<%BANNER%>

Control of Hexavalent Chromium Emission from Welding Using Silica Precursors as Shield Gas Additives

Permanent Link: http://ufdc.ufl.edu/UFE0042177/00001

Material Information

Title: Control of Hexavalent Chromium Emission from Welding Using Silica Precursors as Shield Gas Additives
Physical Description: 1 online resource (96 p.)
Language: english
Creator: Topham, Nathan
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: chromium, fume, hexavalent, precursor, silica, sorbent, welding
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Hexavalent chromium (Cr6+) emitted from arc welding poses serious health risks to workers exposed to welding fumes in occupational settings. Stainless steel gas tungsten arc welding (GTAW) and gas metal arc welding (GMAW) produce aerosols that contain Cr6+, nickel, manganese, and gaseous pollutants such as ozone and nitrogen oxides (NOx) that lead to a number of respiratory and neurological ailments as well as cancer. Traditional welding technologies employ shield gas or material incorporated into the welding wire to minimize fume formation. However, significant amounts of hazardous air pollutants (HAPs) are generated during gas shielded welding processes. In this study, tetraethyloxysilane (TEOS) and tetramethylsilane (TMS) were added to GTAW and GMAW shield gas to provide a two-fold approach at controlling HAPs produced during stainless steel welding. Silica precursors acted as reducing agents when they decomposed in the high temperature welding arc, thereby limiting ozone and Cr6+ formation. Additionally, an amorphous film of silica (SiO2) was deposited on welding fume particles. This film insulated the aerosols, preventing subsequent oxidation of surface chromium and masking the toxic effects of all of the metals contained in the particles. Cr6+ in GTAW fumes was reduced by 45% when 26.3% of the shield gas was used as TEOS carrier gas. A 53% reduction in NOx emissions was observed, indicating a reduction in reactive oxygen species. Tranmission electron microscope (TEM) imagery showed a film of SiO2 on welding fume particles. The SiO2 film could insulate GTAW aerosols, masking the toxic effects of all metals within the fume particles. Experimental results demonstrated that total shield gas flow rate impacted the effectiveness of TMS as a GMAW shield gas additive. Low shield gas flow rates led to premature thermal decomposition of TMS and no observable change in Cr6+ mass in fumes. Increasing shield gas flow rate led to significant reductions in Cr6+ mass when TMS was used. When 4.2% of the 30 liter per minute (Lpm) shield gas flow rate was used as TMS carrier gas, Cr6+ mass in GMAW fumes was reduced by at least 92.9%. Adding TMS to GMAW shield gas increased fume particle size beyond the nanometer range in all shield gas flow rates tested. Geometric mean particle size increased from 40-60 nanometers (nm) under baseline conditions to 180-300 nm when TMS was added. SiO2 particles formed from decomposition of silica precursors scavenged nanoparticles through intercoagulation. TEM imagery provided visual evidence of an amorphous film of SiO2 on some fume particles along with the presence of large amorphous SiO2 agglomerates. These results showed that vapor phase sorbents may be capable of reducing the health risks posed by welding fumes through elimination of nanoparticles and reduction of Cr6+.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Nathan Topham.
Thesis: Thesis (M.E.)--University of Florida, 2010.
Local: Adviser: Wu, Chang-Yu.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042177:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042177/00001

Material Information

Title: Control of Hexavalent Chromium Emission from Welding Using Silica Precursors as Shield Gas Additives
Physical Description: 1 online resource (96 p.)
Language: english
Creator: Topham, Nathan
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: chromium, fume, hexavalent, precursor, silica, sorbent, welding
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Hexavalent chromium (Cr6+) emitted from arc welding poses serious health risks to workers exposed to welding fumes in occupational settings. Stainless steel gas tungsten arc welding (GTAW) and gas metal arc welding (GMAW) produce aerosols that contain Cr6+, nickel, manganese, and gaseous pollutants such as ozone and nitrogen oxides (NOx) that lead to a number of respiratory and neurological ailments as well as cancer. Traditional welding technologies employ shield gas or material incorporated into the welding wire to minimize fume formation. However, significant amounts of hazardous air pollutants (HAPs) are generated during gas shielded welding processes. In this study, tetraethyloxysilane (TEOS) and tetramethylsilane (TMS) were added to GTAW and GMAW shield gas to provide a two-fold approach at controlling HAPs produced during stainless steel welding. Silica precursors acted as reducing agents when they decomposed in the high temperature welding arc, thereby limiting ozone and Cr6+ formation. Additionally, an amorphous film of silica (SiO2) was deposited on welding fume particles. This film insulated the aerosols, preventing subsequent oxidation of surface chromium and masking the toxic effects of all of the metals contained in the particles. Cr6+ in GTAW fumes was reduced by 45% when 26.3% of the shield gas was used as TEOS carrier gas. A 53% reduction in NOx emissions was observed, indicating a reduction in reactive oxygen species. Tranmission electron microscope (TEM) imagery showed a film of SiO2 on welding fume particles. The SiO2 film could insulate GTAW aerosols, masking the toxic effects of all metals within the fume particles. Experimental results demonstrated that total shield gas flow rate impacted the effectiveness of TMS as a GMAW shield gas additive. Low shield gas flow rates led to premature thermal decomposition of TMS and no observable change in Cr6+ mass in fumes. Increasing shield gas flow rate led to significant reductions in Cr6+ mass when TMS was used. When 4.2% of the 30 liter per minute (Lpm) shield gas flow rate was used as TMS carrier gas, Cr6+ mass in GMAW fumes was reduced by at least 92.9%. Adding TMS to GMAW shield gas increased fume particle size beyond the nanometer range in all shield gas flow rates tested. Geometric mean particle size increased from 40-60 nanometers (nm) under baseline conditions to 180-300 nm when TMS was added. SiO2 particles formed from decomposition of silica precursors scavenged nanoparticles through intercoagulation. TEM imagery provided visual evidence of an amorphous film of SiO2 on some fume particles along with the presence of large amorphous SiO2 agglomerates. These results showed that vapor phase sorbents may be capable of reducing the health risks posed by welding fumes through elimination of nanoparticles and reduction of Cr6+.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Nathan Topham.
Thesis: Thesis (M.E.)--University of Florida, 2010.
Local: Adviser: Wu, Chang-Yu.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042177:00001


This item has the following downloads:


Full Text

PAGE 1

1 C ONTROL OF HEXAVALENT CHROMIUM EMISSION FROM WELDING USING SILICA PRECURSORS AS SHIELD GAS ADDITIVES By NATHAN E. TOPHAM A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 Nathan Topham

PAGE 3

3 To my family

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my graduate advisor and committee chair, Dr. Chang Yu Wu, for his unending patience a nd support during my studies. Without his guidance and hard work, this project would not have been possible. I would also like to thank my committee members, Dr. Jean Claude Bonzongo and Dr. Vito Ilacqua, for their suggestions and support. d like to thank my graduate mentor, Dr. Yu Mei Hsu, for her support during my studies. Her leadership helped inspire me to pursue a graduate degree. I would also like to extend my gratitude to all of my fellow undergraduate and graduate lab mates who helpe better half, Chelsea, for supporting me during the difficult times during my studies and helping me find the strength to continue striving towards my goal when I faltered.

PAGE 5

5 TABLE OF CONTENTS pag e ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 10 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 13 Fundamentals of Arc Welding ................................ ................................ ................. 13 Physical and Chemical Properties of Welding Fumes ................................ ............ 14 Health Effects of Wel ding Fume Exposure ................................ ............................. 15 Welding Fume Sampling and Analysis ................................ ................................ .... 18 Past Attempts to Control Welding Fume Exposure ................................ ................. 19 Sorbents in Combustion Systems ................................ ................................ ........... 20 Objective of Study ................................ ................................ ................................ ... 22 2 EXPERIMENTAL METHODS ................................ ................................ ................. 24 Overview of Experimental Methods ................................ ................................ ........ 24 GTAW Fume Generation and Sampling ................................ ................................ 25 GTAW Experimental Conditions ................................ ................................ ............. 26 GTAW Sample Analysis ................................ ................................ .......................... 26 GMAW Fume Generation and Sampling ................................ ................................ 27 GMAW Experimental Conditions ................................ ................................ ............ 29 GMAW Sample Analysis ................................ ................................ ......................... 30 3 RESULTS AND DISCUSSION GAS TUN GSTEN ARC WELDING (GTAW) ....... 35 Results ................................ ................................ ................................ .................... 35 Discussion ................................ ................................ ................................ .............. 36 Cost Analys is ................................ ................................ ................................ .......... 37 4 RESULTS AND DISCUSSION GAS METAL ARC WELDING (GMAW) .............. 39 TMS Cr 6+ and Particle Size Distribution Results ................................ ..................... 39 30 Lpm Shield Gas Flow Rate ................................ ................................ .......... 39 30 Lpm Shield Gas Flow Rate Using High Voltage ................................ .......... 42 25 Lpm Shield Gas Flow Rate ................................ ................................ .......... 43

PAGE 6

6 20 Lpm Shield Gas Flow Rate ................................ ................................ .......... 44 Effects of Shield Gas Flow Rate ................................ ................................ ....... 45 TMS Cost Analysis ................................ ................................ ................................ 46 Regulatory Compliance ................................ ................................ ........................... 47 5 CONCLUSIONS ................................ ................................ ................................ ..... 61 APPENDIX A DETERMINATION OF THEORETICAL MINIMUM PRECURSOR CONCENTRATION IN FUME PARTICLE FORMATION AREA ............................. 63 B DETERMINATION OF TMS AND TEOS VAPOR PRESSURE .............................. 66 C DETERMINATION OF PRECURSOR CONCENTRATION IN SATURATED CARRIER GAS ................................ ................................ ................................ ....... 67 D COST ANALYSIS ................................ ................................ ................................ ... 69 GTAW Operating Cost of TEOS Addition ................................ ............................... 69 GMAW Operating Cost of TMS Cost Addition ................................ ........................ 69 E RAW DATA FOR H EXAVALENT CHROMIUM ................................ ....................... 71 F SCANNING MOBILITY PARTICLE SIZER PARTICLE SIZE DISTRIBUTION DATA ................................ ................................ ................................ ...................... 73 G CASCADE IMPACTOR CUT SIZE AND PART ICLE SIZE DISTRIBUTION DATA ................................ ................................ ................................ ...................... 87 LIST OF REFERENCES ................................ ................................ ............................... 89 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 96

PAGE 7

7 LIST OF TABLES Ta ble page 2 1 GMAW experimental conditions. ................................ ................................ ........ 34 2 2 TMS molar flow rates used during GMAW sampling. ................................ ......... 34 4 1 Qualitative elemental analysis of Figure 4 6 fume particles. ............................... 54 4 2 Fume generation rate using various TMS feed rates. ................................ ......... 55 4 3 Baseline particle size characteristics at different shield gas flow rates. .............. 59 A 1 Particle size data used to calculate minimum precursor conc entration, based on particle size distribution from literature (Zimmer and Biswas 2001). .............. 64 B 1 Antoine coefficients for TMS and TEOS. ................................ ............................ 66 D 1 Cost analysis data for GMAW with TMS as a shield gas additive. ...................... 70 E 1 Raw Cr 6+ data for baseline and TMS samples. ................................ ................... 71 F 1 SMPS particle size data for baseline 20 Lpm shield gas flow rate. ..................... 74 F 2 SMPS particle size data for 20 Lpm shield gas flow rate with 1.05% TMS carrier gas. ................................ ................................ ................................ .......... 74 F 3 SMPS particle size data for 20 Lpm shield gas flow rate with 2.1% TMS carrier gas. ................................ ................................ ................................ .......... 74 F 4 SMPS particle size data for 20 Lpm shield gas flow rate wit h 6.3% TMS carrier gas. ................................ ................................ ................................ .......... 75 F 5 SMPS particle size data for baseline 25 Lpm shield gas flow rate. ..................... 75 F 6 SMPS particle size data for 25 Lpm shield gas flow rate with 0.84% TMS carrier gas. ................................ ................................ ................................ .......... 75 F 7 SMPS particle size data for 25 Lpm shield gas flow rate with 1.68% TMS carrier gas. ................................ ................................ ................................ .......... 76 F 8 SMPS particle size data for 25 Lpm shield gas flow rate with 5.04% TMS carrier gas. ................................ ................................ ................................ .......... 76 F 9 SMPS particle size data for baseline 30 Lpm shield gas flow rate. ..................... 76 F 10 SMPS particle size data for 30 Lpm shield gas flow rate with 0.7% TMS carrier gas. ................................ ................................ ................................ .......... 77

PAGE 8

8 F 11 SMPS particle size data for 30 Lpm shield gas flow rate with 1.4% TMS carrier gas. ................................ ................................ ................................ .......... 77 F 12 SMPS particle size data for 30 Lpm shield gas flow rate with 4.2% TMS carrier gas. ................................ ................................ ................................ .......... 77 F 13 Raw SMPS particle size distribution data for 20 Lpm shield gas flow rate. ........ 78 F 14 Raw SMPS particle size distribution data for 25 Lpm shield gas flow rate. ........ 80 F 15 Raw SMPS particle size distribution data for 30 Lpm shield gas flow rate. ........ 83 G 1 Impactor data used for cut size calculations. ................................ ...................... 87 G 2 Particle size data measured using cascade impactor. ................................ ........ 88

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Welding torches. ................................ ................................ ................................ 23 1 2 Formation of welding particles with and without TMS addition. .......................... 23 2 1 GTAW fume generation and sampling system. ................................ .................. 31 2 2 GTAW torch modified to allow TEOS addition. ................................ ................... 32 2 3 GMAW fume generation and sampling system. ................................ .................. 33 3 1 GTAW Cr 6+ and nitrate data before and after TEOS addition. ............................ 37 3 2 TEM images of GTAW fume particles. ................................ ............................... 38 4 1 Average Cr 6+ mass as a function of TMS carrier gas flow rate at 30 Lpm total shield gas flow rate. ................................ ................................ ............................ 47 4 2 GMAW fume particle size distributions at 30 Lpm shield gas flow rate. .............. 48 4 3 GMAW fume particle size distribution under 4.2% TMS flow rate measured with cascade impactor. ................................ ................................ ....................... 49 4 4 GMAW fume agglomerate without SiO 2 coating. ................................ ................ 50 4 5 GMAW fume agglomerate with SiO 2 coating indicated by arrows. ..................... 51 4 6 GMAW primary particles coated with SiO 2 ................................ ........................ 52 4 7 Amorphous SiO 2 agglomerate fume particle. ................................ ...................... 53 4 8 Intercoagulation between metal particles and SiO 2 agglomerate. ...................... 54 4 9 Average Cr 6+ mass in GMAW fumes using globular metal transfer mode. ......... 55 4 10 Average Cr 6+ mass as a fun ction of TMS carrier gas flow rate at 25 Lpm total shield gas flow rate. ................................ ................................ ............................ 56 4 11 GMAW fume particle size distributions at 25 Lpm shield gas flow rate. .............. 57 4 12 Average Cr 6+ mass as a function of TMS carrier gas flow rate at 20 Lpm total shield gas flow rate. ................................ ................................ ............................ 58 4 13 GMAW fume particle size distributions at 20 Lpm shield gas flow rate. .............. 59 4 14 Particle size trends at different shield gas and TMS carrier gas flow rates. ........ 60

PAGE 10

10 LIST OF ABBREVIATION S AWS American Welding Society cfm Cubic feet per minute Cr 6+ Hexavalent chromium EDS Energy dispersive x ray spectroscopy GMAW Gas metal arc welding GTAW Gas tungsten arc welding IC Ion chromatography ICP AES Inductively coupled plasma atomic emission spectrometry Lpm Liters per minute m mHg Millimeter mercury mM Millimolar MMD Mass median diameter NIOSH National Institute for Occupational Safety and Health nm Nanometer OSHA Occupational Safety and Health Administration ppm Parts per million SMPS Scanning mobility particle sizer TEM Transm ission electron microscopy TEOS Tetraethyloxysilane (CAS # 78 10 4) TMS Tetramethylsilane (CAS # 75 76 3) g Microgram g/m 3 Microgram per cubic meter of air m Micrometer UV Ultraviolet

PAGE 11

11 Abstract of Thesis Presented to the Graduate School of the Universit y of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering CONTROL OF HEXAVALENT CHROMIUM EMISSION FROM WELDING USING SILICA PRECURSORS AS SHIELD GAS ADDITIVES By Nathan E. Topham August 2010 Chair: C hang Yu Wu Major: Environmental Engineering Sciences Hexavalent chromium (Cr 6+ ) emitted from arc welding poses serious health risks to workers exposed to welding fumes in occupational settings. Stainless steel gas tungsten arc welding (GTAW) and gas metal arc welding (GMAW) produce aerosols that contain Cr 6+ nickel, manganese, and gaseous pollutants such as ozone and nitrogen oxides (NO x ) that lead to a number of respiratory and neurological ailments as well as cancer. Traditional welding technolog ies emp loy shield gas or material incorporated into the welding wire to minimize fume formation. However, significant amounts of hazardous air pollutants ( HAPs ) are generated during gas shielded welding processes. In this study, tetraethyloxysilane (TEOS) and te tramethylsilane (TMS) were added to GTAW and GMAW shield gas to provide a two fold approach at controlling HAPs produced during stainless steel welding. Silica precursors act ed as reducing agent s when they decompose d in the high temperature welding arc, th ereby limiting ozone and Cr 6+ formation. Additionally, an amorphous film of silica ( SiO 2 ) was deposited on welding fume particles. This film insulate d the aerosols, preventing subsequent

PAGE 12

12 oxidation of surface chromium and masking the toxic effects of all of the metals contained in the particles. Cr 6+ in GTAW fumes was reduced by 45% when 26.3% of the shield gas was used as TEOS carrier gas. A 53% reduction in NO x emissions was observed, indicating a reduction in reactive oxygen species. Tranmission electron microscope ( TEM ) imagery showed a film of SiO 2 on welding fume particles. The SiO 2 film could insulate GTAW aerosols, masking the toxic effects of all metals within the fume particles. Experimental results demonstrated that total shield gas flow rate imp acted the effectiveness of TMS as a GMAW shield gas additive Low shield gas flow rates led to premature thermal decomposition of TMS and no observable change in Cr 6+ mass in fumes. Increasing shield gas flow rate led to significant reductions in Cr 6+ mass when TMS was used. When 4.2% of the 30 liter per minute ( Lpm ) shield gas flow rate was used as TMS carrier gas, Cr 6+ mass in GMAW fumes was reduced by at least 92.9%. Adding TMS to GMAW shield gas increased fume particle size beyond the nanometer range i n all shield gas flow rates tested. Geometric mean particle size increased from 40 60 nanometers ( nm ) under baseline conditions to 180 300 nm when TMS was added. SiO 2 particles formed from decomposition of silica precursors scavenged nanoparticles through intercoagulation. TEM imagery provided visual evidence of an amorphous film of SiO 2 on some fume particles along with the presence of large amorphous SiO 2 agglomerates. These results show ed that vapor phase sorbents may be capable of reducing the health ri sks posed by welding fumes through elimination of nanoparticles and reduction of Cr 6+

PAGE 13

13 CHAPTER 1 INTRODUCTION Fundamentals of Arc Welding Gas metal arc welding (GMAW) and gas tungsten arc welding (GTAW) are commonly used welding process es that use mild or stainless steel wire as filler material to join pieces of metal. Shield gas is used to protect the superheated weld site from gaseous species in air that degrade the mechanical properties of the weld and destabilize the welding arc. The operator controls welding parameters such as voltage, shield gas composition, shield gas flow rate, and wire feed speed to obtain desired weld characteristics. As shown in Figure 1 1 the GTAW electrode is non consumable tungsten and a separate rod is used for filler materi al while in GMAW the consumable steel wire acts as the electrode and filler material Operating parameters influence the properties of the welding arc and welded material. Voltage and metal transfer mode are related, with low voltage conditions creating s hort circuit and high voltage creating globular and spray transfer (Soderstrom and Mendez 2008). During globular metal transfer mode, droplet detachment from the electrode is chaotic and large amounts of spatter and fumes can result from operating under th is metal transfer mode (Simpson 2009; Pires et al. 2007). Choice of shield gas can also impact the characteristics of the weld. The addition of carbon dioxide ( CO 2 ) to GMAW shield gas requires higher voltage to maintain arc stability and weld penetration d ue to loss of arc heat by conduction by CO 2 (Pires et al. 2007). The operating parameters impact fume formation as well as weld characteristics which creates a challenge for industry to balance maximum product quality with minimum emissions

PAGE 14

14 Physical and C hemical Properties of Welding Fumes The intense energy of the welding process results in the formation of fumes containing a high number concentration of nanosized and sub micron particles containing toxic metals as well as a number of gaseous species incl uding ozone and nitrogen oxides (Jenkins et al. 2005; Hewett 1995a ; Liu et al. 2007 ). Fume aerosols generated during arc welding pro cesses are typically less than 1 micrometer ( ) in diameter and the primary particles are often in the nanometer range (Jen kins et al. 2005; Biswas and Wu 2005; Zimmer et al. 2002; Stephenson et al. 2003). The fume characteristics are influenced by a variety of parameters, including shield gas composition, welding wire composition, voltage, and metal transfer mode (Zimmer et a l. 2002; Hovde and Raynor 2007). Increasing the shield gas flow rate yields higher hexavalent chromium ( Cr 6+ ) formation while increasing the amount of CO 2 in shield gas reduces Cr 6+ formation (Dennis et al. 1997). Fume formation increases as welding voltag e is increased (Hovde and Raynor 2007). GMAW processes are more commonly used and have much higher fume formation rates than GTAW (Serageldin and Reeves 2009). Thus, fume formation from GMAW has been studied much more thoroughly than GTAW. The chemical co mposition of welding aerosols depends on the composition of the welding wire used as well as the operating parameters. Mild steel welding fumes generally consist of iron and manganese oxides (Minni et al. 1984; Jenkins and Eagar 2005). Stainless steel weld ing fumes contain manganese and iron; however, they also contain chromium and nickel oxides (Castner and Null 1998; Heung et al. 2007). About 1 5% of the chromium found in GMAW fumes is in the hazardous hexavalent state (Heung et al. 2007 ; Serageldin and R eeves 2009 ). These metals are layered in a

PAGE 15

15 core shell structure Fume particles typically contain a core made up primarily of iron with other metals making up the outer layers along with surface enrichment by lighter elements such as silicon, chlorine, and fluorine (Konarski et al. 2003a,b ; Maynard et al. 2004 ). Welding uses a shield gas to maintain arc stability produce desirable weld penetration, and reduce fume formation rate (Dennis et al. 1997 ; Ebrahimnia et al. 2009 ). The shield gas chosen has an imp act on ultraviolet ( UV ) light, ozone generated particle size distribution of aerosols and the amount of Cr 6+ (Zimmer et al. 2002; Dennis et al. 1997). The use of shield gas in GMAW decreases overall fume formation rate; however, it may increase the forma tion rate of some of the toxic species in welding fumes. Dennis et al. (1997) studied the effect of shield gas flow rate on formation rates of Cr 6+ UV light, ozone, and total fume. As the shield gas flow rate increase d the total fume formation rate decre ase d However, the formation rates of UV light, ozone, and Cr 6+ all increase d with shield gas flow rate. The absence of shield gas leads to low UV light production and rapid decomposition of ozone through reactions with nitrogen. As ozone formation increas es, the Cr 6+ generation rate also increases as ozone is considered the dominant oxidizer of chromium. The toxic specie s present in welding fumes lead to many adverse health effects following occupational exposure. Health Effects of Welding Fume Exposure Pr ofessional welders are a group that has been the subject of many epidemiological studies linking occupational exposures to aerosols with adverse health effects including a wide variety of respiratory and cardiovascular ailments (Lillienberg et al. 2008; Lo ukzadeh et al. 2009; Fang et al. 2009; Antonini 2003). A number of the metallic species present in welding fumes are potentially detrimental to worker health

PAGE 16

16 and ambient air quality. Hexavalent chromium causes decreased lung function, asthma, and a number of other respiratory ailments (Bagchi et al. 2002; Pascal and Tessier 2004). Cr 6+ is also a known human carcinogen, with extensive human and animal data available (Holmes et al. 2008; IARC 1990; Seel et al. 2007). Studies of welding emissions in Californi a found that welding is the primary source of airborne Cr 6+ in the state (Chang et al. 2004). Nickel is also a known human carcinogen (IARC 1990), which is linked with oxidative stress after inhalation (Luo et al. 2009). Manganese (Mn) exposure can cause a number of adverse neurological effects, like disorder known as manganism (Smargiassi et al. 2000; Yuan et al. 2006; Antonini et al. 2006; Bowler et al. 2006; Halatek et al. 2005; Bowler et al. 2007; Flynn and Susi 2009). Although this disorder creates symptoms similar to The prevalence of neurological effects a mong workers exposed to welding fume manganese is as high as one in three (Park et al. 2006b). The high risk of developing serious neurological impairment from exposure to manganese in welding fumes will likely lead to tightening regulations for this metal Elimination of metals in welding fumes from the body is a slow process. The half lives of chromium and nickel in urine are 730 days and 610 days, respectively (Schaller et al. 2007). The health effects from welding aerosols are compounded by the fact tha t exposure to multiple potentially toxic constituents usually occurs at the same time. The particulate metal species are not the only harmful constituents of welding fumes. Ozone is present in welding fumes and enhances the toxic effects of the metals

PAGE 17

17 in welding fumes by inhibiting defense mechanisms in the human respiratory system (Cohen et al. 2003). Ozone is produced within a few seconds when welding starts and persists up to 10 minutes after welding is stopped (Liu et al. 2007). This complicates the us e of local ventilation because pumps are often turned off after welding stops since particles are only generated during welding while ozone persists beyond the cessation of welding. The size and shape of welding aerosols also play a key role in toxicity. A large percentage of inhaled welding fume particles can deposit in the lungs, typically in the alveoli (Hewett 1995b). Few welding fume particles deposit in the upper respiratory system, while deposition in the lower regions of the respiratory system occu rs more readily (Yu et al. 2000; Kleinstreuer et al. 2008). T oxicity of nanoparticles is unique because the solid particles can be translocated across pathways other than the respiratory system that are typically not considered for larger particles (Biswas and Wu 2005). For example, manganese nanoparticles can be translocated through the olfactory nerve directly to the brain where manganese expresses neurotoxic effects (Elder et al. 2006). Once inhaled, the behavior of the species present in welding fume pa rticles in biological systems becomes important. Solubility is an important factor in determining toxicity. Welding fumes have been divided into soluble and insoluble fractions, and the toxic effects of each fraction have been examined separately. The solu ble fraction of welding fumes is responsible for toxicity due to oxidative stress and free radical production (McNeilly et al. 2004; Taylor et al. 2003). It also causes more loss of lung macrophage viability than the insoluble components (Antonini et al. 1 999). The insoluble fraction of welding particles is also

PAGE 18

18 responsible for pulmonary damage. Welding particles differ from other metal aerosols in that the insoluble and soluble fractions of the fumes are both significant contributors to toxicity (Antonini et al. 2004). The serious adverse health effects of welding fumes have led regulatory agencies to develop a number of sampling and analytical techniques to measure the toxic constituents of welding fumes in occupational settings. Welding Fume Sampling and Analysis The process of sampling and analyzing welding fumes is complicated with many variables that must be controlled. The American Welding Society (AWS) has developed a method of generating and collecting welding fumes (AWS 1999) This technique uses a welding torch fixed inside a chamber in which fumes are generated and welding fumes are collected on a glass fiber filter. The collection efficiency of the glass fiber filters used in this method has been tested and verified (Quimby and Ulrich 1999). The AWS hood allows for a number of operating parameters to be fixed, making it ideal for generating welding fumes while only varying certain operating parameters. Personal samplers are a common method for collecting welding fumes in actual occupational setti ngs. These methods typically involve a filter placed near the breathing zone that is connected to a personal pump. Recent research has shifted to placing personal samplers inside welding masks to gather samples that accurately represent particles present i n the welder s breathing zone (Lidn and Surakka 2009). While there are a number of National Institute for Occupational Safety and Health ( NIOSH ) and Occupational Safety and Health Administration ( OSHA ) methods for personal sampling of Cr 6+ there is no s ignificant difference between the results obtained by these methods (Boiano et al 2000).

PAGE 19

19 Analytical techniques used to measure components of welding fumes focus on measuring Cr 6+ NIOSH has developed multiple methods for analyzing Cr 6+ in occupational ai r that rely on ion chromatography (IC) NIOSH Method 7604 (NIOSH 1994) uses ion chromatography with conductivity detection while Method 7605 (NIOSH 2003) uses ion chromatography with UV detection. Method 7605 is an improvement over Method 7604 in that the detection limit is reduced from 3.5 micrograms ( g ) per sample using 7604 to 0.02 g per sample using 7605. However, m etals such as iron and nickel that are present in welding fumes can potentially interfere with Method 7605 and lead to overestimation of C r 6+ concentration (NIOSH 2003). OSHA Method ID 215 attempts to address the possible interferences by precipitating interfering metals out of the extraction solution while employing the same analytical technique as NIOSH Method 7605 (Ku and Eide 1998). The sampling and analysis methods for Cr 6+ in welding fumes are necessary to protect workers because of significant adverse health effects that can result from exposure Past Attempts to Control Welding Fume Exposure OSHA recently reduced the permissible expos ure limit for Cr 6+ in occupational air from 52 micrograms per cubic meter of air ( g/m 3 ) to 5 g/m 3 (OSHA 2006) This change placed pressure on industry to develop new control technologies to limit emissions of Cr 6+ from arc welding. The simplest method of controlling welding fume exposure is removing the fumes from the breathing zone of the welder. Studies have shown that using large vacuum pumps can reduce exposure to metals in welding fumes by about half if the hose for the pump is placed 2 3 inches from the welding torch (Meeker et al. 2007). This technology can be invasive and inc onvenient for workers and only removes some fume particles (Flynn and Susi 2009)

PAGE 20

20 Previous research has demonstrated that the use of reducing agents as shield gas additives ca n reduce ozone formation in the welding fume (Dennis et al. 2002). Similarly, the addition of reactive metals, such as zinc and aluminum, to welding wires can reduce Cr 6+ formation (Dennis et al. 1996). However, the use of reactive metals was not effective for controlling Cr 6+ formation in some operating conditions. Reactive metal additives caused an increase in Cr 6+ formation in some high voltage operating conditions. These technologies have not controlled formation of Cr 6+ effectively enough to gain accep tance in industry Furthermore nickel and manganese are present in different oxidation states that have varying solubility and toxicity (Minni et al. 1984; Jenkins and Eagar 2005). Therefore, traditional shield gas additives that act solely as reducing ag ents may not be as effective at limiting the toxicity of these metals in welding fumes. Mruczek et al. (2008) attempted to reduce manganese exposure by developing welding wire with reduced Mn content. This study found that Mn content in welding fumes could be reduced with this process modification; however, the change in composition of the wire resulted in unacceptable degradation of weld properties. This study demonstrates the difficulty in controlling hazardous components in welding fumes; they cannot sim ply be removed without degrading the needed mechanical properties of the weld. Therefore, it is necessary to develop a control technology that will limit exposure to these pollutants while maintaining the integrity of the welded metal. Sorbents in Combusti on Systems Sorbent injection into combustion systems has been examined as a means for removing trace metals. Sorbents are capable of scavenging a variety of vapor phase metals in combustion systems to prevent nucleation of metallic aerosols (Biswas and Wu 1998). Thermodynamic analyses have been carried out that have shown sorbents

PAGE 21

21 to be an effective means for removing vanadium, molybdenum, and arsenic metal vapors in combustion environments (Lee and Wu 2002; Cho and Wu 2004; Wu and Barton 2001 ; Iida et al. 2003 ). Injection of bulk solid phase sorbents is not feasible for welding systems. Fortunately, vapor phase sorbents have recently emerged as an alternative to solid phase materials. Numerous studies have tested the use of vapor phase silica precursors for control of metal emissions in generic combustion systems. During combustion, the precursor molecule is broken apart at high temperature This process leads to the formation of silica (SiO 2 ) along with carbon dioxide and water. Silica, formed from pyrolys is and subsequent oxidation of precursor chemicals, has been proven effective as a sorbent for lead produced during combustion (Biswas and Zachariah 1997; Owens and Biswas 1996a,b). Silica is capable of forming an amorphous web that effectively captures ul trafine metal aerosols and increases their particle size (Biswas and Zachariah 1997; Owens and Biswas 1996a,b; Lee et al. 2005). The increase in aggregate particle size is beneficial for improved performance of filtration systems. The silica formed from th e injection of silica precursors into combustion systems is amorphous phase silica rather than crystalline silica (Owens and Biswas 1996; Jang et al. 2006) Amorphous silica does not cause the toxic effects observed from inhalation of crystalline silica (R euzel et al. 1991). In regards to welding fume control, t his technology is superior to traditional shield gas additives because the silica coating formed around the metallic aerosols insulates all of the metals in the aerosols as seen in Figure 1 2, rathe r than simply manipulating oxidation state of chromium while failing to address the toxicity of manganese and nickel During welding, o xygen that is able to penetrate the shield gas is consumed

PAGE 22

22 during the pyrolysis process, thereby reducing the concentrati on of reactive oxygen species (O, O 3 ) that would otherwise lead to the formation of Cr 6+ The reactions in Equations 1 1 and 1 2 are simplified versions of the stoichiometric pyrolysis processes that lead to the formation of SiO 2 using two common silica pr ecursors, tetraethyl oxysilane (TEOS) and tetramethylsilane (TMS) respectively In reality, there are many possible intermediate reactions leading to the end products (Phadungsukanan et al. 2009; Herzler et al. 1997). Si(OC 2 H 5 ) 4 + Oxidant SiO 2 + 8CO 2 + 1 0H 2 O ( 1 1 ) Si(CH 3 ) 4 + Oxidant SiO 2 + 4CO 2 + 6H 2 O ( 1 2 ) Objective of Study The overall goal of this stu dy was to investigate the use of silica precursor s as shield gas additive s to limit exposure to hazardous metals in fumes from GTAW and GMAW The study was conducted in two phases. In phase I, T EOS was tested as an additive to GTAW shield gas to lower Cr 6+ concentration in fumes Cr 6+ nitrate and total Cr formation in fumes were measured and the ratio of Cr 6+ /Cr was compared for baseline weldi ng and welding with TEOS. In phase II, TMS was tested as an additive to GMAW shield gas because this welding process is more commonly used and responsible for far more emissions than GTAW The total shield gas flow rate was varied to investigate its effect on the efficacy of Cr 6+ control The flow rate of TMS carrier gas used also was varied in each of the total shield gas flow rates tested to determine the optimal feed rate Welding voltage was altered to study whether the precursor s were effective under d ifferent operating conditions. Success in this phase of the experiment was defined as a 90% reduction in the formation of Cr 6+ in welding fume particles

PAGE 23

23 Figure 1 1. Welding torches. A) GTAW torch and welding rod. B) GMAW torch. Figure 1 2. Formation o f welding particles with and without TMS addition.

PAGE 24

24 CHAPTER 2 EXPERIMENTAL METHODS Overview of Experimental Methods Two types of welding were used in this study. The first method discussed in this chapter is gas tungsten arc welding (GTAW) The methodology used for gas metal arc welding (GMAW) is presented afterwards. GTAW fumes were generated by a welder operated by a user with welding experience who used a specific mass of welding rod in each sample. The user maintained similar speeds when consuming weldi ng rod between samples. However, there was some variation between welding times in different samples. During GTAW sampling, the welding was performed outdoors with a hood placed directly above the welding. Ion chromatography and atomic emission spectrometr y (ICP AES) were used for characterizi ng how precursors impacted the amount of Cr 6+ emitted from GTAW. Tetraethyloxysilane ( TEOS, Acros Organics, 98%, CAS # 78 10 4) was tested during this phase of the experiment. GMAW fumes were generated automatically by a torch placed inside a chamber. In GMAW samples, welding time was held constant as well as the wire feed rate to consume the same mass of wire between samples. During GMAW sampling, the welding was performed indoors inside a closed chamber. The closed ch amber used in the GMAW experiments allowed for Cr 6+ mass to be compared directly between samples via ion chromatography, eliminating the need for measurement of total chromium. Tetramethylsilane (TMS, Acros Organics, 99.9%, CAS # 75 76 3) was tested during this phase of the experiment.

PAGE 25

25 GTAW Fume Generation and Sampling Figure 2 1 shows the schematic of the welding and sampling system. A Miller Maxstar 150 STL g as t ungsten a rc w elder was used for producing the welding aerosols. ER 308L stainless steel weldin g rods were used, which ha ve an average chromium content of 19.5% 22.0%. This rod also contain ed 9.0 11.0% nickel, 1.0 2.5% manganese, as well as 0.35 0.65% silica added to stabilize the iron in the weld. The remainder of the welding rod was iron. Forty fi ve grams of welding rod were consumed per sample. Mild steel base metal was used for sampling in order to avoid interference from chromium emitted from the base sheet metal. The welding hoses were modified to allow injection of TEOS shown in Figure 2 2 A Y fitt i ng was added to the welding torch where the hose joined the welding gun. Wires were connected to the end of the welder gas hose and the base o f the welding torch to maintain the electric current after the Y fitting was added. The gas hose leaving the shield gas cylinder was connected to a Y fitting to split the gas flow into two parts. The main gas flow was passed through the welder as in normal operation while the remainder of the shield gas was passed over a reservoir of liquid TEOS The TEOS fe ed rate was controlled by varying the flow rate of carrier gas that was passed over the liquid TEOS reservoir, maintained at 100 degrees Celsius ( C ) with a mass flow controller. The recommended minimum gas flow rate designated for this model of welder is 7.1 liters per minute ( Lpm ) which was adopted in this study. A 6 12 rectangular hood was placed directly above the mild stee l base metal on which the welding occurred. A 90 millimeter ( mm ) diameter glass fiber filter (Millipore APFA 090) was placed at the top of the hood and connected to a high volume (Hi Vol, 44 cfm) pump for collecting the aerosols.

PAGE 26

26 Ultra high purity argon ( Airgas, 99.999%) was used as the welding shield gas. This is the shield gas most commonly used in GTAW to maintain arc stability during welding operations. It acted as the needed carrier gas without requiring a secondary carrier gas. The vapor pressure of TEOS at room temperature is only 2 millimeters mercury ( mmHg ) which was too low to meet the theoretical amount of Si needed to produce a mono atomic layer on the welding aerosols, determined in Appendix A A water bath (100 C) was used to heat the reserv oir and raise the vapor pressure of the liquid TEOS to 86 mmHg (Stull 1947) in order to generate sufficient vapor with a reasonable shield gas flow rate passing over the liquid. GTAW Experimental Conditions Baseline samples using only argon as a shield g as were run to determine emissions of Cr 6+ relative to total chromium fume formation during normal welding operation. This scenario represent ed the standard welding method currently used in industrial environments. The baseline samples ensured that the sam pling time was long enough to provide concentrations needed for the analytical methods. For the experimental set, the TEOS carr ier gas feed rate was set at 26.3 % of the total 7.1 Lpm shield gas flow rate, resulting in a TEOS molar flow rate of 6.96 10 3 gr am mol per minute ( gmol/min ) Three TEOS samples and six baseline samples were collected. GTAW Sample Analysis Inductively coupled plasma with atomic emission spectroscopy (ICP AES, Perkin Elmer Plasma 3200) was used to measure total chromium. This analyti cal method does not differentiate between trivalent and hexavalent chromium. Therefore, IC ( Dionex ICS 1500 IonPac AS9 HC analytical column DS6 conductivity detector ) was used to measure the soluble hexavalent chromium species, chromate (CrO 4 2 ). IC was also

PAGE 27

27 used concurrently for nitrate analysis. Using these methods in conjunction provided the needed data to determine the reduction of Cr 6+ species using the TEOS additive. Prior to analysis, samples were extracted from filter media using techniques speci fic to the analytical method used. Sample extraction for IC followed a modified NIOSH Method 7604 (NIOSH 1994). Soluble Cr 6+ species were extracted using a 9 millimolar ( mM ) sodium carbonate solution and were placed in a water bath at 100 C for 1 hour. For ICP AES analysis, acid digestion was used, based on previously verified techniques (Karanasiou et al. 2005). Aerosols and silica coatings, along with the filter media, were dissolved using 9 milliliters ( mL ) of 65% nitric acid and 1 mL o f 48% hydrofluoric acid. The acidic extraction was placed in a heat block for five hours at 150 C. Afterwards, 1 mL of 10% boric acid was added to neutralize the free fluorine in the solution. Additionally, transmission electron microscopy (TEM, Model 201 0F, JEOL) was used to provide images that depicted the SiO 2 coating formed on fume particles. Specialty grids designed for TEM (Pelco, Lacey Carbon Type A, 300 mesh) were held directly in the fumes to collect particles for the TEM analysis. GMAW Fume Gener ation and Sampling Figure 2 3 shows the schematic setup of the welding fume generation and sampling system used during baseline and TMS welding. This system follow ed a base, A Lincoln PowerMIG 140C welder wa s used for producing welding fumes. ER 308L stainless steel welding wire wa s used, which has an average chromium content of 19.5% 22.0%. This wire also contain s 9.0 11.0% Ni, 1.0 2.5% Mn, and 0.35 0.65% Si. Mild steel base metal

PAGE 28

28 wa s used for sampling in order to avoid interference from chromium emitted by the base sheet metal. The welding torch hose was modified to allow injection of TMS. A Y fitting was inserte d to connect the torch and the gas hose. A longer wire guide was installed to ensure it was long enough to reach the contact tip. The trigger was removed and the wires inside the trigger were extended outside the chamber to allow remote operation of the we lding gun. The hose from the shield gas cylinder was connected to a T fitting to allow the gas flow to be separated into two parts. The main shield gas flow wa s passed through the welder as in normal operation while the remainder of the gas was used as TMS carrier gas The recommended minimum shield gas flow rate designated for this model of welder is 15 Lpm. An initial shield gas flow rate of 3 0 Lpm was chosen to produce quality welds after preliminary testing. The silica precursor feed rate wa s controlled by varying the flow rate of carrier gas that wa s passed over the liquid TMS reservoir, maintained at 0 C, with a mass flow controller. The carrier gas saturated with TMS vapor wa s then rejoined to the main shield gas flow prior to reaching the welding torch. The base metal wa s placed on a rotating turntable (MK Products Aircrafter T 25) to maintain a const ant weld speed. The turntable wa s placed in a hood in which the fumes we re generated and collected. A 90 mm diameter glass fiber filter (Millipore APFA 090) wa s placed at the top of the hood and connected to a high volume (Hi Vol 45 cubic feet per minute ( cfm ) ) pump for collecting the aerosols. A mixture of 75% argon and 25% CO 2 wa s used as the welding shield gas. The shield gas act ed as the needed carrier gas for TMS without requiring a secondary carrier gas. TMS is a highly volatile compound with a vap or pressure at room temperature of 598 mmHg (Aston et al 1941). Appendix B contains information used to

PAGE 29

29 calculate vapor pressure. The TMS reservoir wa s placed in an ice bath to lower the vapor pressure to about 270 mmHg. This wa s done to slow down evapora tion of the TMS while still producing enough TMS vapor to exceed the minimum theoretical TMS concentration needed to coat the aerosols in a mono atomic layer of SiO 2 See Appendix A for minimum TMS concentration calculation and Appendix C for calculations and rationale behind the actual operating TMS concentration produced in the carrier gas. GMAW Experimental Conditions The experimental conditions in this study are summarized in Table 2 1. Baseline samples using 75% Ar/25% CO 2 as a shield gas we re run to determine emissions of Cr 6+ during normal welding operation using short circuit metal transfer This scenario represent ed the standard welding method currently used in industrial environments. The baseline samples ensure d that the sampling time wa s long en ough to produce enough Cr 6+ to demonstrate 90% reduction without being limited by the detection limits of the analytical methods. Experimental set A wa s performed to determine the effect of TMS carrier gas flow rate on Cr 6+ formation. In experimental set B, high voltage was tested to account for different emission rates of pollutants under globular metal transfer mode In experimental set s C and D total shield gas flow rate was decreased to 25 and 20 Lpm respectively to test the effect of shield gas flo w rate on the efficacy of this technology The molar flow rates of TMS that correspond to the TMS carrier gas flow rates used in experimental sets A B, C, and D we re calculated by multiplying the flow rate of carrier

PAGE 30

30 gas by the concentration of TMS in sat urated carrier gas, found in Appendix C The values for molar flow rate s tested during GMAW sampling are presented in Table 2 2. GMAW Sample Analysis IC (Dionex ICS 1500 CS5A a nalytical c olumn, DS6 conductivity detector ) wa s used to measure the soluble he xavalent chromium species, chromate (CrO 4 2 ). Prior to analysis, samples we re extracted from filter media using techniques specific to the analytical method used. Sample extraction for IC follow ed a modified NIOSH Method 7604 (NIOSH 1994). Soluble Cr 6+ spe cies we re extracted using a 5 mM sodium carbonate/ 1 mM sodium bi carbonate solution and heated in a water bath to 100 C for one hour Additionally, TEM and energy dispersive x ray spectroscopy (EDS) ( Model 2010F, JEOL) were used to provide images that depict ed the SiO 2 coating formed on fume particles along with elemental analysis of those particles Specialty grids desi gned for TEM (Pelco, Lacey Carbon Type A, 300 mesh) were loaded with fume particles for analysis. A scanning mobility particle sizer (SMPS, TSI Model 3081 Long DMA) wa s used to obtain aerosol size distribution data between 10 nanometers ( nm ) and 515 nm W elding was performed for 10 seconds at which point the SMPS pump was turned on. After the SMPS completed its 135 second sampling run, the hi vol pump was turned on to clear the chamber of particles. A cascade impactor (U of W Mark III Source Test Cascade Impactor) was used to obtain particle size distribution dat a between 0.1 m and 10 m for 30 Lpm shield gas flow rate with 4.2% TMS Welding was performed in 10 second increments. After each 10 second increment, the cascade impactor pump was turned on for 135 seconds after which the hi vol pump was used to clear the chamber of particles. This process was

PAGE 31

31 repeated 15 times. A sampling flow rate of 20 Lpm through the impactor was used to produce desired cut sizes calculated in Appendix D Impaction plates we re coated with Apiezon grease to minimize particle bounce. The grease was mixed with toluene to facilitate spreading on impaction plates Impaction plates were painted with the mixture of grease and toluene and they were baked at 100 C for one hour followed by desiccation for 24 hours to remove toluene Plates were measured gravimetrically (Sartorius MC 210 S, +/ 10 g) before and after loading to measure fume mass at each cut size. Fume loading on filters used for IC analysis was al so measured gravimetrically for 30 Lpm shield gas flow rate samples. The glass fiber filter used for fume collection was weighed before and after loading to determine total fume mass. Figure 2 1 GTAW fume generation and sampling system.

PAGE 32

32 Figure 2 2. GTAW torch modified to allow TEOS addition. Wire Y fitting Gas Hose

PAGE 33

33 Figure 2 3 GMAW fume generation and sampling system. A) Without TMS. B) With TMS.

PAGE 34

34 Table 2 1 GMAW experimental conditions. Experimental c ondition Precursor carrier g as f low r ate (% of t otal s hield g as f low r ate) Total s hield g as f low r ate (Lpm) Voltage Silica p recursor Baseline 0 20, 25, 30 Low None A 1.4 4.2 3 0 Low TMS B 0, 4.2 35 High TMS C 5.0 25 Low TMS D 2.1 6.3 20 Low TMS Table 2 2 TMS molar flow rates used during GMAW sampling. Exper imental c ondition Precursor carrier g as f low r ate (% of t otal s hield g as f low r ate) Total s hield g as f low r ate (Lpm) TMS molar flow rate (mol/min) A 1.4, 2.8, 3.5, 4.2 3 0 0.007, 0.013, 0.017, 0.020 B 4.2 30 0.020 C 5.0 25 0.020 D 2.1, 4.2, 6.3 2 0 0.007, 0.013, 0.020

PAGE 35

35 CHAPTER 3 RESULTS AND DISCUSSI ON GAS TUNGSTEN ARC WEL DING ( GTAW ) Results Data pertinent to chromium, nitrogen oxides, and ozone formation are presented. As shown in Figure 3 1 the average Cr 6+ /Cr ratio decreased by 45% from the average baseline Cr 6+ /Cr ratio (from 0.170 to 0.094) when TEOS was added Figure 3 1 also shows the nitrate concentrations at baseline and TEOS sampling conditions. Nitrate is the end product of oxidation of atmospheric nitrogen, which is oxidized by the same reactive oxygen species as chromium. Therefore, it can be used as an indicator of the formation of reactive oxygen species and oxidation potential within welding fumes during the welding process (Dennis et al. 2002). Upon the addition of TEOS, nitrate concentration decreased from an average baseline value of 83 parts per million ( ppm ) to 39 ppm. This 53% decrease from baseline conditions is in line with the Cr 6+ reduction. This result demonstrates an oxidation potential of chromium within the welding f ume. Previous studies used other reducing agents as shield gas additives. For example, Dennis et al. (2002) used 3% ethene and 3% nitrogen oxide and obtained 42% and 45% reduction in Cr 6+ formation, respectively, which was slightly lower than what was achi eved in this study. There was large variation in the values shown in Figure 3 1 GTAW is a manual process in which inconsistencies when operating the welder can impact emission rates. This can contribute to errors in measuring welding fume emissions that a re present even in automatic systems (Serageldin and Reeves 2009). Figure 3 2 shows TEM images of fume aerosols collected. The metal vapors formed during welding quickly condense and form primary particles a few nanometers in

PAGE 36

36 diameter. These nanometer pri mary particles at very high number concentration undergo rapid coagul ation to form aggregates (Fig. 3 2 a). Fig. 3 2 b shows an aggregate particle with TEOS addition. In this image, the metallic aerosol is coated in a thin SiO 2 film. Metal aerosols appear ve ry dark on the images because the electrons passed through the sample to generate the image are less able to penetrate dense material like chromium. Silica is less dense than the metal aerosols and produces a lighter colored film around the dark aerosols. Figure 3 2 c is a closer look at higher magnification. A primary particle with distinct layering of different metals coated with SiO 2 is clearly seen. These images visually demonstrate that primary and aggregate metal particles formed during TIG welding wer e covered in a film of amorphous SiO 2 Discussion These results support previous studies that showed vapor phase sorbent precursors increase effective particle size, prevent formation of the ultrafine mode of the aerosols, and aid in collection efficiency of traditional control tech nologies (Owens and Biswas 1996 ; Biswas and Zachariah 1997). Therefore, the use of less obtrusive respirators or dust masks may become a more appealing option for workers exposed to welding fumes. The SiO 2 coating prevents or slo ws down the dissolution of chromium as nanometer SiO 2 takes weeks or more to dissolve in lung fluids (Reuzel et al. 1991). Since the soluble portions of welding fumes are the most toxic, decreased solubility decreases the effective toxicity of the aerosols (Antonini et al. 1999). This silica layer also makes silica precursor compounds more effective reducing agents than other gaseous species previously studied because oxidation of the metals is prevented long after formation of the aerosols due to the silic a layer.

PAGE 37

37 Cost Analysis The increase in operating costs when using TEOS as a shield gas additive is calculated in Appendix D Using lab quality TEOS as an additive at the feed rate tested in this study creates an additional cost of about $ 3.80 per hour of welding time Based on an estimate of consumption rate used in this study (3.6 grams of consumable used per minute), the cost of TEOS addition would be roughly $17.60 per kg of welding rod used. The cost can further be reduced using industrial quality chem ical. Figure 3 1 GTAW Cr 6+ and nitrate data before and after TEOS addition.

PAGE 38

38 Figure 3 2 TEM images of GTAW fume particles. A) Baseline a gglomerate particle without SiO 2 coating B) Agglomerate particle with SiO 2 coating when welding with TEO S. C) Prim ary particle with SiO 2 coating when welding with TEOS

PAGE 39

39 CHAPTER 4 RESULTS AND DISCUSSI ON GAS METAL ARC WELDIN G ( GMAW ) TMS Cr 6+ and Particle Size Distribution Results 30 Lpm Shield Gas Flow Rate A total shield gas flow rate of 30 Lpm was tested to dete rmine the effect of TMS addition on Cr 6+ formation in GMAW fumes. The results indicate adding TMS to shield gas reduced formation of Cr 6+ as seen in Figure 4 1. When 4.2% of the shield gas was used as TMS carrier gas, the mass of Cr 6+ was reduced to below the IC detection limit of roughly 4.5 g for all replicate samples. Since it was impossible to determine the exact mass for samples that were below detection, the mass for all non detectable samples was set equal to the detection limit. This was a reduction in Cr 6+ mass of at least 92.9% compa red to baseline conditions. The reduction in Cr 6+ mass achieved when using TMS as a shield gas additive exceeded the results seen in previous studies that incorporated reducing agents into shield gas or welding wire (Dennis et al. 1996; Dennis et al. 2002 ). The differences in Cr 6+ mass between TMS feed rates were statistically significant (p = 6.8 10 11 ANOVA single factor, = 0.05). SMPS particle size data using a total shield gas flow rate of 30 Lpm demonstrated that adding silica precursors to shield gas increased the particle size of fume particles, as seen in Figure 4 2. When welding witho ut TMS, the peak in the nanometer size range at about 20 nm was very large relative to the peak seen at 200 300 nm. The baseline mass median diameter ( MMD ) calculated from SMPS data was 380 nm which agreed very well with previous research that studied this characteristic (Jenkins et al. 2005).

PAGE 40

40 As increasing amounts of TMS were fed into the system, the metal nanoparticles were scavenged by SiO 2 agglomerates and the particle size distribution shifted towards larger particle sizes. The count geometric mean dia meter increased from 60 nm out of the nanometer range to 180 nm as TMS feed rate was increased (Appendix F). The increase in particle size observed when TMS was added agreed with previous studies that utilized silica precursors to control lead emissions fr om combustion systems. In those studies, adding silica precursors increased particle size out of the nanometer range (Owens and Biswas 1996a; Owens and Biswas 1996b). The mass size distribution of welding fume particles with 4.2% TMS additive collected usi ng a cascade impactor is displayed in Figure 4 3. There was evidence of overloading on the lower stages of the impactor. This might lead to re entrainment of some fume particles that would otherwise not have made it to the final filter. Nevertheless, the c ascade impactor data show the right tail of the particle size distribution that was too large to be measured using SMPS. The presence of particles between 1 10 m in diameter further indicated that SiO 2 particles were assisting in coagulation and increasing fume particle size because baseline welding fume particles formed through nucleation and coagulation in this size range were absent in previous research (Jenk ins et al. 2005). TEM images were obtained to determine whether a SiO 2 coating was present on fume particles when TMS was used as an additive. The images demonstrate that in some cases a coating of SiO 2 was formed but there were many particles without a d istinct coating, such as those seen in Figure 4 4. The image showed that some surface enrichment with lighter elements did occur which mirrors results of previous studies of

PAGE 41

41 welding fume particle structure that showed aerosol surface enrichment with silico n chlorine and fluorine (Maynard et al. 2004). Figure 4 5 shows an agglomerate particle that had a coating of SiO 2 that encapsulated the entire agglomerate rather than surface enrichment of individual primary particles. Figure 4 6 displays individual prima ry particles with thick SiO 2 coatings. This coating in Figures 4 5 and 4 6 was more significant than the surface enrichment seen on individual primary particles under baseline conditions in previous research (Maynard et al. 2004; Konarski et al. 2003). Ele mental analysis of the particles in Figure 4 6 was performed using EDS, shown in Table 4 1. About 17.5% of the particle mass was silicon which was much higher than surface enrichment with silicon (about 5%) in baseline welding fumes measured in previous r esearch (Minni et al. 1984). Figure 4 7 shows an amorphous agglomerate particle composed primarily of SiO 2 The TEM imageries obtained show that coating by SiO 2 was not uniform for all fume particles. Some particles were not coated, some primary particles were thickly coated with SiO 2 some agglomerate particles were covered in a thin layer of SiO 2 and some separate amorphous agglomerates composed mostly of SiO 2 were all present. Figures 4 5 and 4 6 indicate that silica precursor additives may be capable o f reducing toxicity of all of the metals in welding fume particles through the formation of a SiO 2 film on metal particles. This phenomenon is similar to that observed in previous research where metal nanoparticles formed in combustion systems can be trapp ed in a web of amorphous SiO 2 (Owens and Biswas 1996). Metal nanoparticles that are coated in SiO 2 are less likely to exhibit toxic effects because the SiO 2 coating could take weeks to dissolve in the respiratory system (Roelofs and Vogelsberger 2004). Thi s would provide

PAGE 42

42 exposure to the toxic metals occurs. In addition, there are agglomerates of metal fume particles with SiO 2 particles formed by intercoagulation, as shown i n Figure 4 8. Total fume mass was measured gravimetrically to evaluate the impact of TMS addition on GMAW fume generation rate, shown in Table 4 2. Although OSHA does not currently regulate total welding fume concentration, it is important to minimize tota l particulate concentration in occupational environments. Meanwhile, NIOSH does have a recommended exposure limit of the lowest feasible concentration of total welding fume. A technology that leads to a drastic increase in fume mass would present problems when trying to produce the lowest feasible concentration of fume particles during GMAW. 30 Lpm Shield Gas Flow Rate Using High Voltage Under high voltage, the welding process undergoes globular metal transfer mode. This process creates violent explosions of metal droplets leaving the tip of the welding wire. The baseline Cr 6+ formation under these conditions was 17.4% higher than using short circuit metal transfer mode. The addition of TMS did not have as dramatic an effect on Cr 6+ formation as under short circuit conditions, as seen in Figure 4 9. This may be due to the violence associated with globular metal transfer. Some molten spatter droplets are ejected far from the arc and may leave the area where TMS scavenges reactive oxygen species and collected during sampling, leading to an increase in measured Cr 6+ mass from fume formation. Although 90% reduction in Cr 6+ mass was not achieved during this type of welding, the 47% reduction was significant (p value <0.05) and could help reduce occupational exposu re. Globular metal transfer is not as widely used as short circuit and spray transfer modes due to the large amount of

PAGE 43

43 spatter and fumes produced during this operating condition. Spatter creates additional labor costs as the spatter particles that deposit on the base metal must be removed with a grinder. Therefore, the success of TMS addition using short circuit metal transfer was more important for reducing Cr 6+ emissions from welding sources than during globular transfer. 25 Lpm Shield Gas Flow Rate Total shield gas flow rate was decreased to 25 Lpm to reduce the consumption rate of shield gas and associated operating costs. There was a decrease in baseline Cr 6+ mass when the shield gas flow rate was decreased from 30 to 25 Lpm. Figure 4 10 shows that when 5.0% of the shield gas flow rate was used as TMS carrier gas, there was a significant decrease in Cr 6+ mass of about 40% with a p value far below 0.05 test, 2 tails, unequal variance). There was some white powder observed inside the head of t he welding torch after sampling. The reduction in Cr 6+ indicated that some of the TMS was surviving until reaching the area where reactive oxygen species were present. However, the decreased efficiency of Cr 6+ reduction coupled with the white powder inside the head of the welding torch indicated that some TMS was likely prematurely decomposing. The welding torch was heating during sampling and the TMS residence time in the hot torch was long enough for it to thermally decompose prior to entering the area wh ere fume particle formation was occurring. Compared with 30 Lpm shield gas flow rate data, this set of experiments shows that total shield gas flow rate impacted effectiveness of TMS and the goal of 90% reduction in Cr 6+ mass was not reached under 25 Lpm s hield gas flow rate. The SMPS particle size data for 25 Lpm shield gas flow rate followed the same pattern as that for 30 Lpm, as displayed in Figure 4 11. When welding without TMS, a

PAGE 44

44 large peak was seen at about 20 nm and a much smaller peak was present at 200 300 nm. As the TMS feed rate was increased, the peak in the nanometer range became smaller relative to the peak larger than 0.1 m. The number concentration for the nanometer range peak at 0.84% TMS carrier gas was higher than the number concentrati on for this peak under baseline conditions. However, the height of this peak relative to the peak at larger particle sizes still follows the same pattern as the 30 Lpm shield gas flow rate data. As TMS feed rate was increased, the peak in the nanometer siz e range shrank relative to the peak above 0.1 m. The aberration in absolute number concentrations was most likely due to variable fume emission rates encountered during welding. The shape of particle size distributions followed the same pattern of change despite the variation in number concentration. 20 Lpm Shield Gas Flow Rate The impact of TMS addition on Cr 6+ generation during GMAW using a total shield gas flow rate of 20 Lpm is shown in Figure 4 12. A large amount of white powder was observed inside the head of the welding torch after sampling. It likely resulted from thermal decomposition of TMS inside the head of the welding torch. When this happened, TMS did not function as a reducing agent or coat fume particles in a SiO 2 film because the TMS had already reacted before reaching the area where f ume particle formation was occurring. There was no statistically significant difference between the amount of Cr 6+ mass in the fumes regardless of the amount of TMS that was added (p = 0.229, ANOVA single factor, = 0.05). The addition of TMS led to an i ncrease in fume particle size just as it had using higher shield gas flow rates, as shown in Figure 4 13. During baseline welding, SMPS measurements showed a large number of particles with an average size of about

PAGE 45

45 20 nm. As increasing amounts of TMS were a dded, this peak decreased and eventually vanished. The particle size distribution shifted towards larger particles with the nanoparticles being replaced with particles larger than 0. m in diameter. These results and TEM imagery (Figure 4 8) are consistent with previous modeling research that found coarse particles such as the large amorphous SiO 2 particles formed here are capable of scavenging nanoparticles through intercoagulation (Lee and Wu 2005). It is also possible that some metal vapor is scavenged by SiO 2 particles, a result demonstrated in previous research studying the impact of sorbents on metals in combustion systems. When SiO 2 is formed in the area where metals are evapor ating, metal vapor in the combustion system is scavenged, which prevents nucleation. Nanomaterials can exhibit unique toxic effects based on their size (Biswas and Wu 2005; Jeng and Swanson 2006). Therefore, scavenging nanoparticles is a beneficial outcom e as it may reduce toxicity of welding fume particles by removing the toxicological mechanisms of these particles that result from their size. The potential reduction in toxicity based on elimination of nanoparticle pathways would occur regardless of wheth er or not metal speciation in the welding fumes changed. Effects of Shield Gas Flow Rate The MMD was between 0.35 0.4 m for baseline samples at all shield gas flow rates tested. The total baseline number concentration decreased and the geometric mean particle size increased as shield gas flow rate increased, as seen in Table 4 3, which agrees with observations reported i n previous studies (Dennis et al. 1997). It is possible that particle size increased with increasing shield gas flow rate because greater dispersion of metal vapor lowered the saturation ratio of metal vapor in the

PAGE 46

46 system. This could have caused the rate o f nucleation to decrease relative to the rate of condensation. As total shield gas flow rate decreased, the reduction of Cr 6+ in welding fumes observed after TMS addition decreased. However, the shift of particle size out of the nanometer range was observe d in all shield gas flow rates tested, as seen in Figure 4 14. This points to intercoagulation as the dominant mechanism in increasing particle size rather than SiO 2 coating on metal agglomerates. If SiO 2 coating was responsible for increased particle size Cr 6+ concentration would have decreased at all shield gas flow rates tested and TEM imagery would have shown consistent SiO 2 coating of fume particles. This means that Cr 6+ reduction was due primarily to TMS scavenging reactive oxygen species during pyro lysis rather than formation of SiO 2 coating on particles containing Cr 6+ Intercoagulation between nanosized fume particles and larger SiO 2 particles occurred whether TMS decomposed in the area where fume particle formation and oxidation occurred or whethe r TMS decomposed before reaching the area of fume particle formation. However, the change in particle size due to intercoagulation would not help meet OSHA standards for Cr 6+ concentration. In order to achieve Cr 6+ reduction using TMS, residence time of th e precursor vapor in the head of the welding torch must be short enough to ensure TMS does not heat enough to decompose before exiting the torch. TMS Cost Analysis The operating cost associated with adding TMS through delivery as a vapor from a liquid rese rvoir is calculated in Appendix D. The cost of TMS addition at the most effective flow rate tested was $78.96 per kg of welding electrode consumed. The cost for TMS addition was high due to the expense of purchasing high purity laboratory

PAGE 47

47 grade chemical. T he added cost can be reduced by purchasing lower purity industrial grade chemical. Regulatory Compliance The baseline welding samples produced at 30 Lpm shield gas flow rate created an average Cr 6+ concentration of 35.1 g/m 3 This concentration was over s even times higher than the OSHA permissible exposure limit of 5 g/m 3 for Cr 6+ (OSHA 2006) When 4.2% of the 30 Lpm shield gas flow rate was used to deliver TMS vapor, the resultant maximum Cr 6+ concentration was 2.5 g/m 3 The use of TMS brought the GMAW process into compliance with OSHA limits for Cr 6+ without the use of any local ventilation. Figure 4 1. Average Cr 6+ mass as a function of TMS carrier gas flow rate at 30 Lpm total shield gas flow rate.

PAGE 48

48 Figure 4 2 GMAW fume particle size distribution s at 30 Lpm shield gas flow rate.

PAGE 49

49 Figure 4 3 GMAW fume particle size distribution under 4.2% TMS flow rate measured with cascade impactor.

PAGE 50

50 Figure 4 4. GMAW fume agglomerate without SiO 2 coating.

PAGE 51

51 Figure 4 5. GMAW fume agglomerate with SiO 2 coati ng indicated by arrows. SiO 2 coating

PAGE 52

52 Figure 4 6. GMAW primary particles coated with SiO 2

PAGE 53

53 Figure 4 7. Amorphous SiO 2 agglomerate fume particle.

PAGE 54

54 Figure 4 8. Intercoagulation between metal particles and SiO 2 agglomerate. Table 4 1. Qualitative elemental anal ysis of Figure 4 6 fume particles. Element Weight % Si K 17.66 Cr K 28.06 Mn K 27.93 Fe K 26.35 SiO 2 Metal Fume Particles

PAGE 55

55 Ta ble 4 2. Fume generation rate using various TMS feed rates. TMS carrier gas flow rate (% of 30 Lpm shield gas flow rate) Average mass of fume parti cles standard deviation (mg) 0 29.515 3.612 1.4 27.728 0.735 2.8 26.82 0 1.551 3.5 25.406 3.288 4.2 22.189 5.307 Figure 4 9. Average Cr 6+ mass in GMAW fumes using globular metal transfer mode.

PAGE 56

56 Figure 4 10. Average Cr 6+ mass as a function of TMS carrier gas flow rate at 25 Lpm total shield gas flow rate.

PAGE 57

57 Figure 4 11. GMAW fume particle size distributions at 25 Lpm shield gas flow rate.

PAGE 58

58 Figure 4 12. Average Cr 6+ mass as a function of TMS carrier gas flow rate at 20 Lpm total shield gas f low rate.

PAGE 59

59 Figure 4 13. GMAW fume particle size distributions at 20 Lpm shield gas flow rate. Table 4 3. Baseline particle size characteristics at different shield gas flow rates. Total shield gas flow rate (Lpm) Total concentration (#/cm 3 ) Count g eometri c mean diameter (nm) 20 8.14E+05 40.7 25 7.10E+05 52.1 30 5.13E+05 60.9

PAGE 60

60 Figure 4 14. Particle size trends at different shield gas and TMS carrier gas flow rates.

PAGE 61

61 CHAPTER 5 CONCLUSIONS The GTAW study conducted as a part of this research showed that vapor phase sorbents can reduce Cr 6+ emissions from this type of welding. Nitrogen oxide concentration s were also reduced during GTAW. Some particles were coated in a film of amorphous SiO 2 This part of the study demonstrated the feasibility of using sili ca precursor vapor to produce in situ silica sorbents for control of Cr 6+ in welding fumes. Mechanical properties of the weld after TMS addition is an important piece of information in evaluating the effectiveness of any welding fume control technology tha t must be address ed prior to any application of vapor phase sorbents to GTAW shield gas. Silica precursors led to reduction in Cr 6+ in GMAW fumes of over 90%. The use of TMS as a shield gas additive brought Cr 6+ concentration down from a level that exceed ed the OSHA permissible exposure limit by over seven times to within regulatory limits. This result may help industry to keep pace with tightening OSHA limits for this hazardous air pollutant. The increase in particle size observed when TMS was added provi des another benefit for using these chemica ls. It may become more feasible to use less obtrusive filtration systems or personal protective equip ment such as dust masks instead of full face respirators to protect wor kers from welding fume particles after th eir size has been increased using silica precursors N ew technologies to scavenge nanoparticles from effluent gas streams will be required as nanomaterial manufacturing spreads. Silica precursors may provide a tool for dealing with this challenging problem TEM imagery demonstrated that some particles were coated in a film of amorphous

PAGE 62

62 SiO 2 All of the metals contained in the particles that were coated in SiO 2 were insulated regardless of speciation. More research is needed to determine how effectively sil ica precursors mask the toxic effects of nickel and manganese through encapsulation in amorphous silica. Research using less expensive silica precursors may lead to a more cost effective process that is more appealing to industry. Further research is neces sary to quantify the effect of silica precursors as shield gas additives on the mechanical properties of the materials being welded.

PAGE 63

63 APPENDIX A DETERMINATION OF THE ORETICAL MINIMUM PRECURSOR CONCENTRAT ION IN FUME PARTICLE FORMAT ION AREA In order to esti mate how much TMS must be added to coat the aerosols produced during GMAW, it was first necessary to estimate the volume and radius of a single SiO 2 molecule using Equation A 1. The molecular mass of SiO 2 divided by the density of SiO 2 was multiplied by Av 1 The diameter of an SiO 2 molecule was calculated using v 1 assuming molecules were spherical. I t was then necessary to calculate the total volume of aerosols produced during GMAW. A typical particle size distribution was obt ained from the literature (Zimmer and Biswas 2001). The next step was to calculate the change in volume that occu r r ed upon adding a mono atomic layer of SiO 2 by adding two diameters of SiO 2 molecules to the aerosol diameter assuming spherical particles F inally, the concentration required to produce the above change in volume was calculated using Equation A 2. ( A 1 ) ( A 2 ) C = Change in concentration required to produce change in volume N = Number concentration (#/c m 3 ) N av = 6.022E23 d p = d p0 + 2d SiO2 v 1 = 4.05E 20 cm 3 d SiO2 = 4.43E 8 cm The concentration calculated using this technique relies on the assumption that coating of uniform thickness occurs on all particles although in reality the coating is

PAGE 64

64 almost certai nly not homogeneous. The values used for the calculation of the minimum precursor concentration are presented in Table A 1 The total concentration derived in Table A 1 was 10 3 10 4 times lower than the concentration of TMS saturated carrier gas used during experiments. The rationale behind the actual operating TMS carrier gas feed rates can be found in Appendix C. Table A 1 Particle size data used to calculate minimum precursor concentration based on particle size distribution from literature (Zimmer and Biswas 2001) N (#/cm 3 ) d po (cm) AV v 1 )] [(d p ) 3 (d po ) 3 ] 3 ) 1.00E+05 1.80E 06 1.29E+01 9.03E 19 1.16E 17 4.48E+05 1.90E 06 5.77E+01 1.00E 18 5.80E 17 6.44E+06 2.00E 06 8.29E+02 1.11E 18 9.20E 16 1.72E+07 2.20E 06 2.22E+03 1.34E 18 2.96E 15 6.05E+07 2.40E 06 7.79E+03 1 .59E 18 1.24E 14 2.55E+08 2.60E 06 3.28E+04 1.86E 18 6.09E 14 1.25E+09 2.80E 06 1.61E+05 2.15E 18 3.46E 13 8.59E+09 3.20E 06 1.11E+06 2.80E 18 3.09E 12 2.31E+06 3.40E 06 2.98E+02 3.15E 18 9.38E 16 1.39E+07 3.60E 06 1.79E+03 3.53E 18 6.30E 15 1.68E+07 3.75E 06 2.16E+03 3.82E 18 8.26E 15 2.41E+07 3.95E 06 3.11E+03 4.24E 18 1.32E 14 4.70E+07 4.30E 06 6.06E+03 5.01E 18 3.04E 14 6.40E+07 4.60E 06 8.24E+03 5.73E 18 4.72E 14 9.88E+07 5.00E 06 1.27E+04 6.76E 18 8.59E 14 1.68E+08 5.40E 06 2.17E+04 7.87E 1 8 1.71E 13 3.09E+08 5.70E 06 3.98E+04 8.76E 18 3.49E 13 5.36E+08 6.00E 06 6.90E+04 9.70E 18 6.69E 13 7.29E+08 6.45E 06 9.39E+04 1.12E 17 1.05E 12 8.88E+08 6.80E 06 1.14E+05 1.24E 17 1.42E 12 1.54E+09 7.40E 06 1.99E+05 1.47E 17 2.93E 12 1.84E+09 7.90E 06 2.37E+05 1.68E 17 3.97E 12 4.10E+09 8.30E 06 5.27E+05 1.85E 17 9.75E 12 9.47E+09 9.00E 06 1.22E+06 2.17E 17 2.65E 11 1.08E+10 9.50E 06 1.39E+06 2.42E 17 3.36E 11 1.98E+10 1.00E 05 2.55E+06 2.68E 17 6.82E 11 2.22E+10 1.10E 05 2.85E+06 3.24E 17 9.24 E 11 3.81E+10 1.20E 05 4.90E+06 3.85E 17 1.89E 10 4.67E+10 1.30E 05 6.01E+06 4.52E 17 2.71E 10 9.89E+10 1.40E 05 1.27E+07 5.24E 17 6.67E 10 1.51E+11 1.55E 05 1.95E+07 6.42E 17 1.25E 09 2.62E+11 1.60E 05 3.38E+07 6.84E 17 2.31E 09

PAGE 65

65 Table A 1. Contin ued. N (#/cm 3 ) d po (cm) AV v 1 )] [(d p ) 3 (d po ) 3 ] 3 ) 1.78E+11 1.70E 05 2.29E+07 7.71E 17 1.77E 09 4.67E+10 1.80E 05 6.01E+06 8.65E 17 5.19E 10 2.77E+10 2.00E 05 3.56E+06 1.07E 16 3.80E 10 1.56E+10 2.10E 05 2.01E+06 1.18E 16 2.37E 10 1.08E+10 2.30E 05 1.39E+06 1 .41E 16 1.96E 10 8.30E+09 2.40E 05 1.07E+06 1.54E 16 1.64E 10 2.57E+09 3.00E 05 3.30E+05 2.40E 16 7.92E 11 2.57E+09 3.10E 05 3.30E+05 2.56E 16 8.45E 11 2.57E+09 3.20E 05 3.30E+05 2.73E 16 9.01E 11 1.84E+09 3.30E 05 2.37E+05 2.90E 16 6.86E 11 1.54E+09 3.40E 05 1.99E+05 3.08E 16 6.12E 11 1.29E+09 4.00E 05 1.66E+05 4.26E 16 7.08E 11 3.87E+08 4.20E 05 4.99E+04 4.69E 16 2.34E 11 1.91E+08 4.15E 05 2.46E+04 4.58E 16 1.13E 11 4.70E+07 4.40E 05 6.06E+03 5.15E 16 3.12E 12 4.83E+06 4.60E 05 6.21E+02 5.63E 1 6 3.50E 13 8.68627E 09 mol/cm 3

PAGE 66

66 APPENDIX B DETERMINATION OF TMS AND TEOS VAPO R PRESSURE The vapor pressures for TMS and TEOS were calculated using the Antoine equation (Equation B 1) The Antoine coefficients used in the calculations are pr esented in Table B 1 (Aston et al. 1941; Stull 1947) Table B 1 Antoine coefficients for TMS and TEOS. A B C TEOS 4.17312 1561.277 67.572 TMS 3.97703 1047.242 36.057 ( B 1 ) P = vapor pressure (bar) T = temperature (K) 273 K for T MS, 373 K for TEOS

PAGE 67

67 APPENDIX C DETERMINATION OF PRECURSOR CONCENTRATION IN SAT URATED CARRIER GAS The following calculations show how the actual TMS feed rates used during this research were calculated. Assuming t he carrier gas leaving the liquid reservoi r was saturated with TMS vapor t he ideal gas law (Equation C 1) was used to calculate the concentration of TMS vapor in the saturated carrier gas. ( C 1 ) ( vapor pressure of TMS vapor at 273 K ) (gas constan t) The TMS molar feed rates were calculated by multiplying the above concentration by the carrier gas flow rates passed over the liquid reservoir, found in Table 2 1 For example, when 4.2% of the 30 Lpm shield gas flow rate was used as TMS carrier gas the calculation would be performed as in Equation C 2. ( C 2 ) The same calculation was performed for TEOS using appropriate values for partial pressure and temperature. The c alculation was done for a temperature of 373 degrees Kelvin ( K ) where vapor pressure was 0.114 atmospheres ( atm ) The corresponding

PAGE 68

68 concentration of TEOS in saturated carrier gas was 3.713 10 6 mol TEOS/cm 3 The actual experimental concentration of TMS in the area of fume formation should be higher than the minimum theoretical concentration calculated in Appendix A because of dilution upon rejoining the main shield gas flow It further dec rease d upon leaving the welder into the ambient air where fume particle formation occurred These factors were taken into account when choosing the carrier gas flow rates to be tested. In previous research that used reducing agents to minimize oxidation of chromium, 4 4 gmol/min nitrogen oxide by Dennis et al. (2002) obtained 42% and 45% reduction in Cr 6+ formation, respectively. Based on this result, the current study used higher TMS feed rates to meet the desired goal of 90% reduction in Cr 6+ formation. The TMS carrier gas flow rates chosen were also influenced by preliminary research performed by Dr. Daniel P. Chang in which 1.6% of the total GMAW shield gas flow rate was used as TMS carrier gas. Cr 6+ was used more of the shield gas as TMS carrier gas than Dr. Chang in an attempt to exceed the 64% reduction attained by Dr. Chang. In experimental set A, t he specific TMS carrier gas percentages var ied from 1.4 4.2 % for a 3 0 Lpm shield gas low rate presented in Table 2 1 Experimental set C tested 5.0% of the 25 Lpm shield gas flow rate to see if a lower shield gas flow rate impacted successful delivery of TMS additive to the welding arc. The results show ed that there was some reduction of Cr 6+ mass but not enough to reach 90%. The shield gas flow rate was further decreased to 20 Lpm and the TMS was varied between 2.1 6.3% in experimental set D

PAGE 69

69 APPENDIX D COST ANALYSIS GTAW Operating Cost of TEOS Additi on The additional operating cost incurred based on the cost of laboratory grade TEOS is calculated below. Th e cost of 98% pure TEOS obtained from Acros Organics used in the calculation is $41.09 U.S. dollars ( USD ) per liter or 0.437 USD per gram TEOS The TEOS carrier gas flow rate used in this study was 1.874 Lpm. The concentration of TEOS in saturated carrier gas was 3.7137 10 3 mol TMS/L carrier gas and the molecular weight of TEOS is 208.33 grams per mole ( g/mol ) The added operating cost of using TEOS as a GTAW shield gas additive is $3.80 USD per hour of welding time as calculated using Equation D 1 GMAW Operating Cost of TMS Cost Addition The cost of using TMS as a GMAW shield gas additive is calculated below. The cost of laboratory grade 99.9% pure TMS obtained from Acros Organics is $80.8 0 USD per 100 g rams of chemical. The molecular weight of TMS is 88.23 g/mol. The concentration of TMS in saturated carrier gas is 1.589 x 10 2 mol TMS/L, as calculated in Appendix C. The calculation in Table D 1 was performed by multiplying the mass flow r ate of TMS by the cost per gram of TMS as shown in Equation D 2 The electrode consumption rate was held constant during experiments so the cost was also calculated in terms of USD per unit mass of electrode consumed. The electrode consumption used in the calculations was roughly 18 grams welding wire per minute of welding time. The

PAGE 70

70 cost of TMS addition was calculated at a number of TMS feed rates and a total shield gas flow rate of 30 Lpm. The condition that provided a >90% reduction in Cr 6+ would require an additional operating cost of $78.96 per kilogram of electrode consumed. This cost is high due to the cost of the high purity laboratory grade chemical used in the experiments. ( D 2 ) Table D 1 Cost analysis data for GMAW with TMS as a shield gas additive. Carrier Gas % of Total Shield Gas Flow Rate Total Shield Gas Flow Rate (Lpm) TMS Carrier Gas Flow Rate (Lpm) TMS Molar Flow Rate (gmol /min) TMS Mass Flow Rate (g/min) Cost of TMS Addition (USD/min welding time) Cost of TM S Addition (USD/kg electrode consumed) 1.4% 30 0.42 0.007 0.589 $0.48 $26.32 2.8% 30 0.84 0.013 1.178 $0.95 $52.64 3.5% 30 1.05 0.017 1.472 $1.19 $65.80 4.2% 30 1.26 0.020 1.766 $1.43 $78.96

PAGE 71

71 APPENDIX E RAW DATA FOR HEXAVAL ENT CHROMIU M Table E 1. Ra w Cr 6+ data for baseline and TMS samples. Sample ID a Shield Gas Flow Rate (Lpm) Precursor Carrier Gas Flow Rate (% of shield gas flow rate) Cr 6+ Mass 20_0_A 20 0 22.156 20_0_B 20 0 53.380 20_0_C 20 0 34.707 20_0_D 20 0 50.913 20_0_E 20 0 40.212 20_0_F 20 0 38.186 20_0_G 20 0 40.476 20_0_H 20 0 25.547 20_0_I 20 0 43.603 20_2.1_A 20 2.1 32.441 20_2.1_B 20 2.1 41.531 20_2.1_C 20 2.1 28.749 20_2.1_D 20 2.1 31.670 20_2.1_E 20 2.1 28.505 20_2.1_F 20 2.1 47.456 20_2.1_G 20 2.1 40.152 20_2.1_H 20 2.1 30.940 20_2.1_I 20 2.1 26.801 20_4.2_A 20 4.2 14.407 20_4.2_B 20 4.2 55.416 20_4.2_C 20 4.2 38.241 20_4.2_D 20 4.2 42.951 20_6.3_A 20 6.3 47.709 20_6.3_B 20 6.3 50.896 20_6.3_C 20 6.3 29.916 20_6.3_D 20 6.3 33.531 20_6.3_E 20 6.3 29.583 25_0_A 25 0 53.724 25_0_B 25 0 46.175 25_0_C 25 0 51.698 25_0_D 25 0 49.710 25_0_E 25 0 52.987 25_5.0_A 25 5.0 33.567 25_5.0_ B 25 5.0 25.6 94 25_5.0_ C 25 5.0 31.735 25_5.0_ D 25 5.0 29.240

PAGE 72

72 Table E 1. Continued. Sample a Shield Gas Flow Rate (Lpm) Precursor Carrier Gas Flow Rate (% of shield gas flow rate) Cr 6+ Mass 25_5.0_ E 25 5.0 30.098 30_0_A 30 0 66.026 30_0_B 30 0 63.212 30_ 0_C 30 0 61.831 30_0_D 30 0 62.521 30_1.4_A 30 1.4 37.588 30_1.4_B 30 1.4 43.546 30_1.4_C 30 1.4 37.916 30_2.8_A 30 2.8 32.463 30_2.8_B 30 2.8 32.354 30_2.8_C 30 2.8 32.517 30_3.5_A 30 3.5 32.071 30_3.5_B 30 3.5 28.719 30_3.5_C 30 3.5 14.732 30_ 3.5_D 30 3.5 11.342 30_3.5_E 30 3.5 20.627 30_3.5_F 30 3.5 23.518 30_3.5_G 30 3.5 26.347 30_3.5_H 30 3.5 28.506 30_4.2_A 30 4.2 4.480 30_4.2_B 30 4.2 4.480 30_4.2_C 30 4.2 4.480 30_4.2_D 30 4.2 4.480 ** 30_0_A_HV 30 0 73.834 ** 30_0_B_HV 30 0 80.86 5 ** 30_0_C_HV 30 0 73.903 ** 30_0_D_HV 30 0 77.938 ** 30_0_E_HV 30 0 65.612 ** 30_4.2_A_HV 30 4.2 38.564 ** 30_4.2_B_HV 30 4.2 38.363 ** 30_4.2_C_HV 30 4.2 32.551 ** 30_4.2_D_HV 30 4.2 39.856 ** 30_4.2_E_HV 30 4.2 48.170 *Sample ID format: ShieldGasFlowr ate(Lpm)_TMS%_Replicate ** HV d enotes high voltage

PAGE 73

73 APPENDIX F SCANNING MOBILITY PA RTICLE SIZER PARTICLE SIZE DISTRI BUTION DATA

PAGE 74

74 Table F 1. SMPS particle size data for baseline 20 Lpm shield gas flow rate Number particle s ize Diameter p article s ize Surf ace particle s ize Mass p article s ize Median (nm) 22.5 218.9 333.1 393.1 Mean (nm) 79.9 228.7 329.9 381.6 Geo. m ean (nm) 40.7 155.8 293.9 360.2 Mode (nm) 18.1 333.8 429.4 552.3 Geo. s t. d ev. 2.89 2.86 1.75 1.45 Total c onc. 8.14e+05(#/cm) 65.0(mm/cm ) 4.67e+10(nm/cm) 3.08e+03(g/m) Table F 2. SMPS particle size data for 20 Lpm shield gas flow rate with 1.05% TMS carrier gas. Number p article s ize Diameter p article s ize Surface p article s ize Mass p article s ize Median (nm) 145.9 290.4 369.1 420.3 Mean (nm) 178.1 293.6 361.4 404 .0 Geo. m ean (nm) 114.8 247.2 333.9 385.7 Mode (nm) 18.1 514 .0 552.3 552.3 Geo. s t. d ev. 2.85 1.97 1.55 1.39 Total c onc. 1.70e+06(#/cm) 303.4(mm/cm) 2.80e+11(nm/cm) 2.02e+04(g/m) Table F 3 SMPS particle size dat a for 20 Lpm shield gas flow rate with 2.1% TMS carrier gas. Number p article s ize Diameter p article s ize Surface p article s ize Mass p article s ize Median (nm) 233 .0 328.4 391.3 433.4 Mean (nm) 245.1 328.1 380.3 416 .0 Geo. m ean (nm) 192 .0 294.3 358.1 400 .3 Mode (nm) 289 .0 532.8 552.3 552.3 Geo. s t. d ev. 2.23 1.68 1.46 1.34 Total c onc. 1.91e+06(#/cm) 469.3(mm/cm) 4.84e+11(nm/cm) 3.68e+04(g/m)

PAGE 75

75 Table F 4 SMPS particle size data for 20 Lpm shield gas flow rate with 6.3% TMS carrier gas. Number p article s ize Diameter p article s ize Surface p article s ize Mass p article s ize Median (nm) 319.2 377.9 421.4 451.8 Mean (nm) 323.8 371.4 407.1 433.9 Geo. m ean (nm) 295.9 350.1 391.1 421.9 Mode (nm) 385.4 552.3 552.3 552.3 Geo. s t. d ev. 1.58 1.44 1.35 1 .28 Total c onc. 2.28e+06(#/cm) 739.1(mm/cm) 8.62e+11(nm/cm) 7.02e+04(g/m) Table F 5. SMPS particle size data for baseline 25 Lpm shield gas flow rate. Number p article s ize Diameter p article s ize Surface p article s ize Mass p article s ize Median (n m) 33.8 220.3 319.9 384.2 Mean (nm) 98.7 234.1 322.1 374.6 Geo. m ean (nm) 52.1 173.3 288.1 352.1 Mode (nm) 17.5 299.6 552.3 552.3 Geo. s t. d ev. 3.06 2.52 1.7 1.46 Total c onc. 7.10e+05(#/cm) 70.1(mm/cm) 5.16e+10(nm/cm) 3.32e+03(g/m) Table F 6. SMPS particle size data for 25 Lpm shield gas flow rate with 0.84% TMS carrier gas. Number p article s ize Diameter p article s ize Surface p article s ize Mass p article s ize Median (nm) 132.1 286.6 367 .0 417.3 Mean (nm) 165.5 291.2 359.1 401.8 Geo. m ean (n m) 97.3 243 .0 331.9 383.3 Mode (nm) 18.1 385.4 552.3 552.3 Geo. s t. d ev. 3.16 2.04 1.55 1.39 Total c onc. 1.71e+06(#/cm) 282.9(mm/cm) 2.59e+11(nm/cm) 1.86e+04(g/m)

PAGE 76

76 Table F 7. SMPS particle size data for 25 Lpm shield gas flow rate with 1.68% TM S carrier gas. Number p article s ize Diameter p article s ize Surface p article s ize Mass p article s ize Median (nm) 182.1 320.9 388.3 430.1 Mean (nm) 201.7 318.4 376.6 413.6 Geo. m ean (nm) 125.9 276.6 353 397.5 Mode (nm) 18.8 429.4 552.3 552.3 Geo. s t. d ev. 3.12 1.87 1.48 1.35 Total c onc. 1.18e+06(#/cm) 238.9(mm/cm) 2.39e+11(nm/cm) 1.80e+04(g/m) Table F 8. SMPS particle size data for 25 Lpm shield gas flow rate with 5.04% TMS carrier gas. Number p article s ize Diameter pa rticle s ize Surface p art icle s ize Mass p article s ize Median (nm) 266.3 346.6 400.9 438.3 Mean (nm) 274.1 344.2 389.3 421.4 Geo. m ean (nm) 226.1 315.8 369.7 407.1 Mode (nm) 333.8 532.8 552.3 552.3 Geo. s t. d ev. 2.07 1.58 1.41 1.32 Total c onc. 1.71e+06(#/cm) 467.5(mm/cm) 5. 06e+11(nm/cm) 3.94e+04(g/m) Table F 9 SMPS particle size data for baseline 30 Lpm shield gas flow rate. Number p article s ize Diameter p article s ize Surface p article s ize Mass particle s ize Median (nm) 59.2 211.8 312 383.5 Mean (nm) 108.6 231.6 31 6.6 372.2 Geo. m ean (nm) 60.9 176.9 281.9 348.1 Mode (nm) 17.5 259.5 461.4 552.3 Geo. s t. d ev. 3.01 2.36 1.71 1.48 Total c onc. 5.13e+05(#/cm) 55.7(mm/cm) 4.05e+10(nm/cm) 2.57e+03(g/m)

PAGE 77

77 Table F 10. SMPS particle size data for 30 Lpm shield gas flow rate with 0.7% TMS carrier gas. Number p article s ize Diameter p article s ize Surface p article s ize Mass p article s ize Median (nm) 110.3 250 343.8 403.5 Mean (nm) 148.1 262.8 339.4 388.8 Geo. m ean (nm) 94 213.4 307.8 367.6 Mode (nm) 18.8 358.7 532. 8 552.3 Geo. s t. d ev. 2.78 2.1 1.63 1.44 Total c onc. 5.42e+05(#/cm) 80.2(mm/cm) 6.62e+10(nm/cm) 4.49e+03(g/m) Table F 11. SMPS particle size data for 30 Lpm shield gas flow rate with 1.4% TMS carrier gas. Number p article s ize Diameter particle s ize Surface p article s ize Mass particle s ize Median (nm) 150.7 254.2 329.8 380.4 Mean (nm) 175.2 262.9 325.1 366.9 Geo. m ean (nm) 130.6 224.1 298.8 348.8 Mode (nm) 224.7 385.4 514 514 Geo. s t. d ev. 2.27 1.88 1.57 1.41 Total c onc. 6.70e+05(#/cm) 117 .4(mm/cm) 9.70e+10(nm/cm) 6.31e+03(g/m) Table F 12. SMPS particle size data for 30 Lpm shield gas flow rate with 4.2% TMS carrier gas. Number p article s ize Diameter particle s ize Surface particle s ize Mass particle s ize Median (nm) 201.1 287.3 349 .8 393.8 Mean (nm) 220.3 290.3 342.4 379.1 Geo. m ean (nm) 182.5 258.8 319.7 363 Mode (nm) 269 461.4 514 514 Geo. s t. d ev. 1.93 1.68 1.49 1.37 Total c onc. 4.96e+05(#/cm) 109.2(mm/cm) 9.96e+10(nm/cm) 6.82e+03(g/m)

PAGE 78

78 Table F 13 Raw SMPS particle s ize distribution data for 20 Lpm shield gas flow rate. Baseline 1.05% TMS 2.1% TMS 6.3% TMS Dia meter ( nm ) #/cm #/cm #/cm #/cm 12.6 5.24E+04 2.42E+04 2.56E+03 3.86E+02 13.1 9.51E+04 3.61E+04 2.07E+03 3.60E+02 13.6 1.22E+05 5.00E+04 1.77E+03 3.9 5E+02 14.1 1.35E+05 5.38E+04 3.64E+03 1.22E+02 14.6 1.73E+05 8.44E+04 6.58E+03 0.00E+00 15.1 2.94E+05 1.22E+05 1.95E+04 0.00E+00 15.7 6.88E+05 2.39E+05 4.57E+04 1.01E+02 16.3 1.65E+06 5.55E+05 1.05E+05 9.55E+01 16.8 2.70E+06 1.06E+06 1.64E+05 2.72E+0 2 17.5 3.56E+06 1.67E+06 2.17E+05 1.72E+02 18.1 3.74E+06 1.98E+06 2.69E+05 5.76E+02 18.8 3.45E+06 1.92E+06 3.29E+05 5.51E+02 19.5 2.88E+06 1.66E+06 3.79E+05 1.33E+03 20.2 2.36E+06 1.38E+06 4.38E+05 1.10E+03 20.9 1.89E+06 1.17E+06 4.28E+05 2.47E+03 2 1.7 1.54E+06 1.01E+06 4.14E+05 1.75E+03 22.5 1.31E+06 9.13E+05 3.83E+05 2.38E+03 23.3 1.14E+06 8.35E+05 3.32E+05 2.30E+03 24.1 9.82E+05 7.51E+05 3.12E+05 2.51E+03 25 .0 8.61E+05 6.77E+05 3.01E+05 2.00E+03 25.9 7.41E+05 6.03E+05 2.70E+05 3.19E+03 26.9 6.18E+05 5.44E+05 2.56E+05 2.36E+03 27.9 5.16E+05 4.92E+05 2.39E+05 3.14E+03 28.9 4.51E+05 4.66E+05 2.26E+05 4.42E+03 30 .0 3.83E+05 4.44E+05 2.18E+05 4.28E+03 31.1 3.40E+05 4.33E+05 2.13E+05 4.70E+03 32.2 3.08E+05 4.34E+05 2.16E+05 6.82E+03 33.4 2.74 E+05 4.32E+05 2.16E+05 7.91E+03 34.6 2.56E+05 4.63E+05 2.24E+05 8.68E+03 35.9 2.39E+05 4.83E+05 2.34E+05 1.09E+04 37.2 2.25E+05 5.16E+05 2.56E+05 1.41E+04 38.5 2.15E+05 5.61E+05 2.82E+05 1.42E+04 40 .0 2.03E+05 5.92E+05 3.06E+05 1.71E+04 41.4 2.02E+05 6.17E+05 3.28E+05 2.07E+04 42.9 1.98E+05 6.51E+05 3.53E+05 2.55E+04 44.5 1.96E+05 6.73E+05 3.77E+05 2.78E+04 46.1 1.91E+05 6.98E+05 4.17E+05 3.26E+04

PAGE 79

79 Table F 13. Continued. Baseline 1.05% TMS 2.1% TMS 6.3% TMS Dia meter ( nm ) #/cm #/cm #/cm #/cm 47.8 1.88E+05 7.32E+05 4.35E+05 3.92E+04 49.6 1.86E+05 7.71E+05 4.83E+05 4.49E+04 51.4 1.83E+05 7.87E+05 5.03E+05 5.09E+04 53.3 1.84E+05 8.11E+05 5.40E+05 5.78E+04 55.2 1.85E+05 8.35E+05 5.72E+05 6.76E+04 57.3 1.80E+05 8.42E+05 6.03E+05 7.76E+ 04 59.4 1.76E+05 8.57E+05 6.15E+05 8.63E+04 61.5 1.82E+05 8.53E+05 6.45E+05 1.00E+05 63.8 1.84E+05 8.70E+05 6.67E+05 1.14E+05 66.1 1.86E+05 8.75E+05 6.91E+05 1.29E+05 68.5 1.82E+05 8.91E+05 7.13E+05 1.43E+05 71 .0 1.83E+05 9.01E+05 7.36E+05 1.65E+05 73.7 1.86E+05 8.89E+05 7.50E+05 1.82E+05 76.4 1.88E+05 8.94E+05 7.94E+05 2.03E+05 79.1 1.86E+05 8.96E+05 8.09E+05 2.31E+05 82 .0 1.90E+05 8.84E+05 8.19E+05 2.54E+05 85.1 1.89E+05 8.93E+05 8.34E+05 2.79E+05 88.2 1.95E+05 9.01E+05 8.46E+05 3.06E+05 91.4 2.01E+05 9.05E+05 8.61E+05 3.34E+05 94.7 2.08E+05 9.19E+05 8.81E+05 3.72E+05 98.2 2.13E+05 9.25E+05 8.94E+05 4.02E+05 101.8 2.19E+05 9.31E+05 9.16E+05 4.52E+05 105.5 2.30E+05 9.60E+05 9.35E+05 4.95E+05 109.4 2.35E+05 9.79E+05 9.75E+05 5.54E+05 113.4 2.45E+05 1.00E+06 1.02E+06 5.98E+05 117.6 2.56E+05 1.02E+06 1.05E+06 6.44E+05 121.9 2.67E+05 1.05E+06 1.10E+06 6.95E+05 126.3 2.78E+05 1.08E+06 1.17E+06 7.59E+05 131 .0 2.85E+05 1.11E+06 1.21E+06 8.28E+05 135.8 3.02E+05 1.15E+06 1.28E+06 9.11E+05 140 .7 3.08E+05 1.20E+06 1.36E+06 1.00E+06 145.9 3.22E+05 1.25E+06 1.42E+06 1.09E+06 151.2 3.26E+05 1.29E+06 1.51E+06 1.18E+06 156.8 3.38E+05 1.33E+06 1.59E+06 1.32E+06 162.5 3.40E+05 1.38E+06 1.69E+06 1.46E+06 168.5 3.39E+05 1.42E+06 1.78E+06 1.60E+06 1 74.7 3.47E+05 1.48E+06 1.88E+06 1.79E+06

PAGE 80

80 Table F 13. Continued. Baseline 1.05% TMS 2.1% TMS 6.3% TMS Dia meter ( nm ) #/cm #/cm #/cm #/cm 181.1 3.52E+05 1.52E+06 1.97E+06 1.98E+06 187.7 3.59E+05 1.57E+06 2.06E+06 2.15E+06 194.6 3.57E+05 1.60E +06 2.15E+06 2.33E+06 201.7 3.71E+05 1.64E+06 2.24E+06 2.52E+06 209.1 3.62E+05 1.66E+06 2.33E+06 2.68E+06 216.7 3.59E+05 1.67E+06 2.42E+06 2.88E+06 224.7 3.52E+05 1.67E+06 2.49E+06 3.07E+06 232.9 3.44E+05 1.67E+06 2.55E+06 3.24E+06 241.4 3.39E+05 1.6 8E+06 2.64E+06 3.44E+06 250.3 3.25E+05 1.68E+06 2.69E+06 3.59E+06 259.5 3.27E+05 1.70E+06 2.73E+06 3.78E+06 269 .0 3.12E+05 1.71E+06 2.77E+06 3.95E+06 278.8 3.00E+05 1.71E+06 2.79E+06 4.06E+06 289 .0 2.94E+05 1.71E+06 2.82E+06 4.23E+06 299.6 2.81E+05 1 .71E+06 2.80E+06 4.35E+06 310.6 2.76E+05 1.68E+06 2.77E+06 4.41E+06 322 .0 2.68E+05 1.63E+06 2.76E+06 4.52E+06 333.8 2.64E+05 1.58E+06 2.76E+06 4.62E+06 346 .0 2.48E+05 1.53E+06 2.68E+06 4.68E+06 358.7 2.42E+05 1.49E+06 2.63E+06 4.71E+06 371.8 2.34E+05 1.44E+06 2.58E+06 4.76E+06 385.4 2.18E+05 1.38E+06 2.52E+06 4.80E+06 399.5 2.09E+05 1.32E+06 2.46E+06 4.73E+06 414.2 1.95E+05 1.27E+06 2.41E+06 4.72E+06 429.4 1.84E+05 1.23E+06 2.35E+06 4.76E+06 445.1 1.69E+05 1.19E+06 2.30E+06 4.71E+06 461.4 1.56E+ 05 1.15E+06 2.25E+06 4.67E+06 478.3 1.48E+05 1.10E+06 2.19E+06 4.65E+06 495.8 1.36E+05 1.07E+06 2.13E+06 4.56E+06 514 .0 1.27E+05 1.05E+06 2.06E+06 4.52E+06 Table F 14. Raw SMPS particle size distribution data for 25 Lpm shield gas flow rate. Baselin e 0.84% TMS 1.68% TMS 5.04% TMS Diameter ( nm ) #/cm #/cm #/cm #/cm 12.6 2.21E+04 5.05E+03 3.89E+03 0.00E+00 13.1 5.46E+04 7.75E+03 3.28E+03 0.00E+00 13.6 9.45E+04 9.60E+03 3.51E+03 2.63E+02

PAGE 81

81 Table F 14. Continued. Baseline 0.84% TMS 1.68% TMS 5.04% TMS Diameter ( nm ) #/cm #/cm #/cm #/cm 14.1 1.25E+05 1.72E+04 7.98E+03 0.00E+00 14.6 1.51E+05 2.46E+04 1.81E+04 1.49E+02 15.1 2.65E+05 6.06E+04 7.08E+04 1.63E+03 15.7 7.01E+05 2.90E+05 3.39E+05 4.52E+03 16.3 1.52E+06 1.06E+06 6.99E+0 5 2.75E+04 16.8 2.39E+06 2.34E+06 1.17E+06 1.19E+05 17.5 3.01E+06 3.54E+06 1.76E+06 2.66E+05 18.1 2.91E+06 3.80E+06 2.07E+06 3.68E+05 18.8 2.43E+06 3.58E+06 2.11E+06 4.13E+05 19.5 1.81E+06 3.14E+06 1.83E+06 3.90E+05 20.2 1.31E+06 2.66E+06 1.49E+06 3. 50E+05 20.9 9.66E+05 2.15E+06 1.16E+06 3.04E+05 21.7 7.47E+05 1.70E+06 8.69E+05 2.72E+05 22.5 6.29E+05 1.33E+06 6.46E+05 2.34E+05 23.3 5.25E+05 1.03E+06 5.12E+05 1.99E+05 24.1 4.50E+05 7.81E+05 3.83E+05 1.74E+05 25 .0 3.99E+05 6.14E+05 3.04E+05 1.53E+ 05 25.9 3.57E+05 5.01E+05 2.44E+05 1.38E+05 26.9 3.35E+05 4.31E+05 2.09E+05 1.15E+05 27.9 3.14E+05 3.76E+05 1.86E+05 1.06E+05 28.9 2.86E+05 3.48E+05 1.62E+05 1.02E+05 30 .0 2.64E+05 3.12E+05 1.48E+05 9.70E+04 31.1 2.50E+05 2.90E+05 1.32E+05 8.81E+04 32.2 2.42E+05 2.79E+05 1.30E+05 9.32E+04 33.4 2.28E+05 2.74E+05 1.22E+05 8.94E+04 34.6 2.27E+05 2.74E+05 1.22E+05 9.46E+04 35.9 2.25E+05 2.83E+05 1.27E+05 9.93E+04 37.2 2.30E+05 2.95E+05 1.27E+05 1.04E+05 38.5 2.29E+05 3.11E+05 1.31E+05 1.08E+05 40 .0 2.31E+05 3.31E+05 1.35E+05 1.18E+05 41.4 2.28E+05 3.43E+05 1.38E+05 1.30E+05 42.9 2.34E+05 3.55E+05 1.48E+05 1.39E+05 44.5 2.38E+05 3.80E+05 1.60E+05 1.47E+05 46.1 2.37E+05 4.04E+05 1.65E+05 1.58E+05 47.8 2.41E+05 4.26E+05 1.80E+05 1.74E+05 49.6 2.3 7E+05 4.43E+05 1.93E+05 1.94E+05 51.4 2.35E+05 4.71E+05 2.11E+05 2.03E+05

PAGE 82

82 Table F 14. Continued. Baseline 0.84% TMS 1.68% TMS 5.04% TMS Dia meter ( nm ) #/cm #/cm #/cm #/cm 53.3 2.36E+05 4.81E+05 2.22E+05 2.16E+05 55.2 2.40E+05 4.98E+05 2.36E +05 2.38E+05 57.3 2.47E+05 5.19E+05 2.43E+05 2.55E+05 59.4 2.38E+05 5.43E+05 2.58E+05 2.81E+05 61.5 2.36E+05 5.56E+05 2.72E+05 2.96E+05 63.8 2.34E+05 5.79E+05 2.82E+05 3.07E+05 66.1 2.35E+05 5.94E+05 2.92E+05 3.27E+05 68.5 2.40E+05 6.15E+05 3.11E+05 3.55E+05 71 .0 2.39E+05 6.36E+05 3.22E+05 3.67E+05 73.7 2.40E+05 6.57E+05 3.38E+05 3.97E+05 76.4 2.43E+05 6.76E+05 3.47E+05 4.22E+05 79.1 2.50E+05 7.02E+05 3.78E+05 4.40E+05 82 .0 2.53E+05 7.13E+05 3.91E+05 4.55E+05 85.1 2.61E+05 7.46E+05 4.15E+05 4.84 E+05 88.2 2.71E+05 7.76E+05 4.48E+05 5.05E+05 91.4 2.70E+05 7.86E+05 4.62E+05 5.26E+05 94.7 2.82E+05 8.17E+05 4.86E+05 5.45E+05 98.2 2.85E+05 8.42E+05 5.06E+05 5.83E+05 101.8 2.93E+05 8.79E+05 5.37E+05 6.11E+05 105.5 3.01E+05 9.22E+05 5.63E+05 6.39E+ 05 109.4 3.10E+05 9.35E+05 5.87E+05 6.84E+05 113.4 3.23E+05 9.78E+05 6.17E+05 7.31E+05 117.6 3.32E+05 1.01E+06 6.46E+05 7.62E+05 121.9 3.49E+05 1.05E+06 6.83E+05 8.14E+05 126.3 3.68E+05 1.10E+06 7.15E+05 8.72E+05 131 .0 3.78E+05 1.14E+06 7.35E+05 9.33 E+05 135.8 3.80E+05 1.17E+06 7.69E+05 1.00E+06 140.7 3.92E+05 1.21E+06 8.04E+05 1.08E+06 145.9 4.12E+05 1.26E+06 8.63E+05 1.17E+06 151.2 4.18E+05 1.30E+06 8.96E+05 1.26E+06 156.8 4.24E+05 1.35E+06 9.35E+05 1.35E+06 162.5 4.29E+05 1.39E+06 9.86E+05 1. 45E+06 168.5 4.32E+05 1.44E+06 1.01E+06 1.55E+06 174.7 4.38E+05 1.46E+06 1.04E+06 1.66E+06 181.1 4.42E+05 1.47E+06 1.07E+06 1.76E+06 187.7 4.44E+05 1.51E+06 1.09E+06 1.85E+06 194.6 4.48E+05 1.53E+06 1.11E+06 1.94E+06

PAGE 83

83 Table F 14. Continued. Basel ine 0.84% TMS 1.68% TMS 5.04% TMS Dia meter ( nm ) #/cm #/cm #/cm #/cm 201.7 4.38E+05 1.55E+06 1.15E+06 2.04E+06 209.1 4.34E+05 1.55E+06 1.17E+06 2.14E+06 216.7 4.23E+05 1.56E+06 1.21E+06 2.24E+06 224.7 4.15E+05 1.58E+06 1.24E+06 2.34E+06 232.9 4 .07E+05 1.61E+06 1.27E+06 2.46E+06 241.4 4.05E+05 1.60E+06 1.29E+06 2.54E+06 250.3 3.88E+05 1.62E+06 1.30E+06 2.63E+06 259.5 3.71E+05 1.62E+06 1.32E+06 2.68E+06 269 .0 3.57E+05 1.64E+06 1.33E+06 2.74E+06 278.8 3.41E+05 1.61E+06 1.33E+06 2.77E+06 289 .0 3.32E+05 1.60E+06 1.35E+06 2.79E+06 299.6 3.27E+05 1.55E+06 1.36E+06 2.83E+06 310.6 3.12E+05 1.52E+06 1.34E+06 2.82E+06 322 .0 2.92E+05 1.46E+06 1.34E+06 2.84E+06 333.8 2.77E+05 1.45E+06 1.34E+06 2.86E+06 346 .0 2.69E+05 1.40E+06 1.32E+06 2.83E+06 358 .7 2.56E+05 1.36E+06 1.32E+06 2.82E+06 371.8 2.40E+05 1.35E+06 1.30E+06 2.78E+06 385.4 2.25E+05 1.31E+06 1.27E+06 2.75E+06 399.5 2.09E+05 1.24E+06 1.25E+06 2.70E+06 414.2 2.03E+05 1.20E+06 1.22E+06 2.66E+06 429.4 1.85E+05 1.15E+06 1.19E+06 2.59E+06 4 45.1 1.74E+05 1.09E+06 1.14E+06 2.51E+06 461.4 1.61E+05 1.05E+06 1.10E+06 2.46E+06 478.3 1.50E+05 9.90E+05 1.05E+06 2.40E+06 495.8 1.42E+05 9.76E+05 1.02E+06 2.35E+06 514 .0 1.25E+05 9.30E+05 9.69E+05 2.29E+06 Table F 1 5 Raw SMPS particle size distri bution data for 30 Lpm shield gas flow rate. Baseline 0.7% TMS 1.4% TMS 4.2% TMS Diameter ( nm ) #/cm #/cm #/cm #/cm 12.6 6.84E+03 3.18E+03 1.10E+03 1.56E+03 13.1 1.39E+04 4.78E+03 1.18E+03 5.71E+02 13.6 3.09E+04 2.96E+03 1.21E+03 5.27E+02 14.1 4.81E+04 2.73E+03 2.72E+03 7.17E+02 14.6 8.28E+04 2.92E+03 2.95E+03 1.50E+03 15.1 1.81E+05 5.62E+03 3.52E+03 1.93E+03

PAGE 84

84 Table F 15. Continued. Baseline 0.7% TMS 1.4% TMS 4.2% TMS Dia meter ( nm ) #/cm #/cm #/cm #/cm 15.7 5.83E+05 1.14E+04 7.72 E+03 2.76E+03 16.3 1.23E+06 5.09E+04 1.87E+04 4.39E+03 16.8 1.58E+06 1.88E+05 3.05E+04 8.03E+03 17.5 1.66E+06 4.49E+05 4.01E+04 1.19E+04 18.1 1.44E+06 6.10E+05 4.68E+04 1.48E+04 18.8 1.18E+06 6.82E+05 4.47E+04 1.48E+04 19.5 9.33E+05 6.56E+05 5.57E+04 1.34E+04 20.2 7.29E+05 6.33E+05 6.81E+04 1.29E+04 20.9 5.84E+05 5.92E+05 8.89E+04 1.38E+04 21.7 4.71E+05 5.20E+05 1.13E+05 1.26E+04 22.5 3.82E+05 4.61E+05 1.29E+05 1.27E+04 23.3 3.30E+05 4.02E+05 1.43E+05 1.00E+04 24.1 2.99E+05 3.66E+05 1.47E+05 9.1 3E+03 25 .0 2.63E+05 3.36E+05 1.46E+05 9.27E+03 25.9 2.50E+05 3.21E+05 1.52E+05 9.97E+03 26.9 2.21E+05 3.10E+05 1.54E+05 1.16E+04 27.9 2.09E+05 2.88E+05 1.64E+05 1.35E+04 28.9 2.02E+05 2.80E+05 1.80E+05 1.62E+04 30 .0 1.90E+05 2.57E+05 2.01E+05 1.63E+0 4 31.1 1.80E+05 2.49E+05 2.23E+05 1.78E+04 32.2 1.80E+05 2.42E+05 2.40E+05 2.37E+04 33.4 1.77E+05 2.37E+05 2.60E+05 2.69E+04 34.6 1.73E+05 2.35E+05 2.84E+05 3.14E+04 35.9 1.71E+05 2.34E+05 2.99E+05 4.00E+04 37.2 1.72E+05 2.34E+05 3.27E+05 4.54E+04 3 8.5 1.74E+05 2.35E+05 3.45E+05 5.31E+04 40 .0 1.72E+05 2.42E+05 3.62E+05 6.03E+04 41.4 1.76E+05 2.38E+05 3.75E+05 7.09E+04 42.9 1.76E+05 2.46E+05 4.04E+05 8.04E+04 44.5 1.77E+05 2.45E+05 4.28E+05 9.02E+04 46.1 1.78E+05 2.50E+05 4.42E+05 1.02E+05 47.8 1.79E+05 2.60E+05 4.67E+05 1.16E+05 49.6 1.82E+05 2.63E+05 4.75E+05 1.29E+05 51.4 1.88E+05 2.67E+05 4.86E+05 1.43E+05 53.3 1.89E+05 2.73E+05 4.94E+05 1.64E+05 55.2 1.91E+05 2.73E+05 4.95E+05 1.72E+05 57.3 1.97E+05 2.73E+05 4.98E+05 1.87E+05

PAGE 85

85 Table F 1 5. Continued. Baseline 0.7% TMS 1.4% TMS 4.2% TMS Dia meter ( nm ) #/cm #/cm #/cm #/cm 59.4 2.00E+05 2.77E+05 4.95E+05 1.96E+05 61.5 2.06E+05 2.76E+05 4.93E+05 2.06E+05 63.8 2.05E+05 2.84E+05 4.97E+05 2.21E+05 66.1 2.11E+05 2.87E+05 4.89E+05 2.36E+05 68.5 2.19E+05 2.92E+05 4.81E+05 2.49E+05 71 .0 2.28E+05 2.99E+05 4.75E+05 2.68E+05 73.7 2.30E+05 2.96E+05 4.76E+05 2.79E+05 76.4 2.37E+05 3.01E+05 4.77E+05 2.99E+05 79.1 2.53E+05 3.04E+05 4.70E+05 3.10E+05 82 .0 2.56E+05 3.01E+05 4.50E+05 3.22 E+05 85.1 2.64E+05 2.97E+05 4.40E+05 3.38E+05 88.2 2.71E+05 2.97E+05 4.26E+05 3.54E+05 91.4 2.74E+05 3.04E+05 4.16E+05 3.66E+05 94.7 2.80E+05 3.15E+05 4.07E+05 3.72E+05 98.2 2.87E+05 3.19E+05 4.16E+05 3.97E+05 101.8 2.94E+05 3.37E+05 4.28E+05 4.01E+0 5 105.5 3.01E+05 3.44E+05 4.33E+05 4.24E+05 109.4 3.08E+05 3.71E+05 4.36E+05 4.41E+05 113.4 3.16E+05 3.79E+05 4.46E+05 4.49E+05 117.6 3.23E+05 3.93E+05 4.58E+05 4.51E+05 121.9 3.30E+05 4.17E+05 4.68E+05 4.59E+05 126.3 3.37E+05 4.24E+05 4.86E+05 4.63E +05 131 .0 3.42E+05 4.37E+05 5.17E+05 4.64E+05 135.8 3.63E+05 4.50E+05 5.47E+05 4.74E+05 140.7 3.69E+05 4.61E+05 5.83E+05 4.87E+05 145.9 3.73E+05 4.68E+05 6.33E+05 4.97E+05 151.2 3.83E+05 4.69E+05 6.60E+05 5.06E+05 156.8 3.82E+05 4.72E+05 6.90E+05 5.0 7E+05 162.5 3.86E+05 4.65E+05 7.10E+05 5.31E+05 168.5 3.85E+05 4.73E+05 7.38E+05 5.37E+05 174.7 3.77E+05 4.67E+05 7.50E+05 5.60E+05 181.1 3.77E+05 4.63E+05 7.49E+05 5.84E+05 187.7 3.69E+05 4.60E+05 7.57E+05 5.84E+05 194.6 3.62E+05 4.59E+05 7.61E+05 6 .09E+05 201.7 3.54E+05 4.65E+05 7.60E+05 6.25E+05 209.1 3.42E+05 4.55E+05 7.61E+05 6.38E+05 216.7 3.33E+05 4.66E+05 7.71E+05 6.50E+05

PAGE 86

86 Table F 15. Continued. Baseline 0.7% TMS 1.4% TMS 4.2% TMS Dia meter ( nm ) #/cm #/cm #/cm #/cm 224.7 3.31E +05 4.61E+05 7.76E+05 6.63E+05 232.9 3.15E+05 4.62E+05 7.70E+05 6.86E+05 241.4 3.09E+05 4.62E+05 7.71E+05 6.90E+05 250.3 2.98E+05 4.53E+05 7.48E+05 6.99E+05 259.5 2.90E+05 4.45E+05 7.32E+05 7.00E+05 269 .0 2.72E+05 4.23E+05 7.08E+05 7.06E+05 278.8 2.5 8E+05 4.16E+05 6.82E+05 6.98E+05 289 .0 2.54E+05 3.99E+05 6.59E+05 6.92E+05 299.6 2.44E+05 3.97E+05 6.50E+05 6.79E+05 310.6 2.35E+05 3.77E+05 6.19E+05 6.72E+05 322 .0 2.19E+05 3.70E+05 5.93E+05 6.58E+05 333.8 2.06E+05 3.53E+05 5.80E+05 6.41E+05 346 .0 1 .94E+05 3.49E+05 5.58E+05 6.04E+05 358.7 1.83E+05 3.39E+05 5.41E+05 5.83E+05 371.8 1.72E+05 3.17E+05 5.27E+05 5.69E+05 385.4 1.71E+05 3.08E+05 5.11E+05 5.41E+05 399.5 1.54E+05 2.94E+05 4.84E+05 5.26E+05 414.2 1.43E+05 2.85E+05 4.56E+05 5.02E+05 429.4 1.42E+05 2.67E+05 4.27E+05 4.90E+05 445.1 1.35E+05 2.49E+05 4.02E+05 4.80E+05 461.4 1.28E+05 2.40E+05 3.82E+05 4.64E+05 478.3 1.14E+05 2.34E+05 3.56E+05 4.44E+05 495.8 1.09E+05 2.13E+05 3.44E+05 4.16E+05 514 .0 9.90E+04 2.02E+05 3.31E+05 4.09E+05

PAGE 87

87 APPENDIX G CASCADE IMPACTOR CUT SIZE AND PARTICLE SIZE DI STRIBUTION DATA Cut sizes for the cascade impactor were calculated using a total flow rate (Q) through the impactor of 20 Lpm. Equations G 1 G 3 were used to calculate the values in Table G 1. The data for cascade impactor measurement with a 30 Lpm shield gas flow rate and 4.2% TMS are presented in Table G 2. The jet velocity is determined by ( G 1 ) where Q j = jet flow rate (Q/n j ), cm 3 /s D j = jet diam eter, cm n j = number of jets The cut size can then be calculated using the following equations: ( G 2 ) 5 cm 2 /s Stk 50 = 0.24 p = 1 g/cm 3 ( G 3 ) Tab le G 1. Impactor data used for cut size calculations. Stage V j n j D j (cm) d p c ) ( m) d p ( m) 1 1.28E+02 1 1.82E+00 23.64 23.56 2 2.11E+02 6 5.79E 01 10.36 10.28 3 5 .95E+02 12 2.44E 01 4.00 3.93 4 7.61E+02 90 7.87E 02 2.01 1.93 5 1.50E+03 110 5.08E 02 1.15 1.07 6 3.28E+03 110 3.43E 02 0.64 0.56 7 7.03E+03 90 2.59E 02 0.38 0.30

PAGE 88

88 Table G 2. Particle size data measured using cascade impactor. Upper cut size Low er cut size Geometric mean d p Fume m ass (g) Mass fraction logd p Fume m ass/ logd p 100.000 23.560 48.539 0.00137 0.018 0.628 0.0022 23.560 10.283 15.565 0.00020 0.003 0.360 0.0006 10.283 3.925 6.353 0.00089 0.012 0.418 0.0021 3.925 1.934 2.755 0.00251 0.033 0.307 0.0082 1.934 1.075 1.442 0.00707 0.093 0.255 0 .0277 1.075 0.561 0.777 0.01135 0.150 0.282 0.0402 0.561 0.302 0.412 0.00743 0.098 0.270 0.0276 0.302 0.030 0.095 0.04491 0.593 1.002 0.0448 Total 0.07574 1 0.1533 The MMD could not be calculated using the Hatch Choate equation. The distribution w as not linear when plotted on log probability paper and over half the mass was found on the final filter, preventing accurate estimation of the geometric standard deviation. However, the MMD could be estimated by plotting only the upper and lower cut sizes of the final filter, where 60% of the mass was present, on log probability paper. The MMD for was estimated at about 265 nm, typical for welding fume particles (Jenkins et al. 2005).

PAGE 89

89 LIST OF REFERENCES American Welding Society. (1999). 199 Laboratory Method for Measuring Fume Generation Rates and Total Fume Emission of Welding and Allied Processes ANSI/AWS; Miami, Florida. Antonini, J.M. (2003). Health Effects of Welding Fumes, Crit. Rev. Toxicol. 33:61 103. Antonini, J.M. Lawryk, N.J. Murthy, G.G.K. and Brain, J.D. (1999). Effect of W elding F ume S olubility on L ung M acrophage V iability and F unction in V itro J. Toxicol. Environ. Health Part A. 58 : 343 363. Antonini, J.M. Taylor, M.D. Zimmer, A.T. and Roberts, J.R. (2004). Pulmonary R esponses to W elding F umes: R ole of M etal C onstituents J Toxicol. Environ. Health Part A. 67 : 233 249. Antonini, J.M. Santamaria, A.B. Jenkins, N.T. Albini, E. and Lucchini, R. (2006). Fate of M anganese A ssociated W ith the I nhalation of W elding F umes: P otential N eurological E ffect s, NeuroToxicology 27 : 304 310. Aston, J.G. Kennedy, R.M. and Messerly, G.H. (1941). The H eat C apacity and E ntropy, H eats of F usion and V aporization and the V apor P ressure of S ilicon T etramethyl J. Am. Chem. Soc. 63 : 2343 23 48. Bagchi, D. Stohs, S.J. Downs, B.W. Bagchi, M. and Preuss, H.G. (2002). Cytotoxicity and O xidative M echanisms of D ifferent F orms of Chromium, Toxicology 180 : 5 22. Biswas, P. and Wu, C.Y. (1995). Control of T oxic M etal E missions F rom C ombustors U si ng S orbents: a R eview J. Air & Waste Manage. Assoc. 48 : 113 127. Biswas, P. and Wu, C.Y. (2005). Nanoparticles and the E nvironment J. Air & Waste Manage. Assoc. 55 : 708 746. Biswas, P. and Zachariah, M. (1997). In S itu I mmobilization of L ead S pecies in C ombustion E nvironments b y I njection of G as P hase S ilica S orbent P recursors Environ. Sci. Technol. 31 : 2455 2463. Boiano, J.M. Wallace, M.E. Sieber, W.K. Groff, J.H. Wang, J. and Ashley, K. (2000). Comparison of T hree S ampling and A nalytical M ethods fo r the D etermination of A irborne H exavalent C hromium J. Environ. Monit. 2 : 329 333. Bowler, R.M. Gysens, S. Diamond, E. Nakagawa, S. Drezgic, M. and Roels, H.A. (2006). Manganese E xposure: N europsychological and N eurological S ymptoms and Ef fects in W e lders NeuroToxiology 27 : 315 326.

PAGE 90

90 Bowler, R.M. Nakagawa, S. Drezgic, M. Roels, H.A. Park, R.M. Diamond, E. Mergler, D. Bouchard, M. Bowler, R.P. and Koller, W. (2007). Sequelae of F ume E xposure in C onfined S pace W elding: a N eurological and N euro psychological C ase S eries NeuroToxicology 28 : 298 311. Castner, H.R. and Null, C.L. (1998). Chromium, N ickel and M anganese in S hipyard W elding F umes Weld. J. 77 : 223 231. Chang, D.P. Heung, W. Yun, M. and Green, D.P. (2004). Improving Welding Toxic Me tal Emission Estimates in California Final Report to the California Air Resources Board 2004. Cho, K. and Wu, C.Y. (2004). Control of M olybdenum E mission by S orbents: E quilibrium A nalysis J. Environ. Eng. 130 : 201 204. Cohen, M.D. Sisco, M. Baker, K. Bowser, D. Chel, L.C. and Schlesinger, R.B. (2003). Impact of C oexposure to O zone on the C arcinogenic P otential of I nhaled C hromium. I. E ffects on R etention and on E xtra and I ntracellular D istribution J. Toxicol. Environ. Health Part A 66 : 39 55. Denn is, J.H. French, M.J. Hewitt, P.J. Mortazavi, S.B. and Redding, C.A.J. (1996). Reduction of H exavalent C hromium C oncentration in F umes from M etal C ored A rc W elding by A ddition of R eactive M etals Ann. Occup. Hyg. 40 : 339 344. Dennis, J.H. Mortazavi, S. B. French, M.J. Hewitt, P.J. and Redding, C.R. (1997). The E ffects of W elding P arameters on U ltra violet L ight E missions, O zone and CrVI F ormation in MIG W elding Ann. Occup. Hyg. 41 : 95 104. Dennis, J.H. French, M.J. Hewitt, P.J. Mortazavi, S.B. and Redding, C.A.J. (2002). Control of E xposure to H exavalent C hromium and O zone in G as M etal A rc W elding of S tainless S teels by U se of a S econdary S hield G as Ann. Occup. Hyg 46 : 43 48. Ebrahimnia, M. Goodarzi, M. Nouri, M. and Sheikhi, M. (2009). Study o f the E ffect of S hielding G as C omposition on the M echanical W eld P roperties of S teel ST 37 2 in G as M etal A rc W elding Mater. Des. 30 : 3891 3895. Elder, A. Gelein, R. Silva, V. Feikert, T. Opanashuk, L. Carter, J. Potter, R. Maynard, A. Ito, Y. Fin kelstein, J. and Oberdrster, G. (2006). Translocation of I nhaled U ltrafine M anganese O xide P articles to the C entral N ervous S ystem Environ. Health Perspect. 114 : 1172 1178. Fang, S.C. Cavallari, J.M. Eisen, E.A. Chen, J.C. Mittleman, M.A. and Christ iani, D.C. (2009). Vascular F unction, I nflammation, and V ariations in C ardiac A utonomic R esponses to P articulate M atter A mong W elders Am. J. Epidemiol. 169 : 848 856. Flynn, M.R. and Susi, P. (2009). Neurological R isks A ssociated W ith M anganese E xposure F r om W elding O perations a L iterature R eview Int. J. Hyg. Environ. Health 212 : 459 469.

PAGE 91

91 Halatek, K. Sinczuk Walczak, H. Szymczak, M. and Rydzynski, K. (2005). Neurological and R espiratory S ymptoms in S hipyard W elders E xposed to M anganese Int. J. Occup. Med. Environ. Health 18 : 265 274. Herzler, J. Manion, J.A. and Tsang, W. (1997). Single P ulse S hock T ube S tudy of the D ecomposition of T etraethoxysilane and R elated C ompounds J. Phys. Chem. A. 101 : 5500 5508. Heung, W. Yun, M.J. Chang, D .P. and Green, P.G. (2007). Emissions of Cr(VI) from A rc W elding J. Air & Waste Manage. Assoc. 57 : 252 260. Hewett, P. (1995). The P article S ize D istribution, D ensity, and S pecific S urface A rea of W elding F umes from SMAW and GMAW M ild and S tainless S teel C onsumables Am. Ind. Hyg. Assoc. J. 56 : 128 135. (a) Hewett, P. (1995). Estimation of R egional P ulmonary D eposition and E xposure for F umes from SMAW and GMAW M ild and S tainless S teel C onsumables Am. Ind. Hyg. Assoc. J. 56 : 136 142. (b) Holmes, A.L. Wise, S.S. and Wise, J.P. (2008). Carcinogenicity of H exavalent C hromium Indian J. Med. Res. 128 : 353 372. Hovde, C.A. and Raynor, P.C. (2007). Effects of V oltage and W ire F eed S peed on W elding F ume C haracteristics J. Occup. Environ. Hyg. 4 : 903 912. IARC. (19 90). Chromium, N ickel and W elding IARC Monographs on the Evaluation of Carcinogenic Risk of Chemicals to Humans 49 Iida, K., Pierman, J., Tolaymat, T., Townsend, T., Wu, C.Y. (2003). Control of Chromated Copper Arsenate Wood Incineration Air Emissions and Ash Leaching Using Sorbent Technology, J. Environ. Eng. 130:184 192. Jang, H.D., Chang, H., Suh, Y., Okuyama, K. (2006). Synthesis of SiO 2 nanoparticles from sprayed droplets of tetraethylorthosilicate by the flame spray pyrolysis, Curr. Appl. Phys. 6S 1:e110 e113. Jeng, H.A. and Swanson, J. (2006). Toxicity of M etal O xide N anoparticles in M ammalian C ells J. Environ. Sci. Health Part A 41 : 2699 2711. Jenkins, N.T. and Eagar, T.W. (2005). Chemical A nalysis of W elding F ume P articles Weld. J. 84 : 87s 93s Jenkins, N.T. Pierce, W.M.G. and Eagar, T.W. (2005). Particle S ize D istribution of G as M etal and F lux C ored A rc W elding F umes Weld. J. 84 : 156s 163s. Karanasiou, A.A., Thomaidis, N.S., Eleftheriadis, K., Siskos, P.A. (2005). Comparative Study of Pretre atment Methods for the Determination of Metals in Atmospheric Aerosol by Electrothermal Atomic Absorption Spectrometry, Talanta 65:1196 1202.

PAGE 92

92 Kleinstreuer, C. Zhang, Z. and Li, Z. (2008). Modeling A irflow and P article T ransport/ D eposition in P ulmonary A irways Respir. Physiol. Neurobiol. 163 : 128 138. Konarski, P. Iwanejko, I. and (2003). Core S hell M orphology of W elding F ume M icro and N anoparticles Vacuum 70 : 385 389. (a) Konarski, P. Iwanejko, I. and Mierzejewska, A. (2003). D epth P rofiling of W orking E nvironment N anoparticles Appl. Surf. Sci. 203 204 : 757 761. (b) Ku, J.C. and Eide, M. (1998). Hexavalent C hromium in the W orkplace A tmosphere, OSHA ID 215 Salt Lake City, Utah, Occupational Safety and Health Administration Lee, M.H. Cho, K. Shah, A.P. and Biswas, P. (2005). Nanostructured S orbents for C aptur e of C admium S pecies in C ombustion E nvironments Environ. Sci. Technol. 39 : 8481 8489. Lee, S.R. and Wu, C.Y. (2005). Size D istribution E volution of F ine A erosols D ue to I ntercoagulation with C oarse A erosols Aerosol Sci. Technol. 39 : 358 370. Lee, S.R. an d Wu, C.Y. (2002). Study of V anadium E mission C ontrol in C ombustion S ystems by T hermodynamic E quilibrium A nalyses Adv. Environ. Res. 7 : 1 10. Lidn, G. and Surakka, J. (2009). A H eadset M ounted M ini S ampler for M easuring E xposure to W elding A erosol in the B reathing Z one Ann. Occup. Hyg. 53 : 99 116. Lillienberg, L. Zock, J.P. Kromhout, H. Plana, E. Jarvis, D. Torn, K. and Kogevinas, M. (2008). A P opulation B ased Study on W elding E xposures at W ork and R espiratory S ymptoms Ann. Occup. Hyg. 52 : 107 115. Liu, H.H. Wu, Y.C. and Chen, H.L. (2007). Production of O zone and R eactive O xygen S pecies A fter W elding Arch. Environ. Contam. Toxicol. 53 : 513 518. Loukzadeh, Z. Sharifian, S.A. Aminian, O. and Shojaoddiny Ardekani, A. (2009). Pulmonary E ffects of S pot W elding in A utomobile A ssembly Occup. Med. 59 : 267 269. Luo, J.C.J. Hsu, K.H. and Shen, W.S. (2009). Inflammatory R esponses and O xidative S tress from M etal F ume E xposure in A utomobile W elders J. Occup. Environ. Med. 51 : 95 103. Maynard, A.D., Ito, Y. Arslan, I., Zimmer, A.T., Browning, N., Nicholls, A. (2004). Examining Elemental Surface Enrichment in Ultrafine Aerosol Particles Using Analytical Scanning Transmission Electron Microscopy, Aerosol Sci. Technol. 38:365 381.

PAGE 93

93 McNeilly, J.D. Heal, M.R. Beverland, I.J. Howe, A. Gibson, M.D. Hibbs, L.R. MacNee, W. and Donaldson, K. (2004). Soluble T ransition M etals C ause the P ro I nflammatory E ffects of W elding F umes in V itro Toxicol. Appl. Pharmacol. 196 : 95 107. Meeker, J.D. Susi, P. and Flynn, M.R (2007). Manganese and W elding F ume E xposure and C ontrol in C onstruction J. Occup. Environ. Hyg. 4 : 943 951. Minni, E. Gustafsson, T.E. Koponen, M. and Kalliomaki, P.L. (1984). A S tudy of the C hemical S tructure of P articles in the W elding F umes of M ild and S tainless S teel J. Aerosol Sci. 15 : 57 68. Mruczek, M.F. Konkol, P.J. Ferree, S. and Sierdzinski, M. (2008). Reducing Mn F umes T hrough E lectrode D evelopment Weld. J. 87 : 25 29. NIOSH. (1994). Method 7604, Issue 2: Chromium, H exavalent by I on C hroma tography N IOSH Manual of Analytical Methods 4 th edition. NIOSH. (2003). Method 7605, Issue 1: Chromium, H exavalent by I on C hromatography N IOSH Manual of Analytical Methods 4 th edition. OSHA. (2006). 29 CFR 1910 1026 Owens, T.M. and Biswas, P. (1996 ). Vapor P hase S orbent P recursors for T oxic M etal E missions C ontrol from C ombustors Ind. Eng. Chem. Res. 35 : 792 798. (a) Owens, T.M. and Biswas, P. (1996). Reactions B etween V apor P hase L ead C ompounds and I n S itu G enerated Si lica P articles at V arious L ea d S ilicon F eed R atios: A pplications to T oxic M etal C apture in C ombustors J. Air & Waste Manage. Assoc. 46 : 530 538. (b) Park, J. Yoo, C.I. Sim, C.S. Kim, J.W. Yi, Y. Shin, Y.C. Kim, D.H. and Kim, Y. (2006). A R etrospective C ohort S s D isease in Korean S hipbuilders NeuroToxicology 27 : 445 449. (a) Park, R.M. Bowler, R.M. Eggerth, D.E. Diamond, E. Spencer, K.J. Smith, D. and Gwiazda, R. (2006). Issues in N eurological R isk A ssessment for O ccupational E xposures: the Bay Bridge W el ders NeuroToxicology 27 : 373 384. (b) Pascal, L.E. and Tessier, D.M. (2004). Cytotoxicity of C hromium and M anganese to L ung E pithelial C ells i n V itro Toxicol. Lett. 147 : 143 151. Phadungsukanan, W. Shekar, S. Shirley, R. Sander, M. West, R.H. and Kr aft, M. (2009). First P rinciples T hermochemistry for S ilicon S pecies in the D ecomposition of T etraethoxysilane J. Phys. Chem. A. 113 : 9041 9049.

PAGE 94

94 Pires, I. Quintino, L. and Miranda, R.M. (2007). Analysis of the I nfluence of S hielding G as M ixtures on the G as M etal A rc W elding M etal T ransfer M odes and F ume F ormation R ate Mater. Des. 28 : 1623 1631. Quimby, B.J. and Ulrich, G.D. (1999). Fume R ormation R ates in G as M etal A rc W elding Weld. J. 78 : 142s 149s. Reuzel, P.G.J. Bruijntjes, J.P. Feron, V.J. and Wou tersen, R.A. (1991). Subchronic I nhalation T oxicity of A morphous Silicas and Q uartz D ust in R ats Fd. Chem. Toxic. 29 : 341 354. Roelofs, F., Vogelsberger, W. (2004). Dissolution Kinetics of Synthetic Amorphous Silica in Biological Like Media and Its Theoret ical Description, J. Phys. Chem. B. 108:11308 11316. Schaller, K.H. Csanady, G. Filser, J. Jngert, B. and Drexler, H. (2007). Elimination K inetics of M etals A fter an A ccidental E xposure to W elding F umes Int. Arch. Occup. Environ. Health. 80 : 635 641. Seel, E.A. Zaebst, D.D. Hein, M.J. Liu, J. Nowlin, S.J. and Chen, P. (2007). Inter Rater Agreement for a Retrospective Exposure Assessment of Asbestos, Chromium, Nickel, and Welding Fumes in a S tudy of Lung Cancer and Ionizing R adiation Ann. Occup. H yg. 51 : 601 610. Serageldin, M. and Reeves, D.W. (2009). Development of W elding E mission F actors for Cr and Cr(VI) with a C onfidence L evel J. Air & Waste Manage. Assoc. 59 : 619 626. Simpson, S.W. (2009). Metal Transfer Instability in Gas Metal Arc Welding, Sci. Technol. Weld. Joining 14:262 273. Smargiassi, A. Baldwin, M. Savard, S. Kennedy, G. Mergler, D. and Zayed, J. (2000). Assessment of E xposure to Manganese in W elding O perations D uring the A ssembly of H eavy E xcavation M achinery A ccessories J. O ccup. Environ. Hyg. 15 : 746 750. Soderstrom, E.J. and Mendez, P.F. (2008). Metal T ransfer D uring GMAW with T hin E lectrodes and Ar CO 2 S hielding G as M ixtures Weld. J. 87 : 124s 133s. Stampfer, M.J. (2009). Welding O ccupations and M D isease and O ther N eurodegenerative D isease A mong United States M en, 1985 1999 J. Occup. Environ. Hyg. 6 : 267 272. Stephenson, D. Seshadri, G. and Veranth, J.M. (2003). Workplace E xposure to S ubmicron P article M ass and N umber C oncentrations f rom M anual A rc W elding of C arbon S teel Am. Ind. Hyg. Assoc. J. 64 : 516 521.

PAGE 95

95 Taylor, M.D. Roberts, J.R. Leonard, S.S. Shi, X. and Antonini, J.M. (2003). Effects of W elding F umes of D iffering C omposition and S olubility on F ree R adical P roduction and A cute L ung I njur y and I nflammation in R ats Toxicol. Sci. 75 : 181 191. Wu, C.Y. and Barton, T. (2001). A T hermodynamic E quilibrium A nalysis to D etermine the P otential S orbent M aterials for the C ontrol of A rsenic E missions from C ombustion S ources Environ. Eng. Sci. 18 : 177 190. Yu, I.J. Kim, K.J. Chang, H.K. Song, K.S. Han, K.T. Han, J.H. Maeng, S.H. Chung, Y.H. Park, S.H. Chung, K.H. Han, J.S. and Chung, H.K. (2000). Pattern of D eposition of S tainless S teel W elding F ume P articles I nhaled i nto the R espiratory sys tems of Sprague Dawley R ats E xposed to a N ovel W elding F ume G enerating S ystem Toxicol. Lett. 116 : 103 111. Yuan, H. He, S. He, M. Niu, Q. Wang, L. and Wang, S. (2006). A comprehensive S tudy on N eurobehavior, N eurotransmitters and L ymphocyte S ubsets A l teration of Chinese M anganese W elding W orkers Life Sci. 78 : 1324 1328. Zimmer, A.T. Baron, P.A. and Biswas, P. (2002). The I nfluence of O perating P arameters on N umber W eighted A erosol S ize D istribution G enerated from a G as M etal A rc W elding P rocess J. A erosol Sci. 33 : 519 531. Zimmer, A.T. and Biswas, P. (2001). Characterization of the Aerosols R esulting from A rc W elding P rocesses J. Aerosol Sci. 32 : 993 1008.

PAGE 96

96 BIOGRAPHICAL SKETCH Nathan Topham was born in Boston, Massachusetts in 1985. He moved to Orlan do, Florida at 5 years of age. Nathan attended high school at Lake Brantley High School in Altamonte Springs. He began attending the University of Florida (UF) in 2003. He performed research in aerosol sampling techniques as an undergraduate student. He ea rned h ciences from the University of Florida in the fall of 2008. Nathan was admitted to graduate school the following semester at UF to pursue a M aster of Engineering degree in environmental engineering s ciences.