentration of 0.011 mg/m³. This is therefore also substantially below the value proposed by NIOSH.

The Dutch parliament charged the Knowledge and Information Centre Risks of Nanotechnology (KIR-nano), in co-operation with the National Institute for Public Health and Environment (RIVM), with the task of examining whether provisional reference values could be derived for nanomaterials. The experts concluded that at the present time, they were not aware of any method better than that of the IFA for deriving provisional reference values for nanomaterials [13]. In August 2010, nano reference values (NRVs) based upon the IFA concept were proposed by the Social and Economic Council (SER), an advisory comittee to the Dutch govenrnment and Parliament, and recommended to companies for application until such time that health-based limit values have been defined [14]. At the same time, the Dutch government sponsored a project for evaluating the facility for implementation of the proposed values. The results of this project and the general suitability of the nano reference values were presented at a workshop in the Hague in September 2011 [15]. It was reported that the IFA's concept for evaluation of the protective measures had proved effective in the Netherlands under the name "Nano Reference Values".

In 2008, BASF SE supplied information to the US Environmental Protection Agency in accordance with the provisions (Section 8e of the Toxic Substances Control Act) on a sub-chronic inhalation study on Wistar rats involving carbon nanotubes [16]. BASF SE states that under the study conditions described, the NOEL must be below 0.1 mg/m³. Based upon this information, Nanocyl in Belgium reports a value of 0.0025 mg/m³ for the MW CNTs which it manufactures [17].

In the CIB of December 2010, NIOSH recommends limiting the concentration of carbon nanotubes in workplace atmospheres to below 0.007 mg/m³, measured in terms of elementary carbon in accordance with NIOSH method 5040. Information on comprehensive protective measures can be found in Chapter 6 of the CIB.

In short, it can be said that the sampling and measurement methods for CNTs are still at the scientific test stage. Suitable, simple methods for practical monitoring of exposure in commercial enterprises do not yet exist. The implementation of protective measures is therefore all the more important.