DEPARTMENT OF HEALTH & HUMAN SERVICES

Transcription

1 DEPARTMENT OF HEALTH & HUMAN SERVICES Public Health Service Centers for Disease Control and Prevention (CDC) National Institute for Occupational Safety and Health (NIOSH) November 26, 2009 Brendan McKenney Director of Operations QuantumSphere, Inc Tech Center Drive Santa Ana, California Subject: Final Report of the NIOSH Nanotechnology Field Research Team Study at QuantumSphere, Inc. (July 2009). Dear Mr. McKenney: Thank you for partnering with NIOSH and providing our Nanotechnology Field Team with the opportunity to collaborate with you to observe and characterize the work practices and processes at your nano-scale metal oxide production facility. Our July 13-16, 2009, site visit to measure the effectiveness of the new flange attached to your local exhaust ventilation (LEV) system was highly successful. Contained in this final report are the results of the study. If you have any questions about the results of the study, please contact me via at: or by phone at (513) Sincerely, cc: PA Schulte CL Geraci L Hodson C Crawford Mark M. Methner, Ph.D., CIH Field Research Team Leader NIOSH Nanotechnology Research Center 4676 Columbia Parkway (MS R-11) Cincinnati, Ohio mmm5@cdc.gov

2 Nanotechnology Research Center National Institute for Occupational Safety and Health Centers for Disease Control and Prevention U.S. Department of Health and Human Services Preface The National Institute for Occupational Safety and Health (NIOSH) is the federal agency that conducts research and makes recommendations for preventing work-related injuries, illnesses, and deaths. The NIOSH Nanotechnology Research Center coordinates the Institute s laboratory, field, and information dissemination activities on the development of tools, practices, and recommendations for the guidance document Approaches to Safe Nanotechnology. ( A key input to the development of that document is field research studies. The NIOSH nanotechnology field research team has the objective of characterizing processes where nanomaterials are produced or used. To do this, the field team: Evaluates the entire material flow of a process and identifies points of potential material emission that can result in worker exposure. Uses an array of instruments and conventional air sampling methods to characterize exposures. Evaluates engineering controls and their effectiveness in reducing emissions and exposures. Evaluates work practices used during the production or use of nanomaterials. Evaluates the use of Personal Protective Equipment in use, if any, including respiratory protection. This report summarizes the effectiveness evaluation of a newly fabricated flange attached to a Local Exhaust Ventilation (LEV) system engineering control at QuantumSphere Inc.

3 Background In 2002, QuantumSphere, Inc. (QSI), was founded in Santa Ana, California, and is engaged primarily in the production of high quality engineered nano-scale materials (ENM s) consisting of metal catalytic materials such as copper, manganese, iron, silver, nickel and cobalt. These ENM s are produced inside stainless steel reactors via a process known as gas-phase condensation. The ENM s are used in the production of energy efficient batteries, fuel cells and photovoltaic devices, to name a few. Currently, the process used by QSI for creating the ENM s is in the research and process development phase, with a long term goal of large-scale commercialization of both the process and equipment. The 7,500 square foot facility consists of a production area, two laboratories and an office area. The production area houses six gas-phase condensation reactors that produce approximately one kilogram of ENM per reactor per day. Two production technicians operate the reactors, harvest material, and perform reactor cleanout and maintenance duties as needed. Each technician works an 8 hour shift, 5 days per week. In January 2007, The National Institute for Occupational Safety and Health (NIOSH) received a request from the management of QSI to be a volunteer participant in a field research study to evaluate process-specific emissions during the production of ENM s (metal and alloy spheres of approximately nanometers [nm] in diameter). An initial assessment of work practices, existing engineering controls, personal protective equipment (PPE), and potential sources of nano-scale emissions was conducted in February, 2007, with an interim report of the results issued to QSI in May, According to the results obtained during the initial assessment, it was determined that, during specific work practices and procedures (reactor cleanout), ENM s were released to the production area atmosphere and, in turn, could result in potential worker exposure. This finding prompted a request from QSI for NIOSH to provide consultative guidance on an engineering control and procedures that could be implemented to reduce/eliminate such releases. In June 2007, QSI purchased a portable fume extractor typically used in the welding industry, to control ENM releases during reactor cleanout procedures. In July 2007, NIOSH returned to QSI and used real-time, direct-reading instrumentation and filter-based air sampling to determine the effectiveness of the implemented control technology. Based on the results of the July 2007 study, it was determined that the local exhaust ventilation (LEV) system was very effective in reducing ENM emissions during the reactor cleanout process. However, the NIOSH project officer felt that fabricating a semi-circular flange that fit between the upper and lower half of the reactor would provide even better particle capture. The NIOSH project officer designed a flange that could be attached to the existing duct of the LEV. Fabrication of the flange (which measures 18 inches by 32 inches) was conducted by the QSI shop foreman, and NIOSH returned to QSI in July 2009 to conduct an effectiveness study on the new LEV system. This report describes the findings of the July 2009 site visit and evaluation study. Survey Overview During the initial assessment, it was determined that the reactor cleanout process emitted a substantial amount of ENM s. To control such an emission, NIOSH suggested using a portable, High Efficiency Particulate Air (HEPA) filtered LEV system commonly used during welding processes to capture fumes (Figure 1). The unit selected for use by QSI

4 consisted of a 6 inch, flexible, flanged duct attached to a HEPA filtered air handler equipped with a carbon pre-filter. This unit is designed to create a maximum exhaust flow rate of 1,000 cubic feet per minute (cfm), which was confirmed using a thermoanemometer. The existing 10.5 inch square flange was removed and the newly fabricated semi-circular aluminum flange was installed and used for the duration of the study (Figures 2,3). In addition to the use of the new flange LEV system, personal protective equipment (PPE) remained identical to that used during the initial assessment (full-body Tyvek suit, nitrile gloves, wrist-to-elbow cotton arm covers, and a 3M L-122 full-face, positive pressure airline respirator). On July 13, 2009, the NIOSH nanotechnology field research team representative arrived at the facility and held an initial meeting with the company management and production staff. During the meeting, the NIOSH representative discussed the proposed effort to evaluate the effectiveness of the newly-fabricated flange for the LEV system and also observed a working demonstration of the new LEV system. Methods: To evaluate the effectiveness of the new LEV system in controlling the release of ENM s, at source (AS) and personal breathing zone (PBZ) air samples were collected with and without the new LEV system for the duration of the reactor cleanout process (Figures 4,5). All AS samples were positioned at the rim of the opening of the reactor (see Figure 5) and adjacent to the inlet of the new flanged LEV. Additionally, a general area air sample was collected for the duration of the work shift each day (in the center of the reactor area and also in the shop/fabrication area, which is located approximately 75 feet away) to serve as an indicator of migration of ENM released during reactor cleanout operations. Filter-based samples were collected using Leland Legacy (SKC Inc., Eighty Four, PA) pumps at a sampling rate of 7.0 liters per minute (lpm). Pumps were calibrated before and after each day of sampling. Air samples to determine mass concentration of ENM s were collected on 37-millimeter (mm) diameter; 0.8 micrometer (µm) pore size, open-face mixed cellulose ester (MCE) membrane filters and analyzed according to NIOSH Manual of Analytical Methods (NMAM) Method A duplicate set of air samples was collected alongside the mass-based air samples and analyzed using a modification of NIOSH Method 7402 [transmission electron microscopy (TEM) with energy dispersive spectroscopy (EDS)]. TEM with EDS allows the microscopist the ability to identify particles in the nanometer size range, morphology of the particles (size, shape, degree of agglomeration), and elemental composition. Filter-based sampling times ranged from 20 minutes to 360 minutes (volume of air sampled ranged from 140 liters to 2,520 liters). The shorter duration samples (20-52 minutes) were collected during reactor cleanout operations and were dependent on the time necessary to complete the cleanout process. The longer duration samples (>290 minutes) were collected as general area samples and matched the duration of the normal work shift. In addition to the filter-based air sampling, two direct-reading, real-time instruments were used to measure particle number concentration and relative size, with and without the new LEV system, during the reactor cleanout process. To maintain consistency with the

5 filter-based air sampling effort, the sampling inlet of each direct-reading instrument was positioned adjacent to the filter-based air sampling media. The first instrument used was a TSI model 3007 (TSI, Inc, Shoreview, MN) handheld condensation particle counter (CPC). The CPC operates by drawing air through a sizeselective inlet, passing it through a heated saturator filled with isopropyl alcohol, and then cooling the air stream via a condenser chamber. In the condenser, the alcohol vapor condenses on the particles which serve as nuclei and grow each particle to approximately 10 micrometers (µm) in diameter. These larger particles then pass through an optical detector where they are counted. The CPC unit measures particles in the size range of 10 to 1000 nm. The data output is expressed as total number of particles per cubic centimeter of sampled air with an upper dynamic range limit of approximately 100,000 particles per cubic centimeter of air (p/cc). The value of this instrument for evaluating ENM emissions is its ability to detect particles in the 10 to 100 nm range, even though it will respond to the presence of larger particles up to 1,000 nm. A second direct-reading, real-time instrument was used to determine the concentration of particulates based on optical counting principles using laser light scattering (HHPC-6, ART Instruments, Grants Pass, Oregon). This instrument can measure the total number of particles per liter of air across 6 specific size cutpoints: 300 nm; 500 nm; 1,000 nm; 3,000 nm; 5,000 nm, and 10,000 nm. Since the size of singular and agglomerated ENM s often vary, it was determined that using these particle sizing and counting instruments would provide a qualitative indication of the number concentration of both species. For example, a high particle number concentration measured by the CPC, in combination with a high number concentration in the lower size range ( nm) on the HHPC-6, would indicate the possible presence of ENM s. Conversely, a low CPC number concentration, in tandem with a high HHPC-6 number concentration in the larger size range (> 1,000 nm) would indicate the presence of larger, possibly agglomerated ENM s. Since these two instruments cannot distinguish between specific types of particulate matter (e.g. ambient dust versus manganese oxide spheres), background particle number concentration measurements were made at the rim of the reactor before and after each cleanout task. The before and after background particle number concentration data were averaged and then subtracted from the measurement made during the reactor cleanout. The background-corrected value is referred to as the adjusted particle number concentration. The adjusted number concentration is intended to serve as a measure of particulate matter emitted during reactor cleanout. Results The ENM s manufactured during this site visit consisted of oxides of five different metals (copper, iron, nickel, manganese, and cobalt), and elemental silver, each in a separate reactor. A total of 36 pairs of filter-based samples (one analyzed for mass, one analyzed via TEM) were collected across 3 days under two different ventilation conditions (with/without new flange LEV system). Additionally, two general area

6 samples, each consisting of a mass-based and TEM filter, were collected each day; one in the center of the reactor area and one in the shop/fabrication area approximately 75 feet away. All general area mass-based air samples were below the analytical limit of detection for each metal evaluated. However, the general area air samples analyzed via TEM showed evidence of the metal oxides/elemental silver handled on that particular day. Air concentrations (PBZ and AS) for the metal oxides and elemental silver ranged from 16 micrograms per cubic meter of air (µg/m 3 ) to 1,467 µg/m 3 when no LEV was used during reactor cleanout. However, air concentrations measured when the new LEV system was operating were substantially lower and ranged from none detected to 239 µg/m 3 (Table 1). The highest mass-based air sampling value was obtained from a PBZ sample during nickel cleanout with no ventilation, while the lowest value was obtained from an AS sample collected during iron cleanout with the new LEV system. Overall, PBZ samples were higher than AS samples regardless of the use of the New LEV system. The single exception to this generalization involved the samples collected during the silver reactor cleanout. During this task, it was noted that the standard work practice of brushing material toward the inlet of the LEV was not consistently followed. Therefore, samples collected during this activity are not indicative of the potential effectiveness of the new LEV system. Subsequent to this observation, the standard work practice was adhered to, and comparisons regarding the effectiveness can be made with more confidence. To determine the percent reduction in emissions when using the new LEV system during reactor cleanout, the following formula was used: ( without LEV - with LEV) x 100 = % Reduction without LEV The percent reduction values for all mass-based filter samples ranged from 83 % to 100 %, with an average (mean) value of 92% (Table 1). A summary of mass-based filter samples, according to sample type (PBZ or AS) appear in Table 2 and Table 3, respectively. In addition to the mass-based filter air samples, emission of ENM s was measured, with and without the new LEV system, using direct-reading, real-time instrumentation. Similar to the results obtained using the mass-based filter air samples, the use of the new LEV system reduced particle number concentrations substantially across the entire size range of both instruments (10 nm to 10 µm). Percent reduction values based on particle number concentration ranged from 95% to 100%, with an average (mean) value of 100% (Table 4). The highest particle number concentration release occurred during the iron reactor cleanout with no LEV. Additional information on the size, shape, morphology and agglomeration state are provided by the TEM images that appear in Figure 6. All samples, regardless of type or ventilation status had ENM present. In addition, all samples analyzed via TEM indicated that the majority of ENM s are emitted as agglomerates in the 0.5 µm to 5 µm size range.

7 Discussion A critical issue when characterizing source emissions and potential exposure using particle number concentration is specificity. Nanoparticles are ubiquitous in many workplaces, originating from sources such as combustion, vehicle emissions, and infiltration of outside air. Direct-reading instruments, as used in this evaluation, are generally insensitive to particle source or composition, making it difficult to differentiate between exogenous and process-related nanoparticles. However, short-term elevations in concentrations above the background that coincided with a particular action, process, or task generally may be ascribed to those actions or processes. Fortunately, potential confounders such as forklifts, gas burners, and other combustion sources were not present in the production area during this survey. Clearly, this air sampling methodology provides an indication of the effectiveness of the new flange LEV system and also supports the finding that, in general, LEV can be used to reduce emissions of ENMs. However, despite the substantial reduction in emissions during reactor cleanout tasks, the fact that ENM s were detected on PBZ, AS, general area samples while the new flange LEV system was in use indicates that the emissions are not completely controlled. In particular, the work practice of reaching up into the reactor during cleanout appears to increase the potential for PBZ exposure (Table 1). Also, on multiple occasions, internal components of the reactor (ceramic boats, conductors, contacts, rings) were removed and brushed down outside the reactor and away from the LEV system. This type of work practice will result in a release of material which could migrate to other areas of the plant (e.g. see TEM images associated with general area samples in Reactor area and Shop/fabrication area in Figure 6). Currently, there are no accepted occupational exposure criteria specific to the engineered nanometer form of a material against which to compare the findings of this survey. The occupational exposure limits for the larger particulate for each metal could serve as a minimal guide. Despite the limitations imposed on this survey by these factors, it can be concluded that the potential for release of ENMs does exist during reactor cleanout operations. However, based on the analysis of the filter-based air samples and the directreading instrumentation, it is clear that a properly maintained LEV system can be highly effective in reducing ENM emissions. This finding, coupled with the current use of PPE, appears to be an acceptable method of reducing the potential for worker exposure. Recommendations Since the air sampling data indicate that LEV is effective in reducing ENM emissions, and appropriate PPE is worn when conducting reactor cleanout operations, no additional protective measures are necessary. However, to further reduce the potential for exposure, the following alterations in current work practices are recommended: 1. Avoid reaching up inside the reactor when performing cleanout operations. An alternate cleaning method might be to use a HEPA-filtered portable vacuum with a brush/wand attachment to allow access to hard to reach areas.

8 2. If ceramic boats and other components must be removed from the reactor for cleaning, this task should be performed inside a laboratory hood. References: NIOSH Manual of Analytical Methods (NMAM) [2006]. Method Elements. In: NIOSH Method of Analytical Methods 4 th ed., Issue 1. Cincinnati, OH: Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health. DHHS (NIOSH) Publication NIOSH Manual of Analytical Methods (NMAM) [2006]. Method Asbestos by TEM. In: NIOSH Method of Analytical Methods 4 th ed., Issue 1. Cincinnati, OH: Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health. DHHS (NIOSH) Publication

9 Table 1: Effectiveness of LEV in Reducing Release of Aerosol during Reactor Cleanout Operations: Air concentrations of Metal Oxides With/Without New Semi-circular Flange Attached to LEV (All Data) - Micrograms/cubic meter (µg/m 3 ) QSI, Inc. Santa Ana, California Air Without New Flange LEV Air With New Flange LEV Percent Reduction in air concentration due to use of New Flange LEV (%)* Sample Operation Type Silver (Ag) Reactor cleanout PBZ 16 33** -106** Silver (Ag) Reactor cleanout AS 21 33** -57** Copper (Cu) Reactor cleanout PBZ Copper (Cu) Reactor cleanout AS Cobalt (Co) Reactor cleanout PBZ NS 209 N/A Cobalt (Co) Reactor cleanout AS NS 15 N/A Nickel (Ni) Reactor cleanout PBZ 1, Nickel (Ni) Reactor cleanout AS Iron (Fe) Reactor cleanout PBZ Iron (Fe) Reactor cleanout AS 278 ND*** 100 Manganese (Mn) Reactor cleanout PBZ Manganese (Mn) Reactor cleanout AS Mean (+/- S.D.) 92 (+/- 8) PBZ = Personal Breathing Zone AS = At Source ND = None detected, below analytical limit of detection of 0.9 µg Fe/sample NS = Operation not sampled; additional material not produced during site visit to allow for the computation of % reduction S.D. = Standard Deviation * Percent reduction was calculated as follows: [(Without New flange LEV - With New flange LEV)/ Without New Flange LEV] x 100 ** Data not used in Mean Percent reduction calculation; change in work practice (brushing ceramic boats and electrodes away from instead of toward LEV inlet) resulted in ineffective use of LEV. *** Analytical limit of detection (0.9 µg Fe/sample) used for % reduction calculation purposes.

10 Table 2: Effectiveness of LEV in Reducing Release of Aerosol during Reactor Cleanout Operations: Air concentrations of Metal Oxides With/Without New Semi-circular Flange Attached to LEV (PBZ Data Only) - Micrograms/cubic meter (µg/m 3 ) QSI, Inc. Santa Ana, California Air Without New Flange LEV Air With New Flange LEV Percent Reduction in Air due to use of New Flange LEV (%)* Sample Operation Type Silver (Ag) Reactor cleanout PBZ 16 33** -106** Copper (Cu) Reactor cleanout PBZ Cobalt (Co) Reactor cleanout PBZ NS 209 N/A Nickel (Ni) Reactor cleanout PBZ 1, Iron (Fe) Reactor cleanout PBZ Manganese (Mn) Reactor cleanout PBZ Mean (+/- S.D.) 86 (+/- 8) PBZ = Personal Breathing Zone NS = Operation not sampled; additional material not produced during site visit to allow for the computation of % reduction N/A = Not applicable S.D. = Standard Deviation * Percent reduction was calculated as follows: [(Without New flange LEV - With New flange LEV)/ Without New Flange LEV] x 100 ** Data not used in Mean percent reduction calculation; change in work practice (brushing ceramic boats and electrodes away from instead of toward LEV inlet) resulted in ineffective use of LEV.

11 Table 3: Effectiveness of LEV in Reducing Release of Aerosol during Reactor Cleanout Operations: Air concentrations of Metal Oxides With/Without New Semi-circular Flange Attached to LEV (AS Data Only) - Micrograms/cubic meter (µg/m 3 ) QSI, Inc. Santa Ana, California Air Without New Flange LEV Air With New Flange LEV Percent Reduction in Air due to use of New Flange LEV (%)* Sample Operation Type Silver (Ag) Reactor cleanout AS 21 33** -57** Copper (Cu) Reactor cleanout AS Cobalt (Co) Reactor cleanout AS NS 15 N/A Nickel (Ni) Reactor cleanout AS Iron (Fe) Reactor cleanout AS 278 ND*** 100 Manganese (Mn) Reactor cleanout AS Mean (+/- S.D.) 98 (+/-2) AS = At Source Analytical Limit of Detection = 0.9 µg Fe/sample NS = Operation not sampled; additional material not produced during site visit to allow for the computation of % reduction N/A = Not applicable S.D. = Standard Deviation * Percent reduction was calculated as follows: [(Without New flange LEV - With New flange LEV)/ Without New Flange LEV] x 100 ** Data not used in Mean percent reduction calculation; change in work practice (brushing ceramic boats and electrodes away from instead of toward LEV inlet) resulted in ineffective use of LEV. *** Analytical limit of detection (0.9 µg Fe/sample) used for % reduction calculation purposes.

12 Table 4: Effectiveness of New Semi-circular Flange LEV in Reducing Release of Aerosol during Reactor Cleanout Operations: Particle number concentrations and Percent Reduction QSI, Inc., Santa Ana, California Particle size (nm) Measured Number (Without New Flange LEV) Average Background Number Adjusted* Number (Without New Flange LEV) Measured Number (With New Flange LEV) Average Background Number Adjusted* Number (With New Flange LEV) Percent Reduction in Air due to use of New Flange LEV (%)** Material and Process Silver (Ag) Reactor cleanout ,343 (a) 173,632 (a) 85,711 (a) 52,670 53, ,851 (a) 30, ,003 (a) 11,416 11, ,000 91,834 (a) 7,473 84,361 (a) 3,839 3, ,000 33, , ,000 21, , ,000 18, , ( ) 45,000 29,000 16,000 *** *** *** N/A Copper (Cu) Reactor cleanout ,169 (a) 54,370 93,799 (a) 149,356 (a) 144,653 (a) 4, ,998 (a) 13,335 84,664 (a) 25,908 25, ,000 62,414 4,851 57,564 7,157 6, ,000 20, , ,000 12, , ,000 10, , ( ) *** *** *** 16,950 16, N/A Cobalt (Co) Reactor cleanout 300 N/S N/S N/S 46,156 46,800 0 N/A 500 N/S N/S N/S 9,198 8, N/A 1,000 N/S N/S N/S 2,728 2, N/A 3,000 N/S N/S N/S N/A 5,000 N/S N/S N/S N/A 10,000 N/S N/S N/S N/A ( ) N/S N/S N/S *** *** 0 N/A Nickel (Ni) Reactor cleanout ,409 (a) 117,988 (a) 115,422 (a) 62,667 65, ,011 (a) 21, ,620 (a) 15,438 16, , ,853 (a) 4,925 96,928 (a) 5,379 5,

13 3,000 38, , ,000 23, , ,000 19, , ( ) 24,000 18,100 5,900 16,000 17, Iron (Fe) Reactor cleanout ,639 (a) 112,637 (a) 309,002 (a) 66,986 61,786 5, ,093 (a) 19, ,420 (a) 15,565 14,221 1, , ,689 (a) 4, ,300 (a) 4,836 4, , ,083 (a) ,889 (a) ,000 79, ,006 (a) ,000 71, ,458 (a) ( ) 20,500 18,865 1,635 22,650 23, Manganese (Mn) ,430 (a) 91,482 (a) 218,948 (a) 48,668 54, ,116 (a) 12, ,124 (a) 11,011 12, , ,702 (a) 3, ,685 (a) 3,363 3, ,000 78,496 (a) ,375 (a) ,000 58, , ,000 53, , ( ) 21,100 19,800 1,300 19,000 24, Mean (+/- S.D.) 100 (+/- 1) Average computed from two measurements (before and after each process/task) * Adjusted Number = Measured Number - Average Background Number. If Average Background Number exceeds the Measured Number, the Adjusted Number value is considered to be zero when calculating percent reduction. ** Percent reduction was calculated as follows: [(Adjusted Without New Flange LEV - Adjusted With New Flange LEV)/Adjusted Without New Flange LEV] x 100. *** Equipment malfunction, data not collected. (a) Indicates measurement exceed the upper dynamic limit of the instrument (70,000 P/liter) and should be interpreted with caution. Note: The number concentration of particles in the size range (300 nm - 10,000 nm) was measured with an ARTI model HHPC-6 optical particle counter and is presented as total particles per liter of air (P/liter) for a specific size range. The number concentration of particles in the size range (10 nm nm) was measured with a TSI Model 3007 condensation particle counter and is presented as total particles per cubic centimeter of air (P/cc). S.D. = Standard Deviation NS = Operation not sampled; additional material not produced during site visit to allow for the computation of % reduction

14 Figure 1: Photo of fume extractor used during reactor cleanout procedure. Figure 2: Fabricated New semi-circular flange

15 Figure 3: New Semi-circular flange, as installed and used during reactor cleanout Figure 4: Reactor cleanout process with no LEV

16 Figure 5: Typical location of New semi-circular flange LEV system and production operator during reactor cleanout activities

17 Figure 6: Transmission Electron Microscopy (TEM) Micrographs of Air Samples Collected Without/With New semi-circular Flange LEV QSI, Inc. Santa Ana, California PBZ Sample - Silver Reactor cleanout using New Flange LEV system PBZ Sample - Silver Reactor cleanout without New Flange LEV system AS Sample - Silver Reactor cleanout using New Flange LEV system AS Sample - Silver Reactor cleanout without New Flange LEV system

18 Figure 6 (Cont d): Transmission Electron Microscopy (TEM) Micrographs of Air Samples Collected Without/With LEV QSI, Inc. Santa Ana, California PBZ Sample - Copper Reactor cleanout using New Flange LEV system PBZ Sample - Copper Reactor cleanout without New Flange LEV system AS sample - Copper Reactor cleanout using New Flange LEV system AS Sample - Copper Reactor cleanout without New Flange LEV system

19 Figure 6 (Cont d): Transmission Electron Microscopy (TEM) Micrographs of Air Samples Collected Without/With LEV QSI, Inc. Santa Ana, California No Sample collected PBZ sample - Cobalt Reactor cleanout using New Flange LEV system PBZ Sample - Cobalt Reactor cleanout without New Flange LEV system No Sample collected AS Sample - Cobalt Reactor cleanout using New Flange LEV system AS Sample - Cobalt Reactor cleanout without New Flange LEV system

20 Figure 6 (Cont d): Transmission Electron Microscopy (TEM) Micrographs of Air Samples Collected Without/With LEV QSI, Inc. Santa Ana, California PBZ sample - Nickel Reactor cleanout using New Flange LEV system PBZ sample - Nickel Reactor cleanout without New Flange LEV system AS sample - Nickel Reactor cleanout using New Flange LEV system AS sample - Nickel Reactor cleanout without New Flange LEV system

21 Figure 6 (Cont d): Transmission Electron Microscopy (TEM) Micrographs of Air Samples Collected Without/With LEV QSI, Inc. Santa Ana, California PBZ sample - Iron Reactor cleanout using New Flange LEV system PBZ sample - Iron Reactor cleanout without New Flange LEV system AS sample - Iron Reactor cleanout using New Flange LEV system AS sample - Iron Reactor cleanout without New Flange LEV system

22 Figure 6 (Cont d): Transmission Electron Microscopy (TEM) Micrographs of Air Samples Collected Without/With LEV QSI, Inc. Santa Ana, California PBZ sample - Manganese Reactor cleanout using New Flange LEV system PBZ sample - Manganese Reactor cleanout without New Flange LEV system AS sample - Manganese Reactor cleanout using New Flange LEV system AS sample - Manganese Reactor cleanout without New Flange LEV system

23 Figure 6 (Cont d): Transmission Electron Microscopy (TEM) Micrographs of Air Samples Collected Without/With LEV QSI, Inc. Santa Ana, California Full-shift General area - Day 1: (Center of reactor area) during cleanout of Silver and Cobalt reactor using New flange LEV system. Image is showing an elemental Silver agglomerate Full-shift General area - Day 1: (Center of reactor area) during cleanout of Copper reactor without New Flange LEV system. Image is showing a Copper oxide agglomerate Full-shift General area - Day 2: (Center of reactor area) during cleanout of Iron reactor with New flange LEV. Image is showing an Iron oxide agglomerate Full-shift General area - Day 2: (Shop/fabrication area) during cleanout of Iron reactor with New flange LEV. Image is showing an Iron oxide agglomerate

24 Figure 6 (Cont d): Transmission Electron Microscopy (TEM) Micrographs of Air Samples Collected Without/With LEV QSI, Inc. Santa Ana, California Full-shift General area - Day 3: (Center of reactor area) during cleanout of Copper, Silver, and Nickel reactors without New flange LEV system - Image is showing a Copper oxide agglomerate Full-shift General area - Day 3: (Shop/Fabrication area) during cleanout of Copper, Silver, and Nickel reactors without New flange LEV system - Image is showing a Copper oxide agglomerate Full-shift General area - Day 4: (Center of reactor area) during cleanout of Iron and Manganese reactors without New flange LEV system. Image is of an Iron oxide agglomerate Full-shift General area - Day 4: (Shop/fabrication area) during cleanout of Iron and Manganese reactors without New flange LEV system. Image is of an Iron oxide agglomerate