Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles during down-stream industrial handling

Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles... Today, engineered nanomaterials are frequently used. Nanosized titanium dioxide (TiO ) has been extensively used for many years and graphene is one type of emerging nanomaterial. Occupational airborne exposures to engineered nanomaterials are important to ensure safe workplaces and to extend the information needed for complete risk assessments. The main aim of this study was to characterize workplace emissions and exposure of graphene nanoplatelets, graphene oxide, TiO nanofibers (NFs) and nanoparticles (NPs) during down-stream industrial handling. Surface contaminations were also investigated to assess the potential for secondary inhalation exposures. In addition, a range of different sampling and aerosol monitoring methods were used and evaluated. The results showed that powder handling, regardless of handling graphene nanoplatelets, graphene oxide, TiO NFs, or NPs, contributes to the highest particle emissions and exposures. However, the exposure levels were below suggested occupational exposure limits. It was also shown that a range of different methods can be used to selectively detect and quantify nanomaterials both in the air and as surface contaminations. However, to be able to make an accurate determination of which nanomaterial that has been emitted a combination of different methods, both offline and online, must be used. ● ● ● ● Keywords Occupational exposure Electron microscopy Thermal-optical carbon analysis Direct-reading instruments PIXE Aerosol Introduction robust, but also flexible, structure rendering it useful in a variety of applications [1]. It can exist in different structures Due to their novel and valuable properties compared with such as graphene, graphene oxide, and graphene nanopla- bulk materials, the use of engineered nanomaterials is telets [2, 3]. Its electrical and thermal properties makes increasing. Graphene, a 2D carbon nanomaterial, has a graphene useful in for example transistors [2, 4] and che- mical sensors [5], and the optical properties can be used in biological sensors [6]. Another application of graphene is as metal surface coatings to inhibit corrosion [7, 8] and to Members of the NanoLund are listed below Acknowledgements. reduce wear and friction on sliding metal surfaces [9, 10]. Multiple reviews [11–15] have generally concluded that * Karin Lovén graphene toxicity depends on the physiochemical properties Karin.loven@design.lth.se of the nanomaterial. However, one of the most widely Ergonomics and Aerosol Technology, Lund University, SE-22100 recognized mechanism for graphene-nanomaterial-induced Lund, Sweden toxicity in living systems is the induction of oxidative stress Solid State Physics, Lund University, SE-22100 Lund, Sweden and production of reactive oxygen species (ROS) [15]. In a more recent review by Fadeel et al. [16], the authors Medical Radiation Physics, Department of Translational Medicine, Lund University, SE-22100 Malmö, Sweden highlighted the need for standardized graphene character- izations, and of robust and validated toxicological assays in Occupational and Environmental Medicine, Lund University, SE- 22100 Lund, Sweden order to advance the field of graphene toxicity. 1234567890();,: 1234567890();,: Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 737 Studies of occupational exposure to nanomaterials are the NPs into other products has been investigated. Different needed to make complete risk assessments. Basinas et al. handling tasks, both with NP dry powder and NP containing [17] showed that many exposure assessments have been liquid, have been shown to constitute an occupational done for carbon nanotubes (CNTs), carbon nanofibers exposure risk [42–45]. The review by Debia et al. [46] (CNFs), and titanium dioxide nanoparticles (TiO NPs), but further strengthen this conclusion. However, studies of TiO 2 2 far less for other engineered nanomaterials, such as gra- NF exposure in occupational settings have, to our knowl- phene. Sanchez et al. [12] stated that there is a need for edge, not yet been conducted. measurements of airborne graphene exposure levels at both The Organisation for Economic Cooperation and research laboratories and full-scale manufacturing facilities. Development (OECD) have suggested a harmonized three- This was further pointed out by Arvidsson et al. [13], who tiered approach for nanomaterial emission and exposure requested that workplace emissions and exposures of gra- assessments [47] and recommendations for measurement phene should be investigated. Only a few emission and strategies and instrument use [48]. exposure measurements during production [18–21] and This study aims to generate new knowledge about handling [18] of graphene nanomaterials have since then emissions and exposures of nanomaterials not extensively been conducted. These studies showed low levels of studied previously. Emissions and exposures, with a focus exposures to graphene. However, according to Lee et al. on different graphene nanomaterials (both nanoplatelets and [18], monitoring of other work tasks including down-stream oxide) and different TiO nanomaterials (both NFs and graphene handling processes is needed for a full under- NPs), were characterized with a multi-metric approach standing of the exposure situation. To our knowledge, no during down-stream industrial handling. Emission and studies of emission and exposure measurements during exposures of carbon black (CB) and copper (Cu) were also down-stream handling processes of graphene nanomaterials, measured in a few cases. Different sampling and aerosol such as manufacturing of ink and surface coatings con- monitoring methods were evaluated to be able to recom- taining graphene, have previously been conducted. mend methods to be used specifically for these nanomater- Another common nanomaterial is TiO . It is used in ials to complement the different tiers and measurement paints and sunscreens [22, 23] as well as in transistors [24], strategies described by OECD. An additional aim was to biosensors [25], cancer treatment [26], and different surface assess the potential of secondary inhalation exposure, caused coatings [27, 28]. Nanosized TiO is commonly found as by resuspension of particles deposited on surfaces. spherical NPs, but can also be produced in other shapes including NFs [29] and nanowires [30]. The toxicity of nanosized TiO , especially spherical ones, has been studied Methods to a greater extent than graphene. As with graphene, the different physiochemical properties of TiO NPs have a Facilities strong influence on the toxicity [31]. Generally though, only moderate effects have been observed, including pulmonary Measurements were performed 2016 and 2017 at two dif- inflammation [32–34] and pathological neural changes [35] ferent workplaces, with 20 and 15 employees respectively, after inhalation/instillation in rodents. Induction of DNA hereafter “Study A” and “Study B”. As recommended by damage has also been observed in lung cell studies [36, 37]. the OECD [47], initial contextual information was gathered A few studies have been carried out on exposed workers. (tier 1) and basic exposure assessments (tier 2) were per- Pelclova et al. [38] showed for example that the leukotriene formed at both workplaces prior to conducting the expert levels in exhaled breath condensate were elevated in exposure assessment (tier 3) described herein. workers exposed to TiO NPs. The toxicity of TiO NFs has During Study A, different nanomaterials including gra- 2 2 not been as thoroughly investigated. Hurbánková et al. [39] phene nanoplatelets, spherical TiO NPs, CB, and Cu were showed the development of serious lung inflammatory and handled. The nanomaterials were used in ink formulations cytotoxic processes after intratracheal instillation of TiO for printing electronics, sensors, and labels. Measurements NFs in rats and Medina-Reyes et al. [40] observed cell were conducted in a chemistry laboratory and a printing cytotoxicity and genomic instability of TiO NF exposure laboratory. In Study B, the nanomaterials handled included on alveolar epithelial cells. In addition, Allegri et al. [41] graphene oxide and TiO NFs for use in friction and wear performed a comparative exposure study on alveolar epi- reducing surface coatings. Measurements were conducted in thelial cells showing that TiO NFs were more toxic than a chemistry laboratory and a test laboratory. Both work- TiO NPs. places were equipped with general ventilation and process Occupational airborne exposure to TiO NPs has been ventilation systems such as fume hoods. The amount of extensively studied. Production of TiO NPs, bagging and TiO handled during the two studies differed by three orders 2 2 handling of the NP dry powder, as well as incorporation of of magnitude (from about 5 kg per day during Study A to 738 K. Lovén et al. about 5 g per day during Study B). The amount of graphene Quantification of EC of the graphene nanomaterials was nanomaterial handled was a few grams per day during both conducted according to the NIOSH NMAM 5040 protocol Study A and B. Similar processes were investigated during for thermal-optical analysis (DRI Model 2001 OC/EC both studies. Carbon Analyzer, Atmoslytic Inc., USA) [50]. Temperature steps for the EC fraction were: 680 °C (EC1), 750 °C (EC2), Work tasks and 900 °C (EC3). The method was modified with an extended oxidation time, 150 s instead of 30 s, at the highest Different work tasks were performed during the two studies temperature, 900 °C, in order to achieve complete oxidation and a thoroughly written logbook documented the specific of all carbonaceous nanomaterials [51]. The limit of activities carried out. Table 1 shows these work tasks with detection (LOD) for EC was determined to be 0.06 µg C/ detailed descriptions. In Study A, three workers performed cm (corresponding to 4-h sampling with a carbon airborne the work tasks and in Study B only one. concentration of 0.5 µg/m ). Quantification of Ti was conducted by Particle-Induced Engineering controls and personal protective X-ray Emission (PIXE) analysis [52]. In PIXE, a 2.55 MeV equipment proton beam is focused on the filter specimen. This renders the atoms in a state of high excitation, which causes inner Different types of exposure control techniques and enclo- shell vacancies. The characteristic X-ray emission lines are sures, as well as different types of personal protection caused by the quickly occurring transition to a state of lower equipment (PPE), were used during the different work tasks energy. When protons are used, the cross section for the (Table 1). creation of an inner shell vacancy is very high, thereby the sensitivity is very high; for instance, the LOD for Ti was Sampling strategy <6 ng/cm . The elemental quantification of Cu was performed by Time resolved and filter based measurements of airborne digestion with 1 ml concentrated nitric acid (Nitric Acid, particles were conducted in four different spatial zones, a Trace metal grade, Fisher Chemicals) in an oven (60 °C) for methodology described in detail in Isaxon et al. [49]. 16 h, followed by dilution to 10 ml with Milli-Q water to a The measurement zones included: (1) emission zone (EZ)— stock solution. Analysis was performed by inductively no more than a few centimeters from a potential source, (2) coupled plasma-mass spectrometry (ICP-MS, iCAP Q, personal breathing zone (PBZ)—within a radius of 30 cm Thermo Scientific, Germany). The LOD for Cu was three from a worker’s nose and mouth, (3) background zone (BZ) times the standard deviation of blank filters i.e. <0.01 µg/ —at least 2–3 m away from any potential particle source, sample. All results were blank filter corrected. and (4) supply air (SA)—in the inflowing air from the general ventilation system. Figure 1 shows the schematics Filter sampling for SEM analysis of the facilities in Study A and B including where the dif- ferent work tasks were performed and where the different To be able to morphologically characterize the workplace measurement zones were located. air, regarding engineered NPs and NFs and their aggregates and agglomerates, during down-stream handling processes, Air sampling methods and analyses total dust fraction on filters according to Nilsson et al. [53] and Vaquero et al. [21] were collected. Time-integrated Filter sampling for elemental composition samples were collected at 2.3 l/min (sampling pump Escort ELF, MSA, USA) by open-face sampling on 25-mm Time-integrated emission zone, personal breathing zone, polycarbonate filters (pore size 0.4 µm, SKC Inc., USA). and background zone samples were collected by open-face The filters were analyzed by Scanning Electron Microscopy sampling on 25-mm filters mounted in filter cassettes. (SEM) using a Hitachi SU8010 Cold Field Emission SEM Quartz filters (SKC Inc., USA) were used to collect samples (Hitachi, Japan) with an acceleration voltage of 10 kV. A for elemental carbon (EC) analysis. Filters made of poly- sputtering tool (Q150T ES, Quorum, UK) was used to coat carbonate (SKC Inc., USA) were used for analysis of tita- the sample with 10 nm of platinum:palladium (Pt:Pd, −5 2 nium (Ti), and of cellulose (pore size 0.45 µm, SKC Inc., 80:20). A minimum of 10 random 1.25 × 10 cm areas of USA) for elemental analysis of Cu. Pumps (Escort ELF, the filter were imaged and used for quantifying the MSA, USA) with a flow rate of 2.3 l/min for the poly- number of particles. During Study A, 17 filter samples carbonate filters and 3.0 l/min for the quartz and cellulose were collected and the highest LOD was determined to be −3 filters—checked before and after sampling—were used for 0.47 cm . During Study B, 13 filter samples were collected −3 all sample collection. and the highest LOD was determined to be 1.91 cm .In Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 739 Table 1 Work tasks performed at the two companies, types of engineering controls and types of personal protective equipment (PPE). Work task number Work task Work task description Location/ Engineering controls PPE Study A A1 Preparation of graphene nanoplatelets ink Weighing of graphene nanoplatelet powder, addition of Chemistry laboratory/Fume hood Half-face respirator (A + P3), lab (weighing, mixing). Note that this is not the same liquid and mixing. coat, protective gloves of nitrile, ink printed with in A3! goggles A2—day 1 Preparation of titanium dioxide nanoparticle ink Weighing of titanium dioxide nanoparticle powder, (weighing, mixing) weighting of liquid. Addition of powder by sieving into the liquid and mixing. A2—day 2 Preparation of titanium dioxide nanoparticle ink (weighing, mixing) A3 Screen printing with graphene ink (HDPlas Transfer of ink to the screen printer, start of printing Printing laboratory/process ventilation Clean room lab coat, hair net, Graphene Ink), 10 sheets process and a number of sheets was printed. The printed protective gloves of nitrile sheets were placed in the drying oven. After the printing A4 Screen printing with carbon black ink (C740), process the ink was removed and the screen printer and 10 sheets its equipment was cleaned. A5 Screen printing with carbon black ink (CXT0641), 10 sheets A6—day 1 Screen printing with carbon black ink (C7102), 20 sheets A6—day 2 Screen printing with carbon black ink (C7102), 22 sheets A7 Screen printing with copper ink (CP-PLS-010715- R1A), 10 sheets Study B B1 Preparation of graphene oxide coating (weighing, Weighing of graphene powder, addition of liquid to the Chemistry laboratory/Parts in open fume hood Half-face respirator (A + P3), lab mixing) powder and mixing. Transferring of the liquid to another (weighing, mixing), parts on open bench coat, protective gloves of nitrile, container followed by addition of glass beads and more (filtration steps) goggles mixing. Finally, 2 steps of filtration of the liquid to remove the glass beads. B2 Spraying and curing graphene oxide coating, and Transfer of the liquid coating to a paint sprayer, cleaning Test laboratory/parts in spray booth, parts in cleaning/washing up after of metal plates, paint spraying of metal plates, placing the open fume hood and sink (cleaning and coated plates in a furnace. Cleaning of paint sprayer in washing up) the spray booth followed by washing up mixing equipment in the fume hood and at the sink. B3 Abrasion test with a metal brush on graphene The coated plate was placed in the abrasion testing Test laboratory/none Laboratory coat, protective gloves oxide coating equipment. A brush, rotated with a certain speed, was of nitrile, goggles placed to have contact with the plate. When the surface B4 Abrasion test with a nylon brush on graphene was abraded the test was ended. oxide coating B5 Preparation of titanium dioxide nanofiber coating Weighing of titanium dioxide nanofiber powder, addition Chemistry laboratory/parts in open fume hood Half-face respirator (A + P3), lab (weighing, mixing) of liquid to the powder and mixing. Transferring of the (weighing, mixing), parts on open bench coat, protective gloves of nitrile, liquid to another container followed by addition of glass (filtration steps) goggles beads and more mixing. Finally, 2 steps of filtration of the liquid to remove the glass beads. B6 Spraying and curing titanium dioxide nanofiber Transfer of the liquid coating to a paint sprayer, cleaning Test laboratory/parts in spray booth, parts in coating, and cleaning/washing up after of metal plates, paint spraying of metal plates, placing the open fume hood and sink (cleaning and coated plates in a furnace. Cleaning of paint sprayer in washing up) the spray booth followed by washing up mixing equipment in the fume hood and at the sink. B7 Abrasion test with a metal brush on titanium The coated plate was placed in the abrasion testing Test laboratory/None Laboratory coat, protective gloves dioxide nanofiber coating equipment. A brush, rotated with a certain speed, was of nitrile, goggles placed to have contact with the plate. When the surface B8 Abrasion test with a nylon brush on titanium was abraded the test was ended. dioxide nanofiber coating 740 K. Lovén et al. Fig. 1 Schematics of the facilities. The location of where the different work tasks were performed, and the placement of the different measurement zones, emission zone (EZ), background zone (BZ), and supply air (SA), are shown during a Study A and b Study B. addition to the areas imaged for particle quantification, a measure the black carbon (BC) mass concentration (as a larger part of the surface was investigated with lower proxy for EC) in the background zone, with a time reso- resolution in order to identify any engineered nanomaterial. lution of 1 min. During Study A, the measurements in the emission zone were supplemented with a fast aerosol Direct reading instruments mobility size spectrometer (DMS Model 500 MkII, Cambustion, UK) and a DustTrak (model DRX 8533, TSI Several different direct-reading time resolved instruments Inc., USA). The DMS measured the particle number size were used in the four different measurement zones distribution in the size range 0.005–1 µm, with a time (Table 2). The aerodynamic particle diameter size dis- resolution of 1 s. The DustTrak measured the particle tribution in the range 0.5–20 µm was obtained by two mass concentration in four size fractions: PM ,PM , 1 2.5 Aerodynamic Particle Sizers (APS, model 3321, TSI Inc., respirable and PM , with a time resolution of 1 s. During USA) with a time resolution of 5 s. Two condensation Study B, a portable aethalometer (model AE51, AethLabs, particle counters (CPC, model 3775 and 3010, respec- USA) was used to measure the BC mass concentration in tively, TSI Inc., USA) were used to measure the total the emission zone, with a time resolution of 10 s. number concentration of particles > 0.007 µm, with a time During both studies, two instruments were carried by the resolution of 1 s. One APS and one CPC measured in the workers to measure in the personal breathing zone. A emission zone, while the second APS and CPC were similar second portable aethalometer to measure the BC measuring in the background zone. An aethalometer mass concentration, with a time resolution of 10 s, and an (model AE33, Magee Scientific, USA) was used to aerosol dosimeter (Partector, Naneos, Switzerland) to Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 741 Table 2 Direct reading instruments used for time resolved studies of particle emissions. Personal breathing zone Emission zone Background zone Supply air Study A Partector, portable aethalometer APS, CPC, DMS, DustTrak APS, CPC, Aethalometer P-Trak, DustTrak Study B Partector, portable aethalometer APS, CPC, portable aethalometer APS, CPC, Aethalometer P-Trak, DustTrak measure the lung deposited surface area (LDSA) con- used during Study A, and Fig. 2b shows the sampled centration, with a time resolution of 1 s. material found in the emission zone during handling of For particle measurements in the supply air, a P-Trak graphene nanoplatelet powder (work task A1). Figure 2c (model 8525, TSI Inc., USA) and a DustTrak (model DRX shows the raw graphene oxide material used during Study 8534, TSI Inc., USA) were used to assess the particle B, and Fig. 2d shows the sampled material found in the number concentration (0.02–1 µm) and the particle mass emission zone during handling of graphene oxide powder concentration (0.1–15 µm), respectively, with a time reso- (work task B1). The amounts of graphene nanomaterial in lution of 1 s. the emission zones (26 and 1.9 µg/m for Study A and B, respectively) were quantified as EC with thermal-optical Surface sampling method and analysis analysis. During Study B, EC (most likely from graphene oxide) was also quantified in the personal breathing zone During Study B, tape samples from different surfaces were (1.3 µg/m ). Two of the direct reading instruments (APS collected according to a tape stripping method described by and portable aethalometer) showed clear particle con- Hedmer et al. [54]. Tape samples were collected at the end centration peaks in the emission zone during the 1-min of the workdays from surfaces in the near-field zone of the weighing event of graphene oxide powder (performed exposure source (<1 m). Surface contaminations in the far- twice) during Study B (work task B1), see Fig. 3. The field zone (>1 m), including two offices and a conference coarse particle number size distribution (count median room, were also studied (Table 4). Two field blank tape aerodynamic diameter (CMD) of ~2 µm) of the initial gra- samples were also obtained. The tape samples were pre- phene oxide particle concentration peak (seen in Fig. 3)is pared and analyzed with the same SEM method as the air shown in Fig. 4. From the SEM images (Fig. 2c, d), the size filter samples. of the graphene oxide particles is estimated to be about 10–20 µm. Results Titanium dioxide nanofiber detection and quantification After the completion of tier 1 (contextual information gathering) and tier 2 (basic exposure assessment), we con- As shown in Table 3, TiO NFs, handled at the workplace cluded that tier 3 expert exposure assessments were needed during Study B, could be detected (Fig. 5) both in the at both companies. The tier 3 measurements are the ones emission zone and in the personal breathing zone during described herein. powder handling (work task B5). Figure 5a shows the raw TiO NF material used during Study B, and Fig. 5b, c shows Filter sampling examples of sampled material found in the emission zone and personal breathing zone, respectively. The sampled Table 3 shows the results from the filter samples collected fiber length was assessed to be 1–15 µm. TiO NF con- during work tasks using CB, graphene nanomaterials and centration in the emission zone was 2.2 µg/m , quantified as TiO NPs and NFs. Note that some of the filters have been Ti with PIXE. The APS in the emission zone showed par- sampled for more than one work task (according to ticle concentration peaks during the two-minute weighing of Table 1). During the one work task where Cu was used the TiO NF powder, see Fig. 6. The coarse particle number (A7), no concentration of Cu was detected in either the size distribution of the concentration peaks seen in Fig. 6 personal breathing zone or the emission zone. revealed a CMD of ~0.7 µm (not shown). Graphene detection and quantification Titanium dioxide nanoparticle detection and quantification Table 3 shows that graphene nanomaterials were detected (Fig. 2) on the emission zone filters at both workplaces. Spherical TiO NP agglomerates, handled at the work- Figure 2a shows the raw graphene nanoplatelet material place during Study A, were detected (Fig. 7)on filters 742 K. Lovén et al. Table 3 Results from the filter based measurements during the work tasks using graphene nanoplatelets, graphene oxide, titanium dioxide (TiO ) nanofibers (NFs) and nanoparticles (NPs) and carbon black (CB). Work task number Nanomaterials handled Sampling SEM analysis Elemental carbon Metal conc. 3 3 during the work task time (min) (µg/m ) Ti (µg/m ) Detection of nanomaterial Number conc. −3 (Yes/No)/Type (cm ) Study A (2016) Personal breathing zone aa A1 Graphene nanoplatelets 42 <LOD – A2—day 1 TiO NPs 83 Yes/TiO NPs 25 – 7.5 2 2 A2—day 2 TiO NPs 45 Yes/TiO NPs 1.8 – 2.1 2 2 A3-6—day 1 CB, graphene 106 Yes/CB 5.6 – A6—day 2 CB 39 –– <LOD – Emission zone A1 Graphene nanoplatelets 43 Yes/graphene 26 – nanoplatelets A2—day 1 TiO NPs 83 Yes/TiO NPs 25 – 70 2 2 A2—day 2 TiO NPs 57 Yes/TiO NPs 7.0 – 28 2 2 aa A3 Graphene 13 <LOD – A4 CB 15 Yes/CB <LOD – A5 CB 15 Yes/CB 98 – A6—day 1 CB 31 Yes/CB 8.2 – A6—day 2 CB 37 –– <LOD – Study B (2017) Personal breathing zone B1–4 Graphene oxide 172 No <LOD 1.3 – B5–6 TiO NFs 100 Yes/TiO NFs – <LOD 2 2 B7-8 TiO NFs 60 No <LOD – <LOD Emission zone B1–2 Graphene oxide 207 Yes/graphene oxide 1.9 – B3 Graphene oxide 32 No <LOD <LOD – aa B4 Graphene oxide 111 <LOD – B5 TiO NFs 54 Yes/TiO NFs – 2.2 2 2 B6 TiO NFs 129 No <LOD – <LOD B7 TiO NFs 23 No <LOD – <LOD B8 TiO NFs 123 No <LOD – 0.2 Background zone B1–4 Graphene oxide 480 No <LOD 0.2 B5–6 TiO NFs 202 No <LOD – <LOD B7–8 TiO NFs 173 No <LOD – <LOD –Not sampled. Not possible to determine. from both the emission zone and the personal breathing Surface contamination zone during handling of the powder on two separate days (work task A2—day 1 and A2—day 2). The concentration Workplace surfaces in the chemical laboratory, test reached 70 and 28 µg/m in the emission zone for the first laboratory, offices, and conference room were tape sam- and second day, respectively (quantified with the PIXE), pled in Study B. In total, 15 tape samples were collected, and 7.5 and 2.1 µg/m in the personal breathing zone the and TiO NFs were detected in 20% (n = 3) and graphene first and second day, respectively. Particle concentration oxide in 20% (n = 3) (Table 4). Surface contamination of peaks during weighing and mixing of the TiO NP pow- both TiO NFs and graphene oxide was found on only 2 2 der were detected by APS and DustTrak in the emission one surface, at the sink in the chemical laboratory which zone during both the first and second day, see Fig. 8. was related to work task B1–2and B5–6. A SEM image Similar to the PIXE results, the particle concentrations of the nanomaterial-related surface contamination can be measured with the APS and DustTrak were lower seen in Fig. 9. The length of the TiO NFs was assessed during the second day. The Partector in the personal to be 10–50 µm. TiO NFswerealsodetectedonthe plate breathing zone showed a particle exposure reaching a with TiO NF based coating after the abrasion test with peak lung deposited surface area concentration of 92 µm / the nylon brush. No surface contamination of nanoma- 3 2 3 cm during the first day and of only 9 µm /cm the second terial was found outside the chemical and test day (not shown). laboratories. Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 743 Fig. 2 SEM images of the different graphene raw materials and the materials measured in the emission zones. a The raw material of graphene nanoplatelets used during Study A, b graphene nanoplatelets found on the emission zone filter sampled during work task A1 during Study A, c the raw material of graphene oxide used during Study B, and d graphene oxide found on the emission zone filter sampled during work task B1–2 during Study B. Fig. 3 Coarse particle number concentration (APS, left y- axis) and black carbon concentration (portable aethalometer, µ-Aeth & aethalometer, Aeth, right y- axis) measured in the emission zone (EZ) and the background zone (BZ). The figure shows the measurements during weighing of the graphene oxide powder during Study B (work task B1). Note the different scales on the y-axes. Discussion of different methods, both offline and online, must be used. This study aimed to characterize workplace emissions and exposure of graphene nanoplatelets, graphene oxide, TiO Measurements of airborne graphene nanomaterials NFs, and NPs, as well as CB and Cu during industrial handling. The results showed that powder handling, The results show that graphene can be detected and quan- regardless of nanomaterial, generates the highest (among tified with several different methods. With the offline the investigated processes) particle emissions and expo- method consisting of filter sampling and SEM analysis we sures. It was shown that a range of different methods were, by comparing with the original nanomaterial, able to successfully can be used to selectively detect and quantify identify both graphene nanoplatelets used during Study A, nanomaterials both in the air and on surfaces, and that, to and graphene oxide used during Study B (Fig. 2). Lee et al. be able to make an accurate determination of which [18] and Vaquero et al. [21] have also reported similar nanomaterial that has been emitted when, a combination graphene-like structures in workplace air samples. 744 K. Lovén et al. According to OECD [48], both SEM and TEM are fre- intensive than the SEM/TEM method. In previous work- quently used for determination of nanomaterial present in place studies of graphene, Lee et al. [18] and Vaquero et al. the workplace air. [21] also used EC as an exposure metric for graphene, while In Fig. 4, the peak particle number size distribution exposure data of EC are missing in the studies by Spinazze during handling of graphene oxide powder (seen in Fig. 3) et al. [19] and Boccuni et al. [20]. In the current study, EC showed a CMD of ~2 µm. In the SEM images of both concentrations collected as total dust fractions were detected collected graphene nanoplatelets and graphene oxide, larger both in emission zone and personal breathing zone filter particles were found (10–20 µm). This is not contradictive samples, for example during weighing and mixing of gra- since 2D materials such as graphene may have an aero- phene oxide. Lee et al. [18] collected respirable fractions of dynamic diameter much smaller than its geometrical EC while Vaquero et al. [21] collected total dust fraction, dimensions [55], a fact that indicates its ability to reach the but none of these studies detected any airborne EC con- alveolar region of the lung when inhaled. centrations during manufacturing of graphene nanomaterial. The EC concentration (assumed to originate from gra- We used on-line measurements of equivalent BC (eBC) as phene) was quantified with thermal-optical analysis (OC/ a proxy for EC by using a portable aethalometer. This allowed EC) of filter samples. EC analysis is much less labor for highly time-resolved measurements and identification of exposure peaks for several work tasks. The combination of OC/EC analysis and portable aethalometer for graphene quantification has previously been used by Lee et al. [18]at a graphene nanoplatelets manufacturer. They found the clearest emission peaks in terms of eBC during graphene weighing. We have shown that these methods can be used also at down- stream handling facilities. The portable aethalometer and the APS (both time-resolved instruments) were used together with a thoroughly written logbook. By this approach we could determine that weighing and mixing the dry powder material generated the highest particle emission and exposure levels. The importance of documenting the activities to be able to match the measured concentration profiles have previously been demonstrated by e.g., Ham et al. [44], Hedmer et al. Fig. 4 Aerodynamic particle size distribution measured in the [51], and Isaxon et al. [49]. emission zone (APS EZ) and the background zone (APS BZ) in Both OC/EC analysis and eBC measurements of gra- Study B. The figure shows the size distribution of the initial peak in Fig. 3. phene nanomaterials is subject to cross sensitivity by other Fig. 5 SEM images of the TiO NF raw material and the material measured in the emission zone and the personal breathing zone in Study B. a The raw material of TiO NFs used, b TiO NFs found on the emission zone filter sampled during work task B5, and c TiO NFs found on the personal breathing zone filter sampled during work task B5–6. Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 745 carbonaceous nanomaterials such as CB and CNTs, but also emission zone compared with simultaneous measurements from soot particle sources, such as diesel exhaust, that may in the background zone during Study B, suggesting that be present in workplace air. Both EC and cross sensitivity to background EC or eBC was low. eBC concentrations were substantially elevated in the The eBC mass concentration was reported assuming the standard instrument settings. This may have affected the accuracy of the method as an offset of the results if the ENMs have an altered instrument response, for example by a different mass absorption cross section compared with the standard value that is based on atmospheric soot. This uncertainty can be tolerated when it comes to ENM expo- sures, where the aim often is to link exposure peaks with specific work tasks. In the literature, an eBC/EC ratio of 0.14 was found during simulated powder handling of CNTs [56]. Even though a combination of offline and online meth- ods is recommended, different methods can be used indi- vidually for specific needs, depending on what type of information that is needed. For graphene, the following can be used: (1) SEM analysis for accurate identification of the Fig. 6 Coarse particle number concentration measured in the nanomaterial, (2) EC analysis of filter samples for an emission zone (APS EZ) and the background zone (APS BZ). The accurate assessment of higher carbon concentrations (rela- figure shows the measurements during the 2-min weighing of the TiO NF powder during Study B (work task B5). tively high LOD), (3) direct reading portable aethalometer Fig. 7 SEM images of the TiO NP raw material and the material measured in the emission zone and the personal breathing zone in Study A. a The used raw material of titanium dioxide NPs, b TiO NPs found on the emission zone filter sampled during work task A2—day 1, c TiO NPs found on the emission zone filter sampled during work task A2— day 2, d TiO NPs found on the personal breathing zone filter sampled during work task A2— day 1, and e TiO NPs found on the personal breathing zone filter sampled during work task A2— day 2. 746 K. Lovén et al. Fig. 8 Coarse particle number concentration (APS, left y- axis) and particle mass concentration (DustTrak, right y-axis) measured in the emission zone (EZ) and the background zone (BZ). The figure shows the measurements during weighing and mixing of the TiO2 NP powder during Study A. a Shows work task A2 —day 1 and b shows work task A2—day 2. for highly time-resolved eBC assessment, and (4) direct phagocytosis and accumulation over time finally causing lung reading APS for low-concentration (although less disease [57]. During Study B, TiO NFs were detected both in specific) detection of particle emissions (low LOD). This the emission zone and the personal breathing zone with the means that with access to one of these methods, a pre- offline filter sampling method followed by SEM analysis liminary emission and exposure assessment should be (Fig. 5). They appear to be present in the workplace air as possible to conduct. bundles, and some of the fibers detected by SEM during Study B were at least 5 µm and could therefore constitute a Measurements of airborne titanium dioxide risk if inhaled. Bianchi et al. [58]haverecentlyshown that nanofibers and nanoparticles long (around 10 µm) TiO NFs induce cell cytotoxicity in vitro and inflammation in vivo, while shortened (around 2 NFs are another type of nanomaterials that are of concern, µm) TiO NFs seem to mitigate the toxic effects, even without especially if they are long (>5 µm) and insoluble, due to their macrophages present in the in vitro cultures. SEM-analysis potential to cause adverse health effects, including frustrated can also be used to quantify the particle concentration in the Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 747 Table 4 An overview of the 15 tape-sampled surface locations and the presence of titanium dioxide nanofibers and graphene oxide particles on different surfaces found during Study B. Sampling location Room Related to work Surface characteristics SEM analysis task number Material Assessed indication Detection of TiO Detection of graphene of roughness nanofiber (Yes/No) oxide particles (Yes/No) Work areas Work areas in fume hood Chemistry laboratory B1, B5 Stainless steel Smooth No No Cover on work area B1, B5 Cardboard Smooth No No Sink B1–2, B5–6 Stainless steel Smooth Yes Yes Desk next to abrasion testing machine Test laboratory B3–4, B7–8 Laminate Smooth No Yes Floors Floor in front of fume hood Chemistry laboratory B1, B5 Epoxy coated concrete Smooth No No Floor in front of spray booth Test laboratory B2, B6 Epoxy coated concrete Smooth No No Floor Conference room – Parquet Smooth No No Floor Office 1 – Laminate Smooth No No Door sill Office 2 – Wood Smooth No No Other surfaces Balance in fume hood Chemistry laboratory B1, B5 Plastic Smooth No Yes Cover under the brush in the abrasion Test laboratory B3–4, B7–8 Metal Smooth Yes No testing machine Plate with graphene based coating, B3 Metal with graphene based Smooth No sampled after abrasion test with coating metal brush Plate with graphene based coating, B4 Metal with graphene based Smooth No sampled after abrasion test with coating nylon brush Plate with TiO nanofiber based B7 Metal with TiO nanofiber Smooth No No 2 2 coating, sampled after abrasion test based coating with metal brush Plate with TiO nanofiber based B8 Metal with TiO nanofiber Smooth Yes No 2 2 coating, sampled after abrasion test based coating with nylon brush –Far-field surfaces. Not possible to determine which type of particles that are detected . 748 K. Lovén et al. second day, this procedure had been adjusted so that the TiO NP powder were directly mixed in with the liquid. This clearly shows that with an easy change in the handling procedure, the particle emissions and exposures can be lowered by as much as a factor 10 (for example seen with the Partector where the lung deposited surface area went 2 3 2 3 from 92 µm /cm during the first day to only 9 µm /cm during the second day). The Partector has also previously been shown to be an important tool for personal exposure assessments to improve workplace monitoring [60–62]. Quantification of Ti from TiO nanomaterials by ICP-MS is Fig. 9 SEM image of TiO NFs and graphene oxide. The image 2 shows materials found as surface contamination at the sink in the not possible due to that TiO is a poorly soluble oxide. Laser- chemical laboratory in Study B. ICP-MS could be one alternative [45]and ICP-OES(optical emission spectrometer) [42] another. The current study has highlighted PIXE as a valid and reliable method for titanium air. However, it is a time consuming process, and during quantification, which has also been shown by Relier et al. [63]. Study B it was shown to be difficult to distinguish the dif- ferent fibers due to agglomeration and therefore to count Surface contamination them. The PIXE analysis was a valid alternative, even though it gives the concentration as a different metric (mass con- Surface contamination of TiO NFs and graphene oxide was centration, µg/m ). Detection and quantification can also be for the first time studied at down-stream handling processes. conducted with direct reading measurements (APS could be Surface contaminations of both TiO NFs and graphene used to determine what specific handling process generated oxide were found on one surface (sink) in the near-field the particles (Fig. 6); weighing the dry powder material), but zone in the chemical laboratory in Study B. The detected to confirm fiber emissions, time-integrated filter sampling surface contamination can probably be related to washing of followed by SEM is necessary. equipment after preparations of both types of coatings. The General NP dry powder handling has previously been surface contamination of TiO NFs and graphene oxide shown to constitute a potential source for worker exposure could probably be resuspended into the workplace air, by e.g., Huang et al. [42]and Curwin andBertke[43]. Lee which if so would cause a risk of secondary inhalation et al. [59] found substantial total mass concentrations exposure. The percentage of TiO NF surface contamination during powder collection of TiO NPs in the pigment in the collected tape samples were lower compared with industry (500–5000 µg/m ), however, no elemental ana- what was previously found in a similar study [54]. This lysis of Ti was carried out. In the current study, this was indicates that there was no widespread nanomaterial con- confirmed, were elevated concentrations of TiO NPs tamination at workplace B. An interesting finding was that were found in both the emission zone and the personal TiO NFs could be detected on a tape sample from one of breathing zone (with SEM, PIXE, APS, DustTrak, and the abrasion-tested metal plates with TiO NF based coat- Partector) during weighing and mixing of dry TiO NP ing. During the abrasion tests, dust containing manufactured powder. Interestingly, the mass concentrations of Ti nanomaterials is generated and could be emitted to the (measured with PIXE) were about a factor 10 lower in the workplace air, especially since this process was openly personal breathing zone than in the emission zone. This performed at the workplace without any engineering reduction factor was most likely achieved by the use of controls. engineering controls in the emission zone (fume hood) and such a factor could be used for other assessments of Proposed occupational exposure limits personal exposure where only measurements in the emission zone can be conducted or vice versa. Mihalache et al. [64] have reviewed the proposed occupa- Another observation was that the particle concentrations tional exposure limits (OELs) for a number of measured with PIXE, APS, DustTrak, and Partector were all different nanomaterials. For TiO NPs, a range between 17 lower during the second day compared with the first day of and 2000 µg/m have been proposed, with probably the weighing and mixing of the TiO NP powder. As shown in most established one by NIOSH of 300 µg/m as an 8 h Fig. 8, the procedure of mixing the powder with the liquid average airborne exposure [65]. The highest TiO con- had been adjusted from the first to the second day. During centration in the personal breathing zone during the current the first day, the TiO NP powder were weighed first and study was 7.5 µg/m , corresponding to an 8 h average of then the liquid was added to form the printing ink. The 1.3 µg/m , well below the suggested OEL. For TiO NFs, 2 Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 749 there is no specific OEL, but the BSI [66] suggested OEL analysis is also a good option, especially when aiming to for fibrous nanomaterials of 0.01 fibers/cm could be a compare the result with existing OEL, if one can ensure suitable OEL to consult. However, in the current study it that no other sources of that chemical compound are was found to be difficult to distinguish the different fibers present in the workplace. For carbon-based materials by SEM due to agglomeration and therefore challenging to (such as graphene) filter sampling with following EC count them. The detected fibers were, furthermore, only analysis can be used. In the current study, open-face filter found during the low resolution scanning-to-identify sampling was used, but it could have been possible to investigation and none were found during the quantifica- instead use a respirable cyclone to get the respirable EC tion imaging, rendering the fiber number concentration concentrations. Then, it would be possible to compare below the LOD. Tsang et al. [67] demonstrated that a measured EC concentrations with the recommended probabilistic approach can be used in the risk assessment of exposure limit of respirable EC set to 1 µg/m as an 8 h exposure scenarios involving production of TiO NPs. time weighted average concentration for CNTs and CNFs Thus, they could identify one out of seven exposure sce- [72]. For TiO nanomaterials, filter sampling with fol- narios with statistically significant level of risk. lowing PIXE analysis would be the recommended che- Neither for graphene nanomaterials are there any legal mical analysis to perform. binding OELs yet, but e.g., Lee et al. [68] have based on data from a subchronical inhalation toxicity study calculated a recommended OEL for graphene nanomaterial to be 18 Conclusions µg/m . Also, health based guidance value for occupational inhalation exposure to graphene nanoplatelets was esti- Down-stream industrial handling of graphene nanomaterials mated to 212 µg/m by Spinazze et al. [69]. However, so far and TiO NFs were, for the first time, investigated for there are only few in vivo inhalation exposure studies particle emissions and exposures into the workplace air. We reported in the literature. Thus, according to Pelin et al. [3] showed that weighing the dry powder material generated OEL for graphene cannot be determined based on the particle emissions, even though the exposure levels were available data. The highest EC concentration in the personal low compared with proposed OELs. Surface contaminations breathing zone, in the current study, was 5.6 µg/m , corre- of both TiO NFs and graphene were found on the sink in sponding to an 8 h average of 1.2 µg/m , well below the the chemical laboratory at workplace B. A range of different suggested OELs. sampling and aerosol monitoring methods was used and evaluated. For a fast and reliable workplace emission and Recommendations for occupational hygienists exposure assessment of graphene and TiO nanomaterials, a combination of different methods, both offline and online, As a general recommendation for occupational hygienists, must be used to ensure the detected emissions contain the the portable aethalometer could be one option for a small specific nanomaterials. If a first exposure assessment should and easy-to-use instrument for personal exposure mon- be performed (according to tier 2 of the OECD guidelines), itoring of carbon-based nanomaterials. However, from the access to only one or a few of these methods is enough. current study, data from the personal sampling portable When a tier 3 exposure assessment is needed, a combination aethalometer could not be used due to too much noise, of multiple of the mentioned methods should be used most likely arising from the short sampling time. This together with additional direct reading instruments such as could possibly be addressed by longer averaging times CPC and APS. [70] as well as noise reduction treatment during the data Acknowledgements The authors would like to acknowledge the fol- post-processing [71]. As a stationary instrument in the lowing persons: Louise Gren for helping out with the on-site work- emission zone, the portable aethalometer worked well and place measurements during Study B. Linus Ludvigsson for helping out can be an alternative to bulkier, more expensive and with the SEM analyses. Jan Pallon for conducting the PIXE analysis. advanced instruments. For non-carbon-based nanomater- Thomas Lundh for conducting the metal analysis with ICP-MS. The project was funded by AFA insurance (Dnr 130122) and NanoLund ials, a suitable personal exposure instrument option would (p38-2013), and was carried out within the framework of NanoLund at instead be a lung deposited surface area instrument. Lund University. However, when it is critical with an accurate determina- tion of which nanomaterial that is emitted, SEM analysis 1 2 1 NanoLund Karin Lovén , Sara M. Franzén , Christina Isaxon , Maria 2 1 1 4 must be performed as a complement to the direct reading E. Messing , Anders Gudmundsson , Joakim Pagels , Maria Hedmer measurements. Open-face filter sampling was used in the current study to morphologically characterize the real Funding The project was funded by AFA insurance (Dnr 130122) and emission and exposure situations of all released NPs and NanoLund (p38-2013), and was carried out within the framework of NanoLund at Lund University. NFs and their aggregates and agglomerates. Chemical 750 K. Lovén et al. Compliance with ethical standards 14. Hu XG, Zhou QX. Health and ecosystem risks of graphene. Chem Rev. 2013;113:3815–35. 15. Guo XQ, Mei N. Assessment of the toxic potential of graphene Conflict of interest The authors declare no conflict of interest. The family nanomaterials. J Food Drug Anal. 2014;22:105–15. project sponsors have only contributed financially, and have not par- 16. Fadeel B, Bussy C, Merino S, Vazquez E, Flahaut E, Mouchet F, ticipated in preparing the research material, writing, reviewing, or et al. Safety assessment of graphene-based materials: focus on approving the submitted paper. human health and the environment. ACS Nano. 2018;12:10582–620. Publisher’s note Springer Nature remains neutral with regard to 17. Basinas I, Jimenez AS, Galea KS, van Tongeren M, Hurley F. A jurisdictional claims in published maps and institutional affiliations. Systematic review of the routes and forms of exposure to engi- neered nanomaterials. Ann Work Expo Heal. 2018;62:639–62. Open Access This article is licensed under a Creative Commons 18. Lee JH, Han JH, Kim JH, Kim B, Bello D, Kim JK, et al. Attribution 4.0 International License, which permits use, sharing, Exposure monitoring of graphene nanoplatelets manufacturing adaptation, distribution and reproduction in any medium or format, as workplaces. Inhal Toxicol. 2016;28:281–91. long as you give appropriate credit to the original author(s) and the 19. Spinazze A, Cattaneo A, Campagnolo D, Bollati V, Bertazzi PA, source, provide a link to the Creative Commons license, and indicate if Cavallo DM. Engineered nanomaterials exposure in the produc- changes were made. The images or other third party material in this tion of graphene. Aerosol Sci Technol. 2016;50:812–21. article are included in the article’s Creative Commons license, unless 20. Boccuni F, Ferrante R, Tombolini F, Lega D, Antonini A, Alvino indicated otherwise in a credit line to the material. If material is not A, et al. Workers’ exposure to nano-objects with different included in the article’s Creative Commons license and your intended dimensionalities in R&D laboratories: measurement strategy and use is not permitted by statutory regulation or exceeds the permitted field studies. Int J Mol Sci. 2018;19:349. use, you will need to obtain permission directly from the copyright 21. Vaquero C, Wendelbo R, Egizabal A, Gutierrez-Cañas C, López holder. To view a copy of this license, visit http://creativecommons. de Ipiña J. Exposure to graphene in a pilot production plant. J org/licenses/by/4.0/. Phys Conf Series. 2019;1323. 22. Khataee A, Moradkhannejhad L, Heydari V, Vahid B, Joo SW. Self-cleaning acrylic water-based white paint modified with different types of TiO2 nanoparticles. Pigm Resin Technol. 2016;45:24–9. References 23. Sharma S, Sharma RK, Gaur K, Torres JFC, Loza-Rosas SA, Torres A, et al. Fueling a hot debate on the application of TiO2 1. Castro Neto AH, Guinea F, Peres NMR, Novoselov KS, Geim nanoparticles in sunscreen. Materials. 2019;12:2317. AK. The electronic properties of graphene. Rev Mod Phys. 24. Dastan D, Gosavi SW, Chaure NB. Studies on electrical properties 2009;81:109–62. of hybrid polymeric gate dielectrics for field effect transistors. 2. Balandin AA. Thermal properties of graphene and nanostructured Macromol Symp. 2015;347:81–6. carbon materials. Nat Mater. 2011;10:569–81. 25. Victorious A, Clifford A, Saha S, Zhitomirsky I, Soleymani L. 3. Pelin M, Sosa S, Prato M, Tubaro A. Occupational exposure to Integrating TiO2 nanoparticles within a catecholic polymeric graphene based nanomaterials: risk assessment. Nanoscale. network enhances the photoelectrochemical response of bio- 2018;10:15894–903. sensors. J Phys Chem C. 2019;123:16186–93. 4. Chen ZH, Lin YM, Rooks MJ, Avouris P. Graphene nano-ribbon 26. Ou G, Li ZW, Li DK, Cheng L, Liu Z, Wu H. Photothermal electronics. Physica E. 2007;40:228–32. therapy by using titanium oxide nanoparticles. Nano Res. 5. Schedin F, Geim AK, Morozov SV, Hill EW, Blake P, Katsnelson 2016;9:1236–43. MI, et al. Detection of individual gas molecules adsorbed on 27. Pi JK, Yang HC, Wan LS, Wu J, Xu ZK. Polypropylene micro- graphene. Nat Mater. 2007;6:652–5. filtration membranes modified with TiO2 nanoparticles for surface 6. Loh KP, Bao QL, Eda G, Chhowalla M. Graphene oxide as a wettability and antifouling property. J Membr Sci. 2016;500:8–15. chemically tunable platform for optical applications. Nat Chem. 28. Tang YJ, Hu XL, Zhang XQ, Guo DL, Zhang JH, Kong FO. 2010;2:1015–24. Chitosan/titanium dioxide nanocomposite coatings: rheological 7. Rafiee J, Mi X, Gullapalli H, Thomas AV, Yavari F, Shi YF, et al. behavior and surface application to cellulosic paper. Carbohyd Wetting transparency of graphene. Nat Mater. 2012;11:217–22. Polym. 2016;151:752–9. 8. Ding R, Li WH, Wang X, Gui TJ, Li BJ, Han P, et al. A 29. Nalbandian MJ, Greenstein KE, Shuai DM, Zhang ML, Choa YH, brief review of corrosion protective films and coatings based on Parkin GF, et al. Tailored synthesis of photoactive TiO2 nanofi- graphene and graphene oxide. J Alloy Compd. 2018; bers and Au/TiO2 nanofiber composites: structure and reactivity 764:1039–55. optimization for water treatment applications. Environ Sci Tech- 9. Berman D, Erdemir A, Sumant AV. Few layer graphene to reduce nol. 2015;49:1654–63. wear and friction on sliding steel surfaces. Carbon. 30. Safajou H, Khojasteh H, Salavati-Niasari M, Mortazavi-Derazkola 2013;54:454–9. S. Enhanced photocatalytic degradation of dyes over graphene/Pd/ 10. Mutyala KC, Wu YMA, Erdemir A, Sumant AV. Graphene— TiO2 nanocomposites: TiO2 nanowires versus TiO2 nano- MoS2 ensembles to reduce friction and wear in DLC-Steel con- particles. J Colloid Interface Sci. 2017;498:423–32. tacts. Carbon. 2019;146:524–7. 31. Wang JX, Fan YB. Lung injury induced by TiO2 nanoparticles 11. Jastrzebska AM, Kurtycz P, Olszyna AR. Recent advances in depends on their structural features: size, shape, crystal phases, graphene family materials toxicity investigations. J Nanopart Res. and surface coating. Int J Mol Sci. 2014;15:22258–78. 2012;14:1320. 32. Grassian VH, O’Shaughnessy PT, Adamcakova-Dodd A, Petti- 12. Sanchez VC, Jachak A, Hurt RH, Kane AB. Biological interac- bone JM, Thorne PS. Inhalation exposure study of titanium tions of graphene-family nanomaterials: an interdisciplinary dioxide nanoparticles with a primary particle size of 2 to 5 nm. review. Chem Res Toxicol. 2012;25:15–34. Environ Health Perspect. 2007;115:397–402. 13. Arvidsson R, Molander S, Sanden BA. Review of potential 33. Pujalte I, Dieme D, Haddad S, Serventi AM, Bouchard M. Tox- environmental and health risks of the nanomaterial graphene. icokinetics of titanium dioxide (TiO2) nanoparticles after inhala- Hum Ecol Risk Assess. 2013;19:873–87. tion in rats. Toxicol Lett. 2017;265:77–85. Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 751 34. Lauvas AJ, Skovmand A, Poulsen MS, Kyjovska ZO, Roursgaard 51. Hedmer M, Isaxon C, Nilsson PT, Ludvigsson L, Messing ME, M, Goericke-Pesch S, et al. Airway exposure to TiO2 nano- Genberg J, et al. Exposure and emission measurements during particles and quartz and effects on sperm counts and testosterone production, purification, and functionalization of Arc-discharge- levels in male mice. Reprod Toxicol. 2019;90:134–40. produced multi-walled carbon nanotubes. Ann Occup Hyg. 35. Wang JX, Liu Y, Jiao F, Lao F, Li W, Gu YQ, et al. Time- 2014;58:355–79. dependent translocation and potential impairment on central ner- 52. Johansson SAE. PIXE – A Novel Technique for Elemental vous system by intranasally instilled TiO2 nanoparticles. Tox- Analysis. Endeavour. 1989;13:48–53. icology. 2008;254:82–90. 53. Nilsson PT, Isaxon C, Eriksson AC, Messing ME, Ludvigsson L, 36. Falck GCM, Lindberg HK, Suhonen S, Vippola M, Vanhala E, Rissler J, et al. Nano-objects emitted during maintenance of Catalan J, et al. Genotoxic effects of nanosized and fine TiO2. common particle generators: direct chemical characterization with Hum Exp Toxicol. 2009;28:339–52. aerosol mass spectrometry and implications for risk assessments. J 37. Di Bucchianico S, Cappellini F, Le Bihanic F, Zhang YN, Dreij Nanopart Res. 2013;15:1–16. K, Karlsson HL. Genotoxicity of TiO2 nanoparticles assessed by 54. Hedmer M, Ludvigsson L, Isaxon C, Nilsson PT, Skaug V, mini-gel comet assay and micronucleus scoring with flow cyto- Bohgard M, et al. Detection of multi-walled carbon nanotubes and metry. Mutagenesis. 2017;32:127–37. carbon nanodiscs on workplace surfaces at a small-scale producer. 38. Pelclova D, Zdimal V, Kacer P, Fenclova Z, Vlckova S, Komarc Ann Occup Hyg. 2015;59:836–52. M, et al. Leukotrienes in exhaled breath condensate and fractional 55. Schinwald A, Murphy FA, Prina-Mello A, Poland CA, Byrne F, exhaled nitric oxide in workers exposed to TiO2 nanoparticles. J Movia D, et al. The threshold length for fiber-induced acute Breath Res. 2016;10:036004. pleural inflammation: shedding light on the early events in 39. Hurbankova M, Cerna S, Kovacikova Z, Wimmerova S, Hraskova asbestos-induced mesothelioma. Toxicol Sci. 2012;128:461–70. D, Marcisiakova J, et al. Effect of TiO2 nanofibres on selected 56. Hashimoto N, Ogura I, Kotake M, Kishimoto A, Honda K. bronchoalveolar parameters in acute and subacute phase— Evaluating the capabilities of portable black carbon monitors and experimental study. Cent Eur J Publ Heal. 2013;21:165–70. photometers for measuring airborne carbon nanotubes. J Nanopart 40. Medina-Reyes EI, Delgado-Buenrostro NL, Deciga-Alcaraz A, Res. 2013;15:1–17. Freyre-Fonseca V, Flores-Flores JO, Hernandez-Pando R, et al. 57. Donaldson K, Murphy FA, Duffin R, Poland CA. Asbestos, carbon Titanium dioxide nanofibers induce angiogenic markers and nanotubes and the pleural mesothelium: a review of the hypothesis genomic instability in lung cells leading to a highly dediffer- regarding the role of long fibre retention in the parietal pleura, entiated and fibrotic tumor formation in a xenograft model. inflammation and mesothelioma. Part Fibre Toxicol. 2010;7:5. Environ Sci-Nano. 2019;6:286–304. 58. Bianchi MG, Campagnolo L, Allegri M, Ortelli S, Blosi M, Chiu 41. Allegri M, Bianchi MG, Chiu M, Varet J, Costa AL, Ortelli S, M, et al. Length-dependent toxicity of TiO2 nanofibers: mitigation et al. Shape-related toxicity of titanium dioxide nanofibres. PLos via shortening. Nanotoxicology. 2019;14:433–52. ONE. 2016;11:e0151365. 59. Lee JH, Kwon M, Ji JH, Kang CS, Ahn KH, Han JH, et al. 42. Huang CH, Tai CY, Huang CY, Tsai CJ, Chen CW, Chang CP, Exposure assessment of workplaces manufacturing nanosized et al. Measurements of respirable dust and nanoparticle con- TiO2 and silver. Inhal Toxicol. 2011;23:226–36. centrations in a titanium dioxide pigment production factory. J 60. Geiss O, Bianchi I, Barrero-Moreno J. Lung-deposited surface Environ Sci Heal A. 2010;45:1227–33. area concentration measurements in selected occupational and 43. Curwin B, Bertke S. Exposure characterization of metal oxide non-occupational environments. J Aerosol Sci. 2016;96:24–37. nanoparticles in the workplace. J Occup Environ Hyg. 61. Todea AM, Beckmann S, Kaminski H, Bard D, Bau S, Clavaguera 2011;8:580–7. S, et al. Inter-comparison of personal monitors for nanoparticles 44. Ham S, Yoon C, Lee E, Lee K, Park D, Chung E, et al. Task- exposure at workplaces and in the environment. Sci Total based exposure assessment of nanoparticles in the workplace. J Environ. 2017;605:929–45. Nanopart Res. 2012;14:1–17. 62. Iavicoli I, Fontana L, Pingue P, Todea AM, Asbach C. Assess- 45. Xu HD, Zhao L, Chen ZJ, Zhou JW, Tang SC, Kong FL, et al. ment of occupational exposure to engineered nanomaterials in Exposure assessment of workplace manufacturing titanium diox- research laboratories using personal monitors. Sci Total Environ. ide particles. J Nanopart Res. 2016;18:1–13. 2018;627:689–702. 46. Debia M, Bakhiyi B, Ostiguy C, Verbeek JH, Brouwer DH, 63. Relier C, Dubreuil M, Garcia OL, Cordelli E, Mejia J, Eleuteri P, Murashov V. A systematic review of reported exposure to engi- et al. Study of TiO2 P25 nanoparticles genotoxicity on lung, blood, neered nanomaterials. Ann Occup Hyg. 2016;60:916–35. and liver cells in lung overload andnon-overload conditionsafter 47. OECD. Harmonized tiered approach to measure and assess the repeated respiratory exposure in rats. Toxicol Sci. 2017;156:527–37. potential exposure to airborne emissions of engineered nano- 64. Mihalache R, Verbeek J, Graczyk H, Murashov V, van Broe- objects and their agglomerates and aggregates at workplaces. khuizen P. Occupational exposure limits for manufactured nano- Series on the safety of manufactured nanomaterials organisation materials, a systematic review. Nanotoxicology. 2017;11:7–19. for economic co-operation and development. J Nanopart Res: 65. NIOSH. US National Institute for Occupational Safety and Health. OECD; 2015. p. 55. US: NIOSH TiO2 OEL 2011.pdf; 2011. 48. OECD. Strategies, techniques and sampling protocols for deter- 66. BSI. Nanotechnologies—Part 2: guide to safe handling and dis- mining the concentrations of manufactured nanomaterials in air at posal of manufactured nanomaterials. PD 6699-2:2007 UK: the workplace. Series on the safety of manufactured nanomaterials British Standards Institution; 2007. organisation for economic co-operation and development. J 67. Tsang MP, Hristozov D, Zabeo A, Koivisto AJ, Jensen ACO, Nanopart Res: OECD; 2017. p. 82. Jensen KA, et al. Probabilistic risk assessment of emerging 49. Isaxon C, Lovén K, Ludvigsson L, Sivakumar S, Gudmundsson materials: case study of titanium dioxide nanoparticles. Nano- A, Messing ME, et al. Workplace emissions and exposures during toxicology. 2017;11:558–68. semiconductor nanowire production, post-production, and main- 68. Lee YS, Sung JH, Song KS, Kim JK, Choi BS, Yu IJ, et al. tenande work. Ann Work Expo Heal. 2020;64:38–54. Derivation of occupational exposure limits for multi-walled car- 50. Birch ME, Cary RA. Elemental carbon-based method for mon- bon nanotubes and graphene using subchronic inhalation toxicity itoring occupational exposures to particulate diesel exhaust. data and a multi-path particle dosimetry model. Toxicol Res. Aerosol Sci Technol. 1996;25:221–41. 2019;8:580–6. 752 K. Lovén et al. 69. Spinazze A, Cattaneo A, Borghi F, Del Buono L, Campagnolo D, 71. Hagler GSW, Yelverton TLB, Vedantham R, Hansen ADA, Rovelli S, et al. Probabilistic approach for the risk assessment of Turner JR. Post-processing method to reduce noise while preser- nanomaterials: a case study for graphene nanoplatelets. Int J Hyg ving high time resolution in aethalometer real-time black carbon Envir Heal. 2019;222:76–83. data. Aerosol Air Qual Res. 2011;11:539–46. 70. AethLabs. microAeth AE51 Operating Manual. US: AethLabs; 72. NIOSH. Occupational exposure to carbon nanotubes and nanofi- 2016. bers. Current Intell Bull. 2013;65. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Exposure Science & Environmental Epidemiology Springer Journals

Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles during down-stream industrial handling

Loading next page...
 
/lp/springer-journals/emissions-and-exposures-of-graphene-nanomaterials-titanium-dioxide-iWZqjjY1bn

References (84)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2020
ISSN
1559-0631
eISSN
1559-064X
DOI
10.1038/s41370-020-0241-3
Publisher site
See Article on Publisher Site

Abstract

Today, engineered nanomaterials are frequently used. Nanosized titanium dioxide (TiO ) has been extensively used for many years and graphene is one type of emerging nanomaterial. Occupational airborne exposures to engineered nanomaterials are important to ensure safe workplaces and to extend the information needed for complete risk assessments. The main aim of this study was to characterize workplace emissions and exposure of graphene nanoplatelets, graphene oxide, TiO nanofibers (NFs) and nanoparticles (NPs) during down-stream industrial handling. Surface contaminations were also investigated to assess the potential for secondary inhalation exposures. In addition, a range of different sampling and aerosol monitoring methods were used and evaluated. The results showed that powder handling, regardless of handling graphene nanoplatelets, graphene oxide, TiO NFs, or NPs, contributes to the highest particle emissions and exposures. However, the exposure levels were below suggested occupational exposure limits. It was also shown that a range of different methods can be used to selectively detect and quantify nanomaterials both in the air and as surface contaminations. However, to be able to make an accurate determination of which nanomaterial that has been emitted a combination of different methods, both offline and online, must be used. ● ● ● ● Keywords Occupational exposure Electron microscopy Thermal-optical carbon analysis Direct-reading instruments PIXE Aerosol Introduction robust, but also flexible, structure rendering it useful in a variety of applications [1]. It can exist in different structures Due to their novel and valuable properties compared with such as graphene, graphene oxide, and graphene nanopla- bulk materials, the use of engineered nanomaterials is telets [2, 3]. Its electrical and thermal properties makes increasing. Graphene, a 2D carbon nanomaterial, has a graphene useful in for example transistors [2, 4] and che- mical sensors [5], and the optical properties can be used in biological sensors [6]. Another application of graphene is as metal surface coatings to inhibit corrosion [7, 8] and to Members of the NanoLund are listed below Acknowledgements. reduce wear and friction on sliding metal surfaces [9, 10]. Multiple reviews [11–15] have generally concluded that * Karin Lovén graphene toxicity depends on the physiochemical properties Karin.loven@design.lth.se of the nanomaterial. However, one of the most widely Ergonomics and Aerosol Technology, Lund University, SE-22100 recognized mechanism for graphene-nanomaterial-induced Lund, Sweden toxicity in living systems is the induction of oxidative stress Solid State Physics, Lund University, SE-22100 Lund, Sweden and production of reactive oxygen species (ROS) [15]. In a more recent review by Fadeel et al. [16], the authors Medical Radiation Physics, Department of Translational Medicine, Lund University, SE-22100 Malmö, Sweden highlighted the need for standardized graphene character- izations, and of robust and validated toxicological assays in Occupational and Environmental Medicine, Lund University, SE- 22100 Lund, Sweden order to advance the field of graphene toxicity. 1234567890();,: 1234567890();,: Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 737 Studies of occupational exposure to nanomaterials are the NPs into other products has been investigated. Different needed to make complete risk assessments. Basinas et al. handling tasks, both with NP dry powder and NP containing [17] showed that many exposure assessments have been liquid, have been shown to constitute an occupational done for carbon nanotubes (CNTs), carbon nanofibers exposure risk [42–45]. The review by Debia et al. [46] (CNFs), and titanium dioxide nanoparticles (TiO NPs), but further strengthen this conclusion. However, studies of TiO 2 2 far less for other engineered nanomaterials, such as gra- NF exposure in occupational settings have, to our knowl- phene. Sanchez et al. [12] stated that there is a need for edge, not yet been conducted. measurements of airborne graphene exposure levels at both The Organisation for Economic Cooperation and research laboratories and full-scale manufacturing facilities. Development (OECD) have suggested a harmonized three- This was further pointed out by Arvidsson et al. [13], who tiered approach for nanomaterial emission and exposure requested that workplace emissions and exposures of gra- assessments [47] and recommendations for measurement phene should be investigated. Only a few emission and strategies and instrument use [48]. exposure measurements during production [18–21] and This study aims to generate new knowledge about handling [18] of graphene nanomaterials have since then emissions and exposures of nanomaterials not extensively been conducted. These studies showed low levels of studied previously. Emissions and exposures, with a focus exposures to graphene. However, according to Lee et al. on different graphene nanomaterials (both nanoplatelets and [18], monitoring of other work tasks including down-stream oxide) and different TiO nanomaterials (both NFs and graphene handling processes is needed for a full under- NPs), were characterized with a multi-metric approach standing of the exposure situation. To our knowledge, no during down-stream industrial handling. Emission and studies of emission and exposure measurements during exposures of carbon black (CB) and copper (Cu) were also down-stream handling processes of graphene nanomaterials, measured in a few cases. Different sampling and aerosol such as manufacturing of ink and surface coatings con- monitoring methods were evaluated to be able to recom- taining graphene, have previously been conducted. mend methods to be used specifically for these nanomater- Another common nanomaterial is TiO . It is used in ials to complement the different tiers and measurement paints and sunscreens [22, 23] as well as in transistors [24], strategies described by OECD. An additional aim was to biosensors [25], cancer treatment [26], and different surface assess the potential of secondary inhalation exposure, caused coatings [27, 28]. Nanosized TiO is commonly found as by resuspension of particles deposited on surfaces. spherical NPs, but can also be produced in other shapes including NFs [29] and nanowires [30]. The toxicity of nanosized TiO , especially spherical ones, has been studied Methods to a greater extent than graphene. As with graphene, the different physiochemical properties of TiO NPs have a Facilities strong influence on the toxicity [31]. Generally though, only moderate effects have been observed, including pulmonary Measurements were performed 2016 and 2017 at two dif- inflammation [32–34] and pathological neural changes [35] ferent workplaces, with 20 and 15 employees respectively, after inhalation/instillation in rodents. Induction of DNA hereafter “Study A” and “Study B”. As recommended by damage has also been observed in lung cell studies [36, 37]. the OECD [47], initial contextual information was gathered A few studies have been carried out on exposed workers. (tier 1) and basic exposure assessments (tier 2) were per- Pelclova et al. [38] showed for example that the leukotriene formed at both workplaces prior to conducting the expert levels in exhaled breath condensate were elevated in exposure assessment (tier 3) described herein. workers exposed to TiO NPs. The toxicity of TiO NFs has During Study A, different nanomaterials including gra- 2 2 not been as thoroughly investigated. Hurbánková et al. [39] phene nanoplatelets, spherical TiO NPs, CB, and Cu were showed the development of serious lung inflammatory and handled. The nanomaterials were used in ink formulations cytotoxic processes after intratracheal instillation of TiO for printing electronics, sensors, and labels. Measurements NFs in rats and Medina-Reyes et al. [40] observed cell were conducted in a chemistry laboratory and a printing cytotoxicity and genomic instability of TiO NF exposure laboratory. In Study B, the nanomaterials handled included on alveolar epithelial cells. In addition, Allegri et al. [41] graphene oxide and TiO NFs for use in friction and wear performed a comparative exposure study on alveolar epi- reducing surface coatings. Measurements were conducted in thelial cells showing that TiO NFs were more toxic than a chemistry laboratory and a test laboratory. Both work- TiO NPs. places were equipped with general ventilation and process Occupational airborne exposure to TiO NPs has been ventilation systems such as fume hoods. The amount of extensively studied. Production of TiO NPs, bagging and TiO handled during the two studies differed by three orders 2 2 handling of the NP dry powder, as well as incorporation of of magnitude (from about 5 kg per day during Study A to 738 K. Lovén et al. about 5 g per day during Study B). The amount of graphene Quantification of EC of the graphene nanomaterials was nanomaterial handled was a few grams per day during both conducted according to the NIOSH NMAM 5040 protocol Study A and B. Similar processes were investigated during for thermal-optical analysis (DRI Model 2001 OC/EC both studies. Carbon Analyzer, Atmoslytic Inc., USA) [50]. Temperature steps for the EC fraction were: 680 °C (EC1), 750 °C (EC2), Work tasks and 900 °C (EC3). The method was modified with an extended oxidation time, 150 s instead of 30 s, at the highest Different work tasks were performed during the two studies temperature, 900 °C, in order to achieve complete oxidation and a thoroughly written logbook documented the specific of all carbonaceous nanomaterials [51]. The limit of activities carried out. Table 1 shows these work tasks with detection (LOD) for EC was determined to be 0.06 µg C/ detailed descriptions. In Study A, three workers performed cm (corresponding to 4-h sampling with a carbon airborne the work tasks and in Study B only one. concentration of 0.5 µg/m ). Quantification of Ti was conducted by Particle-Induced Engineering controls and personal protective X-ray Emission (PIXE) analysis [52]. In PIXE, a 2.55 MeV equipment proton beam is focused on the filter specimen. This renders the atoms in a state of high excitation, which causes inner Different types of exposure control techniques and enclo- shell vacancies. The characteristic X-ray emission lines are sures, as well as different types of personal protection caused by the quickly occurring transition to a state of lower equipment (PPE), were used during the different work tasks energy. When protons are used, the cross section for the (Table 1). creation of an inner shell vacancy is very high, thereby the sensitivity is very high; for instance, the LOD for Ti was Sampling strategy <6 ng/cm . The elemental quantification of Cu was performed by Time resolved and filter based measurements of airborne digestion with 1 ml concentrated nitric acid (Nitric Acid, particles were conducted in four different spatial zones, a Trace metal grade, Fisher Chemicals) in an oven (60 °C) for methodology described in detail in Isaxon et al. [49]. 16 h, followed by dilution to 10 ml with Milli-Q water to a The measurement zones included: (1) emission zone (EZ)— stock solution. Analysis was performed by inductively no more than a few centimeters from a potential source, (2) coupled plasma-mass spectrometry (ICP-MS, iCAP Q, personal breathing zone (PBZ)—within a radius of 30 cm Thermo Scientific, Germany). The LOD for Cu was three from a worker’s nose and mouth, (3) background zone (BZ) times the standard deviation of blank filters i.e. <0.01 µg/ —at least 2–3 m away from any potential particle source, sample. All results were blank filter corrected. and (4) supply air (SA)—in the inflowing air from the general ventilation system. Figure 1 shows the schematics Filter sampling for SEM analysis of the facilities in Study A and B including where the dif- ferent work tasks were performed and where the different To be able to morphologically characterize the workplace measurement zones were located. air, regarding engineered NPs and NFs and their aggregates and agglomerates, during down-stream handling processes, Air sampling methods and analyses total dust fraction on filters according to Nilsson et al. [53] and Vaquero et al. [21] were collected. Time-integrated Filter sampling for elemental composition samples were collected at 2.3 l/min (sampling pump Escort ELF, MSA, USA) by open-face sampling on 25-mm Time-integrated emission zone, personal breathing zone, polycarbonate filters (pore size 0.4 µm, SKC Inc., USA). and background zone samples were collected by open-face The filters were analyzed by Scanning Electron Microscopy sampling on 25-mm filters mounted in filter cassettes. (SEM) using a Hitachi SU8010 Cold Field Emission SEM Quartz filters (SKC Inc., USA) were used to collect samples (Hitachi, Japan) with an acceleration voltage of 10 kV. A for elemental carbon (EC) analysis. Filters made of poly- sputtering tool (Q150T ES, Quorum, UK) was used to coat carbonate (SKC Inc., USA) were used for analysis of tita- the sample with 10 nm of platinum:palladium (Pt:Pd, −5 2 nium (Ti), and of cellulose (pore size 0.45 µm, SKC Inc., 80:20). A minimum of 10 random 1.25 × 10 cm areas of USA) for elemental analysis of Cu. Pumps (Escort ELF, the filter were imaged and used for quantifying the MSA, USA) with a flow rate of 2.3 l/min for the poly- number of particles. During Study A, 17 filter samples carbonate filters and 3.0 l/min for the quartz and cellulose were collected and the highest LOD was determined to be −3 filters—checked before and after sampling—were used for 0.47 cm . During Study B, 13 filter samples were collected −3 all sample collection. and the highest LOD was determined to be 1.91 cm .In Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 739 Table 1 Work tasks performed at the two companies, types of engineering controls and types of personal protective equipment (PPE). Work task number Work task Work task description Location/ Engineering controls PPE Study A A1 Preparation of graphene nanoplatelets ink Weighing of graphene nanoplatelet powder, addition of Chemistry laboratory/Fume hood Half-face respirator (A + P3), lab (weighing, mixing). Note that this is not the same liquid and mixing. coat, protective gloves of nitrile, ink printed with in A3! goggles A2—day 1 Preparation of titanium dioxide nanoparticle ink Weighing of titanium dioxide nanoparticle powder, (weighing, mixing) weighting of liquid. Addition of powder by sieving into the liquid and mixing. A2—day 2 Preparation of titanium dioxide nanoparticle ink (weighing, mixing) A3 Screen printing with graphene ink (HDPlas Transfer of ink to the screen printer, start of printing Printing laboratory/process ventilation Clean room lab coat, hair net, Graphene Ink), 10 sheets process and a number of sheets was printed. The printed protective gloves of nitrile sheets were placed in the drying oven. After the printing A4 Screen printing with carbon black ink (C740), process the ink was removed and the screen printer and 10 sheets its equipment was cleaned. A5 Screen printing with carbon black ink (CXT0641), 10 sheets A6—day 1 Screen printing with carbon black ink (C7102), 20 sheets A6—day 2 Screen printing with carbon black ink (C7102), 22 sheets A7 Screen printing with copper ink (CP-PLS-010715- R1A), 10 sheets Study B B1 Preparation of graphene oxide coating (weighing, Weighing of graphene powder, addition of liquid to the Chemistry laboratory/Parts in open fume hood Half-face respirator (A + P3), lab mixing) powder and mixing. Transferring of the liquid to another (weighing, mixing), parts on open bench coat, protective gloves of nitrile, container followed by addition of glass beads and more (filtration steps) goggles mixing. Finally, 2 steps of filtration of the liquid to remove the glass beads. B2 Spraying and curing graphene oxide coating, and Transfer of the liquid coating to a paint sprayer, cleaning Test laboratory/parts in spray booth, parts in cleaning/washing up after of metal plates, paint spraying of metal plates, placing the open fume hood and sink (cleaning and coated plates in a furnace. Cleaning of paint sprayer in washing up) the spray booth followed by washing up mixing equipment in the fume hood and at the sink. B3 Abrasion test with a metal brush on graphene The coated plate was placed in the abrasion testing Test laboratory/none Laboratory coat, protective gloves oxide coating equipment. A brush, rotated with a certain speed, was of nitrile, goggles placed to have contact with the plate. When the surface B4 Abrasion test with a nylon brush on graphene was abraded the test was ended. oxide coating B5 Preparation of titanium dioxide nanofiber coating Weighing of titanium dioxide nanofiber powder, addition Chemistry laboratory/parts in open fume hood Half-face respirator (A + P3), lab (weighing, mixing) of liquid to the powder and mixing. Transferring of the (weighing, mixing), parts on open bench coat, protective gloves of nitrile, liquid to another container followed by addition of glass (filtration steps) goggles beads and more mixing. Finally, 2 steps of filtration of the liquid to remove the glass beads. B6 Spraying and curing titanium dioxide nanofiber Transfer of the liquid coating to a paint sprayer, cleaning Test laboratory/parts in spray booth, parts in coating, and cleaning/washing up after of metal plates, paint spraying of metal plates, placing the open fume hood and sink (cleaning and coated plates in a furnace. Cleaning of paint sprayer in washing up) the spray booth followed by washing up mixing equipment in the fume hood and at the sink. B7 Abrasion test with a metal brush on titanium The coated plate was placed in the abrasion testing Test laboratory/None Laboratory coat, protective gloves dioxide nanofiber coating equipment. A brush, rotated with a certain speed, was of nitrile, goggles placed to have contact with the plate. When the surface B8 Abrasion test with a nylon brush on titanium was abraded the test was ended. dioxide nanofiber coating 740 K. Lovén et al. Fig. 1 Schematics of the facilities. The location of where the different work tasks were performed, and the placement of the different measurement zones, emission zone (EZ), background zone (BZ), and supply air (SA), are shown during a Study A and b Study B. addition to the areas imaged for particle quantification, a measure the black carbon (BC) mass concentration (as a larger part of the surface was investigated with lower proxy for EC) in the background zone, with a time reso- resolution in order to identify any engineered nanomaterial. lution of 1 min. During Study A, the measurements in the emission zone were supplemented with a fast aerosol Direct reading instruments mobility size spectrometer (DMS Model 500 MkII, Cambustion, UK) and a DustTrak (model DRX 8533, TSI Several different direct-reading time resolved instruments Inc., USA). The DMS measured the particle number size were used in the four different measurement zones distribution in the size range 0.005–1 µm, with a time (Table 2). The aerodynamic particle diameter size dis- resolution of 1 s. The DustTrak measured the particle tribution in the range 0.5–20 µm was obtained by two mass concentration in four size fractions: PM ,PM , 1 2.5 Aerodynamic Particle Sizers (APS, model 3321, TSI Inc., respirable and PM , with a time resolution of 1 s. During USA) with a time resolution of 5 s. Two condensation Study B, a portable aethalometer (model AE51, AethLabs, particle counters (CPC, model 3775 and 3010, respec- USA) was used to measure the BC mass concentration in tively, TSI Inc., USA) were used to measure the total the emission zone, with a time resolution of 10 s. number concentration of particles > 0.007 µm, with a time During both studies, two instruments were carried by the resolution of 1 s. One APS and one CPC measured in the workers to measure in the personal breathing zone. A emission zone, while the second APS and CPC were similar second portable aethalometer to measure the BC measuring in the background zone. An aethalometer mass concentration, with a time resolution of 10 s, and an (model AE33, Magee Scientific, USA) was used to aerosol dosimeter (Partector, Naneos, Switzerland) to Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 741 Table 2 Direct reading instruments used for time resolved studies of particle emissions. Personal breathing zone Emission zone Background zone Supply air Study A Partector, portable aethalometer APS, CPC, DMS, DustTrak APS, CPC, Aethalometer P-Trak, DustTrak Study B Partector, portable aethalometer APS, CPC, portable aethalometer APS, CPC, Aethalometer P-Trak, DustTrak measure the lung deposited surface area (LDSA) con- used during Study A, and Fig. 2b shows the sampled centration, with a time resolution of 1 s. material found in the emission zone during handling of For particle measurements in the supply air, a P-Trak graphene nanoplatelet powder (work task A1). Figure 2c (model 8525, TSI Inc., USA) and a DustTrak (model DRX shows the raw graphene oxide material used during Study 8534, TSI Inc., USA) were used to assess the particle B, and Fig. 2d shows the sampled material found in the number concentration (0.02–1 µm) and the particle mass emission zone during handling of graphene oxide powder concentration (0.1–15 µm), respectively, with a time reso- (work task B1). The amounts of graphene nanomaterial in lution of 1 s. the emission zones (26 and 1.9 µg/m for Study A and B, respectively) were quantified as EC with thermal-optical Surface sampling method and analysis analysis. During Study B, EC (most likely from graphene oxide) was also quantified in the personal breathing zone During Study B, tape samples from different surfaces were (1.3 µg/m ). Two of the direct reading instruments (APS collected according to a tape stripping method described by and portable aethalometer) showed clear particle con- Hedmer et al. [54]. Tape samples were collected at the end centration peaks in the emission zone during the 1-min of the workdays from surfaces in the near-field zone of the weighing event of graphene oxide powder (performed exposure source (<1 m). Surface contaminations in the far- twice) during Study B (work task B1), see Fig. 3. The field zone (>1 m), including two offices and a conference coarse particle number size distribution (count median room, were also studied (Table 4). Two field blank tape aerodynamic diameter (CMD) of ~2 µm) of the initial gra- samples were also obtained. The tape samples were pre- phene oxide particle concentration peak (seen in Fig. 3)is pared and analyzed with the same SEM method as the air shown in Fig. 4. From the SEM images (Fig. 2c, d), the size filter samples. of the graphene oxide particles is estimated to be about 10–20 µm. Results Titanium dioxide nanofiber detection and quantification After the completion of tier 1 (contextual information gathering) and tier 2 (basic exposure assessment), we con- As shown in Table 3, TiO NFs, handled at the workplace cluded that tier 3 expert exposure assessments were needed during Study B, could be detected (Fig. 5) both in the at both companies. The tier 3 measurements are the ones emission zone and in the personal breathing zone during described herein. powder handling (work task B5). Figure 5a shows the raw TiO NF material used during Study B, and Fig. 5b, c shows Filter sampling examples of sampled material found in the emission zone and personal breathing zone, respectively. The sampled Table 3 shows the results from the filter samples collected fiber length was assessed to be 1–15 µm. TiO NF con- during work tasks using CB, graphene nanomaterials and centration in the emission zone was 2.2 µg/m , quantified as TiO NPs and NFs. Note that some of the filters have been Ti with PIXE. The APS in the emission zone showed par- sampled for more than one work task (according to ticle concentration peaks during the two-minute weighing of Table 1). During the one work task where Cu was used the TiO NF powder, see Fig. 6. The coarse particle number (A7), no concentration of Cu was detected in either the size distribution of the concentration peaks seen in Fig. 6 personal breathing zone or the emission zone. revealed a CMD of ~0.7 µm (not shown). Graphene detection and quantification Titanium dioxide nanoparticle detection and quantification Table 3 shows that graphene nanomaterials were detected (Fig. 2) on the emission zone filters at both workplaces. Spherical TiO NP agglomerates, handled at the work- Figure 2a shows the raw graphene nanoplatelet material place during Study A, were detected (Fig. 7)on filters 742 K. Lovén et al. Table 3 Results from the filter based measurements during the work tasks using graphene nanoplatelets, graphene oxide, titanium dioxide (TiO ) nanofibers (NFs) and nanoparticles (NPs) and carbon black (CB). Work task number Nanomaterials handled Sampling SEM analysis Elemental carbon Metal conc. 3 3 during the work task time (min) (µg/m ) Ti (µg/m ) Detection of nanomaterial Number conc. −3 (Yes/No)/Type (cm ) Study A (2016) Personal breathing zone aa A1 Graphene nanoplatelets 42 <LOD – A2—day 1 TiO NPs 83 Yes/TiO NPs 25 – 7.5 2 2 A2—day 2 TiO NPs 45 Yes/TiO NPs 1.8 – 2.1 2 2 A3-6—day 1 CB, graphene 106 Yes/CB 5.6 – A6—day 2 CB 39 –– <LOD – Emission zone A1 Graphene nanoplatelets 43 Yes/graphene 26 – nanoplatelets A2—day 1 TiO NPs 83 Yes/TiO NPs 25 – 70 2 2 A2—day 2 TiO NPs 57 Yes/TiO NPs 7.0 – 28 2 2 aa A3 Graphene 13 <LOD – A4 CB 15 Yes/CB <LOD – A5 CB 15 Yes/CB 98 – A6—day 1 CB 31 Yes/CB 8.2 – A6—day 2 CB 37 –– <LOD – Study B (2017) Personal breathing zone B1–4 Graphene oxide 172 No <LOD 1.3 – B5–6 TiO NFs 100 Yes/TiO NFs – <LOD 2 2 B7-8 TiO NFs 60 No <LOD – <LOD Emission zone B1–2 Graphene oxide 207 Yes/graphene oxide 1.9 – B3 Graphene oxide 32 No <LOD <LOD – aa B4 Graphene oxide 111 <LOD – B5 TiO NFs 54 Yes/TiO NFs – 2.2 2 2 B6 TiO NFs 129 No <LOD – <LOD B7 TiO NFs 23 No <LOD – <LOD B8 TiO NFs 123 No <LOD – 0.2 Background zone B1–4 Graphene oxide 480 No <LOD 0.2 B5–6 TiO NFs 202 No <LOD – <LOD B7–8 TiO NFs 173 No <LOD – <LOD –Not sampled. Not possible to determine. from both the emission zone and the personal breathing Surface contamination zone during handling of the powder on two separate days (work task A2—day 1 and A2—day 2). The concentration Workplace surfaces in the chemical laboratory, test reached 70 and 28 µg/m in the emission zone for the first laboratory, offices, and conference room were tape sam- and second day, respectively (quantified with the PIXE), pled in Study B. In total, 15 tape samples were collected, and 7.5 and 2.1 µg/m in the personal breathing zone the and TiO NFs were detected in 20% (n = 3) and graphene first and second day, respectively. Particle concentration oxide in 20% (n = 3) (Table 4). Surface contamination of peaks during weighing and mixing of the TiO NP pow- both TiO NFs and graphene oxide was found on only 2 2 der were detected by APS and DustTrak in the emission one surface, at the sink in the chemical laboratory which zone during both the first and second day, see Fig. 8. was related to work task B1–2and B5–6. A SEM image Similar to the PIXE results, the particle concentrations of the nanomaterial-related surface contamination can be measured with the APS and DustTrak were lower seen in Fig. 9. The length of the TiO NFs was assessed during the second day. The Partector in the personal to be 10–50 µm. TiO NFswerealsodetectedonthe plate breathing zone showed a particle exposure reaching a with TiO NF based coating after the abrasion test with peak lung deposited surface area concentration of 92 µm / the nylon brush. No surface contamination of nanoma- 3 2 3 cm during the first day and of only 9 µm /cm the second terial was found outside the chemical and test day (not shown). laboratories. Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 743 Fig. 2 SEM images of the different graphene raw materials and the materials measured in the emission zones. a The raw material of graphene nanoplatelets used during Study A, b graphene nanoplatelets found on the emission zone filter sampled during work task A1 during Study A, c the raw material of graphene oxide used during Study B, and d graphene oxide found on the emission zone filter sampled during work task B1–2 during Study B. Fig. 3 Coarse particle number concentration (APS, left y- axis) and black carbon concentration (portable aethalometer, µ-Aeth & aethalometer, Aeth, right y- axis) measured in the emission zone (EZ) and the background zone (BZ). The figure shows the measurements during weighing of the graphene oxide powder during Study B (work task B1). Note the different scales on the y-axes. Discussion of different methods, both offline and online, must be used. This study aimed to characterize workplace emissions and exposure of graphene nanoplatelets, graphene oxide, TiO Measurements of airborne graphene nanomaterials NFs, and NPs, as well as CB and Cu during industrial handling. The results showed that powder handling, The results show that graphene can be detected and quan- regardless of nanomaterial, generates the highest (among tified with several different methods. With the offline the investigated processes) particle emissions and expo- method consisting of filter sampling and SEM analysis we sures. It was shown that a range of different methods were, by comparing with the original nanomaterial, able to successfully can be used to selectively detect and quantify identify both graphene nanoplatelets used during Study A, nanomaterials both in the air and on surfaces, and that, to and graphene oxide used during Study B (Fig. 2). Lee et al. be able to make an accurate determination of which [18] and Vaquero et al. [21] have also reported similar nanomaterial that has been emitted when, a combination graphene-like structures in workplace air samples. 744 K. Lovén et al. According to OECD [48], both SEM and TEM are fre- intensive than the SEM/TEM method. In previous work- quently used for determination of nanomaterial present in place studies of graphene, Lee et al. [18] and Vaquero et al. the workplace air. [21] also used EC as an exposure metric for graphene, while In Fig. 4, the peak particle number size distribution exposure data of EC are missing in the studies by Spinazze during handling of graphene oxide powder (seen in Fig. 3) et al. [19] and Boccuni et al. [20]. In the current study, EC showed a CMD of ~2 µm. In the SEM images of both concentrations collected as total dust fractions were detected collected graphene nanoplatelets and graphene oxide, larger both in emission zone and personal breathing zone filter particles were found (10–20 µm). This is not contradictive samples, for example during weighing and mixing of gra- since 2D materials such as graphene may have an aero- phene oxide. Lee et al. [18] collected respirable fractions of dynamic diameter much smaller than its geometrical EC while Vaquero et al. [21] collected total dust fraction, dimensions [55], a fact that indicates its ability to reach the but none of these studies detected any airborne EC con- alveolar region of the lung when inhaled. centrations during manufacturing of graphene nanomaterial. The EC concentration (assumed to originate from gra- We used on-line measurements of equivalent BC (eBC) as phene) was quantified with thermal-optical analysis (OC/ a proxy for EC by using a portable aethalometer. This allowed EC) of filter samples. EC analysis is much less labor for highly time-resolved measurements and identification of exposure peaks for several work tasks. The combination of OC/EC analysis and portable aethalometer for graphene quantification has previously been used by Lee et al. [18]at a graphene nanoplatelets manufacturer. They found the clearest emission peaks in terms of eBC during graphene weighing. We have shown that these methods can be used also at down- stream handling facilities. The portable aethalometer and the APS (both time-resolved instruments) were used together with a thoroughly written logbook. By this approach we could determine that weighing and mixing the dry powder material generated the highest particle emission and exposure levels. The importance of documenting the activities to be able to match the measured concentration profiles have previously been demonstrated by e.g., Ham et al. [44], Hedmer et al. Fig. 4 Aerodynamic particle size distribution measured in the [51], and Isaxon et al. [49]. emission zone (APS EZ) and the background zone (APS BZ) in Both OC/EC analysis and eBC measurements of gra- Study B. The figure shows the size distribution of the initial peak in Fig. 3. phene nanomaterials is subject to cross sensitivity by other Fig. 5 SEM images of the TiO NF raw material and the material measured in the emission zone and the personal breathing zone in Study B. a The raw material of TiO NFs used, b TiO NFs found on the emission zone filter sampled during work task B5, and c TiO NFs found on the personal breathing zone filter sampled during work task B5–6. Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 745 carbonaceous nanomaterials such as CB and CNTs, but also emission zone compared with simultaneous measurements from soot particle sources, such as diesel exhaust, that may in the background zone during Study B, suggesting that be present in workplace air. Both EC and cross sensitivity to background EC or eBC was low. eBC concentrations were substantially elevated in the The eBC mass concentration was reported assuming the standard instrument settings. This may have affected the accuracy of the method as an offset of the results if the ENMs have an altered instrument response, for example by a different mass absorption cross section compared with the standard value that is based on atmospheric soot. This uncertainty can be tolerated when it comes to ENM expo- sures, where the aim often is to link exposure peaks with specific work tasks. In the literature, an eBC/EC ratio of 0.14 was found during simulated powder handling of CNTs [56]. Even though a combination of offline and online meth- ods is recommended, different methods can be used indi- vidually for specific needs, depending on what type of information that is needed. For graphene, the following can be used: (1) SEM analysis for accurate identification of the Fig. 6 Coarse particle number concentration measured in the nanomaterial, (2) EC analysis of filter samples for an emission zone (APS EZ) and the background zone (APS BZ). The accurate assessment of higher carbon concentrations (rela- figure shows the measurements during the 2-min weighing of the TiO NF powder during Study B (work task B5). tively high LOD), (3) direct reading portable aethalometer Fig. 7 SEM images of the TiO NP raw material and the material measured in the emission zone and the personal breathing zone in Study A. a The used raw material of titanium dioxide NPs, b TiO NPs found on the emission zone filter sampled during work task A2—day 1, c TiO NPs found on the emission zone filter sampled during work task A2— day 2, d TiO NPs found on the personal breathing zone filter sampled during work task A2— day 1, and e TiO NPs found on the personal breathing zone filter sampled during work task A2— day 2. 746 K. Lovén et al. Fig. 8 Coarse particle number concentration (APS, left y- axis) and particle mass concentration (DustTrak, right y-axis) measured in the emission zone (EZ) and the background zone (BZ). The figure shows the measurements during weighing and mixing of the TiO2 NP powder during Study A. a Shows work task A2 —day 1 and b shows work task A2—day 2. for highly time-resolved eBC assessment, and (4) direct phagocytosis and accumulation over time finally causing lung reading APS for low-concentration (although less disease [57]. During Study B, TiO NFs were detected both in specific) detection of particle emissions (low LOD). This the emission zone and the personal breathing zone with the means that with access to one of these methods, a pre- offline filter sampling method followed by SEM analysis liminary emission and exposure assessment should be (Fig. 5). They appear to be present in the workplace air as possible to conduct. bundles, and some of the fibers detected by SEM during Study B were at least 5 µm and could therefore constitute a Measurements of airborne titanium dioxide risk if inhaled. Bianchi et al. [58]haverecentlyshown that nanofibers and nanoparticles long (around 10 µm) TiO NFs induce cell cytotoxicity in vitro and inflammation in vivo, while shortened (around 2 NFs are another type of nanomaterials that are of concern, µm) TiO NFs seem to mitigate the toxic effects, even without especially if they are long (>5 µm) and insoluble, due to their macrophages present in the in vitro cultures. SEM-analysis potential to cause adverse health effects, including frustrated can also be used to quantify the particle concentration in the Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 747 Table 4 An overview of the 15 tape-sampled surface locations and the presence of titanium dioxide nanofibers and graphene oxide particles on different surfaces found during Study B. Sampling location Room Related to work Surface characteristics SEM analysis task number Material Assessed indication Detection of TiO Detection of graphene of roughness nanofiber (Yes/No) oxide particles (Yes/No) Work areas Work areas in fume hood Chemistry laboratory B1, B5 Stainless steel Smooth No No Cover on work area B1, B5 Cardboard Smooth No No Sink B1–2, B5–6 Stainless steel Smooth Yes Yes Desk next to abrasion testing machine Test laboratory B3–4, B7–8 Laminate Smooth No Yes Floors Floor in front of fume hood Chemistry laboratory B1, B5 Epoxy coated concrete Smooth No No Floor in front of spray booth Test laboratory B2, B6 Epoxy coated concrete Smooth No No Floor Conference room – Parquet Smooth No No Floor Office 1 – Laminate Smooth No No Door sill Office 2 – Wood Smooth No No Other surfaces Balance in fume hood Chemistry laboratory B1, B5 Plastic Smooth No Yes Cover under the brush in the abrasion Test laboratory B3–4, B7–8 Metal Smooth Yes No testing machine Plate with graphene based coating, B3 Metal with graphene based Smooth No sampled after abrasion test with coating metal brush Plate with graphene based coating, B4 Metal with graphene based Smooth No sampled after abrasion test with coating nylon brush Plate with TiO nanofiber based B7 Metal with TiO nanofiber Smooth No No 2 2 coating, sampled after abrasion test based coating with metal brush Plate with TiO nanofiber based B8 Metal with TiO nanofiber Smooth Yes No 2 2 coating, sampled after abrasion test based coating with nylon brush –Far-field surfaces. Not possible to determine which type of particles that are detected . 748 K. Lovén et al. second day, this procedure had been adjusted so that the TiO NP powder were directly mixed in with the liquid. This clearly shows that with an easy change in the handling procedure, the particle emissions and exposures can be lowered by as much as a factor 10 (for example seen with the Partector where the lung deposited surface area went 2 3 2 3 from 92 µm /cm during the first day to only 9 µm /cm during the second day). The Partector has also previously been shown to be an important tool for personal exposure assessments to improve workplace monitoring [60–62]. Quantification of Ti from TiO nanomaterials by ICP-MS is Fig. 9 SEM image of TiO NFs and graphene oxide. The image 2 shows materials found as surface contamination at the sink in the not possible due to that TiO is a poorly soluble oxide. Laser- chemical laboratory in Study B. ICP-MS could be one alternative [45]and ICP-OES(optical emission spectrometer) [42] another. The current study has highlighted PIXE as a valid and reliable method for titanium air. However, it is a time consuming process, and during quantification, which has also been shown by Relier et al. [63]. Study B it was shown to be difficult to distinguish the dif- ferent fibers due to agglomeration and therefore to count Surface contamination them. The PIXE analysis was a valid alternative, even though it gives the concentration as a different metric (mass con- Surface contamination of TiO NFs and graphene oxide was centration, µg/m ). Detection and quantification can also be for the first time studied at down-stream handling processes. conducted with direct reading measurements (APS could be Surface contaminations of both TiO NFs and graphene used to determine what specific handling process generated oxide were found on one surface (sink) in the near-field the particles (Fig. 6); weighing the dry powder material), but zone in the chemical laboratory in Study B. The detected to confirm fiber emissions, time-integrated filter sampling surface contamination can probably be related to washing of followed by SEM is necessary. equipment after preparations of both types of coatings. The General NP dry powder handling has previously been surface contamination of TiO NFs and graphene oxide shown to constitute a potential source for worker exposure could probably be resuspended into the workplace air, by e.g., Huang et al. [42]and Curwin andBertke[43]. Lee which if so would cause a risk of secondary inhalation et al. [59] found substantial total mass concentrations exposure. The percentage of TiO NF surface contamination during powder collection of TiO NPs in the pigment in the collected tape samples were lower compared with industry (500–5000 µg/m ), however, no elemental ana- what was previously found in a similar study [54]. This lysis of Ti was carried out. In the current study, this was indicates that there was no widespread nanomaterial con- confirmed, were elevated concentrations of TiO NPs tamination at workplace B. An interesting finding was that were found in both the emission zone and the personal TiO NFs could be detected on a tape sample from one of breathing zone (with SEM, PIXE, APS, DustTrak, and the abrasion-tested metal plates with TiO NF based coat- Partector) during weighing and mixing of dry TiO NP ing. During the abrasion tests, dust containing manufactured powder. Interestingly, the mass concentrations of Ti nanomaterials is generated and could be emitted to the (measured with PIXE) were about a factor 10 lower in the workplace air, especially since this process was openly personal breathing zone than in the emission zone. This performed at the workplace without any engineering reduction factor was most likely achieved by the use of controls. engineering controls in the emission zone (fume hood) and such a factor could be used for other assessments of Proposed occupational exposure limits personal exposure where only measurements in the emission zone can be conducted or vice versa. Mihalache et al. [64] have reviewed the proposed occupa- Another observation was that the particle concentrations tional exposure limits (OELs) for a number of measured with PIXE, APS, DustTrak, and Partector were all different nanomaterials. For TiO NPs, a range between 17 lower during the second day compared with the first day of and 2000 µg/m have been proposed, with probably the weighing and mixing of the TiO NP powder. As shown in most established one by NIOSH of 300 µg/m as an 8 h Fig. 8, the procedure of mixing the powder with the liquid average airborne exposure [65]. The highest TiO con- had been adjusted from the first to the second day. During centration in the personal breathing zone during the current the first day, the TiO NP powder were weighed first and study was 7.5 µg/m , corresponding to an 8 h average of then the liquid was added to form the printing ink. The 1.3 µg/m , well below the suggested OEL. For TiO NFs, 2 Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 749 there is no specific OEL, but the BSI [66] suggested OEL analysis is also a good option, especially when aiming to for fibrous nanomaterials of 0.01 fibers/cm could be a compare the result with existing OEL, if one can ensure suitable OEL to consult. However, in the current study it that no other sources of that chemical compound are was found to be difficult to distinguish the different fibers present in the workplace. For carbon-based materials by SEM due to agglomeration and therefore challenging to (such as graphene) filter sampling with following EC count them. The detected fibers were, furthermore, only analysis can be used. In the current study, open-face filter found during the low resolution scanning-to-identify sampling was used, but it could have been possible to investigation and none were found during the quantifica- instead use a respirable cyclone to get the respirable EC tion imaging, rendering the fiber number concentration concentrations. Then, it would be possible to compare below the LOD. Tsang et al. [67] demonstrated that a measured EC concentrations with the recommended probabilistic approach can be used in the risk assessment of exposure limit of respirable EC set to 1 µg/m as an 8 h exposure scenarios involving production of TiO NPs. time weighted average concentration for CNTs and CNFs Thus, they could identify one out of seven exposure sce- [72]. For TiO nanomaterials, filter sampling with fol- narios with statistically significant level of risk. lowing PIXE analysis would be the recommended che- Neither for graphene nanomaterials are there any legal mical analysis to perform. binding OELs yet, but e.g., Lee et al. [68] have based on data from a subchronical inhalation toxicity study calculated a recommended OEL for graphene nanomaterial to be 18 Conclusions µg/m . Also, health based guidance value for occupational inhalation exposure to graphene nanoplatelets was esti- Down-stream industrial handling of graphene nanomaterials mated to 212 µg/m by Spinazze et al. [69]. However, so far and TiO NFs were, for the first time, investigated for there are only few in vivo inhalation exposure studies particle emissions and exposures into the workplace air. We reported in the literature. Thus, according to Pelin et al. [3] showed that weighing the dry powder material generated OEL for graphene cannot be determined based on the particle emissions, even though the exposure levels were available data. The highest EC concentration in the personal low compared with proposed OELs. Surface contaminations breathing zone, in the current study, was 5.6 µg/m , corre- of both TiO NFs and graphene were found on the sink in sponding to an 8 h average of 1.2 µg/m , well below the the chemical laboratory at workplace B. A range of different suggested OELs. sampling and aerosol monitoring methods was used and evaluated. For a fast and reliable workplace emission and Recommendations for occupational hygienists exposure assessment of graphene and TiO nanomaterials, a combination of different methods, both offline and online, As a general recommendation for occupational hygienists, must be used to ensure the detected emissions contain the the portable aethalometer could be one option for a small specific nanomaterials. If a first exposure assessment should and easy-to-use instrument for personal exposure mon- be performed (according to tier 2 of the OECD guidelines), itoring of carbon-based nanomaterials. However, from the access to only one or a few of these methods is enough. current study, data from the personal sampling portable When a tier 3 exposure assessment is needed, a combination aethalometer could not be used due to too much noise, of multiple of the mentioned methods should be used most likely arising from the short sampling time. This together with additional direct reading instruments such as could possibly be addressed by longer averaging times CPC and APS. [70] as well as noise reduction treatment during the data Acknowledgements The authors would like to acknowledge the fol- post-processing [71]. As a stationary instrument in the lowing persons: Louise Gren for helping out with the on-site work- emission zone, the portable aethalometer worked well and place measurements during Study B. Linus Ludvigsson for helping out can be an alternative to bulkier, more expensive and with the SEM analyses. Jan Pallon for conducting the PIXE analysis. advanced instruments. For non-carbon-based nanomater- Thomas Lundh for conducting the metal analysis with ICP-MS. The project was funded by AFA insurance (Dnr 130122) and NanoLund ials, a suitable personal exposure instrument option would (p38-2013), and was carried out within the framework of NanoLund at instead be a lung deposited surface area instrument. Lund University. However, when it is critical with an accurate determina- tion of which nanomaterial that is emitted, SEM analysis 1 2 1 NanoLund Karin Lovén , Sara M. Franzén , Christina Isaxon , Maria 2 1 1 4 must be performed as a complement to the direct reading E. Messing , Anders Gudmundsson , Joakim Pagels , Maria Hedmer measurements. Open-face filter sampling was used in the current study to morphologically characterize the real Funding The project was funded by AFA insurance (Dnr 130122) and emission and exposure situations of all released NPs and NanoLund (p38-2013), and was carried out within the framework of NanoLund at Lund University. NFs and their aggregates and agglomerates. Chemical 750 K. Lovén et al. Compliance with ethical standards 14. Hu XG, Zhou QX. Health and ecosystem risks of graphene. Chem Rev. 2013;113:3815–35. 15. Guo XQ, Mei N. Assessment of the toxic potential of graphene Conflict of interest The authors declare no conflict of interest. The family nanomaterials. J Food Drug Anal. 2014;22:105–15. project sponsors have only contributed financially, and have not par- 16. Fadeel B, Bussy C, Merino S, Vazquez E, Flahaut E, Mouchet F, ticipated in preparing the research material, writing, reviewing, or et al. Safety assessment of graphene-based materials: focus on approving the submitted paper. human health and the environment. ACS Nano. 2018;12:10582–620. Publisher’s note Springer Nature remains neutral with regard to 17. Basinas I, Jimenez AS, Galea KS, van Tongeren M, Hurley F. A jurisdictional claims in published maps and institutional affiliations. Systematic review of the routes and forms of exposure to engi- neered nanomaterials. Ann Work Expo Heal. 2018;62:639–62. Open Access This article is licensed under a Creative Commons 18. Lee JH, Han JH, Kim JH, Kim B, Bello D, Kim JK, et al. Attribution 4.0 International License, which permits use, sharing, Exposure monitoring of graphene nanoplatelets manufacturing adaptation, distribution and reproduction in any medium or format, as workplaces. Inhal Toxicol. 2016;28:281–91. long as you give appropriate credit to the original author(s) and the 19. Spinazze A, Cattaneo A, Campagnolo D, Bollati V, Bertazzi PA, source, provide a link to the Creative Commons license, and indicate if Cavallo DM. Engineered nanomaterials exposure in the produc- changes were made. The images or other third party material in this tion of graphene. Aerosol Sci Technol. 2016;50:812–21. article are included in the article’s Creative Commons license, unless 20. Boccuni F, Ferrante R, Tombolini F, Lega D, Antonini A, Alvino indicated otherwise in a credit line to the material. If material is not A, et al. Workers’ exposure to nano-objects with different included in the article’s Creative Commons license and your intended dimensionalities in R&D laboratories: measurement strategy and use is not permitted by statutory regulation or exceeds the permitted field studies. Int J Mol Sci. 2018;19:349. use, you will need to obtain permission directly from the copyright 21. Vaquero C, Wendelbo R, Egizabal A, Gutierrez-Cañas C, López holder. To view a copy of this license, visit http://creativecommons. de Ipiña J. Exposure to graphene in a pilot production plant. J org/licenses/by/4.0/. Phys Conf Series. 2019;1323. 22. Khataee A, Moradkhannejhad L, Heydari V, Vahid B, Joo SW. Self-cleaning acrylic water-based white paint modified with different types of TiO2 nanoparticles. Pigm Resin Technol. 2016;45:24–9. References 23. Sharma S, Sharma RK, Gaur K, Torres JFC, Loza-Rosas SA, Torres A, et al. Fueling a hot debate on the application of TiO2 1. Castro Neto AH, Guinea F, Peres NMR, Novoselov KS, Geim nanoparticles in sunscreen. Materials. 2019;12:2317. AK. The electronic properties of graphene. Rev Mod Phys. 24. Dastan D, Gosavi SW, Chaure NB. Studies on electrical properties 2009;81:109–62. of hybrid polymeric gate dielectrics for field effect transistors. 2. Balandin AA. Thermal properties of graphene and nanostructured Macromol Symp. 2015;347:81–6. carbon materials. Nat Mater. 2011;10:569–81. 25. Victorious A, Clifford A, Saha S, Zhitomirsky I, Soleymani L. 3. Pelin M, Sosa S, Prato M, Tubaro A. Occupational exposure to Integrating TiO2 nanoparticles within a catecholic polymeric graphene based nanomaterials: risk assessment. Nanoscale. network enhances the photoelectrochemical response of bio- 2018;10:15894–903. sensors. J Phys Chem C. 2019;123:16186–93. 4. Chen ZH, Lin YM, Rooks MJ, Avouris P. Graphene nano-ribbon 26. Ou G, Li ZW, Li DK, Cheng L, Liu Z, Wu H. Photothermal electronics. Physica E. 2007;40:228–32. therapy by using titanium oxide nanoparticles. Nano Res. 5. Schedin F, Geim AK, Morozov SV, Hill EW, Blake P, Katsnelson 2016;9:1236–43. MI, et al. Detection of individual gas molecules adsorbed on 27. Pi JK, Yang HC, Wan LS, Wu J, Xu ZK. Polypropylene micro- graphene. Nat Mater. 2007;6:652–5. filtration membranes modified with TiO2 nanoparticles for surface 6. Loh KP, Bao QL, Eda G, Chhowalla M. Graphene oxide as a wettability and antifouling property. J Membr Sci. 2016;500:8–15. chemically tunable platform for optical applications. Nat Chem. 28. Tang YJ, Hu XL, Zhang XQ, Guo DL, Zhang JH, Kong FO. 2010;2:1015–24. Chitosan/titanium dioxide nanocomposite coatings: rheological 7. Rafiee J, Mi X, Gullapalli H, Thomas AV, Yavari F, Shi YF, et al. behavior and surface application to cellulosic paper. Carbohyd Wetting transparency of graphene. Nat Mater. 2012;11:217–22. Polym. 2016;151:752–9. 8. Ding R, Li WH, Wang X, Gui TJ, Li BJ, Han P, et al. A 29. Nalbandian MJ, Greenstein KE, Shuai DM, Zhang ML, Choa YH, brief review of corrosion protective films and coatings based on Parkin GF, et al. Tailored synthesis of photoactive TiO2 nanofi- graphene and graphene oxide. J Alloy Compd. 2018; bers and Au/TiO2 nanofiber composites: structure and reactivity 764:1039–55. optimization for water treatment applications. Environ Sci Tech- 9. Berman D, Erdemir A, Sumant AV. Few layer graphene to reduce nol. 2015;49:1654–63. wear and friction on sliding steel surfaces. Carbon. 30. Safajou H, Khojasteh H, Salavati-Niasari M, Mortazavi-Derazkola 2013;54:454–9. S. Enhanced photocatalytic degradation of dyes over graphene/Pd/ 10. Mutyala KC, Wu YMA, Erdemir A, Sumant AV. Graphene— TiO2 nanocomposites: TiO2 nanowires versus TiO2 nano- MoS2 ensembles to reduce friction and wear in DLC-Steel con- particles. J Colloid Interface Sci. 2017;498:423–32. tacts. Carbon. 2019;146:524–7. 31. Wang JX, Fan YB. Lung injury induced by TiO2 nanoparticles 11. Jastrzebska AM, Kurtycz P, Olszyna AR. Recent advances in depends on their structural features: size, shape, crystal phases, graphene family materials toxicity investigations. J Nanopart Res. and surface coating. Int J Mol Sci. 2014;15:22258–78. 2012;14:1320. 32. Grassian VH, O’Shaughnessy PT, Adamcakova-Dodd A, Petti- 12. Sanchez VC, Jachak A, Hurt RH, Kane AB. Biological interac- bone JM, Thorne PS. Inhalation exposure study of titanium tions of graphene-family nanomaterials: an interdisciplinary dioxide nanoparticles with a primary particle size of 2 to 5 nm. review. Chem Res Toxicol. 2012;25:15–34. Environ Health Perspect. 2007;115:397–402. 13. Arvidsson R, Molander S, Sanden BA. Review of potential 33. Pujalte I, Dieme D, Haddad S, Serventi AM, Bouchard M. Tox- environmental and health risks of the nanomaterial graphene. icokinetics of titanium dioxide (TiO2) nanoparticles after inhala- Hum Ecol Risk Assess. 2013;19:873–87. tion in rats. Toxicol Lett. 2017;265:77–85. Emissions and exposures of graphene nanomaterials, titanium dioxide nanofibers, and nanoparticles. . . 751 34. Lauvas AJ, Skovmand A, Poulsen MS, Kyjovska ZO, Roursgaard 51. Hedmer M, Isaxon C, Nilsson PT, Ludvigsson L, Messing ME, M, Goericke-Pesch S, et al. Airway exposure to TiO2 nano- Genberg J, et al. Exposure and emission measurements during particles and quartz and effects on sperm counts and testosterone production, purification, and functionalization of Arc-discharge- levels in male mice. Reprod Toxicol. 2019;90:134–40. produced multi-walled carbon nanotubes. Ann Occup Hyg. 35. Wang JX, Liu Y, Jiao F, Lao F, Li W, Gu YQ, et al. Time- 2014;58:355–79. dependent translocation and potential impairment on central ner- 52. Johansson SAE. PIXE – A Novel Technique for Elemental vous system by intranasally instilled TiO2 nanoparticles. Tox- Analysis. Endeavour. 1989;13:48–53. icology. 2008;254:82–90. 53. Nilsson PT, Isaxon C, Eriksson AC, Messing ME, Ludvigsson L, 36. Falck GCM, Lindberg HK, Suhonen S, Vippola M, Vanhala E, Rissler J, et al. Nano-objects emitted during maintenance of Catalan J, et al. Genotoxic effects of nanosized and fine TiO2. common particle generators: direct chemical characterization with Hum Exp Toxicol. 2009;28:339–52. aerosol mass spectrometry and implications for risk assessments. J 37. Di Bucchianico S, Cappellini F, Le Bihanic F, Zhang YN, Dreij Nanopart Res. 2013;15:1–16. K, Karlsson HL. Genotoxicity of TiO2 nanoparticles assessed by 54. Hedmer M, Ludvigsson L, Isaxon C, Nilsson PT, Skaug V, mini-gel comet assay and micronucleus scoring with flow cyto- Bohgard M, et al. Detection of multi-walled carbon nanotubes and metry. Mutagenesis. 2017;32:127–37. carbon nanodiscs on workplace surfaces at a small-scale producer. 38. Pelclova D, Zdimal V, Kacer P, Fenclova Z, Vlckova S, Komarc Ann Occup Hyg. 2015;59:836–52. M, et al. Leukotrienes in exhaled breath condensate and fractional 55. Schinwald A, Murphy FA, Prina-Mello A, Poland CA, Byrne F, exhaled nitric oxide in workers exposed to TiO2 nanoparticles. J Movia D, et al. The threshold length for fiber-induced acute Breath Res. 2016;10:036004. pleural inflammation: shedding light on the early events in 39. Hurbankova M, Cerna S, Kovacikova Z, Wimmerova S, Hraskova asbestos-induced mesothelioma. Toxicol Sci. 2012;128:461–70. D, Marcisiakova J, et al. Effect of TiO2 nanofibres on selected 56. Hashimoto N, Ogura I, Kotake M, Kishimoto A, Honda K. bronchoalveolar parameters in acute and subacute phase— Evaluating the capabilities of portable black carbon monitors and experimental study. Cent Eur J Publ Heal. 2013;21:165–70. photometers for measuring airborne carbon nanotubes. J Nanopart 40. Medina-Reyes EI, Delgado-Buenrostro NL, Deciga-Alcaraz A, Res. 2013;15:1–17. Freyre-Fonseca V, Flores-Flores JO, Hernandez-Pando R, et al. 57. Donaldson K, Murphy FA, Duffin R, Poland CA. Asbestos, carbon Titanium dioxide nanofibers induce angiogenic markers and nanotubes and the pleural mesothelium: a review of the hypothesis genomic instability in lung cells leading to a highly dediffer- regarding the role of long fibre retention in the parietal pleura, entiated and fibrotic tumor formation in a xenograft model. inflammation and mesothelioma. Part Fibre Toxicol. 2010;7:5. Environ Sci-Nano. 2019;6:286–304. 58. Bianchi MG, Campagnolo L, Allegri M, Ortelli S, Blosi M, Chiu 41. Allegri M, Bianchi MG, Chiu M, Varet J, Costa AL, Ortelli S, M, et al. Length-dependent toxicity of TiO2 nanofibers: mitigation et al. Shape-related toxicity of titanium dioxide nanofibres. PLos via shortening. Nanotoxicology. 2019;14:433–52. ONE. 2016;11:e0151365. 59. Lee JH, Kwon M, Ji JH, Kang CS, Ahn KH, Han JH, et al. 42. Huang CH, Tai CY, Huang CY, Tsai CJ, Chen CW, Chang CP, Exposure assessment of workplaces manufacturing nanosized et al. Measurements of respirable dust and nanoparticle con- TiO2 and silver. Inhal Toxicol. 2011;23:226–36. centrations in a titanium dioxide pigment production factory. J 60. Geiss O, Bianchi I, Barrero-Moreno J. Lung-deposited surface Environ Sci Heal A. 2010;45:1227–33. area concentration measurements in selected occupational and 43. Curwin B, Bertke S. Exposure characterization of metal oxide non-occupational environments. J Aerosol Sci. 2016;96:24–37. nanoparticles in the workplace. J Occup Environ Hyg. 61. Todea AM, Beckmann S, Kaminski H, Bard D, Bau S, Clavaguera 2011;8:580–7. S, et al. Inter-comparison of personal monitors for nanoparticles 44. Ham S, Yoon C, Lee E, Lee K, Park D, Chung E, et al. Task- exposure at workplaces and in the environment. Sci Total based exposure assessment of nanoparticles in the workplace. J Environ. 2017;605:929–45. Nanopart Res. 2012;14:1–17. 62. Iavicoli I, Fontana L, Pingue P, Todea AM, Asbach C. Assess- 45. Xu HD, Zhao L, Chen ZJ, Zhou JW, Tang SC, Kong FL, et al. ment of occupational exposure to engineered nanomaterials in Exposure assessment of workplace manufacturing titanium diox- research laboratories using personal monitors. Sci Total Environ. ide particles. J Nanopart Res. 2016;18:1–13. 2018;627:689–702. 46. Debia M, Bakhiyi B, Ostiguy C, Verbeek JH, Brouwer DH, 63. Relier C, Dubreuil M, Garcia OL, Cordelli E, Mejia J, Eleuteri P, Murashov V. A systematic review of reported exposure to engi- et al. Study of TiO2 P25 nanoparticles genotoxicity on lung, blood, neered nanomaterials. Ann Occup Hyg. 2016;60:916–35. and liver cells in lung overload andnon-overload conditionsafter 47. OECD. Harmonized tiered approach to measure and assess the repeated respiratory exposure in rats. Toxicol Sci. 2017;156:527–37. potential exposure to airborne emissions of engineered nano- 64. Mihalache R, Verbeek J, Graczyk H, Murashov V, van Broe- objects and their agglomerates and aggregates at workplaces. khuizen P. Occupational exposure limits for manufactured nano- Series on the safety of manufactured nanomaterials organisation materials, a systematic review. Nanotoxicology. 2017;11:7–19. for economic co-operation and development. J Nanopart Res: 65. NIOSH. US National Institute for Occupational Safety and Health. OECD; 2015. p. 55. US: NIOSH TiO2 OEL 2011.pdf; 2011. 48. OECD. Strategies, techniques and sampling protocols for deter- 66. BSI. Nanotechnologies—Part 2: guide to safe handling and dis- mining the concentrations of manufactured nanomaterials in air at posal of manufactured nanomaterials. PD 6699-2:2007 UK: the workplace. Series on the safety of manufactured nanomaterials British Standards Institution; 2007. organisation for economic co-operation and development. J 67. Tsang MP, Hristozov D, Zabeo A, Koivisto AJ, Jensen ACO, Nanopart Res: OECD; 2017. p. 82. Jensen KA, et al. Probabilistic risk assessment of emerging 49. Isaxon C, Lovén K, Ludvigsson L, Sivakumar S, Gudmundsson materials: case study of titanium dioxide nanoparticles. Nano- A, Messing ME, et al. Workplace emissions and exposures during toxicology. 2017;11:558–68. semiconductor nanowire production, post-production, and main- 68. Lee YS, Sung JH, Song KS, Kim JK, Choi BS, Yu IJ, et al. tenande work. Ann Work Expo Heal. 2020;64:38–54. Derivation of occupational exposure limits for multi-walled car- 50. Birch ME, Cary RA. Elemental carbon-based method for mon- bon nanotubes and graphene using subchronic inhalation toxicity itoring occupational exposures to particulate diesel exhaust. data and a multi-path particle dosimetry model. Toxicol Res. Aerosol Sci Technol. 1996;25:221–41. 2019;8:580–6. 752 K. Lovén et al. 69. Spinazze A, Cattaneo A, Borghi F, Del Buono L, Campagnolo D, 71. Hagler GSW, Yelverton TLB, Vedantham R, Hansen ADA, Rovelli S, et al. Probabilistic approach for the risk assessment of Turner JR. Post-processing method to reduce noise while preser- nanomaterials: a case study for graphene nanoplatelets. Int J Hyg ving high time resolution in aethalometer real-time black carbon Envir Heal. 2019;222:76–83. data. Aerosol Air Qual Res. 2011;11:539–46. 70. AethLabs. microAeth AE51 Operating Manual. US: AethLabs; 72. NIOSH. Occupational exposure to carbon nanotubes and nanofi- 2016. bers. Current Intell Bull. 2013;65.

Journal

Journal of Exposure Science & Environmental EpidemiologySpringer Journals

Published: Jun 16, 2020

Keywords: Occupational exposure; Electron microscopy; Thermal-optical carbon analysis; Direct-reading instruments; PIXE; Aerosol

There are no references for this article.