Worker’s Personal Exposure to PM 0.1 and PM 4 Titanium Dioxide Nanomaterials during Packaging

For the appropriate hygienic management of workplaces where workers handled nano-TiO 2 products, the status of the personal exposure of workers to respirable dust (PM 4 ) and nanoparticles (PM 0.1 ) was investigated. Using a cyclone sampler for PM 4 and a personal sampler for PM 0.1 , PM 4 and PM 0.1 exposure levels were evaluated to discuss them in relation to worker’s duties. The number of particles of 0.01–10 µ m was also monitored online in order to examine the short-term fluctuation in the concentration and size distribution of particles. The 8h-time-weighted average (TWA) and 95% upper limit for respirable dust exposure were below the occupational exposure limit specified by the Japan Society for Occupational Health and the recommended exposure limits for TWA by NIOSH. The action level was exceeded during the filling of a flexible container bag. More than 70% of particles in the breathing zone was coarse agglomerates of > 1 µ m, while it may be influenced by powder properties and the handling process as well as the management of local ventilation. The maximum PM 0.1 concentration (31.3 µ g m –3 ) occurred in a powder filling booth without air ventilation. The operation of a gasoline powered forklift temporarily increased the concentration of ultrafine particles. Most of TiO 2 powder was suspended as micron-order agglomerates in the breathing zone. However, since PM 0.1 exposure was much larger than those in outdoor environment particularly under insufficient cares to aerosolized powder and air ventilation, PM 0.1 exposure should also be monitored.


INTRODUCTION
According to the development of nanomaterial manufacturing technologies (Ding et al., 2017), existing procedures for occupational exposure to airborne particulates have been requested to be updated by the nanomaterials industries (Leidel et al., 1977;British Occupational Hygiene Society, 1996;BSI, 1996;MHLW, 2010;AIHA, 2015;JSOH, 2015).The reason for this is because inhaled aerosol nanoparticles can penetrate into the deeper interstitial regions of the lungs where they are retained for long periods of time.This suggests that particles could be retained in pulmonary tissue, leading to lung injury, abnormal heart rate variability and/or enhanced risks of developing cancer (Shi et al., 2013;Ichihara et al., 2016).Because the effects have been found to be the result of exposure to the lungs, pulmonary inflammation, which has an adverse effect, even at the lowest concentrations can be regarded as the endpoint of the assessment (Nakanishi et al., 2011).The exposure to nano-sized TiO2 particles, e.g., the most widely used nanomaterial and the main focus of this study, has been investigated so far from various standpoints.In a recent review, Shakeel et al. (2016) stated that TiO2 nanoparticles (NPs) through pulmonary exposure can be translocated to the brain or the systemic circulation, depending on where they reach other organs, such as the kidney, liver, etc. Liao et al. (2009) found a particle size-related health risk for anatase nano-TiO2 (10-30 nm primary size) that increased the probability of cytotoxicity and inflammatory responses for TiO2 production workers.Pogribna et al. (2020) reported on the effects of exposure to nano-TiO2 on DNA methylation in multiple cell types in animal models.Lee et al. (2020) reported on higher urinary 8-hydroxy-2'-deoxyguanosin (8-OHdG) concentrations in salespersons both of cosmetics and clothing who had a higher co-exposure index of ZnO and TiO2 NPs.
The Japan Ministry of Health, Labour and Welfare (MHLW) initiated a discussion of this in 2004 on industrial nanomaterials with regard to an evaluation of their toxicity, a protocol for measuring the extent of exposure to workers and effective methods for controlling this exposure (Kanno et al., 2005).These issues are of interest in terms of preventing damage to the health of workers and the MHLW produced a notification "On an immediate action to the prevention of exposure to nanomaterials" in 2008 (Japan Ministry of Health, Labour and Welfare, 2008a).The MHLW also organized a committee of experts on nanomaterials to discuss the risk assessment and issued "On the preventative action on the prevention of exposure to nanomaterials" (Japan Ministry of Health, Labour and Welfare, 2008b).The survey on the exposure of nano-titanium oxide (TiO2) to workers in factories that produce more than 500 kg y -1 of nano-TiO2 was conducted from 2012 to obtain information for the current and actual situations regarding respirable exposure to workers.Specifically, the data used to analyze the behavior and characteristics of nanomaterials and the current methods that are used to control this type of exposure.From environmental surveys of manufacturing processes such as surface treatment through filtering, drying, micronizing, and bagging by the New Energy and Industrial Technology Development Organization (NEDO) for the risk assessment of nanomaterials, or, TiO2, fullerene and carbon nanotubes, particle concentrations have been found to increase in these types of works.Additional testing showed that the emitted particles mainly consist of aggregate/agglomerate particles in sizes ranging from several hundred nm to several µm (Nakanishi et al., 2011;Ding et al., 2017;Kylafis et al., 2019).
It is essential to determine both the concentration and composition of fine and NPs in the breathing zone of a worker with regards to aerodynamic particle size, which affects the regional deposition of particles inhaled into the human respiratory system.However, for nano-TiO2, many studies reported on the exposure to NPs only by agglomerates with sizes above 100 nm of (e.g., van Broekhuizen et al., 2011;Buonanno et al., 2012) or examined exposure to NPs not in the breathing zone (Liao et al., 2009;Curwin and Bertke, 2011;Koivisto et al., 2011), in which the number concentration of nano-TiO2 particles (10 nm of primary size) was monitored using a scanning mobility particle sizer and an optical particle counter (Curwin and Bertke, 2011).Hence, very few case studies of breathing zone sampling of workers based on the actual measurement of worker's exposure to NPs have been reported for nano-TiO2 although Ichihara et al. (2016) evaluated the personal exposure to nano-TiO2 using a Siutas cascade impactor for particles < 250 nm.The reason for this is that only a few portable tools are available for measuring the degree of exposure to nanoparticles (< 100 nm) in the breathing zone of a worker and only a few commercial products capable of measuring nanoparticles are currently available (Furuuchi et al., 2010a;Tsai et al., 2011Tsai et al., , 2012;;Young et al., 2013;Thongyen et al., 2015;Asbach et al., 2017).
In this study, under the commission of the research project entitled "Investigation for Risk Assessment of Hazardous Chemical Substances in the Work Environment" since the Fiscal Year 2013 supported by the MHLW, a titanium oxide (TiO2) manufacturing factory was selected as a site to examine the status of personal exposure of nanomaterials to workers and in workplaces of TiO2 plants for two different types of products between Fiscal Year 2013 and 2014, namely, an evaluation of the exposure of workers to nanoparticles on a particle mass basis (PM0.1).For the evaluation of nanoparticle exposure, a personal air sampler for nanoparticles was developed by the authors (Furuuchi et al., 2010a;Thongyen et al., 2015), which can be used to collect particles smaller than 0.1 µm in the breathing zone of a worker, and was used in conjunction with a cyclone sampler for collecting respirable dust (PM4).The time-weighted average (TWA) and short term exposure limit (STEL) as well as exposure to PM0.1 were discussed in relation to worker's duties such as, as powder filling, bag mouth closing and bag carrying and also to the characteristics of the workplace environment, including ventilation of local areas and spaces.The respirable exposure (RPE) for TiO2 was also discussed based on the mass of titanium (Ti) analyzed.The number concentration of particles in the 10 nm-10 µm size range was monitored online using a scanning mobility particle sizer and an optical particle counter at selected locations including various locations in the workplace in order to evaluate fluctuations in the concentration and size distribution of particles in relation to workers' activities.

Sampling and Monitoring Schedule
Schedules for measuring the exposure of each worker and the sampling and monitoring at each location are summarized in Table 1 along with information on sampling locations and the PSM-O2-6 Ave.Temp., Ave.RH, Ave.WD, Ave.WS: Average value of air temperature, air relative humidity, wind direction, and wind speed at Yokkaichi meteorological observatory (Japan Meteorological Agency, 2014).RPE: respirable personal exposure sampling using a respirable sampler (RPS) (SKC GS-3).NPE: nanoparticle personal exposure sampling using a personal nanoparticle sampler (PNS).NSS: nanoparticle spot sampling using an ambient nanoparticle sampler (ANS) or a personal nanoparticle sampler (PNS).PSM: particle spot monitoring of number concentration using a scanning mobility particle sizer (SMPS) and an optical particle sizer (OPS).* a and b are notations to designate two different monitoring spots in Workplace-A (1F).corresponding sample IDs.The measurements were conducted on Dec. 14, 2012, Jan. 11, 2013and Jan. 10, 2014.The working period for each worker is classified here as AM and PM respectively corresponding to 8:00-12:00 and 13:00-16:00.The last figure (1-6) of the sample ID was added to designate the date and working period.The meteorological conditions for the corresponding days at the nearest meteorological observatory (Yokkaichi) of the Japan Meteorological Agency are listed in Table 1.

Target Processes, Characteristics of Workplaces and Assigned Duties for Workers
In the identified factory, two different types of TiO2, or, rutile and anatase nano-powders, were manufactured.Rutile nano-powder is commonly used in cosmetics, paint and as an additive for toners by a general chloride process that involves neutralization and hydrolysis of a titanium salt.Anatase is used as denitration catalysts and photo-catalysts by a sulfate process by the sulfation and hydrolysis of ilmenite and slag.Related information on each TiO2 product is summarized in Table 2, along with the roles of the workers assigned to these tasks.In both processes, TiO2 powder was manufactured through the above chemical processes in a sealed system and was stored in a hopper after surface treatment and drying.The most likely process where workers would be exposed to TiO2 powder would be during the packing of the powder into bags for shipping and related processes such as the transport of bags.

Workplace-A for Rutile TiO2 Nanopowder Production
Workplace-A was located in a building where rutile TiO2 is manufactured.Fig. 1 shows a plane view of Workplace-A, which consisted of three areas that are classified for different purposes in a 5 story plant building with a void space connecting all of the floors: 1) a small booth located on the ground floor where worker "A1" was assigned to fill paper bags with nano-TiO2 powder, 2) an area also located on the ground floor and next to the filling booth for weighing and shipping preparation of TiO2 packed bags where Worker-A2 was assigned to transport filled bags of powder along a belt-conveyer and stack them on a pallet to be handled by a forklift after weighing and checking for possible metal impurities, and 3) an area located on floor 5 just above the previous two-areas on the ground floor where Worker-A3 was assigned to feed an intermediate TiO2 product into a hopper connected to the final stage of TiO2 nano powder production that was located below the hopper.
Fig. 2(a) shows an illustration of Worker-A1 working in the packing booth screened by transparent vinyl sheets (W2.6 × L2.6 × H2.2 m 3 ) into which clean air was fed from the booth   3, where the velocity was measured by means of a hot-wire based anemometer (KANOMAX, ANEMOMASTER MODEL 6113) at the inlet of a ventilation hood.Worker-A2 received a filled bag from Worker-A1 outside the packing booth then closed the bag by hemming the bag mouth.Worker-A2 transported the powder filled bags on a conveyer belt between the packing booth and the bag stacking site.After checking the weight and metal content automatically using a balance and a metal detector installed at the middle of the conveyer, Worker-A2 then stacked the checked bags on a pallet, which was moved by a forklift powered by a gasoline engine ( 2013) or an electric battery motor (2014) for shipping.
Worker-A3 handled an intermediate product of TiO2 (500 kg) in a container (1 m 3 ), which was lifted through the void space of the building then mounted to a hopper of a mill that was connected to the final TiO2 nanopowder process site, which was located on the floor.Worker-A3 operated an overhead travelling crane and adjusted the configuration of the container and hopper.In order to avoid the dispersion of powder during the filling procedure, a gap between the container outlet and the hopper inlet was covered by a fabric sheet and the air suction flow to the hopper was used.
Both dispersed and aerosolized TiO2 powder were observed during the powder filling, bag hemming and bag stacking procedures while nothing could be visually observed during the intermediate product filling.Although not related to the TiO2 powder, the emission of fine particulate matter from the gasoline engine of the forklift could not be neglected, as described below.
Clean filtered air was supplied from a ventilator at a ceiling of the filling booth and was removed via a local exhaust ventilation system.The pressure inside the booth was kept slightly positive by adjusting the clean air flow rate.The windows and entries of the plant building were closed during the working period except when a forklift was being used to ship the products stacked on a pallet.Fans were installed in the roof of the plant building (5 story) and operated constantly.An electric fan was also installed beside a hopper inlet at the 5 th floor and was operated at a ventilation rate of 233 m 3 min -1 only during the intermediate product filling operation.The entries were opened rather frequently in 2013 but were mostly closed in 2014.All windows were closed during both years of this investigation.When the entries were closed (2014), the pressure in the workplace was kept slightly below the ambient pressure and draft air flow was not very serious.The presence of an air supply by draft air flow, therefore, may play an important role here.The plant floor was cleaned with a vacuum cleaner so that resuspended powder may have made some contributions, although it was not quantitatively evaluated in the present study.

Workplace-B for Producing Anatase TiO2 Nanopowder
Workplace-B was located in a building that manufactured anatase TiO2.Figs. 3 and 4 respectively show a map of the floor of Workplace-B and an illustration of Worker-B1 working in a packing booth.Workplace-B was located on the ground floor of the plant that had a wide entry on one side which was fully opened to the outside during the powder filling procedure.The packing  booth, which was screened by transparent vinyl curtains (2014), had a local ventilation system that operated at the suction velocity shown in Table 2, where the velocity was measured at the inlet of the ventilator hood.As briefly summarized in Table 1, Worker-B1 put 600 kg of the TiO2 nanopowder into a flexible intermediate bulk container (FIBC).Since the FIBC was inflated by the powder filling, Worker-B1 adjusted the bag inflation by manipulating an air pressure adjuster at a bag opening.The FIBC was closed at the top by Worker-B2 using a drawstring then moved outside the plant building by a forklift with a diesel engine.Worker-B3 processed several powder samples a day at a sampling port of the powder filling facility for checking the quality of the product.
The TiO2 powder could be visually observed to be aerosolized from the bag opening when Worker-B1 adjusted the FIBC inflation.Emissions from a diesel forklift may also have made some contribution to the number of fine particles in the workplace.Since the plant building had an opening to the outside, Workplace-B should be well ventilated by air from outside.

Evaluation of Personal Exposure to Respirable TiO2 Particles (RPE)
Respirable personal exposure (RPE) samples were collected using a commercial PM4 sampler (SKC, GS-3 cyclone connected to a SKC AirCheck2000 pump) attached to a worker's breathing zone as show in Fig. 5.A mixed cellulose ester filter (MCEF) (Japan Millipore, 0.8 µm, ϕ 37 mm) was used downstream of the cyclone that was operated at a flow rate of 2.75 L min -1 .In order to evaluate the time weighed average (TWA) and the short time exposure limit (STEL), samples obtained at both long (> 1 h) and short (15 min) periods were collected as a set of RPE samples, where RPE samplers attached to workers and the samplers temporarily installed near workers were used for long and short samplings, respectively.RPE samplers that were attached to workers were also retained during short breaks but were detached during lunch time since they were in locations where they were not exposed to TiO2.

Evaluation of Personal Exposure to Nanoparticles (NPE)
For the evaluation of the personal exposure of workers to aerosol nanoparticles, a personal sampler for nanoparticles (PNS), which was originally developed by Furuuchi et al. (2010a) and Thongyen et al. (2015), was used both in the 2013 and 2014 studies.The PNS consists of a PM10/4 pre-cut impactor and a PM0.45/0.1 inertial filter stage and can be operated at a total pressure drop of 5-6 kPa for 5 L min -1 rate of air sampling using a commercial battery pump (e.g., Hario Sci., HSP-5000).The surface of an impaction plate of the 1-st stage of a pre-cut impactor (PM10) was uniformly coated with a thin layer (approximately 0.2 mm) of silicon-grease (Dow Corning, 03253589) while a ϕ 10 mm glass fiber filter (Pallflex, T60A20) was attached on an impaction plate of the 2-nd stage (PM4).A pre-cut inertial filter for PM0.45 particles consisting of webbed SUS fibers (Nippon Seisen Co. Ltd., felt type, SUS-304, df = 9.8 µm) packed in a ϕ 4.75 mm circular nozzle (5.5 mm length) through a resin cartridge was used upstream from the PM0.1 inertial filter.The main inertial filter for obtaining PM0.1 on a backup filter consists of 5-layered mesh TEM grids (Glider, G600HSS) sandwiched by 6-spacers with circular holes (ϕ 1.9 mm, spacer thickness t = 30 µm) held in a nozzle through an aluminum cartridge (ϕ 3 mm, 9 mm nozzle length).The PM0.1 was collected on a ϕ 47 mm filter (Pallflex, quartz fiber filter (QFF), 2500QAT-UP).The PNS was attached to a worker as shown in Fig. 5.

Evaluation of Particle Mass Concentration by Nanoparticle Spot Sampling (NSS)
An ambient sampler for nano-particles (ANS) developed by Furuuchi et al. (2010b) was used for the nanoparticle spot sampling at fixed sites in Workplace-A and outdoors.This sampler can collect PM10/2.5/1/0.5/0.1 at a flow rate of 40 L min -1 at a total pressure drop of 10-15 kPa.Sampling locations are shown in Figs. 1 and 3, or, NSS-A1, -A2, and -A3 in Workplace-A (Fig. 1) and NSS-B1 in Workplace-B (Fig. 3).The background mass concentration was evaluated both for particles and TiO2.However, a set of PM4 cyclones and NSS were just installed in 2014, while the PM4 cyclone was installed alone in 2012 and 2013.These changes made a difference between the 2012/2013 and 2014 sampling conditions and are limitations associated with this study.The background mass concentration sampling was at an outdoor monitoring site, located beside an administrative office building that was located 10 m away from the main roads in the factory.These are Fig. 6.Instruments for nanoparticle spot sampling (NSS-A1) and the particle spot monitoring (PSM-A1) installed outside the powder filling booth with sampling tubes connected to inside the booth.designated as RPE-O and NSS-O respectively for PM4 cyclone and NS where "O" denotes the outdoor site.The same filter sets, or, QFFs were used along with the same protocol for NS and PNS while mixed cellulose ester filters were used for the cyclones.The height of sampling inlets and ports of air samplers and monitoring instruments described below were set at a distance of 1.5 m from the ground by using tripods as shown in Fig. 6 except for the case of NS in the powder filling booth (the height was about 30 cm from the booth floor), which was necessary due to space limitations in the booth.

Monitoring of Particle Number Concentration
The particle number concentration was continuously monitored at 2-3 minute intervals using a scanning mobility particle sizer (SMPS) (TSI 3910 Nanoscan, detectable size range of 10-420 nm by a scanning period of 1 min.)and an optical particle counter (OPC) (TSI 3330, detectable size range of 0.3-10 µm) at selected locations in the workplaces: respectively designated as PSM-A1, A2 and A3 in Workplace-A, PSM-B1 in Workplace-B and PSM-O in Outdoors.

Evaluation of TiO2 Mass in Samples
Since the personal sampler used for nanoparticle collection had a limitation of pressure drop through a filter, it was used only to evaluate the "respirable dust" concentration, and this represents a limitation regarding evaluating nanoparticles in this research.Normally, the NIOSH 7300, 7302, and 7304 protocols should be used for the mass of TiO2 particles.However, TiO2 in a collected particulate sample was analyzed following an original protocol.TiO2 extracted from a cellulose ester filter was heated in a solvent mixture of nitric, hydrofluoric, and the sulphuric acids on a hotplate at 120°C for 20 min.The extract was then heated at 240°C for 60 min to dissolve the silica and other coating materials that control the hydrophilic properties of the products.Titanium (Ti) was analyzed by graphite acids furnace atomic absorption spectrometry (GFAAS) (HITACHI Z-5010) at 364.3 nm and a pyro-coated graphite atomization tube A (HITACHI) under charring at 1000°C and a final atomizing temperature of 2800°C.The titanium recovery rate was checked for solutions with various water-repellents on TiO2 for various types of solutions, and the recovery was confirmed to be over 95% in every case so that the mass of TiO2 evaluated by the protocol was successfully used in the following discussion.The lower limit of quantitation was estimated to be 0.001 mg m -3 , assuming an air volume of 960 L sampled at a flow rate of 2 L min -1 for 480 min.The limit was far lower than 1/10 th level of the OEL value of 0.3 mg m -3 .

SMPS and OPS for particle spot monitoring (PSM)
Ambient nano-particle sampler (ANS) for nano-particle spot sampling (NSS) Booth Tube connected into the booth

Statistical Indices of TiO2 RPE
The present sampling period for total dust was as described in the NIOSH Occupational Exposure Sampling Strategy Manual (NIOSH, 1977) so that the corresponding results were valid for a period of 6 out of 8 h and the sampled portion should cover 70% to 80% of the full period.The 8-hour average for respirable exposure to TiO2 (8h-TWA) was then calculated by a following equation: 8h-TWA = (CpiTpi)/8h (1) where Cpi denotes the collected TiO2 respirable exposure concentration during a sampling period and Tpi represents the measured sampling period in hours for the TiO2 sampling.The 95% upper limit (95% UL) was also calculated using a following equation defining the upper 95 th percentile of the log-normal distribution: log(95%UL) = log(GM) + 1.65 log(GSD) (2) where GM and GSD are the geometric mean and geometric standard deviation of the focused RPEs, respectively.The condition of 95%UL < the occupational exposure limit (OEL, 300 µg m -3 ) was used to define "the control measure" corresponding to an acceptable exposure in accordance with the concept of exposure risk defined by the NIOSH Occupational Exposure Sampling Strategy Manual (NIOSH, 1977;AIHA, 2015;Hashimoto et al., 2018).The log-normality for the collected data was confirmed by using Kolmogorov-Smirnov test (Massey, 1951).

Respirable TiO2 Mass Concentrations
The mass concentration of TiO2 estimated from an analysis of the titanium content in respirable dust is summarized in Table 4.The time-weighted average (TWA) for Worker-B1 (powder filling to the FIBC, PM) and Worker-B2 (bag closing and transfer of the FIBC, AM) were found to represent the largest and the second-largest concentrations respectively, with values of 668 and 349 µg m -3 .These values exceeded the occupational exposure limit (OEL) of the Japan Society for Occupational Health (JSOH), 300 µg m -3 , which is the same as the recommended exposure limit of time-weighted average (REL-TWA) by NIOSH.The OEL for the TiO2 nano-particles was not discussed, since it has not been defined so far.However, the 8 hour time-weighted average (8h-TWA) for Worker-B1 and -B2 in 2012 were, respectively, 247 and 149 µg m -3 , each of which cleared the OEL and REL-TWA but exceeded the action level (AL, 150 µg m -3 ) defined as 50% of the OEL for Worker-B1, where the AL is generally defined as 50% of the permissible-exposure level (PEL) (NIOSH, 1975).The lower exposure to Worker-B1 and -B2 in 2014 may be attributed to an improvement in the powder filling process both that reduced the powder dispersion by air leakage from FIBC during the powder filling operation.Since 2014, the packing booth has been protected by transparent vinyl curtains so that it had a local ventilation system that was operated at the suction velocity that was measured at the inlet of the ventilator hood.Therefore, when Worker-B1 added 600 kg of the TiO2 nano-powder into a flexible intermediate bulk container (FIBC), and Worker-B2 then moved it outside the plant building, the ventilation system could reduce the respirable TiO2 mass concentration in this plant.
Statistical information on the RPE in Workplace-A and -B is summarized in Table 5.The lognormality was confirmed for Workplace-A (n = 5, p = 0.496 ≥ 0.1) and Workplace-(A + B) (n = 11, p = 0.523 ≥ 0.1) although it was slightly worse for Workplace-B (n = 6, p = 0.096 < 0.1) because of the larger RPE level of B1.The 95% UL was, therefore, evaluated separately for all cases in Workplace-A and -B, where the values for Workplace-B should be regarded as advisory values.The 95% UL was evaluated separately for Workplace-A and -B as well as for all data in AM and PM.The 95% UL of 8h-TWA was 188 µg m -3 in Workplace-A, which was less than the OEL of JSOH and REL-TWA by NIOSH while the RPE in AM and 8h-TWA as well as the maximum STEL and 95% UL exceeded the AL value.In Workplace-B, the RPE (TWA) exceeded OEL both in AM (349 µg m -3 ) and PM (668 µg m -3 ) while the RPE (8h-TWA) (247µg m -3 ) and the 95% UL (278 µg m -3 ) were below the OEL.Statistical values for all data (A + B) were also calculated as a representative RPE of the  factory regarding the powder filling as the worst similar exposure group (SEG) (NIOSH, 1977;AIHA, 2015;Hashimoto et al., 2018) since there were other cases of less exposure in manufacturing processes in the study plants as closed processes of nano-TiO2 production.Both 8h-TWA and 95% UL were also below the OEL for this case.Although the management of task by task could be another choice in some cases, it may provide other issues as the management of OEL for multi tasks by a single worker and should be discussed in a future study.
Since as described above, the RPE to TiO2 nano powder was below that for 8h-TWA and the 95% upper limit while it exceeded the AL in some cases, suitable counter measures for reducing exposure during the powder filling process from an engineering point of view, such as protectors for respirable dust, the management of the work period should be needed.

Mass Concentrations of Size Fractionated Particles Exposed to Workers and those in the Workplace
The particle mass concentration of different size fractions, or, total suspended (TSP), respirable (PM4) and nano (PM0.1)particles, is summarized in Table 6, where the PM4 mass for NSS was interpolated from the mass concentration in the vicinity because of the absence of a PM4 stage in NSS.The particle mass concentration for different size fractions is shown in Figs.7(a)-7(c) and In time range from 8:00 am to 12:00 am.PM: In time range from 13:00 am to 16:00 pm.TWA: Time-weighted average for 8 hours.STEL: Short term exposure limit (time-weighted average concentration measured over a limited sampling period (usually 15 minutes unless otherwise noted)).Table 6.Estimated and measured mass concentrations of particulate matter for total suspended (TSP), respirable (PM4) and nano (PM0.1)from nano-particle personal exposure (NPE) and spot sampling (NSS) at Workplace-A, -B and outdoors (O2).was calculated using averaged values for 2013 and 2014 in cases where the values were available for both years.

Work-place
The cumulative particle size distribution on a mass basis is shown in Figs.8(a) and 8(b) respectively for values obtained by air samplers (sample ID of NPE and NSS) and by SMPS (< 400 nm) and OPS (> 400 nm) (sample ID of PSM).The distributions shown in Fig. 8(b) are averages for the monitoring period and converted from the number basis by assuming a spherical particle with a uniform density and light scattering characteristics.
Based on the above results, the conditions and worker behavior in each workplace are discussed below:

Workplace-A (1F)
As seen from Fig. 7(a) for Workplace-A, all of the RPE values (NPE-A1, A2 and A3) as well as TSP were not only larger than the corresponding particle mass concentrations at the outdoor sites (NSS-O2-56, PSM-O2-5 and 6 and PSM-O1-4) but also larger than that at a peripheral site in    Workplace-A (NSS-A2b-56b).The RPE values were similar to the PM4 measured at the site in the powder filling booth (NSS-A1) while a coarser fraction (> 4 µm) contributed more to the exposure to workers.This suggests that the powder filling and bag handling result in coarse agglomerates being dispersed in the workers' breathing zone.This is also clear from the size distribution data shown in Fig. 8(a), indicating that more than 70% of the particles was agglomerates larger than 1 µm in spite of the fact that the average primary size of the target TiO2 product was less than 100 nm (See Table 1).This tendency in the worker's breathing zone was more significant than those at spot monitoring sites and ambient conditions (See Fig. 8(a)).Since the powder concentration in the booth was controlled by the ventilation fan, larger amounts of coarser agglomerates may actually be dispersed during the powder filling operation.These results were different from data reported in a previous study (Curwin and Bertke, 2011) that characterized the exposure to various metal oxide nanoparticles including nano-TiO2 in production and handling material sites, in which they found the predominant particle size to be between 0.1 and 1 µm by using air samplers and other aerosol instruments installed at fixed locations in areas close to potential sources.Although different operations for different materials cannot be directly compared, such differences might be due to powder properties (sizes and the morphology of primary particles, aggregates and agglomerates, and surface treatment etc.) and the handling process being used as well as the management of local ventilation.In the outdoor environment, there are various contributions from other sources especially anthropogenic ones in urban and industrial areas so that the particle size distribution could be influenced more by fine particles, as shown in Figs.8(a) (NSS-O2-56) and 8(b) (PMS-O1, PMS-O2).The larger outdoor PM0.1 concentration (NSS-O2-56) may be explained, therefore, by a large contribution from local emission sources such as heavy vehicle traffic through the main roads in the factory, which were located near the site while the estimated PM0.1 (PSM-O1-4) was small and comparable to that for the TiO2 concentration (RBE-O1-3, Table 4) at the outside a plant building of Workplace-A, which is located further from main roads.As a reference, the PM0.1 mass concentration was typically in the range of 2-5 µg m -3 in residential areas in Japan according to a previous report by the authors (Hata et al., 2013).
The exposure to PM0.1 by Worker-A1 in the packing booth and a corresponding PM0.1/PM4 mass ratio (0.7-0.8%) were smaller than that for another worker as well as than the concentration in outdoor (NSS-O2) locations and in the booth (NSS-A1) sites.There are two possible reasons for these findings: 1) Filtered clean air was supplied from the ceiling ventilator in the packing booth, 2) Most of the particles that were dispersed from the powder filling operation may have been coarse agglomerates, which can be easily and quickly removed by the local ventilation system before dispersing the fraction of fine particles by fragmentation.Regarding the PM0.1 concentration at the site in the booth (NSS-A1), exposure to workers was much larger, which might be attributed partially to the presence of re-suspended fine particles from huge agglomerates, which were observed to be spread over the booth floor.These results suggest that the design of ventilation systems and the management of powder on the floor are important issues that need to be addressed.
Figs. 9(a) and 9(b) show time fluctuations for particle number concentration and the geometric mean size monitored at the site inside the powder filling booth, respectively on Jan 11, 2013 and Jan 10, 2014.As seen from Fig. 9(a), spikes in the number concentration were observed for both coarse (0.3-4 µm) and ultrafine (0.01-0.1 µm) fractions at 8:12, 9:35 and 10:09.These spikes corresponded to the operation of a forklift (gasoline powered vehicle) by Worker-2 at 8:10, 9:35 and 10:06, or, concentration spikes always appeared just after the operation.Although the powder filling booth was screened with plastic curtains, penetrating particles may also have been detected.The geometric mean size of particles in the range of OPC (0.3-4 µm), which dominate the particle mass, was nearly stable so that the mass of emitted particles caused by the forklift may not be so important.However, there was an important contribution to a fine fraction measured by SMPS, which dominated the particle number.Since this type of emission of fine particles should be avoided from the point of view of contamination by nanoparticles, in 2014, the gasoline forklift was changed to an electric one, leading to no spikes in the particle concentration related to the forklift operation (Fig. 9(b)).

Workplace-A (5F)
The exposure to Worker-A3 (RPE-A3, PM4 of NPE-A3-3) was much larger than those to other workers in Workplace-A (1F) (Fig. 7 (85%) emitted from intermediate products probably during the period with the product was fed into the hopper of the mill.Since this result is not consistent with the visual observations of the activities of Worker-3 and the management of the previously described ventilation facility, accidental exposure mainly by a coarse fraction of > 1 µm (See Fig. 8(a)) dispersed from granular intermediate products before the milling process is possible.The PM0.1 exposure (PM0.1 of NPE-A3-3) of a moderate value (-10 µg m -3 ) was similar to that for Worker-A2 (NPE-A2-4 as a reference) and less than the PM0.1 concentration in the powder filling booth (NSS A1-34).It is interesting to note that the PM4 concentration at the fixed monitoring site on 5F (NSS-A3-6, PSM-A3-3) was larger than that on the ground floor site (NSS-A1-34, PSM-A2-4).Although influence of the degree of door opening by persons entering the building was not clear, ventilation air flow from the bottom to the top of the building by ceiling fans and an associated vertical temperature distribution, which increased with the elevation along a utility void through the plant building might also be influencing factors.

Workplace-B
As shown in Table 6 and Fig. 7(c), PM0.1 evaluated for Worker-B1 (NPE-B1-5) was very similar to that at the monitoring site (NSS-B1-5).The PM4 at the worker's breathing zone, however, was almost twice that at the monitoring site.These particle concentrations (PM4 and PM0.1) in Workplace-B were smaller than those at the outdoor site (NSS-B1-5) because of heavy vehicle traffic around the plant building of Workplace-B had much less influence on the data.The large concentration of PM4 supports the conclusion that the aerosolized TiO2 powder from a FIBC bag opening was observed while Worker-B1 adjusted the FIBC inflation, thus leading to a considerable level of exposure to TiO2, as shown in Tables 4 and 5.The dispersion of ultrafine particles could be controlled to a moderate concentration by properly managing the local ventilation at the powder filling spot of FIBC (Fig. 4) but also by appropriate ventilation of the open space near Workplace-B(< -10 µg m -3 ).Diesel soot emission from a forklift may also have some influence.However, even if the diesel engine made some contribution, its influence was still less than that from the street traffic at the outdoor site (NSS-B1-5).

CONCLUSIONS
The 8h-TWA and 95% upper limit of estimated TiO2 concentration of respirable dust exposure were below the occupational exposure limit (OEL) specified by the Japan Society for Occupational Health (JSOH) (300 µg m -3 ) and the recommended exposure limits of time-weighted average (REL-TWA) by NIOSH (300 µg m -3 ) both in Workplace-A (filling and carrying of 10 kg paper bags or handling of a middle product) and -B (filling and carrying of 600 kg flexible containers) while the action level (AL, 150 µg m -3 ) was exceeded in Workplace-B during the powder filling.However, these data were not sufficient to conclude that the worker's long-term exposure is less than the OEL.Suitable counter measures for decreasing the extent of exposure during the powder filling process from an engineering point of view, would include protectors for respirable dust, appropriate management of the working period, and these should be considered.The powder filling and bag handling, including adjusting the bag opening and bag transport would result in the dispersion of coarse agglomerates, more than 70% of which was larger than 1 µm, especially in the workers' breathing zone.Such results may be influenced by powder properties and the handling process as well as the management of local ventilation.The maximum PM0.1 concentration (31.3 µg m -3 ) was found in a powder filling booth that had no air ventilation while it was decreased to below an outdoor level in the case of a ventilated booth.The operation of a gasoline forklift temporarily increased the concentration of particles particularly in the ultrafine range (0.01-0.1 µm), but these disappeared when an electric-powered forklift was used.
Most of TiO2 powder was suspended as micron-order agglomerates in the breathing zone.However, since PM0.1 exposure was much larger than those in outdoor environment particularly under insufficient cares to aerosolized powder and air ventilation, PM0.1 exposure should also be monitored.An increase in particle concentration with elevation along a utility void from the bottom to the top of the plant building, which may be related to air ventilation and vertical temperature distribution along the void, should be investigated as a remaining issue.

Fig. 2 .
Fig. 2. Schematics of (a) powder filling by Worker-A1 and (b) air flow by local exhaust ventilation.

Fig. 3 .
Fig. 3.A floor map of the Workplace-B.

Fig. 4 .
Fig. 4. Schematic of powder filling by Worker-B1 and air flow by local exhaust ventilation.
246 min average including a time for preparation (~30 min) in Workplace-A and a short break in a worker's office.*2 : Assuming no exposure to TiO2 outside Workplace-A.*3 : 180 min average including a temporal duty in Workplace-A (~30 min) and a short break in a worker's office.*4 : Outdoor (background).AM: In time range from 8:00 am to 12:00 am.PM: In time range from 13:00 am to 16:00 pm.

Fig 8 .
Fig 8. Cumulative size distributions of particles on a mass basis: (a) exposure samples (NPE) and spot samples (NSS), (b) calculated from the number base size distributions at spot monitoring sites (PSM).

Fig. 9 .
Fig. 9. Time fluctuations for particle number concentration and the geometric mean size of particles monitored at the site inside the powder filling booth (PSM-A1-1): (a) Jan 11, 2013 and (b) Jan 10, 2014.

Table 1 .
Schedule of personal exposure evaluation and monitoring of the workplace environment.

Table 5 .
Summary of statistical values of the workers' respirable personal exposure (RPE) in Workplace-A and -B.