**4. Discussion**

## *4.1. Quantitative Assessment*

On-site measurements showed the lowest mean number particle concentration on the background trial, as expected, since the printer was not yet operating. After the AM operation started, the highest mean number particle concentration was obtained while the worker removed the part from the printer and cleaned it with a brush (task 2), as shown in Table 3. This number is very close to the one measured during the first task (printing). In reality, when analyzing Figure 1, it is possible to verify that the highest values of the number of particles occurred during the printing process, and not during the subsequent tasks. This result may be an indicator that, although the metal parts are printed in a closed chamber, there is still emission of matter during the process that may be released into the work atmosphere. In fact, the real-time measurement of air velocity near the door of the printer indicated 0.17 m/s, as shown in Table 2, endorses this possibility, since it is significantly higher than the background measurement (<0.01 m/s). Regardless this finding, several studies showing results of workplace airborne matter measurements during metal 3D printing do not consider the printing process [5,6,8]. In view of these results, further investigation is needed in this field, to verify if currently containment conditions are enough to prevent workers' exposure to nanomaterials during printing processes, or if containment improvement is required and/or if safety-by-design measures are needed at the printer manufacturing stage.

The results of the SMPS shown in Table 4 are consistent with the ones from the CPC (Table 3). When comparing these results to the previously mentioned recommended value of 20,000 nanoparticles/cm<sup>3</sup> for an 8-h exposure time (mean number of particles between 10 and 100 nm lower than 9300 particles/cm<sup>3</sup> for all tasks), it is possible to conclude that the results are consistently lower, which does not mean an absence of risk. In Figure 2, it is possible to confirm that SMPS measurements indicate that the smaller particles are released during the printing activity.

Another finding of this quantitative approach, by using the EDS technology, was that there was no significant change in the chemical composition of the powder after laser action (Figures 4 and 6). The same results were achieved in similar studies [7]. The results of SEM analysis to the airborne samples (Figures 7 and 9) indicate the presence of agglomerates/aggregates of nanometer-scale particles, with an anisotropic shape.

This quantitative approach gives good insights on number particle concentration, size and shape of airborne matter, chemical composition, and environmental conditions.

#### *4.2. Qualitative Assessment*

Qualitative assessments present risk levels as a result and allow the user to access information on recommended controls. Additionally, opposite to quantitative analysis, this approach does not require access to measuring equipment.

Table 5 summarizes the application of CB Nanotool 2.0 to this case study. Since stainless steel 316L is an alloy of iron and chromium and contains a significant quantity of nickel (≈12.5%), nickel inorganic compounds' OEL was considered as PM OEL, since it is the lowest one amount the significant components of this alloy. According to the material safety data sheet, the metal powder used is carcinogenic (H351) and skin sensitizing (H317), so PM carcinogenicity and PM dermal toxicity factors were scored as yes (this last one considering a precautionary approach). All nanomaterial related factors were classified as "unknown" since there is no information available for these airborne incidental nanomaterials. These considerations lead to a severity score of 63 (high band) for all tasks performed.

Regarding probability band, the amount of powder used in each task is similar (always more than 100 mg). So is the number of employees exposure and the frequency of the operation, thus scores were the same. Only the duration of the operation is different, so the probability score for task 1 (the longest one) is 85 (Probable band) and for task 2 and 3 the score is 70 (Likely band). According to these results, for task 1 it is recommended to seek specialist advice since risk level is the highest possible. For task 2 and 3 the recommendation is containment since the Risk Level is 3.

These results may be considered unexpected, since the highest risk level is usually associated with handling tasks, like sieving and cleaning [27]. Another observation of the CB Nanotool results is related to the recommended controls. Containment may not be adequate for task 2 and 3 since it may not be viable when carrying the part to remove it from the chamber of the machine and when removing the remains of powder.

However, Stoffenmanager Nano 1.0 lead to different results, as presented in Table 6 since it is a source-receptor model [28]. The criteria for the hazard band were the same for all tasks: dry powder with very high dustiness, small concentration of nanocomponents and unknown characteristics of the nanomaterials (concentration and inhalation hazard). In the factor related to OECD components, the option "other MNOs" was selected in the absence of another specific for incidental NM and, in the last factor, it was necessary to establish a relation between the current hazard identification and the one considered in this method, defined in Annex III of European Union Directive 67/548/EEC, which is no longer in force (replaced by CLP Regulation No 1272/2008). Hazard band E, the hazardous one, was therefore the result for the 3 tasks. In this method, hazard band E is assigned when the parental material is classified for carcinogenicity, mutagenicity, reproduction toxicity, or sensitization [16].

Concerning the exposure band, the duration of each task was considered, as well as distance to the breathing zone of the worker and specific existing local control measures for each operation. Thus, exposure band 1 was the result for task 1 (lowest band) and exposure band 2 for the other tasks. Despite the different exposure bands, the overall risk level for all tasks was RL 1 (highest priority).

Subsequent controls recommended for each task are listed in Table 6 and they are different for tasks 1 as it shows lower exposure. The recommendations for printing operation include automation of tasks and enclosure of the source, which are already implemented. It also mentions controls that are not viable for this operation, such as wetting the powder or eliminating this task, since it would compromise all the manufacturing process. For task 2 and 3, recommendations also mention already implemented controls, such as respiratory protection, and not suitable solutions, like using a spraying booth or wetting the powder.

When applying Stoffenmanager Nano 1.0 to this case study, the hazard band E was obtained for all tasks, therefore risk level 1 was the corresponding final result by default. In view of these results, it is possible to conclude that, although this method considers relevant inputs for incidental NM and considers some control measures already implemented, it is not a suitable method for metal AM workstations, since it does not differentiate the level of risk of different tasks performed and it does not provide tailored control actions aiming at the reduction of the exposure risk in these workplaces.

### *4.3. IN Nanotool*

Considering all results and limitations from the previous described qualitative and quantitative approaches, IN Nanotool was designed for managing the exposure to incidental NM in metal 3D printing workstations and it was applied in this case study. The results of this application are presented in Table 10, in which it is possible to verify that the results obtained by using the IN Nanotool are significantly different from the ones achieved by the other approaches.

When analyzing the hazard band, the first six factors were provided by the properties of the metal powder present in the MSDS of the product, being clear that it is a powder with carcinogenicity and other associated health hazards. The lowest OEL criteria was the same as the one used for CB Nanotool 2.0 application. According to its MSDS the powder is water insoluble, and the average particle size range is between 20 and 53 μm. The two remaining factors to define the hazard band (shape and size) were possible to score due to the results of SEM analysis (Figures 7 and 9). If these SEM results would not be available, the score of these two factors would be 18.75 (unknown), which would increase the hazard band, since a precautionary approach is intended. The hazard band obtained for all three tasks is Medium (47 points), since the material used is the same throughout all 3D printing process.

Regarding the exposure band in this case study, material and operation conditions were determined by observing the conditions in situ and consulting the organization and the workers involved. The outcome was an exposure score of 46.5 (medium band) for task 1, mainly because it was considered that there is containment of the source and high dustiness, even though the time of exposure is higher, and no PPE were used during this period. For tasks 2 and 3 the exposure score was 70, meaning the exposure band is high. In this case, although the worker uses filter mask FFP3 and protective clothing, no eye protection is used and there is no containment of the source or isolation of the worker, when dustiness is high.

Using the risk matrix from Figure 11, it is possible to conclude that the printing process represents a Risk Level 1 and the other two tasks a Risk Level 3. These results are different from the ones obtained by applying CB Nanotool 2.0 and Stoffenmanager Nano 1.0. Using IN Nanotool, distinct risk levels are obtained for considerably different operations and the results seem to support the belief that not contained manual handling processes are the ones with higher risk [26].

It should be highlighted that in this case study using IN Nanotool the highest risk level (RL4) was not assigned to any of the tasks under study. This is aligned with the quantitative results, that show that the measured number particle concentration was not high when comparing to other metal 3D printing case studies [6,8] and to previously mentioned nano reference values.

Finally, according to the IN Nanotool, additional risk control measures should be considered. Critically analyzing the recommended controls for task 1 (see Table 10), in addition to the already containment of source, mechanical ventilation can be installed in the room, the operation conditions can change (for example, by reducing the frequency and/or duration), additional PPE can be used by precaution and/or internal procedures can be improved. For tasks 2 and 3, it is possible to clean the part with a brush and to sieve the powder in a glove box or bag, to install local exhaust ventilation or fume hood and/or to change operation conditions.

#### **5. Conclusions**

The difficulties to manage the risk of exposure to incidental nanomaterials and the lack of information on this matter have been recently discussed and are a cause of concern. Quantitative assessments require access to specific measurement equipment and don't provide control recommendations, requiring expert knowledge to assess and control the risk. On the other hand, limiting the risk management approach to the existing qualitative

tools focused on ENM may be biased. Using those methods for incidental NM represents a significant difficulty in background data gathering, as shown in this case study.

The main objective of this study was to explore and highlight these difficulties and to design and test a tool to manage the risk of exposure to metal incidental NM in 3D printing processes. IN Nanotool redefined the inputs of CB approaches for incidental NM and added quantitative ones. Unlike quantitative approaches, this method does not necessary require special measurement equipment and it is not dependent from reference or limit values. Moreover, this method culminates in risk control recommendations, allowing to manage the risk of exposure to airborne incidental NM originated in metal AM processes, without the need to resort to a specialist. This tool was designed to enable this risk management, by providing a comprehensive and accessible approach to OSH professionals, including non-experts. However, there are limitations to this method. For instance, if the user does not have access to majority of background information, the method allows to score factors as unknown, resulting in a high risk level. This precautionary result may lead to the suggestion of exaggerated control measures in relation to the real risk. Additionally, this tool requires additional testing and further validation. Regardless of its limitations, the IN Nanotool application to the present case study led to reliable results that are more in line with the state-of-the-art, showing its potential to fill the lack of methods for incidental NM.

**Author Contributions:** Conceptualization, M.S., P.A. and F.S.; methodology, M.S., P.A. and F.S.; validation, P.A. and F.S.; investigation, M.S.; writing—review and editing, M.S., P.A. and F.S.; supervision, P.A. and F.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** No new data were created or analyzed in this study. Data sharing is not applicable to this article.

**Acknowledgments:** The results of this study would not be possible to obtain without the support of the organization where measurements were performed. We wish to acknowledge the help provided by the board of the organization and the workers involved in the study. Additionally, we express our acknowledgments to CTCV (Technological Center of Ceramics and Glass) for the measuring equipment providing during the monitoring campaign.

**Conflicts of Interest:** The authors declare no conflict of interest.
