<%BANNER%>

Characterization of Reaerosolization in an Effort to Improve Sampling of Airborne Viruses

Permanent Link: http://ufdc.ufl.edu/UFE0022331/00001

Material Information

Title: Characterization of Reaerosolization in an Effort to Improve Sampling of Airborne Viruses
Physical Description: 1 online resource (103 p.)
Language: english
Creator: Riemenschneider, Lindsey
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: 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: Airborne virus outbreaks, including the influenza pandemic of 1918, the recent SARS pandemic, and the anticipated H5N1 outbreaks, plus the perceived threat of bioterrorism have led to heightened concern about the prevalence and potential effects of airborne viruses. However, current bioaerosol sampling methods are unable to effectively sample airborne viruses (typically 20 to 300 nanometers). To address this problem, a novel Bioaerosol Amplification Unit has been designed and constructed to increase the size of the virus particles by condensational growth, thereby enhancing sampling recovery. In order to evaluate the Bioaerosol Amplification Unit, reaerosolization of viral particles from the impinger needs to be investigated to assess its impact on the capability of the new device. Reaerosolization as a function of flow rate and impinger collection liquid concentration has been characterized. An impinger containing a known concentration of particles (MS2 bacteriophage or polystyrene latex particles) was operated at various flow rates with sterile air, and a scanning mobility particle sizer was used to determine the reaerosolization rates. Results indicate that reaerosolization increases as flow rate increases, due to the additional energy introduced to the system. However, increased concentration does not necessarily lead to an increase in reaerosolization for virus particles. Rather, reaerosolization increases as concentration increases until it reaches a concentration of approximately 106 PFU/mL, at which point reaerosolization begins to decrease. The phenomenon likely results from the aggregation of viral particles or the increase of surface tension or viscosity at high concentrations. Adjusting the surface tension by adding soap and increasing viscosity by adding a layer of heavy white mineral oil decreased reaerosolization. In summary, reaerosolization from an impinger could compromise the improved collection capability of the new BAU and is a major mode of loss in airborne virus sampling with impingers in certain scenarios, but reaerosolization can be minimized by sampling over shorter periods of time.
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 Lindsey Riemenschneider.
Thesis: Thesis (M.E.)--University of Florida, 2008.
Local: Adviser: Wu, Chang-Yu.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022331:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022331/00001

Material Information

Title: Characterization of Reaerosolization in an Effort to Improve Sampling of Airborne Viruses
Physical Description: 1 online resource (103 p.)
Language: english
Creator: Riemenschneider, Lindsey
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: 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: Airborne virus outbreaks, including the influenza pandemic of 1918, the recent SARS pandemic, and the anticipated H5N1 outbreaks, plus the perceived threat of bioterrorism have led to heightened concern about the prevalence and potential effects of airborne viruses. However, current bioaerosol sampling methods are unable to effectively sample airborne viruses (typically 20 to 300 nanometers). To address this problem, a novel Bioaerosol Amplification Unit has been designed and constructed to increase the size of the virus particles by condensational growth, thereby enhancing sampling recovery. In order to evaluate the Bioaerosol Amplification Unit, reaerosolization of viral particles from the impinger needs to be investigated to assess its impact on the capability of the new device. Reaerosolization as a function of flow rate and impinger collection liquid concentration has been characterized. An impinger containing a known concentration of particles (MS2 bacteriophage or polystyrene latex particles) was operated at various flow rates with sterile air, and a scanning mobility particle sizer was used to determine the reaerosolization rates. Results indicate that reaerosolization increases as flow rate increases, due to the additional energy introduced to the system. However, increased concentration does not necessarily lead to an increase in reaerosolization for virus particles. Rather, reaerosolization increases as concentration increases until it reaches a concentration of approximately 106 PFU/mL, at which point reaerosolization begins to decrease. The phenomenon likely results from the aggregation of viral particles or the increase of surface tension or viscosity at high concentrations. Adjusting the surface tension by adding soap and increasing viscosity by adding a layer of heavy white mineral oil decreased reaerosolization. In summary, reaerosolization from an impinger could compromise the improved collection capability of the new BAU and is a major mode of loss in airborne virus sampling with impingers in certain scenarios, but reaerosolization can be minimized by sampling over shorter periods of time.
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 Lindsey Riemenschneider.
Thesis: Thesis (M.E.)--University of Florida, 2008.
Local: Adviser: Wu, Chang-Yu.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022331:00001


This item has the following downloads:


Full Text

PAGE 1

1 CHARACTERIZATION OF REAEROSOLIZA TION IN AN EFFORT TO IMPROVE SAMPLING OF AIRBORNE VIRUSES By LINDSEY RIEMENSCHNEIDER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Lindsey Riemenschneider

PAGE 3

3 To my family and friends

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank m y family and friends fo r all of their support, encouragement, and understanding. I would also like to thank Dr. Wu for his patience, dependability, and generosity and for serving as an inspiration on how to interact with others. I would also like to express my gratitude to Dr. Lundgren for al ways providing so many ideas along with a smile to help improve both my project and my mood. I am also ve ry grateful to have Dr. Wander for his encouragement and willingness to provide tr emendous help, along with the many ideas and suggestions he provided. I also feel very fortunate and proud to belong to my research group and thank each of my lab mates, past and present, for all of their kindness, assi stance, and hard work. It has truly been a pleasure to work along side all of them every day for the past two years. I would also like to recognize the sources of fina ncial support which helped me get through graduate school: Camp Dresser McKee for the CDM Fellowship; the UF Environmental Engineering Department for Teaching and Research Assistantships; and the US Air Force (Project No. FA8650-06-C-5913) fo r the research project.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 LIST OF ABBREVIATIONS........................................................................................................ 12 ABSTRACT...................................................................................................................................13 CHAP TER 1 INTRODUCTION..................................................................................................................15 Introduction................................................................................................................... ..........15 Bioaerosol Sampling Methodologies...................................................................................... 15 Bioaerosol Amplification Unit............................................................................................... 16 Description of Technology.............................................................................................. 17 Conceptual Implementation of In-lin e Bioaerosol Am plification Device...................... 20 Reaerosolization.....................................................................................................................22 Objective.................................................................................................................................28 2 EXPERIMENTAL METHODOLOGY..................................................................................30 Reaerosolization with PSL.....................................................................................................31 Reaerosolization with MS2..................................................................................................... 32 Effect of Surface Tension and Visco sity on Reaerosolization............................................... 35 Reaerosolization with BioSampler......................................................................................... 36 3 RESULTS...............................................................................................................................37 Reaerosolization with PSL.....................................................................................................37 Reaerosolization with MS2..................................................................................................... 38 Effect of Surface Tension and Visco sity on Reaerosolization............................................... 43 Reaerosolization with BioSampler......................................................................................... 45 4 DISCUSSION.........................................................................................................................47 Reaerosolization with PSL.....................................................................................................47 Reaerosolization with MS2..................................................................................................... 50 Effect of Surface Tension and Visco sity on Reaerosolization............................................... 54 Reaerosolization with BioSampler......................................................................................... 55 5 CONCLUSION..................................................................................................................... ..58

PAGE 6

6 APPENDIX A BIOAEROSOL AMPLIFICATION UNIT INFORMATION............................................... 60 Bioaerosol Amplification Unit Design................................................................................... 60 Experimental Methodology.................................................................................................... 61 Inert Particle Amplification............................................................................................. 62 Physical Collection Challenge.........................................................................................64 Viable Collection Challenge............................................................................................65 Preliminar y Results ............................................................................................................ .....66 B REAEROSOLIZATION WI TH PSL DATA SETS ............................................................... 68 Test 1: 03/07/2008a.........................................................................................................68 Test 2: 03/07/2008b.........................................................................................................69 Test 3: 04/02/2008...........................................................................................................70 C REAEROSOLIZATION WITH MS2 DATA SETS.............................................................. 71 Flow Rate: 3 Lpm............................................................................................................... ....71 Test 1: 03/15/2007a.........................................................................................................71 Test 2: 03/15/2007b.........................................................................................................72 Summary of 3 Lpm.......................................................................................................... 73 Flow Rate: 6 Lpm............................................................................................................... ....74 Test 1: 03/15/2007...........................................................................................................74 Test 2: TBD.....................................................................................................................75 Summary of 6 Lpm.......................................................................................................... 75 Flow Rate: 9 Lpm............................................................................................................... ....76 Test 1: 02/14/2008a.........................................................................................................76 Test 2: 02/14/2008b.........................................................................................................77 Test 3: 03/15/2007...........................................................................................................78 Summary of 9 Lpm.......................................................................................................... 79 Flow Rate: 12.5 Lpm............................................................................................................ ..80 Test 1: 01/24/2008...........................................................................................................80 Test 2: 01/28/2008...........................................................................................................81 Test 3: 02/09/2008...........................................................................................................82 Summary of 12.5 Lpm..................................................................................................... 83 Flow Rate: 15 Lpm.............................................................................................................. ...84 Test 1: 02/17/2008a.........................................................................................................84 Test 2: 02/17/2008b.........................................................................................................85 Test 3: TBD.....................................................................................................................86 Summary of 15 Lpm........................................................................................................ 86 Mode Size of MS2 Experiments......................................................................................87 D EFFECT OF SURFACE TENSION AND VISCOSITY DATA SETS ................................. 88 Effect of Surface Tension...................................................................................................... .88 Test 1: 03/21/2008a.........................................................................................................88

PAGE 7

7 Test 2: 03/21/2008b.........................................................................................................89 Test 3: 03/24/2008...........................................................................................................90 Summary of Surface Tension.......................................................................................... 91 Effect of Viscosity..................................................................................................................92 Test 1: 03/31/2008...........................................................................................................92 Test 2: 04/02/2008a.........................................................................................................93 Test 3: 04/02/2008b.........................................................................................................94 Summary of Viscosity.....................................................................................................95 E REAEROSOLIZATION WITH BIOSAMPLER DATA SETS ............................................ 96 Test 1: 03/15/2008...........................................................................................................96 Test 2: 03/19/2008a.........................................................................................................97 Test 3: 03/19/2008b.........................................................................................................98 Summary of BioSampler................................................................................................. 99 LIST OF REFERENCES.............................................................................................................100 BIOGRAPHICAL SKETCH.......................................................................................................103

PAGE 8

8 LIST OF TABLES Table page 1-1 Nomenclature for mass transfer of particles between bubbles and liquid ......................... 23 2-1 Summary of tests and co rresponding purpose to characte rize viral reaerosolization from impingers................................................................................................................. ..30 2-2 Experimental matrix to determine the effect of flow rate and concentration on reaerosolization ..................................................................................................................33 3-1 Percent of PSL particles reaerosolized.............................................................................. 37 3-2 Comparison of mode sizes for PSL partic les under different expe rim ental conditions.... 45 4-1 Comparison of reaerosolization for bacteria -sized particles to that for virus-sized particles..............................................................................................................................48 4-2 Estimated droplet volume generated from impinger......................................................... 50 4-3 Comparison of present work to past research on reaerosolization from AGI-30 Impinger and BioSampler.................................................................................................. 56 A-1 Summary of tests and corres ponding purpose to evaluate BAU ....................................... 62 A-2 Preliminary results from vi ral aerosol sam pling using BAU............................................. 67

PAGE 9

9 LIST OF FIGURES Figure page 1-1 Bioaerosol amplificatio n conceptual schem atic................................................................. 21 1-2 Coefficients of mass transfer due to inertia and diffusion for transfer of particles from bubbles to surrounding liquid.................................................................................... 24 1-3 Coefficients of mass transfer due to inertia and diffusion for transfer of particles from liquid to bubbles........................................................................................................ 25 2-1 Reaerosolization experimental setup................................................................................. 31 2-2 The BioSampler............................................................................................................. ....36 3-1 Reaerosolization as a function of impinger con centration for PSL particles at 12.5 Lpm....................................................................................................................................38 3-2 Size distribution of reaeroso lized particles as a function of flow rate at an im pinger collection liquid c oncentration of 1x102 PFU/mL............................................................. 39 3-3 Size distribution of reaerosolized par tic les as a function of collection liquid concentration at a flow rate of 9 Lpm................................................................................ 40 3-4 Reaerosolization as a function of flow and im pinger concentration for MS2 viral particles..............................................................................................................................41 3-5 Comparison of the reaero solization of PSL at 12.5 Lp m with and without soap present........................................................................................................................ ........44 3-6 Comparison of the reaeroso lization of PSL at 12.5 Lpm with and without oil present..... 44 3-7 Comparison of the reaerosolization of PSL at 12.5 Lpm from impinger and BioSampler........................................................................................................................46 4-1 Baseline at 12.5 Lpm from impinger with oil and without oil present. ............................. 55 A-1 BAU prototype schematic.................................................................................................. 61 A-2 Experimental setup for in ert particle amplification. .......................................................... 62 A-3 Experimental setup for physical collection efficiency testing. .......................................... 64 B-1 Size distribution and average concentr ation of PSL at 12.5 Lpm (03/07/08a). .................68 B-2 Size distribution and average concentr ation of PSL at 12.5 Lpm (03/07/08b). .................69 B-3 Size distribution and average concen tration of PSL at 12.5 Lpm (04/02/08). ...................70

PAGE 10

10 C-1 Size distribution and adjusted avg con centration of MS2 at 3 Lpm (03/15/07a).............. 71 C-2 Size distribution and adjusted avg con centration of MS2 at 3 Lpm (03/15/07b).............. 72 C-3 Summary of adjusted average concentration of MS2 at 3 Lpm. ........................................73 C-4 Size distribution and adjusted avg con centration of MS2 at 6 Lpm (03/15/07)................ 74 C-5 Summary of adjusted average concentration of MS2 at 6 Lpm. ........................................75 C-6 Size distribution and adjusted avg con centration of MS2 at 9 Lpm (02/14/08a).............. 76 C-7 Size distribution and adjusted avg con centration of MS2 at 9 Lpm (02/14/08b).............. 77 C-8 Size distribution and adjusted avg con centration of MS2 at 9 Lpm (03/15/07)................ 78 C-9 Summary of adjusted average concentration of MS2 at 9 Lpm. ........................................79 C-10 Size distribution and adjusted avg con centration of MS2 at 12.5 Lpm (01/24/07)........... 80 C-11 Size distribution and adjusted avg con centration of MS2 at 12.5 Lpm (01/28/07)........... 81 C-12 Size distribution and adjusted avg con centration of MS2 at 12.5 Lpm (02/09/07)........... 82 C-13 Summary of adjusted average co ncentration of MS2 at 12.5 Lpm .................................... 83 C-14 Size distribution and adjusted avg con centration of MS2 at 15 Lpm (02/17/08a)............ 84 C-15 Size distribution and adjusted avg con centration of MS2 at 15 Lpm (02/17/08b)............ 85 C-16 Summary of adjusted average concentration of MS2 at 15 Lpm .......................................86 C-17 Average mode size as a function of flow and im pinger concentration for MS2 viral particles..............................................................................................................................87 D-1 Size distribution and adjusted avg con centration of PSL at 12.5 Lpm (03/21/08a).......... 88 D-2 Size distribution and adjusted avg con centration of PSL at 12.5 Lpm (03/21/08b).......... 89 D-3 Size distribution and adjusted avg con centration of PSL at 12.5 Lpm (03/21/08b).......... 90 D-4 Summary of adjusted average concentration of PSL at 12.5 Lpm ....................................91 D-5 Size distribution and adjusted avg con centration of PSL at 12.5 Lpm (03/31/08)............ 92 D-6 Size distribution and adjusted avg con centration of PSL at 12.5 Lpm (04/02/08a).......... 93 D-7 Size distribution and adjusted avg con centration of PSL at 12.5 Lpm (04/02/08b).......... 94 D-8 Summary of adjusted average concentration of PSL at 12.5 Lpm ....................................95

PAGE 11

11 E-1 Size distribution and adjusted avg con centration of PSL at 12.5 Lpm (03/15/08)............ 96 E-2 Size distribution and adjusted avg con centration of PSL at 12.5 Lpm (03/19/08a).......... 97 E-3 Size distribution and adjusted avg con centration of PSL at 12.5 Lpm (03/19/08b).......... 98 E-4 Summary of adjusted average concentration of PSL at 12.5 Lpm ....................................99

PAGE 12

12 LIST OF ABBREVIATIONS BAU Bioaeroso l Amplification Unit SMPS Scanning Mobility Particle Sizer PSL Polystyrene Latex nm Nanometer m Micrometer AGI Ace Glass, Inc. ppm Parts per million PFU Plaque forming unit cP Centipoise

PAGE 13

13 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering CHARACTERIZATION OF REAEROSOLIZA TION IN AN EFFORT TO IMPROVE SAMPLING OF AIRBORNE VIRUSES By Lindsey Riemenschneider August 2008 Chair: Chang-Yu Wu Major: Environmental Engineering Sciences Airborne virus outbreaks, including the in fluenza pandemic of 1918, the recent SARS pandemic, and the anticipated H5N1 outbreaks, plus the perceived threat of bioterrorism have led to heightened concern about the prevalence and pote ntial effects of airborne viruses. However, current bioaerosol sampling methods are unable to effectively sample airb orne viruses (typically 20 to 300 nanometers). To address this problem a novel Bioaerosol Amplification Unit has been designed and constructed to increase the size of the virus particles by condensational growth, thereby enhancing sampling recover y. In order to evaluate the Bioaerosol Amplification Unit, reaerosolization of viral pa rticles from the impinger needs to be investigated to assess its impact on the capability of the new device. Reaerosolization as a function of flow rate and impinger collection liquid concentration has been characterized. An impinger containing a known concentration of particles (MS2 bacteriophage or polystyrene latex particles) was operated at various flow rates with sterile air, and a scanning mobility particle sizer was used to determine the reaerosolization rates. Results indicate that reaerosolization increases as fl ow rate increases, due to the additional energy introduced to the system. However, increased concentration does not necessarily lead to an increase in reaerosolization for virus particles. Rather, reaerosolization increases as

PAGE 14

14 concentration increases until it reaches a concentration of approximately 106 PFU/mL, at which point reaerosolization begins to decrease. Th e phenomenon likely results from the aggregation of viral particles or the increase of surface tension or viscosity at high concentrations. Adjusting the surface tension by adding soap and increasi ng viscosity by adding a layer of heavy white mineral oil decreased reaerosolization. In summ ary, reaerosolization from an impinger could compromise the improved collection capability of the new BAU and is a major mode of loss in airborne virus sampling with impingers in ce rtain scenarios, but reaerosolization can be minimized by sampling over shorter periods of time.

PAGE 15

15 CHAPTER 1 INTRODUCTION Introduction The perceived threat of bioterrorism and ai rborne virus outbreaks, including historical epidem ics of influenza and more recent occurren ces of SARS and various strains of influenza, have led to heightened concern about the prevalence and potential e ffects of airborne viruses. Although there is increased attention to these potentially deadly microorganisms, current bioaerosol sampling methods are unable to eff ectively sample airborne viruses, which are typically in the 20 nanometer range (Madigan et al. 2003). Sampling efficiency of current bioaerosol sampling methods is less than 10% for the most challenging sizes of 30 nm (Hogan et al. 2005), a significant concern with re spect to the common size of a virus. If sampling methodologies do not provide accurate resu lts, the discrepancy between measured and actual virus concentratio ns could potentially lead to disast rous decision errors because the infectivity of viruses is measured as a minimum threshold. Bioaerosol Sampling Methodologies Bioaerosol sam pling methods must physically collect the bioa erosols to subject them to viability analysis. The viability determines th e infectivity of the particular air stream and provides critical health effect info rmation. Therefore, the challenge in bioaerosol sampling is to have a system that provides high physical collec tion efficiency while maintaining high viable collection efficiency, such that viable organisms are not inactivated during the collection process. A typical sampling system for airborne viruses utilizes an impactor, liquid impinger, or filter (Tseng and Li 2005). Liquid impingement is often an advantageous choice, as the method lends itself nicely to viral enumerati on techniques (Terzieva et al. 1 996; Hogan et al. 2005). As previously mentioned, recent experiments with liquid impingers, including an AGI-30 impinger,

PAGE 16

16 a BioSampler, and a frit bubbler, demonstrated that less than 10% of part icles in the 30 nm range were collected (Hogan et al. 2005). Colle ction efficiency is highl y dependent on particle size for the impingement methods. Collection effici ency increased with increasing particle size for particles greater than 100 nm due to the increased inertia (H ogan et al. 2005). On the other end of the spectrum, the collection efficiency of particles less th an 30 nm increased as particle size decreased, likely due to increased diffusi on. The particles in the 30 nm range had neither sufficient inertia nor suffi cient diffusion to be collected we ll in the impinger. Reponen et al. (2001) report that the 50% cutpoint for the AGI-30 is 0.31 m, at least ten times larger than viruses on the lower end of the viral size range. Along with particle size, othe r characteristics of the airborne virus affect physical collection efficiency. Specifica lly, the hydrophobicity of the particle plays a significant role. Viruses are classified as envel oped if a lipid layer surrounds th e nucleocapsid (Madigan et al. 2003). The presence of a lipid layer generally indicates that a virus will be hydrophobic, while the absence conversely indicates a hydrophilic nature (Vidaver et al. 1973; Madigan et al. 2003; Tseng and Li 2005). Comparison of collection e fficiency in several different bioaerosol samplers for four different bacteriophages in dicated that hydrophilic vi ruses are collected 10 100 times better than hydrophobic viruses (Tseng and Li 2005). This issue makes sampling hydrophobic viruses even more challenging than simply the issue of particle size, and these viral characteristics are important considerati ons for airborne virus sampling methodology. Bioaerosol Amplification Unit In review, airborne virus sam pling is limited by the extremely small size of viruses as well as certain particle characteristics such as hydrop hobicity. To address the limitations due to size, a novel device known as the Bioaerosol Amplifi cation Unit (BAU) has recently been designed

PAGE 17

17 and constructed to grow the ultrafine virus particles by condensationa l amplification, thus enhancing sampling recovery in impingers and imp roving the accuracy of measured results. The term amplification to describe the effective increase in particle size is consistent with other work in this field (Vonnegut 1954; Hering and St olzenburg 2005). Particle amplification by condensational growth is an established method th at has been previously applied to many inert ultrafine particles (O kuyama et al. 1984; Sioutas and Kout rakis 1996; Sioutas et al. 1999; Demokritou et al. 2002; Hering and Stolzenburg 2005). The ubi quitous condensation nuclei counter is a well known applicati on of condensational growth to improve sampling and detection (Hinds 1999). Although the commercially available condensation nucle i counter is highly suitable for determining physical counts of aeros ols, the unit lacks that necessary ability to provide viability information importa nt for bioaerosol characterization. The use of condensational growth to amplif y bioaerosols for improved physical collection is a new application of a proven technology. By implementing th e BAU prior to the selected bioaerosol sampling method, the effective particle size and subsequent co llection efficiency of airborne viruses should increase significantly, as found in other research evaluating the use of condensational growth on iner t particles (Okuyama et al. 1984; Sioutas and Koutrakis 1996; Demokritou et al. 2002). Description of Technology This novel bioaerosol collection m ethod has been developed and disclosed through the University of Florida Office of Technol ogy Licensing (oral disclosure UF#12430, Provisional Patent Application 60/956,316 file d 8/16/07). The main function of the device is to amplify the size of ultrafine bioaerosols, so as to make the particle size sufficiently large (>1.0 m) to be collected efficiently using standard bioaerosol collection techniques. The device employs condensational deposition ont o the ultrafine particle to the extent that the particle has a new

PAGE 18

18 effective diameter that is large enough for high sampling efficiency. This is accomplished by using an established process in which a supe rsaturated vaporair mixture condenses onto condensation nuclei, which conti nue to grow until a vaporliquid equilibrium is reached (Hinds 1999; Wu and Biswas 1998; Friedlander 2003). Conde nsation nuclei can be in either solid or liquid phase and serve primarily as a host for th e saturated vapor. In the case of ultrafine bioaerosols such as viruses, indi vidual or clusters of airborne viruses can serve as condensation nuclei in the presence of supersaturated water vapor, initiating water condensation on the virus particle surface. Provided sufficient time a nd appropriate conditions, the virus or other bioaerosol is capable of growi ng to micron-sized droplets and then can be collected efficiently using standard sampling techniques. To successfully apply this concept specifically to viruses will requir e a substantial amount of amplification in the BAU. As discussed pr eviously, Hogan et al. ( 2005) reported less than 10% collection efficiency for 30-nm particle s in the AGI-30. Lin et al. (1997) reported approximately 70% collection efficiency during 30 minutes of sampling for 0.51-m PSL in the AGI-30. That amount increased to nearly 90% collection efficiency fo r 1.02-m PSL particles in the AGI-30 impinger for 30 minutes of sampli ng or less (Lin et al. 1997). Therefore, amplifying particle size from the lower end of the viral particle size range (30 nm) to 1 m would increase collection efficiency from less than 10% to almost 90%. The time required to achieve this by the condensational amplificati on process can be estimated using Equation 1-1 (Hinds 1999). d d pp pT p T p dR MD dt dd4 (1-1) The equation describes the rate of particle growth, where dp is particle size (amplified from 0.03 m to 1.0 m) and t is time to be calculated. Dv is the diffusion coefficient of the vapor

PAGE 19

19 (2.35x10-5 m2/s at 15oC), M is the molecular weight of the liquid (18 g/gmole), p is the liquid density (1 g/cm3), and T is the surrounding temperature in the condensation chamber (15oC). pd is the partial pressure of the va por at the particle surface, assume d to be equal to saturation vapor pressure at 15oC (0.017 atm), calculated using Equation 1-2 (Hinds 1999). Td is the temperature at the particle surface. p is the partial pressure of vapor in the surrounding air, which will be equal to the vapor pressure in th e saturation chamber (0.072 atm at 40oC) as calculated by Equation 1-2. kPa T ps 37 4060 7.16 exp (1-2) For simplicity, this estimate will not account for the effect on Td due to the release of latent heat of vaporization during condensation. Equation 1-1 can be simp lified and integrated to solve for the time required to grow the particle to the desired level unde r specific conditions. Equation 1-3 is the result. t T pp R MD ddd p ptp 82 0, 2 (1-3) An estimate indicates that the time for a 30-nm part icle to grow to 1.0 m in the condenser under such conditions would be 1.2x10-4 seconds, indicating that the am plification process is nearly instantaneous if the proper condi tions are provided in the BAU. Using the condensational amplification pro cess, Okuyama et al. (1984) found that ZnCl2 aerosols in the 0.005to 0.05m range were all amplified to larger than 0.3 m, sufficient for detection by contemporary optical techniques. Similarly, Siou tas and Koutrakis (1996) were able to sufficiently amplify 0.05m inert particles to achieve 89.3% collection efficiency in a virtual impactor with a 50% cut size of 1.4 m. The virtual impactor operated at 8 Lpm and had

PAGE 20

20 an acceleration nozzle diameter of 0.16 cm and a collection nozzle diameter of 0.24 cm. Their system used a water bath at 50oC followed by a cooling tube with a thermoelectric cooler at 8oC to obtain such results. Residence time in th e saturator was approximately 10 seconds, while residence time was less than 1 second in the co ndenser. Demokritou et al. (2002) used the condensational amplification method for high-volume collection of ultrafine inert particles for inhalation studies. With a saturator temperature of 34oC and a condenser temperature of 16oC, the research team was able to co llect 91% of all the ultrafine part icles in a system with a 50% cut size of 1.0 m, suggesting that nearly all of the ultrafine particles were grown to supermicron sizes. The system still maintained 88% collec tion in similar conditi ons (saturator at 33oC and condenser at 15oC) but with a 50% cut size of 1.8 m. The fact that this method achieved nearly identical collection efficiencies regardless of the two different cu t size values indicates that the ultrafine particles in this system frequently grew to amplified sizes greater than 1.8 m, significantly larger than the desired 1.0m size required for the virtua l impactor used in their first experiment (Demokritou et al. 2002). Conceptual Implementation of In-l ine Bioaerosol Amplification Device A conceptual device used to achieve the condensational phenomenon comprises two essential components: a humidifi cation section, in which the bi oaerosols are introduced into a saturated water vapor atmosphere and a condensation section, in which the atmosphere becomes supersaturated with vapor and leads to condensat ion on the biological nuclei. Alcohol and other low-vapor-pressure liquids are often selected to be used as the vapor for the condensational amplification process of inert particles. Achiev ing high supersaturation le vels is easier for lowvapor-pressure liquids with low molecular diffu sivity (Hering and St olzenburg 2005). If the molecular diffusivity is significantly slower than the thermal diffusivity, then the temperature in

PAGE 21

21 the condenser drops much faster than the saturation level, leading to high supersaturation levels. Achieving high supersaturation levels is more difficult when water is used as the condensing vapor because its high molecular diffusivity decreases the supersaturation levels quickly. However, water is the appropriate choice in this study to preserve the viability of the bioaerosols, as alcohol and other liquids can inactivate microorganisms and ther efore cannot be utilized as the vapor. Figure 1-1 illustrate s the conceptual design. Vacuum Figure 1-1. Bioaerosol amplif ication conceptual schematic. Here, a bioaerosol sample flow is introduced into the humidification stage, in which a heated pool of water is used to create a virtuall y saturated water vapor at mosphere at a slightly elevated temperature (Hinds 1999; Friedlander 2003). After the sample volume passes through the saturator, it enters the condenser, which consists simply of a cooled environment. This section lowers the water vapor temperature enoug h to reach supersaturat ed conditions. As the

PAGE 22

22 vapor becomes supersaturated, water condenses onto the bioaerosol nucl ei, and the biological particle subsequently begins to amplify in diameter. The effectiv e size of the amplified particle is now significantly larger than th e actual particle size of airborne virus. The grown sample exits the BAU and can be collected using common bioaerosol collection methods such as impingement or impaction. Reaerosolization Besides size and hydrophobicity, reaerosolizatio n is another factor that can affect the collection efficiency. Although the BAU can mi nimize the limitations of airborne virus sampling due to size, the nature of impingers a llows for significant potential for reaerosolization (Willeke et al. 1995; Lin et al. 1997). Although impingers are currently the established method for airborne virus sampling, the performance characteristics of the impinger are still lacking. Reaerosolization occurs when collected particles are re-entrained into the air stream; this leads to decreased collection efficiency and is a comm on concern in impinger operation (Willeke et al. 1995; Grinshpun et al. 1997). The turbulence associated with impinger operation provides enough energy for particles to become re-entrain ed into rising bubbles due to high operational flow rates. The particles re-entrained in the rising bubbles are then reaerosolized (Grinshpun et al. 1997). Theoretical models for aerosol transfer be tween bubbles and surrounding liquid are largely based on Fuchs work presented in The Mechanics of Aerosols (1964), which describes the movement of particles from the bubble to the surrounding liquid, similar to the mass transfer used in a wet scrubber. The theory sums th e contributions of mass transfer from multiple removal mechanisms, of which diffusion and inertia are the most relevant to the present research. For small, submicron particles, Brownian mo tion and thermophoresis are the main removal mechanisms. For larger particles, inertia and sedimentation become increasingly effective for

PAGE 23

23 removal of aerosols from a bubble to the surrounding liquid (Ghiaasiann and Yao 1997). The coefficients of deposition due to diffusion (d) and inertia (i) are presented in the Equations 1-4, 1-5, 1-6, and 1-7, and a comp lete list of nomenclature is provided in Table 1-1. 22 9B B iR U (1-4) 9 22 ppR (1-5) 2/1 38.1 BB dUR D (1-6) p CGR CkT D6 (1-7) Table 1-1. Nomenclature for mass transf er of particles between bubbles and liquid Symbol Title Units iCoefficient of deposition due to inertia 1/m dCoefficient of deposition due to diffusion 1/m UB Bubble velocity m/s Relaxation time s RB Bubble radius m RP Particle size m pParticle density kg/m3 Gas viscosity Pa-s k Boltzmann constant 1.3x10-23 J/K TG Temperature of gas K Cc Cunningham's correction factor Dimensionless D Aerosol diffusivity m2/s Although his work does not specifica lly refer to mass transfer in a scenario such as that in the impinger, it provides some insight into what can be expected. By analyzing these theoretical expressions, it becomes apparent that inertial removal of aerosols from bubbles into a surrounding liquid clearly plays a larger role as particle size (Rp) increases and as bubble velocity ( UB) increases. Similarly, diffusive removal of particles from a bubble into a surrounding liquid

PAGE 24

24 increases as the diffusivity coefficient increases as Rp decreases, as the bubble radius ( RB) decreases, and as UB decreases, which allows more time for diffusion to occur. The overall effect of mass transfer of aerosols from a bubble to a liquid is very similar to the passage of aerosols through filters (Fuchs 1964; Pich and Schutz 1991). The deposition of aerosols from a bubble into a liquid as a function of particle size will have a clear minimum, with large aerosols collected well by inertial transfer and small particles collected well by diffusive transfer. For a scenario with a bubble velocity of 25 cm/s and bubble radius of 0.5 cm at standard conditions (20oC and 1 atm), the calculated coefficients are displayed in Figure 1-2. 0.00E+00 5.00E-03 1.00E-02 1.50E-02 2.00E-02 2.50E-02 0.010.1110 Particle Size ( dp, micron)Coefficient of Deposition alpha,i alpha,d Total Figure 1-2. Coefficients of mass transfer due to inertia and diffusion for transfer of particles from bubbles to surrounding liquid. Both methods of mass transfer must be exam ined to study the reverse effect, such that particles in the liquid are tran sferred from the liqui d into the bubble, as is the case during reaerosolization. The case for diffusion will be a si milar situation. Mass transfer from the liquid into the bubble will increase as diffusivity of the particle increases and more time is allowed for

PAGE 25

25 diffusion to occur. The case for inertial mass transf er will be nearly opposite. In this scenario, the amount of inertia required to re -entrain a particle in to a bubble will incr ease as particle size increases. Therefore, mass transfer of aerosols into the bubble due to in ertia will increase as particle size decreases. The ma ss transfer of particles into bubbles from liquid to enable reaerosolization will be significantly larger fo r smaller particles because diffusion will be increased and the inertial requirement to re-entrain the particles will be smaller. Therefore, the diffusional deposition coefficient wi ll have the same equation, while the inertial coefficient will be inverted. For a scenario with a bubble velo city of 25 cm/s and bubble radius of 0.5 cm at standard conditions (20oC and 1 atm), the calculated coefficients are displayed in Figure 1-3. Note that the i values are predominant, and the Total trend is dictated by them. The total diffusive transfer would also depend on the c oncentration of particles in the liquid. 0.00E+00 5.00E+04 1.00E+05 1.50E+05 2.00E+05 2.50E+05 3.00E+05 0.01 0.1 1 10 Particle Size ( dp, micron)Coefficient of Deposition alpha,i alpha,d Total Figure 1-3. Coefficients of mass transfer due to inertia and diffusion for transfer of particles from liquid to bubbles.

PAGE 26

26 A study on Bacillus cereus bacterial spore reaerosoliza tion from liquid impingers by Grinshpun et al. (1997) observed that intense bubbling occurred even at 10 Lpm, although no bacterial reaerosolization was recorded. At the normal operation of 12.5 Lpm, the research group reported that several superm icrometer particles were detected downstream of the impinger outlet, and the collection effici ency decreased from 100% at 10 Lp m to 80% at 12.5 Lpm. A separate experiment noted that there was significantly more r eaerosolization of the bacterial spore as flow rate increased from 5 to 12.5 Lp m. The observation and subsequent experiment led them to conclude that the decrease in impinger collection efficiency of the Bacillus cereus bacterial spore at 12.5 Lpm was due to reaerosolization. Theoretical models presented by Fuchs (1964) and Ghiaasiaan and Yao (1997) predict this observation, by theorizing that increased bubble rise velocity introduces mo re inertia into the system, th ereby enhancing the removal of aerosols from a liquid. Lin et al. (1997) found that re aerosolization of polystyrene latex (PSL) particles in size range of 0.51 m to 1.60 m increased with sampling time due to the increased concentration of collected particles in the impinger collection liquid. After approximately 30 minutes of sampling, reaerosolization started to steadily incr ease as collection liquid was removed from the impinger. After 60 minutes of sampling, the am ount of reaerosolization significantly affected the collection efficiency of the impingers. They note that reaerosolization is much more dependent on the concentration in the collection liquid than the in coming airborne concentration. This trend can be explained with the theoretical di scussions due to the fact that the concentration gradient between the liquid and bubble will incr ease as the concentration in the collection liquid increases, thereby increasi ng diffusive mass transfer into the bubble and enhancing reaerosolization. The collection of 0.51m particles was much more impacted by

PAGE 27

27 reaerosolization than the collection of 1.02m or 1.60m particles; nearly 10% of all 0.51m particles were reaerosolized af ter 60 minutes of sampling. This experimental observation th at reaerosolization increases as particle size decreases is a physical representation of the theoretical mass tr ansfer expressions discussed previously. As particle size decreases, diffusive mass transfer increases while th e inertia required to reaerosolize the particle is less. Because of these two combin ed effects, total reaeroso lization is expected to increase significantly as particle size decreases. Regardless of the past work done on the subject of reaerosolization for specific particle types, it should be stressed that the extent of reaerosolization is of ten dependent on particle characteristics (Grinshpun et al. 1997; Lin et al. 1997; Tseng and Li 2005). Size plays a role because of increased diffusion and lower inertia re quired to reaerosolize sma ll particles, as seen by the increased reaerosolization of the 0.51m particles in comparison with the 1.02m or 1.60m particles in the study conduct ed by Lin et al. (1997). Hydr ophobicity is also one of the key components that will determine reaerosolizat ion. As previously mentioned, Tseng and Li (2005) found that hydrophilic viruses were collected 10 times more effectively in bioaerosol samplers than were hydrophobic viruses. Alt hough they did not diffe rentiate between the contributing factors to the low collection of hydrophobic viruses, it is reasonable to assume that hydrophobicity affects initial collection as well as reaerosolization; hydrophobicity makes initial collection more challenging while reaerosolization occurs more easily. Aerosolization of the collection liquid is depe ndent on its characteris tics, especially the viscosity and surface tension (Russell and Si ngh 2006). Viscosity and surface tension play a crucial role in the amount of liquid that is aerosolized, and subsequently in the amount of particles that are reaerosolized (Hogan et al. 2005; Lin et al. 1999). For inst ance, 20 mL of water

PAGE 28

28 (viscosity of 0.89 cP at 25oC) will be evaporated or aerosolized from an impinger after approximately 1.5 hours (Lin et al. 1997; Willeke et al. 1998). Similarly, 20 mL of water will be evaporated from a BioSampler after approximately 2 hours of operation, while 10 mL of glycerol (viscosity of ~930 cP at 25oC) (White 2003) in a BioSampler ha s negligible loss after 8 hours of sampling (Willeke et al. 1998). A decrease in surface tension or viscosity should result in higher aerosolization, while an increase in surface tension or viscosity should result in lower aerosolization. The addition of a surfactant to the collection liquid will decrease surface tension (Weissenborn 2006) and likely result in higher reaero solization. The addition of a more viscous, insoluble liquid on the surface of the collection liquid has the potential to suppress the bubbling due to the energy addition during the normal operati on of the impinger. H eavy white mineral oil has been found to support microbial viability well enough to serve as a pote ntial collection liquid for the BioSampler (Lin et al. 1999), but higher viscosity liquids cannot be used alone with the impinger because of the formation of large bubbles th at prevent the transfer of airborne particles to the bulk liquid (Willeke et al. 1998). Reaerosolization due to aerosoliz ation of the impinger collecti on liquid can be expected to reduce airborne virus collection efficiency. Analysis of reaerosolization will determine the extent to which the phenomenon competes with the improved physical collection efficiency of the new BAU. Objective The objective of the project was to investigate reaerosolization of viral particles from the impinger to assess the impact of this mode of loss on the capability of the new Bioaerosol Amplification Unit. The present work focuse s on characterizing reae rosolization of viral particles from impingers as a f unction of flow rate and impinger collection liquid concentration. The impact of collection liquid surface tensi on and viscosity was also explored, and virus

PAGE 29

29 reaerosolization from the AGI-30 impinger was co mpared to that from the BioSampler. The Bioaerosol Amplification Unit is introduced pr esently to demonstrate the motivation for the current work. Experimental methods and preliminary results for the BAU are included in Appendix A, and a full characterization study of the BAU will be addressed in future work.

PAGE 30

30 CHAPTER 2 EXPERIMENTAL METHODOLOGY Four sets of experim ents were conducted to analyze reaerosoli zation of viral particles from the AGI-30 impinger (Ace Glass, Inc., Vine land, NJ, USA). Table 2-1 summarizes the experimental conditions and the co rresponding purpose for each test. Table 2-1. Summary of tests and corresponding purpose to char acterize viral reaerosolization from impingers Test No. Test Name Purpose 1 Reaerosolization with PSL particles Determine the effect of liquid impinger concentration on reaero solization at 12.5 Lpm 2 Reaerosolization with MS2 Determine the effect of flow rate and liquid impinger concentration on reaerosolization 3 Effect of Surface Tension and Viscosity Determine the effect of collection liquid viscosity and surface tension on reaerosolization 4 Reaerosolization from BioSampler Determine extent of reaerosolization from BioSampler in comparison to AGI-30 impinger Experiments to determine reaerosolization as a function of flow rate and impinger concentration followed experimental methods si milar to those used in previous work to characterize bacterial reaerosolization (Willeke et al. 1998; Lin et al. 2000; Hogan et al. 2005). Figure 2-1 shows a schematic of the experimental system. A known concen tration of particles was placed in the impinger collection liquid, and th e impinger was operated with sterile air. The flow rate through the system was controlled by a rotameter. The flow exiting the impinger carried any aerosolized droplets an d reaerosolized particles downstr eam. A slip stream directed 0.6 Lpm of the impinger exhaust flow through a diffusion dryer to remove excess moisture and finally into the Scanning Mobility Particle Sizer (SMPS, Model 3936, Shoreview, Minn., USA), where reaerosolized particles were measured. A baselin e test using pure deionized water (0 PFU/mL) in the impinger collection liquid was used to confirm that the experimental setup was operating properly prior to every experi ment. The nebulizer in the baseline test hypothetically produced pure water droplets, in which case the diffusion dryer removed any

PAGE 31

31 moisture, and the SMPS registered negligible aerosol particles. Any aerosols detected in the baseline were due to low levels of salt present in the deionized water and the total reaerosolized concentrations were accordingly adjusted. Cylinder Air Figure 2-1. Reaerosolization experimental setup. Reaerosolization with PSL The first experiment utilized polystyrene late x particles (PSL, Duke Scientific, Palo Alto, Calif., 3030A, nominal 30-nm particles, density 1.05 g/cm3) in the impinger liquid to provide a preliminary analysis of reaerosolization of viral-sized particles. Airbor ne virus particles are difficult to distinguish from residual solute part icles caused by the liquid medium, and the liquid medium typically dictates the airborne virus particle size distributi on (Hogan et al. 2005). Viruses also have inherent microbial uncertainties, including loss of viability and microbial interactions, which can influence their behavior. The use of PSL particles in deionized water eliminated these uncertainties and provided a simp ler and more straightforward test. Therefore, very explicit results regarding aerosolization of viral-sized particles were obtained.

PAGE 32

32 Deionized water was used to dilute the PSL particles to the desired concentrations. The test was conducted for a range of concentrations (0.1, 1.0, 10, 100 ppm by mass or volume; =1.05 g/cm3) at a flow rate of 12.5 Lp m only, which is the recomm ended operational flow rate for the impinger (Ace Glass Inc. 2008). Reaerosolization with MS2 Although PSL particles provided explicit information regardi ng the reaerosolization of virus-sized particles, the test needed to be conducted with MS2 to see if the results from PSL translated to actual virus. Thus, the reaeroso lization experiment was then conducted much more thoroughly with MS2 virus ( Escherichia coli bacteriophage ATCC 15597-B1) in the impinger liquid to obtain results specific to the virus. MS2 is a bacteriophage that is often used as a surrogate pathog en for airborne virus testing and is an appropriate choice for use as a surrogate human pathogenic virus (AranhaCreado and Brandwein 1999). The nominal size of the MS2 b acteriophage is 27.5 nm (Golmohammadi et al. 1993); thus, it will serve as a suitable challenge for the BAU in future studies because typical collection efficiencies in the im pinger at this particle size are less than 10% (Hogan et al. 2005). MS2 is classified as a hydrophilic virus because of the absence of a lipid envelope surrounding the nucleocapsid (Vidaver et al. 1973; Madigan et al. 2003; Tseng and Li 2005). To determine the effect of flow rate a nd impinger collection liquid concentration on reaerosolization, a matrix was used with several different flow rates and concentrations as shown in Table 2-2. Twenty-five tests were run in total for the experiment, and each is numbered separately in the matrix. A few samples were also examined at an impinger collection liquid concentration of 101 PFU/mL.

PAGE 33

33 Preliminary infectivity tests were run during so me of the initial reaerosolization tests, but the results were not sensitive enough to assess any trends to the resu lts. Although extensive operation of the impinger can decrease the viabil ity of MS2 after approx imately 30 minutes of sampling (Hogan et al. 2005), the time span of the reaerosolization tests (<15 minutes) did not appear to be long enough to substantially affect viability. Any minor difference in the level of infectivity before and after the reaerosolization test could easily have b een attributed to many other circumstances that are known to affect microbiological enumeration techniques. The infectivity component was ther efore stopped because the resu lts were not providing clear information. Table 2-2. Experimental matrix to determine the effect of flow rate and concentration on reaerosolization Flow Rate (Lpm) Impinger Liquid Concentration (PFU/mL) 3 6 9 12.5 15 Baseline No. 1 No. 2 No. 3 No. 4 No. 5 102 No. 6 No. 7 No. 8 No. 9 No. 10 104 No. 11 No. 12 No. 13 No. 14 No. 15 106 No. 16 No. 17 No. 18 No. 19 No. 20 108 No. 21 No. 22 No. 23 No. 24 No. 25 Lin et al. (1997) found that increased impinge r collection liquid con centration significantly increased the amount of reaerosoli zation for PSL particles in bacter ial size ranges. Grinshpun et al. (1997) observed that reaeroso lization of bacterial spores was initiated as impinger flow rate increased. Theoretical models predict that mass transfer from the collection liquid to the bubble will increase as particle size decreases due to two reasons: (1) diffusivity will increase as particle size decreases; and (2) the inertial requirement to re-entrain particles will decrease as particle

PAGE 34

34 size decreases. The models also predict that di ffusive mass transfer from the collection liquid to the bubble will increase as the concentration gradient betwee n the collection liquid and the bubble increases. Therefore, logical hypotheses for virus partic les were that reaerosol ization would increase as both flow rate and accumulative impinger coll ection liquid concentration increased, and that reaerosolization for virus-sized par ticles would be greater than that for bacteria-sized particles. While operational flow rate is often establishe d by sampling protocol, the accumulative impinger concentration increases with sampling time. The re sults from these experiments will be able to provide recommended sampling time limits based on accumulative impinger concentration that can be established to minimize the effects of reaerosolization. Hydrophobic viruses will have different considerations than MS2, and futu re work may need to address this issue. One concern for the reaerosolization tests with MS2 was the decision to use phosphate buffered saline (PBS) or pure deionized water as the impinger liquid. Although PBS (1.8 g KH2PO4, 15.2 g K2H2PO4, and 85 g NaCl in 1L of deionized water) is often used in bioaerosol sampling to maintain bioaerosol viability, the salt aerosols formed when the liquid is aerosolized can mask the magnitude of reaerosolized virus part icles. Similarly, it is difficult to distinguish aerosolized viruses from residual solute particles formed from aer osolized liquid media, and the media typically dictates the air borne virus particle size distri bution (Hogan et al. 2005). The experiments attempted to minimize the effects of background media such as PBS and virus stock medium, but the issue was inherent to the task at hand regardless of these efforts. 0.02 mL of PBS was inherently included in 20 mL of viru s solution in the proce ss of the impinger liquid preparation.

PAGE 35

35 Effect of Surface Tension and Viscosity on Reaerosolization Once reaerosolization for virus-sized particle s was fully characterized, experiments were conducted to analyze the effects of surface tens ion and viscosity, which were hypothesized to play a crucial role in the amount of liquid that is aerosolized, and subsequently the amount of particles that were reaerosolized (Lin et al 1999; Hogan et al. 2005; Russell and Singh 2006). Two experiments were conducted to explore the effects of viscosity and surface tension for various concentrations of PSL particles (0.1, 1.0, 10, 100 ppm) at 12.5 Lpm. To determine the effect of surface tension, a surfactant (Palmolive concentrated dish liquid, ColgatePalmolive Company, New York, N.Y.) was introduced to redu ce the surface tension of the collection liquid. 0.2 mL of a diluted soap solution (composed of approximately 0.1 mL in 10 mL of deionized water) was used in each of the triplicate experiments, such that approxim ately 0.002 mL of the concentrated dish liquid was added to the impi nger liquid. The amount was sufficient to cause significant bubbling without overwhelming the impinger wi th soap bubbles under normal operation. Although the condition was predicted to increase reae rosolization, the experiment was conducted to verify the effect of surfactants. The addition of a more viscous insoluble liquid on the surface of the collection liquid has the potential to suppress the bubbling during the normal operation of the impinger. To investigate this, a small layer (0.3 mL) of heavy white mineral oil (Mineral Oil, NDC 00030559-33, E.R. Squibb and Sons, Inc. Princeton, N.J. ) was added to provide a more viscous layer, and the results were analyzed to determine wh ether reaerosolization decreased. Heavy white mineral oil was selected because it was found to ma intain bacterial viability over 8 hours (Lin et al. 1999).

PAGE 36

36 Reaerosolization with BioSampler The last experiment was a test comparing reaerosolization from the AGI-30 impinger to that from the BioSampler (SKC, Inc., Eighty-Four, P.A.). Previous research indicated that the BioSampler had significantly less reaerosolization for particles in the bacterial size range due to the swirling motion during collection (Willeke et al. 1998; Lin et al. 2000). This experiment attempted to confirm that this was also the case for particles in the virus size range. Figure 2-2 shows the BioSampler and the sw irling collection mechanism. Reaerosolization of 30-nm PSL particles from the BioSampler was compared to that from the AGI-30 impinger using conditions identical to the first experiment. The expe rimental flow rate was 12.5 Lpm, and the concentrations used were 0.1, 1.0, 10, and 100 ppm. Figure 2-2. The BioSampler. A) A schematic. B) Collection mechanism. (Lin et al. 2000)

PAGE 37

37 CHAPTER 3 RESULTS Reaerosolization with PSL Polystyrene latex (PSL) particles (30 nm) were tested at various concentrations in the impinger collection liquid at 12.5 Lpm, which is the standard operational flow rate recommended for impingers (Ace Glass, Inc. 2008). Five test s at each scenario were measured by the SMPS over a 12-minute period. Deioni zed water was the impinger collection liquid, and the results were appropriately adjusted for any particles measured during the baseline experiment (pure deionized water represented as 100 PFU/mL for graphing purposes). The measured concentrations for the subsequent tests are th erefore displayed as adjusted reaerosolized concentration. The amount of PSL particle reaerosolization c ontinued to increase as the collection liquid concentration increased, as shown in Figure 3-1. Data sets for each of the triplicate tests are included in Appendix B. An analysis of variance (ANOVA) test shows that each increase at the higher concentrations (from 1 to 10 ppm and from 10 to 100 ppm) was deemed significant ( p<0.05), while the increases between the lower c oncentrations (from 0 ppm to 0.1 ppm and from 0.1 ppm to 1 ppm) were not as significant ( p>0.05). Table 3-1 displa ys the percentages of particles reaerosolized. Less th an 1.0% of the particles c ontained in the impinger were reaerosolized at any concentration over the sampling period. Table 3-1. Percent of PSL particles reaerosolized Impinger Concentration (ppm) % Particles Reaerosolized 0.0 N/A 0.1 0.84% 1.0 0.10% 10.0 0.03% 100.0 0.01%

PAGE 38

38 Impinger Concentration (ppm) 0.01 0.1 1 10 100 Adjusted Reaerosolized Concentration (#/cm3) 2e+4 4e+4 6e+4 8e+4 1e+5 Figure 3-1. Reaerosolization as a function of impinger concentration for PSL particles at 12.5 Lpm. The lower end of the box represents the 25th percentile, the middle line represents the median, and the uppe r end of the box represents the 75th percentile. Reaerosolization with MS2 The dependence of the reaerosolization of MS2 bacteriophage particles on flow rate and impinger liquid concentration was evaluated using a diffusion dryer followed by the SMPS. Five tests at each scenario were run over a 12-minute period. Result s indicated that increasing the flow rate significantly increased the number of vi rus particles reaerosolized. A typical trend is displayed in Figure 3-2, which presents data obtained from th e SMPS at a constant impinger collection liquid c oncentration (1x102 PFU/mL) and varying flow rate.

PAGE 39

39 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 3 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 6 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 9 Lpm 10^2 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 12.5 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 15 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) Figure 3-2. Size distribu tion of reaerosolized particles as a f unction of flow rate at an impinger collection liquid c oncentration of 1x102 PFU/mL. Increased concentration, however, did not necessarily lead to an increase in reaerosolization for virus particles. Rather, the count of reaerosoliz ed virus particles increased as concentration increased until it reached a concentration of approximately 106 PFU/mL, at which

PAGE 40

40 point the reaerosolized count began to decrease. This trend is shown in Figure 3-3. These data sets maintained a constant impinger flow rate, but the concentration of MS2 in the collection liquid varied. Baseline 10 100 1000 0 5e+4 1e+5 2e+5 2e+5 3e+5 3e+5 Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10^2 10 100 1000 0 5e+4 1e+5 2e+5 2e+5 3e+5 3e+5 Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10^4Particle Diameter, dp (nm) 10 100 1000 dN/dlog(d p ) (#/cm 3 ) 0 5e+4 1e+5 2e+5 2e+5 3e+5 3e+5 10^6 10 100 1000 0 5e+4 1e+5 2e+5 2e+5 3e+5 3e+5 dN/dlog(d p ) (#/cm 3 )Particle Diameter, dp (nm) 10^8 10 100 1000 0 5e+4 1e+5 2e+5 2e+5 3e+5 3e+5 dN/dlog(d p ) (#/cm 3 )Particle Diameter, dp (nm) Figure 3-3. Size distri bution of reaerosolized particles as a function of collection liquid concentration at a flow rate of 9 Lpm.

PAGE 41

41 Results for the final reaerosolized concentrati on as a function of concentration for each of the five flow rates tested (3, 6, 9, 12.5, 15 Lpm) are displayed in Figure 3-4. Three experiments were conducted at each of the three higher flow rates (9, 12.5, and 15 Lpm), while only two experiments were conducted at the two lower flow rates (3 and 6 Lpm). The trends at the lower flow rates followed the same pattern, but the ove rall level of reaerosoliz ation was minimal in comparison to the higher flow rates. This agre es with Fuchs work, the application of which predicted that increasing flow rate will increase reaerosolization. 0.E+00 1.E+05 2.E+05 3.E+05 4.E+05 5.E+05 6.E+05 1.E+001.E+021.E+041.E+061.E+08 Impinger Concentration (PFU/mL)Adjusted Average Reaerosolized Concentration (#/cm^3) 3 Lpm 6 Lpm 9 Lpm 12.5 Lpm 15 Lpm Figure 3-4. Reaerosolization as a function of flow and impinger concentration for MS2 viral particles. Some reaerosolized particles detected were at tributed to low salt concentrations in the deionized water. These salt aeros ols were represented as the concentrations measured during the baseline experiments. The subsequent test s were adjusted by subtracting the baseline concentration from each. The resulting values ar e displayed as adjusted average reaerosolized concentration. Complete data sets for each of the presented results are included in Appendix C.

PAGE 42

42 Small variations in the experime ntal sets displayed in Appendix C can likely be attributed to microbiological interactions, virus age, and sma ll variations in virus stock compositions. Statistical analysis was run on the tests us ing the ANOVA test. The difference between reaerosolized concentrations is generally significant ( p<0.05) between the baseline (100 PFU/mL) and the middle concentrations (102, 104, 106 PFU/mL). The three middle concentrations (102, 104, 106 PFU/mL) are generally not significantly different ( p>0.05). The reaerosolized amount is generally significant (p <0.05) between the middle concentrations (102, 104, 106 PFU/mL) and the highest concentration (108 PFU/mL). The significance of the difference between the middle and hi ghest concentrations is stronger at the higher flow rates. As an example, at 9 Lpm, the difference between the baseline and 102 PFU/mL is significant ( p=0.0008), the difference between 102 PFU/mL and 108 PFU/mL is also significant ( p=0.03), but the difference between the three middle concentrations (102, 104, 106 PFU/mL) is not significant ( p=0.42). In general, reaerosolization increased significan tly as flow rate increased. The increasing trend agreed with the hypothesis that extra inertia would reaerosolize more particles. The unexpected result was the decrease in reae rosolization at concentrations higher than approximately 106 PFU/mL. For all flow rates, the highest levels of concentration resulted in a decrease in reaerosolization. Although this was not expected, the observation may be explained by changes in aggregation and surf ace tension due to changes in concentration of proteins and salt in the deionized water and virus stock. Th is will be discussed in more detail in the Discussion section. These results make it clear that reaerosolizati on of viral particles from an impinger occurs and may potentially compete with the BAU if an impinger is the bioaerosol sampler in use.

PAGE 43

43 Future testing should aim to establish recommended sampling time limits based on influent airborne virus concentrations. This will minimize the effects of reaerosolization by preventing the accumulative impinger concentration from reaching a certain threshold. Effect of Surface Tension and Viscosity on Reaerosolization Viscosity and surface tension play a crucial role in the amount of liquid that is aerosolized, and subsequently the amount of particles that are reaerosolized (Hogan et al. 2005; Lin et al. 1999; Russell and Singh 2006). Figure 3-5 displays the results for the adjust ed average reaerosolized concentration as a function of concentration for this experiment in comparison to the previous PSL experiment without soap. The percent change from without soap to with soap is also displayed for each concentration. As shown, PSL reaerosoliza tion decreased with th e soap under identical conditions within the same sampling timeframe. Co mplete data sets for each of the triplicate tests are included in Appendix D. A possible solution to reduce reaerosolization would be the addition of a more viscous, insoluble collection liquid on top of the typical collection liquid. Figure 3-6 displays the results for the adjusted average reaerosolized concentr ation as a function of concentration for this experiment in comparison to the previous experiment without oil. The percent change from without oil to with oil is also displayed for each concentration. As expected, the addition of heavy white mineral oil led to a decrease in reaerosolized particles in comparison to PSL reaerosolization under identical c onditions without the oil within the same sampling timeframe. Complete particle size distributions and reaerosoli zed concentrations for each of the three tests are included in Appendix E.

PAGE 44

44 0.E+00 1.E+04 2.E+04 3.E+04 4.E+04 5.E+04 6.E+04 0.00.11.010.0100.0Impinger Concentration (ppm solids)Adjusted Average Reaerosolized Concentration (#/cm^3) Without Soap With Soap-70% -77% -63% -48% Figure 3-5. Comparison of the reaerosolization of PSL at 12.5 Lpm with and without soap present. 0.E+00 1.E+04 2.E+04 3.E+04 4.E+04 5.E+04 6.E+04 0.00.11.010.0100.0Impinger Concentration (ppm solids)Adjusted Average Reaerosolized Concentration (#/cm^3) Without Oil With Oil-45% -65% -47% -85% Figure 3-6. Comparison of the r eaerosolization of PSL at 12.5 Lp m with and without oil present. The mode sizes for these tests are presented in Table 3-2. The PSL particles reaerosolized in the test with oil were generally larger than those reaerosolized from the pure water or water

PAGE 45

45 with soap tests. As heavy white mineral oil is no t as volatile as water, it is possible that a small layer of oil remained after the evaporative components of the aerosol had been evaporated. This could resulted in larger mode sizes. Table 3-2. Comparison of mode sizes for PSL particles under different experimental conditions Impinger Concentration (ppm) Average Mode with no Additives (nm) Average Mode with Soap (nm) Average Mode with Oil (nm) 0.0 12 11 81 0.1 14 23 70 1.0 33 25 88 10.0 37 36 37 100.0 54 58 66 Reaerosolization with BioSampler Reaerosolization of 30-nm polys tyrene latex (PSL) particle s from the BioSampler at various concentrations in the co llection liquid was compared to that from the AGI-30 impinger. The flow rate was 12.5 Lpm, which is the standard operational flow rate for both the AGI-30 and the BioSampler (Ace Glass Inc. 2008; SKC, Inc. 2008). The results are shown in Figure 3-7 as adjusted average reaerosolized concentration from the three tests. Percent change from impinger to BioSampler at each concentration is also di splayed. The level of reaerosolization of PSL particles from the BioSampler was significan tly lower than that from the impinger. Reaerosolization from the BioSampler was genera lly two orders of magnitude lower than that from the impinger under identical conditions. Complete data sets for each of the three experiments are included in Appendix F.

PAGE 46

46 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 0.00.11.010.0100.0Impinger Concentration (ppm solids)Average Average Reaerosolized Concentration (#/cm^3) Impinger BioSampler-99.99% -100% -99.7% -99.98% Figure 3-7. Comparison of the reaerosoliza tion of PSL at 12.5 Lpm from impinger and BioSampler.

PAGE 47

47 CHAPTER 4 DISCUSSION Reaerosolization with PSL Analysis of the reaerosolization of 30-nm PSL particles prov ided very direct documentation of viral-sized particle reaerosoliza tion. Reaerosolization of virus particles is complicated by proteins, salts, and complex microbial interactions, all of which make it difficult to distinguish pure reaerosolized virus particles. In contrast, the results from the PSL particles in deionized water are very explicit: reaerosolizatio n of viral-sized particle s occurs and can be a mode of loss. This agrees with previous rese arch on reaerosolization completed for particles in the bacterial size range (Grinshpun et al. 1997; Lin et al. 1997; Willeke et al. 1998). Another benefit of conducting the experiment wi th PSL particles is that comparison with previous work can be made. The level of reaero solization for virus-sized particles was compared to previous work done on reaeroso lization for particles in the b acterial size range under similar conditions. Willeke et al. (1998) used 0.5and 1.0-m PSL particles in the AGI-30 impinger with 20 mL of deionized water, and Lin et al. (2000) used Pseudomonas fluorescens vegetative cells ( da of 0.8 m) and Bacillus subtilis spores ( da of 1.0 m) in 20 mL of deionized water. The study with PSL particles had 108 particles/mL in the impinger liquid and obtained peak mode concentrations of reaerosolized particles less than 60 particles/cm3. The study with bacteria had 108 particles/mL of the respect ive species in the impinger liquid. The experiment with P. fluorescens obtained reaerosolized peak mode co ncentrations less than 5 particles/cm3, and the experiment with B. subtilis obtained peak mode concentrations less than 55 particles/cm3. B. subtilis was more hydrophobic than P. fluorescens, which may have caused the increased reaerosolization (Lin et al. 2000). In co mparison, the present research used 109 particles/mL (0.1 ppm) and obtained ma ximum modes around 104 particles/cm3 downstream of the impinger.

PAGE 48

48 A simple analysis made by normalizing these results for comparison shows that a 30-nm particle is much more likely to be reaerosolized than a 0.5or 1.0-m particle at similar number concentrations. Table 4-1 displays the comparison. Note that the simple analysis normalizes the peak mode concentration to the impinger concentration, as total reaerosolized concentrations for the previous literature were not reported. This corroborates the expectation based on Fuchs work (1964) that smaller particles are more lik ely to be reaerosolized because the energy requirements to re-entrain them are lowe r and their diffusivity is higher. Table 4-1. Comparison of reaerosolization for b acteria-sized particles to that for virus-sized particles Current work (0.1 ppm) Current work (100 ppm) Willeke et al. 1998 Lin et al. 2000 Lin et al. 2000 Particle Type PSL PSL PSL P. fluorescens B. subtilis Particle Size (m) 0.03 0.03 0.5, 1.0 0.8 1.0 Approximate Number Concentration in Impinger (particles/mL) 1x109 1x1012 1x108 1x108 1x108 Approximate Mass Concentration in Impinger (g/mL) 1x10-7 1x10-4 5x10-5 2x10-5 5x10-5 Peak Mode Concentration 8,900 120,000 60 5 55 Normalized Peak Mode Concentration 8.9x10-6 1.2x10-7 6.0x10-7 5.0x10-8 5.5x10-7

PAGE 49

49 While it is clear that there is more reaerosolization of the 30-nm particles for similar number concentrations, the analysis was also conducted for similar mass concentrations, assuming unit densities for the ba cterial species. The 100 ppm (1012 particles/mL) scenario in the present research had a mass concentration sim ilar to that in Willeke et al. 1998. Based on mass concentrations, the normalized reaerosolization is similar. This could potentially be explained by aggregation of the PSL pa rticles at very high concentrations. Although reaerosolization occurs and results in downstream total concentrations that are seemingly high, it should be noted that the amount of reaerosolization was less than 1% for all of the present work, as previously displayed in Table 3-1 in the Results section. This number cannot be compared to previous research for bacteria-sized partic les because the total concentrations for those scenarios were not provided in the literature. While the overall percentage of reaerosolization for the present work is not substantial, it is important to remember that these tests spanned only 12 minutes or less. Reaerosolization is expected to increase with sampli ng time (Lin et al. 1997). However, the present work indicates that reaerosolization may not be significant (<1% ) over short sampling periods Therefore, if the BAU can overcome initial physical collection lim itations due to partic le size, significant improvements in airborne virus sampling can be achieved. Further analysis of the PSL reaerosolization data attempted to provide more insight into the aerosolization from the impinger. Although the im pinger is not considered to be a traditional aerosol generator like the nebuliz er or atomizer, it can be considered to be one due to the formation of liquid aerosols during operation. Equation 4-1 can be used to determine whether reaerosolization was volumetrically proportional to the impinger concentration (Hinds 1999). VdPFVV (4-1)

PAGE 50

50 The equation states that aerosols with particle volume concentration, Vp, can be generated based on the droplet volume concentration, Vd, produced by the impinger and the volume fraction of solid material in the impinger liquid, Fv. Table 4-2 provides th e volumetric analysis. Table 4-2. Estimated droplet volume generated from impinger Droplet Volume, nm3/cm3 Impinger Concentration (ppm) Experiment 1 Experiment 2 Experiment 3 Average 0.1 1.1x1015 1.7x1015 6.2x1014 1.1x1015 1.0 3.0x1014 2.7x1014 9.3x1013 2.2x1014 10.0 2.0x1014 1.8x1014 1.7x1014 1.8x1014 100.0 2.3x1014 2.5x1014 1.3x1014 2.0x1014 As shown, the consistency between experiment s was strong. At the lowest concentration (0.1 ppm), the average droplet volume was signifi cantly higher than the higher concentrations (1.0, 10, 100 ppm). This was likely due to the str onger effect of the residu al concentration in the deioninzed water on the lowest concentration. The higher concentrations showed strong consistency, with approximately 2x1014 nm3/cm3 of liquid generated fr om the impinger at 12.5 Lpm. This confirms the hypothesis that the im pinger was essentially operating as an aerosol generator and was able to consistently generating a specific droplet volume. Reaerosolization with MS2 The results from the reaerosolization of MS 2 convey the extent to which this phenomenon is a concern for bioaerosol samplers, especially in the submicrometer size range. The increase in reaerosolization as a function of flow rate is not surprising; th e result is in agreement with previous work for bacteria by Grinshpun et al. (1997) as well as theoretical models based on Fuchs (1964) work on aerosol transfer between bubbles and surrounding li quid, which indicates

PAGE 51

51 that increased bubble rise velocity introduces more inertia into the sy stem, thereby causing reaerosolization. In contrast, the trends observed as concentr ation increased were unexpected. The initial increase in reaerosolization with increased concen tration seems logical, but the final decrease in reaerosolization was unexpected based on the litera ture review. Most bioaerosols in natural systems do not generally exist in very high concen trations, and the time limitations on impingers due to the evaporative nature of the collection liquid prevent long-term sampling (Lin et al. 1997). Thus, impinger liquid concentrations as high as 108 PFU/mL are unlikely in most sampling scenarios. Regardless, the phenomenon is interesting, and science compels further explanation. A possible explan ation was that the viruses might have been present in an aggregated state at the highest concentrations, thereby increasing the effective particle size and making reaerosolization more difficu lt. Another possibili ty was that the addition of virus stock media to the collection liquid affected the su rface tension or viscosity of the liquid. Investigation into the subject of viral aggregation provide s information about the frequent state of aggregation based on viral and external factors. In aqueous scenarios, it is common for viruses to aggregate at high concentrations (Floyd and Sharp 1977; Grant 1994). External factors, such as salt concentration and pH, can also affect th e level of aggregation (Floyd and Sharp 1977; Floyd and Sharp 1978). Generally, aggregation is dependent on the virus concentration, ionic strength, and pH of the liquid. Virus concentration in an aqueous solution c ontributes to the leve l of aggregation. Aggregation occurred when Floyd and Shar p (1977) prepared re ovirus and poliovirus suspensions with 1:10 dilutions in dei onized water from stock solutions of 7x1011 and 2x1012 particles/mL, respectively, as physically counted by electron micr oscopy. Any further dilution

PAGE 52

52 (1:100 or 1:1000) resulted in dispersed viruses. The concentration of viruses in natural water is typically very low, but this is not always the case in laboratory scenarios, including the concentrations used in the present work. The 108 PFU/mL impinger liquid concentrations possibly had total particle concentrat ion sufficient to cause aggregation. Literature suggests that an increase in salt concentration can significantly decrease aggregation. Floyd and Sharp (1977) found that the level of ionic strength required for their version of PBS to prevent aggregation was appr oximately 10 mM for poliovirus; they concluded that aggregation could occur even with appreciabl e salts present. As only a very small amount of PBS was added (0.01 mL 10X PBS) in the presen t study, there probably wa s not sufficient salt concentration to prevent viral aggregation. Hydrophobic interactions between proteins fr om neighboring viruses also lead to the formation of aggregates. Although MS2 is labele d a hydrophilic virus because of the absence of a lipid envelope surrounding the nucleocapsi d, it can still experience these hydrophobic interactions in aqueous solutions (Thomas et al. 1998; Hogan et al. 2004). In fact, Shields and Farrah (2002) found that MS2 experienced stro ng hydrophobic interactions during adsorption to solids, even though the absence of the lipid envelope indicates general hydrophilic behavior. The hydrophobic tendencies of the MS2 might be espe cially manifested at high concentrations. The effect of pH on aggregation is based on th e isoelectric point of the virus. Generally, pH values below the isoelectric point result in aggregation, while pH values above generally do not (Floyd and Sharp 1977; van Voorthuizen et al. 2001). The isoel ectric point of MS2 bacteriophage is 3.9 (van Voorthuizen et al. 200 1), and the pH of the impinger liquid was close to neutral. Thus, the pH in this scenario probab ly did not contribute to potential aggregation of viral particles as much as the other factors.

PAGE 53

53 As previously discussed in the Introducti on, solution viscosity and surface tension can affect the amount of liquid aerosolized during impinger operation (Hogan et al. 2005; Lin et al. 1999). Therefore, viscosity and su rface tension also affect the numb er of particles reaerosolized. The addition of salts to a solu tion often results in an increase in surface tension and viscosity (Weissenborn 2006). The addition of salts required for preservation of the virus stock solution at the highest concentrations could have increased the surface tension or viscosity of the solution, thereby reducing aerosoli zation of the liquid. In summary, the high concentration of MS2, the limited salt, and the tendency for MS2 to initiate hydrophobic interactions with one another indicates that vi ral aggregation in the impinger liquid is a plausible reason to explain the phe nomenon of decreased re aerosolization at high (>106 PFU/mL) impinger liquid concen trations. Another possible r eason is that the components of the virus stock solution could have affected th e surface tension or viscosity of the solution and decreased reaerosoliz ation at high (>106 PFU/mL) impinger liquid concentrations. Although reaerosolization was observed throughout the du ration of these tests, reaerosolization is expected to increase with sa mpling time. This is due to an increase in accumulative concentration in the impinger collec tion liquid as operation time increases and the amount of collection liquid decreases because of evaporation and aerosolization (Lin et al. 1997). This places yet another constraint to the use of the AGI-30 impinge r for airborne virus sampling. The likelihood of reaching very high impinger concentrations (>105 PFU/mL) is also unlikely before all of the 20 mL of initial collection li quid evaporates. Thus, reaerosolization of MS2 may not be as significant a contri buting factor in poor airborne vi rus sampling as the effect due to virus particle size.

PAGE 54

54 Effect of Surface Tension and Viscosity on Reaerosolization Although the surfactant did not dire ctly result in an increase of reaerosolized particles as predicted by the literatu re review, there was more bubbling with in the impinger collection liquid. One possible reason the observed increase in bub bling did not translate to an increase in reaerosolized particles was that the presence of thin layers of soap film formed prevented the transfer of particles. The th in layers of soap moved thr ough the impinger vessel towards the exhaust, similar to the movement of soap through a bubble meter. These thin layers of soap film would trap reaerosolized particles. Although the impinger cannot be successfully op erated with only highe r viscosity liquids as the collection liquid (Willeke et al. 1998), the addition of an insoluble, more viscous layer was thought to potentially be able to suppress bubbling. Unexpectedly, the high energy and turbulence associated with the traditional impinge r resulted in the emulsification of the heavy white mineral oil shortly after impinger operatio n ensued. Although the addition of the more viscous layer might have decreased reaerosoliz ation, the background concentration increased significantly at higher particle si zes and resulted in a bimodal distribution. This trend is displayed in Figure 4-1. The la rger distribution is likely due to the aerosolization of the nonvolatile oil, which could not be dried by the diffusion dryer prior to entering the SMPS for measurement. However, this distribution is co mpletely overwhelmed when high concentrations of PSL particles are present, as displayed in Appendix D.

PAGE 55

55 10 100 1000 0.0 1.0e+3 2.0e+3 3.0e+3 4.0e+3 With Oil 12.5 Lpm 0 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+3 2.0e+3 3.0e+3 4.0e+3 Without Oil 12.5 Lpm 0 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) Figure 4-1. Baseline at 12. 5 Lpm from impinger with oil and without oil present. Reaerosolization with BioSampler The significant decrease in reae rosolization for 30-nm PSL part icles from a BioSampler in comparison to an AGI-30 impinger corroborated past work done on the subj ect for bacterial size particles. Willeke et al. (1998) used 0.5and 1.0 m PSL pa rticles in 20 mL of deionized water in both the AGI-30 impinger and the BioSampler to compare reaerosolization from the two methods of liquid impingement. The BioSampler performed significantly better, with a peak mode concentration of reaerosolized particles lower than 5 particles/cm3. In comparison, the AGI-30 had a peak mode concentration nearly 60 particles/cm3. Similarly, Lin et al. (2000) used Pseudomonas fluorescens vegetative cells ( da of 0.8 m) and Bacillus subtilis spores ( da of 1.0 m) in 20 mL of deionized water to compare r eaerosolization between the BioSampler and the AGI-30. For both bacterial cells, the peak mode concentration from the BioSampler was about 20% of that from the AGI-30. Ta ble 4-3 displays results from th e current work as well as the previous work using bacter ia-sized particles. The improved performance in reduced reaeroso lization by the BioSampler is attributed to the swirling motion in which the air travels, as sh own in Figure 2-2. The entrance of the air in

PAGE 56

56 the AGI-30 is perpendicular to th e base of the collection vessel which allows for much more bubbling and aerosolization of the collec tion liquid (Willeke et al. 1998). Table 4-3. Comparison of present work to past research on reaerosolization from AGI-30 Impinger and BioSampler Current work Willeke et al. 1998 Lin et al. 2000 Lin et al. 2000 Particle Type PSL PSL P. fluorescens B. subtilis Particle Size (m) 0.03 1.0 0.8 1.0 Impinger Concentration (particles/mL) 1x1012 1x108 1x108 1x108 Peak Mode Concentration from AGI-30 (particles/cm3) 120,000 60 5 55 Peak Mode Concentration from BioSampler (particles/cm3) 1,600 4 1 10 Ratio of Peak Mode Concentrations from BioSampler to AGI-30 0.013 0.067 0.200 0.182 Although Hogan et al. (2005) noted that even the BioSampler cannot exceed 10% physical collection in the 20 nm size rang e, the use of a BioSampler minimizes reaerosolization of virus-sized particles. Thus the low physical coll ection efficiency for virus particles seen in the BioSampler is likely due to insufficient initial phy sical collection rather than reaerosolization. This is likely due to the ability of nanosized particles to tolerate the centrifugal collection motion and escape collection. Particles greater than 0.5 m are collected well (>80% collection efficiency) with the BioSampler (Willeke et al. 1998). However, Willeke et al. (1998) confirmed that physical collection efficiency of 0.3-m particles was higher for the AGI-30 than for the

PAGE 57

57 BioSampler, indicating that the AGI-30 impinger has better physical collection of the smaller particles. Thus, the low physical collection of the AGI-30 impinger shown by Hogan et al. (2005) can apparently be attr ibuted to both poor physical collection and higher amounts of reaerosolization. Based on this information, the use of the BAU in conjunction with the BioSampler should significantly improve ai rborne virus sampling by increasing physical collection and minimizing reaeroso lization issues, respectively.

PAGE 58

58 CHAPTER 5 CONCLUSION Reaerosolization occurs for viru s-s ized particles and is more of a concern for viral-sized particles than for bacterial-sized particles, as demonstrated explicitly by the use of 30-nm PSL particles. Reaerosolization increases as flow rate increases, due to the additional energy introduced to the system. However, increased concentration does not ne cessarily lead to an increase in reaerosolization fo r virus particles. Rather, reaerosolization increases as concentration increases until it reaches a concentration of approximately 106 PFU/mL, at which point reaerosolization begins to decrease. Althou gh such high concentrations are unlikely due to typical airborne virus concentrations and liq uid impingement sampling limitations, science compels further exploration. The observed phe nomenon likely results from the aggregation of viral particles or the increase of surface tension or viscosity at high concentrations. Further investigation into the effects of surface tension and viscosity on reaerosolization indicates that both properties affect aerosolization from the impinger. While the addition of soap as a surfactant increases bubbling, it decreases reae rosolization over the time frame studied, possibly due to the formation of thin soap films that prevents reaerosolization. The addition of heavy white mineral oil to provide a viscous surface decrea ses reaerosolization. Further work to verify the effects of the soap and oil on vi rus viability needs to be conducted. In summary, airborne virus sampling is lim ited by primary partic le size, hydrophobicity, and reaerosolization. While hydrophobicity is not specifically addressed in this work, sampling limitations caused by particle size and reaerosolizati on are addressed in this research. The use of the Bioaerosol Amplification Unit has the potenti al to minimize the issue related to size by amplifying the particle size. Reaerosolization of virus-sized particles doe s not appear to be a

PAGE 59

59 significant mode of loss during 15 minutes of sampling for most t ypical sampling scenarios, and it can be minimized by preventing high impinger concentrations and using the BioSampler. Recommendations to improve airborne virus sampling based on the present work include the use of the BAU in conjunction with a Bi oSampler. Based on previous work, it is recommended that sampling last no more than approximately 30 minutes. Although much more work is required to drastically improve airbor ne virus sampling, the current work provides a solid base for the future of the fiel d by characterizing reaerosolization.

PAGE 60

60 APPENDIX A BIOAEROSOL AMPLIFICATION UNIT INFORMATION The Bioaero sol Amplification Unit (BAU) serves as the main motivation for the entire project. Reaerosolization needed to be characte rized to understand how it will affect the BAU; because of this, reaerosolization therefore served as the primary focus this paper. However, significant work has also been completed on the design, construction, and preliminary evaluation of the BAU and is provided here to give more details about the motivation and future of the project. Bioaerosol Amplification Unit Design A prototype of the aforementi oned Bioaerosol Amplification Unit was developed with the intention of its evaluation in subs equent studies. The prototype design consists of two parallel aluminum square tubes, 1 inch in diameter and 3.01 ft in length. The length of the tubes was determined by ensuring that sufficient cooling occu rred to decrease a 12 -Lpm air stream from 40oC to 25oC if the condensation tube surface was 10oC. The saturator tube and the condenser tube are heated and cooled, respectively, using tw elve 8-W Peltier thermoelectric heat pumps. The base temperature is monitored by thermocouples in both the hot and co ld bases, and the tube surface is considered to be equal to the base te mperature. By varying the amount of voltage and current supplied to the Peltier array, the temperature differences between the humidification and condensation chambers can be controlled. A schematic diagram of the first prototype B AU is shown in Figure A-1. The air sample enters the saturation chamber, where water vapor is transferred to the air stream from the water chamber via a porous hydrophilic evaporative mate rial produced by Porex, Inc. The saturation chamber is designed to achieve a relative humidit y of approximately 90% to sufficiently prepare it for the condensation chamber. Humidity is measured before and after the saturation chamber

PAGE 61

61 to ensure this goal was achieved. As the air sample passes into the condensation chamber, the temperature drop decreases the vapor pressure and induces su persaturation conditions. Once supersaturation is achieved, the water vapor pref erentially condenses onto the bioaerosols, which serve as condensation nuclei. The particles ar e subsequently amplified system and can be collected more effectively by the pr eferred bioaerosol sampling method. A B Insulation 1" Porous evaporative surface Water supply channel Peltier heat pump H C Humidifier/Saturator Condenser Ulrafine Bioaerosols Grown Bioaerosols Variable Power Supply Pump & reservoir Humidifier/SaturatorCondenser THotTColdRH Figure A-1. BAU prototype schema tic. A) Overview of system. B) Cross-sectional view of humidification and condensation chambers. Experimental Methodology Three experiments will be conducted to evalua te the BAU. The tests will evaluate the success of the BAU and provide insight into how this novel unit can be applied for future research and public health uses. Table A-1 su mmarizes the experimental conditions and the corresponding purpose for each test.

PAGE 62

62 Table A-1. Summary of tests and corresponding purpose to evaluate BAU Test No. Test Name Purpose 1 Inert Particle Amplification Conf irm successful design and construction 2 Physical Collection Challenge Confirm improved physical collection efficiency 3 Viable Collection Challenge Ev aluate airborne virus sampling Inert Particle Amplification Firstly, the amplification of inert aerosols in the BAU will be evaluated. PSL and sodium chloride aerosols will be used as challenge ae rosols to confirm that the design and construction of the system was properly completed. By using inert particles, the viab ility issue associated with bioaerosols will be eliminated. The experimental setup is displayed in Figure A-2. Humidifier/Saturator Condenser Vent Cylinder Air RH meter Figure A-2. Experimental setup fo r inert particle amplification.

PAGE 63

63 A six-jet Collison nebulizer (Model # CN25, BGI, Inc.) with a flow rate of 5.5 Lpm will generate the inert aerosols that will serve as the challenge aerosols. The aerosol sizes will be measured before and after the BAU by the scan ning mobility particle si zer (SMPS, Model 3936, Shoreview, Minn., USA) to determine the degree of amplification. Given that the limits of the SMPS configuration are in the subm icron range, a different particle sizer may be required if the particles grow to supermicron sizes. A control will be run to measure particle sizes with the BAU in line but not in operation. Three PSL particle sizes will be used to evaluate the success of the system at multiple sizes: 30-nm, 64-nm, and 100-nm particles. The sodi um chloride aerosols will also challenge the system at similar sizes. The sodium chloride ae rosols will be generated by nebulizing a sodium chloride solution. The droplets of nebulized solu tion will be dried in th e diffusion dryer and will form sodium chloride aerosols. Equation A-1 will be used to estimate the concentration of sodium chloride required in solution to produce the desired aerosol size (Hinds 1999). 3/1 vdpFdd (A-1) The equation states that aerosols with diameter, dp, can be generated based on the droplet diameter, dd, produced by the nebulizer and the volume fraction of solid material in solution, Fv. The six-jet Collison nebulizer operating at a flow rate of 5.5 Lpm will generate an aerosol approximately 2.1 m in size (Hinds 1999). For example, to obtain a sodium chloride aerosol approximately 350 nm in size, 3.5 g/L of sodium chloride will be dissolved in the nebulizer solution to achieve an Fv of approximately 0.0035. The Inert Particle Amplification test will de termine the amplification capability of the BAU. Since the PSL particles are hydrophobi c and the NaCl aerosols are hydrophilic, the experiment will also provide insight into the effect of hydrophobicity on the process.

PAGE 64

64 Physical Collection Challenge Once the amplification process has been d eemed successful in the Inert Particle Amplification test, the next test will evaluate the improved physic al collection efficiency of the BAU. Figure A-3 displays the experimental setu p for the Physical Collection Challenge. The same challenge aerosols (PSL and sodium chloride ) will be used, but an impinger will sample the aerosols downstream of the BAU, and physical collection efficiency will be determined. Humidifier/Saturator Condenser Vent Cylinder Air RH meter Figure A-3. Experimental setup for phys ical collection efficiency testing. For the PSL particles, physical collection efficiency will be determined by using the SMPS to measure particle concentr ation upstream of the BAU, CUP, and downstream of the impinger,

PAGE 65

65 CDOWN. The physical collection effici ency for the PSL particles, EC, will be calculated with Equation A-2 (Willeke et al. 1998). UP DOWN UP DOWN UP CC C C CC E 1 (A-2) The physical collection efficiency of the sodium chloride aerosols will be determined by using an ion chromatograph (ICS-1500, Dionex Co rporation) to measure the concentration of sodium in the impinger collection liquid. The control for each test aerosol will follow the same experimental procedure but will not utilize the BAU, such that th e flow will still go through the unit although it will not be in operation. The improved physical collection efficiency of as a result of the BAU will be evaluated by comparing the control results with the aerosol challenge results Viable Collection Challenge Once improved physical collection efficiency is confirmed, the next step will be to use bioaerosols as the challenge aerosol and introduce the variability associat ed with microorganisms into the system. Using the same experimental se tup as the physical collection efficiency test shown in Figure A-3, MS2 virus (Escherichia coli bacteriophage ATCC 15597-B1) will serve as the challenge aerosol viruses. Physi cal collection efficiency will be determined by SMPS measurements of the inlet and outlet stream s (Equation A-2), and viral enumeration of the impinger collection liquid will determine viable collection efficiency. The improved viable collection efficiency of the BAU will be eval uated as collection enrichment. Collection enrichment, CE will be calculated with Equation A-3, which compares the viable collection of the BAU ( PFUBAU) to that of the control ( PFUControl). Control BAUPFU PFU CE (A-3)

PAGE 66

66 MS2 is a bacteriophage that is often used as a surrogate pathog en for airborne virus testing and is an appropriate choice for use as a surrogate human pathogenic virus (AranhaCreado and Brandwein 1999). The size of the MS2 bacteriophage at 27.5 nm (Golmoha mmadi et al. 1993) is a suitable challenge for the BAU because typical collection efficiencies in the impinger at this particle size are less than 10% (Hogan et al. 2005). The virus stock suspension is obtained by combining freeze-dried MS2 bact eriophage with approximately 10 mL of filtered deionized water to reach a stock concentration of 1089 PFU/mL. Approximately 0.1.2 mL of the virus stock suspension will be added to 50 mL of sterile deionized water, which has an approximate concentration of 1056 PFU/mL. This solution will be placed in the Collison nebulizer and will be used to generate the airborne virus for the experiments. The MS2 medium for the viral enumeration analysis will be prepared by gently mixing 1.0 g tryptone, 0.1 g yeast extract, 0.1 g D-glucose, 0.8 g NaCl, and 0.022 g CaCl2 into 100 mL of distilled water in a 250-mL flask. The medium will be autoclaved at 121oC for 30 minutes to ensure sterility. The MS2 agar will be prepar ed by gently mixing 3.0 g tryptone, 0.3 g yeast extract, 0.3 g glucose, 2.4 g NaCl, 0.066 g CaCl2, and 0.3 g of Bacto-agar into 300 mL of distilled water in a 500-mL flask. The mixed agar is autoclaved at 121oC for 30 minutes to achieve sterility. Preliminary Results Although only recently initiated, preliminary re sults from the evaluation of the BAU are displayed in Table A-2, which di splays viable collection efficien cy. Samples were collected in an AGI-30 impinger operated at 12.5 Lpm. When the BAU was in operation, the temperature difference was 20oC between the saturator tube surface (approximately 40oC) and the condenser tube surface (approximately 20oC). Residence time was approximately 2.8 seconds in each

PAGE 67

67 segment. For the nebulizer soluti on, 0.2 mL of MS2 virus stock (1x1010 PFU/mL) was added to 100 mL of sterile deionized wate r. Therefore, the nebulizer solution was approximately 2x107 PFU/mL. Relative humidity of the system at the time is unknown due to technical difficulties with the RH meters. Table A-2. Preliminary results fr om viral aerosol sampling using BAU Sampling time (30 min) Bioaerosol Amplification Unit Status Concentration (PFU) Mean Concentration (PFU) Control Off 76,000 Control Off 80,000 78,000 Experiment On 166,000 Experiment On 131,000 Experiment On 143,000 146,000 The preliminary results from the airborne MS2 virus sampling with the BAU in operation are promising, nearly doubling th e viable collection efficiency with an 87% increase. Future experiments will continue to address all phases of the evaluation process explained previously.

PAGE 68

68 APPENDIX B REAEROSOLIZATION WI TH PSL DATA SETS Test 1: 03/07/2008a 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 1.0E+04 2.0E+04 3.0E+04 4.0E+04 5.0E+04 6.0E+04 7.0E+04 8.0E+04 0.00.11.010.0100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure B-1. Size distribution and average concentration of PSL at 12.5 Lpm (03/07/08a).

PAGE 69

69 Test 2: 03/07/2008b 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 1.0E+04 2.0E+04 3.0E+04 4.0E+04 5.0E+04 6.0E+04 0.0 0.1 1.0 10.0 100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure B-2. Size distribution and average concentration of PSL at 12.5 Lpm (03/07/08b).

PAGE 70

70 Test 3: 04/02/2008 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 5.0E+03 1.0E+04 1.5E+04 2.0E+04 2.5E+04 3.0E+04 3.5E+04 4.0E+04 4.5E+04 0.0 0.1 1.0 10.0 100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration(#/cm^3) Figure B-3. Size distribution and average concentration of PSL at 12.5 Lpm (04/02/08).

PAGE 71

71 APPENDIX C REAEROSOLIZATION WI TH M S2 DATA SETS Flow Rate: 3 Lpm Test 1: 03/15/2007a 10 100 1000 0.0 5.0e+2 1.0e+3 1.5e+3 2.0e+3 2.5e+3 3 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+2 1.0e+3 1.5e+3 2.0e+3 2.5e+3 3 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+2 1.0e+3 1.5e+3 2.0e+3 2.5e+3 3 Lpm 10^4Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+2 1.0e+3 1.5e+3 2.0e+3 2.5e+3 3 Lpm 10^6Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+2 1.0e+3 1.5e+3 2.0e+3 2.5e+3 3 Lpm 10^8Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 1.0E+02 2.0E+02 3.0E+02 4.0E+02 5.0E+02 6.0E+02 7.0E+02 8.0E+02 9.0E+02 1.0E+03 1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+08 Impinger Collection Liquid Concentration (PFU/mL)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure C-1. Size distri bution and adjusted avg concentra tion of MS2 at 3 Lpm (03/15/07a).

PAGE 72

72 Test 2: 03/15/2007b 10 100 1000 0.0 1.0e+3 2.0e+3 3.0e+3 4.0e+3 3 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+3 2.0e+3 3.0e+3 4.0e+3 3 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+3 2.0e+3 3.0e+3 4.0e+3 3 Lpm 10^4Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+3 2.0e+3 3.0e+3 4.0e+3 3 Lpm 10^6Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+3 2.0e+3 3.0e+3 4.0e+3 3 Lpm 10^8Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 5.0E+02 1.0E+03 1.5E+03 2.0E+03 2.5E+03 1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+08 Impinger Collection Liquid Concentration (PFU/mL)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure C-2. Size distribution and adjusted avg concentrati on of MS2 at 3 Lpm (03/15/07b).

PAGE 73

73 Summary of 3 Lpm Impinger Concentration (PFU/mL) 1e+01e+11e+21e+31e+41e+51e+61e+71e+81e+9 Adjusted Final Reaerosolized Concentration (#/cm^3) 1e+0 1e+1 1e+2 1e+3 1e+4 Figure C-3. Summary of adjusted aver age concentration of MS2 at 3 Lpm.

PAGE 74

74 Flow Rate: 6 Lpm Test 1: 03/15/2007 10 100 1000 0.0 5.0e+3 1.0e+4 1.5e+4 2.0e+4 2.5e+4 3.0e+4 6 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+3 1.0e+4 1.5e+4 2.0e+4 2.5e+4 3.0e+4 6 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+3 1.0e+4 1.5e+4 2.0e+4 2.5e+4 3.0e+4 6 Lpm 10^4Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+3 1.0e+4 1.5e+4 2.0e+4 2.5e+4 3.0e+4 6 Lpm 10^6Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+3 1.0e+4 1.5e+4 2.0e+4 2.5e+4 3.0e+4 6 Lpm 10^8Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 2.0E+03 4.0E+03 6.0E+03 8.0E+03 1.0E+04 1.2E+04 1.4E+04 1.6E+04 1.8E+04 2.0E+04 1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+08 Impinger Collection Liquid Concentration (PFU/mL)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure C-4. Size distribution and adjusted avg concentrati on of MS2 at 6 Lpm (03/15/07).

PAGE 75

75 Test 2: TBD Summary of 6 Lpm Impinger Concentration (PFU/mL) 1e+01e+11e+21e+31e+41e+51e+61e+71e+81e+9 Adjusted Final Reaerosolized Concentration (#/cm^3) 1e+0 1e+1 1e+2 1e+3 1e+4 1e+5 1e+6 Figure C-5. Summary of adjusted aver age concentration of MS2 at 6 Lpm.

PAGE 76

76 Flow Rate: 9 Lpm Test 1: 02/14/2008a 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 9 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 9 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 9 Lpm 10^4Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 9 Lpm 10^6Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 9 Lpm 10^8Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 1.0E+04 2.0E+04 3.0E+04 4.0E+04 5.0E+04 6.0E+04 1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+08 Impinger Collection Liquid Concentration (PFU/mL)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure C-6. Size distri bution and adjusted avg concentra tion of MS2 at 9 Lpm (02/14/08a).

PAGE 77

77 Test 2: 02/14/2008b 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 9 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 9 Lpm 10^1Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 9 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 9 Lpm 10^4Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 9 Lpm 10^6Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+4 1.0e+5 1.5e+5 2.0e+5 2.5e+5 3.0e+5 9 Lpm 10^8Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 1.0E+04 2.0E+04 3.0E+04 4.0E+04 5.0E+04 6.0E+04 7.0E+04 8.0E+04 1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+08 Impinger Collection Liquid Concentration (PFU/mL)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure C-7. Size distribution and adjusted avg concentrati on of MS2 at 9 Lpm (02/14/08b).

PAGE 78

78 Test 3: 03/15/2007 9 Lpm Baseline 10 100 1000 0.0 5.0e+4 1.0e+5 1.5e+5 2.0e+5 2.5e+5 3.0e+5 Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 9 Lpm 10^2 10 100 1000 0.0 5.0e+4 1.0e+5 1.5e+5 2.0e+5 2.5e+5 3.0e+5 Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 9 Lpm 10^4Particle Diameter, dp (nm) 10 100 1000 dN/dlog(d p ) (#/cm 3 ) 0.0 5.0e+4 1.0e+5 1.5e+5 2.0e+5 2.5e+5 3.0e+5 9 Lpm 10^6 10 100 1000 0.0 5.0e+4 1.0e+5 1.5e+5 2.0e+5 2.5e+5 3.0e+5 dN/dlog(d p ) (#/cm 3 )Particle Diameter, dp (nm) 9 Lpm 10^8 10 100 1000 0.0 5.0e+4 1.0e+5 1.5e+5 2.0e+5 2.5e+5 3.0e+5 dN/dlog(d p ) (#/cm 3 )Particle Diameter, dp (nm) 0.0E+00 2.0E+04 4.0E+04 6.0E+04 8.0E+04 1.0E+05 1.2E+05 1.4E+05 1.6E+05 1.8E+05 1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+08 Impinger Collection Liquid Concentration (PFU/mL)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure C-8. Size distribution and adjusted avg concentrati on of MS2 at 9 Lpm (03/15/07).

PAGE 79

79 Summary of 9 Lpm Impinger Concentration (PFU/mL) 1e+01e+11e+21e+31e+41e+51e+61e+71e+81e+9 Adjusted Final Reaerosolized Concentration (#/cm^3) 1e+0 1e+1 1e+2 1e+3 1e+4 1e+5 1e+6 Figure C-9. Summary of adjusted average conc entration of MS2 at 9 Lpm. The lower end of the box represents the 25th percentile, the middle line represents the median, and the upper end of the box represents the 75th percentile.

PAGE 80

80 Flow Rate: 12.5 Lpm Test 1: 01/24/2008 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 1.0e+6 12.5 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 1.0e+6 12.5 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 1.0e+6 12.5 Lpm 10^4Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 1.0e+6 12.5 Lpm 10^6Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 1.0e+6 12.5 Lpm 10^8Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 1.0E+05 2.0E+05 3.0E+05 4.0E+05 5.0E+05 6.0E+05 1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+08 Impinger Collection Liquid Concentration (PFU/mL)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure C-10. Size distribution a nd adjusted avg concentration of MS2 at 12.5 Lpm (01/24/07).

PAGE 81

81 Test 2: 01/28/2008 10 100 1000 0.0 5.0e+3 1.0e+4 1.5e+4 2.0e+4 2.5e+4 3.0e+4 12.5 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+3 1.0e+4 1.5e+4 2.0e+4 2.5e+4 3.0e+4 12.5 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+3 1.0e+4 1.5e+4 2.0e+4 2.5e+4 3.0e+4 12.5 Lpm 10^4Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+3 1.0e+4 1.5e+4 2.0e+4 2.5e+4 3.0e+4 12.5 Lpm 10^6Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 5.0e+3 1.0e+4 1.5e+4 2.0e+4 2.5e+4 3.0e+4 12.5 Lpm 10^8Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 2.0E+03 4.0E+03 6.0E+03 8.0E+03 1.0E+04 1.2E+04 1.4E+04 1.6E+04 1.E+001.E+011.E+021.E+031.E+ 041.E+051.E+061.E+071.E+08 Impinger Collection Liquid Concentration (PFU/mL)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure C-11. Size distribution a nd adjusted avg concentration of MS2 at 12.5 Lpm (01/28/07).

PAGE 82

82 Test 3: 02/09/2008 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 1.0e+6 12.5 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 1.0e+6 12.5 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 1.0e+6 12.5 Lpm 10^4Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 1.0e+6 12.5 Lpm 10^6Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+5 4.0e+5 6.0e+5 8.0e+5 1.0e+6 12.5 Lpm 10^8Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 1.0E+04 2.0E+04 3.0E+04 4.0E+04 5.0E+04 6.0E+04 7.0E+04 8.0E+04 1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+08 Impinger Collection Liquid Concentration (PFU/mL)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure C-12. Size distribution a nd adjusted avg concentration of MS2 at 12.5 Lpm (02/09/07).

PAGE 83

83 Summary of 12.5 Lpm Impinger Concentration (PFU/mL) 1e+01e+11e+21e+31e+41e+51e+61e+71e+81e+9 Adjusted Final Reaerosolized Concentration (#/cm^3) 1e+0 1e+1 1e+2 1e+3 1e+4 1e+5 1e+6 Figure C-13. Summary of adjusted average conc entration of MS2 at 12.5 Lpm. The lower end of the box represents the 25th percentile, the middle line represents the median, and the upper end of the box represents the 75th percentile.

PAGE 84

84 Flow Rate: 15 Lpm Test 1: 02/17/2008a 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm 10^1Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm 10^4Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm 10^6Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm 10^8Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 5.0E+04 1.0E+05 1.5E+05 2.0E+05 2.5E+05 3.0E+05 3.5E+05 1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+08 Impinger Collection Liquid Concentration (PFU/mL)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure C-14. Size distribution a nd adjusted avg concentration of MS2 at 15 Lpm (02/17/08a).

PAGE 85

85 Test 2: 02/17/2008b 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm BaselineParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm 10^1Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm 10^2Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm 10^4Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm 10^6Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+5 2.0e+5 3.0e+5 4.0e+5 5.0e+5 6.0e+5 15 Lpm 10^8Particle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 5.0E+04 1.0E+05 1.5E+05 2.0E+05 2.5E+05 3.0E+05 3.5E+05 4.0E+05 1.E+001.E+011.E+021.E+031.E+041.E+051.E+061.E+071.E+08 Impinger Collection Liquid Concentration (PFU/mL)Adjusted AverageReaerosolized Concentration (#/cm^3) Figure C-15. Size distribution and adjusted avg concentratio n of MS2 at 15 Lpm (02/17/08b).

PAGE 86

86 Test 3: TBD Summary of 15 Lpm Impinger Concentration (PFU/mL) 1e+01e+11e+21e+31e+41e+51e+61e+71e+81e+9 Adjusted Final Reaerosolized Concentration (#/cm^3) 1e+0 1e+1 1e+2 1e+3 1e+4 1e+5 1e+6 1e+7 Figure C-16. Summary of adjusted average concen tration of MS2 at 15 Lpm. The lower end of the box represents the 25th percentile, the middle line represents the median, and the upper end of the box represents the 75th percentile.

PAGE 87

87 Mode Size of MS2 Experiments 0 20 40 60 80 100 120 1.E+001.E+021.E+041.E+061.E+08 Impinger Concentration (PFU/mL)Average Mode (nm) 3 Lpm 6 Lpm 9 Lpm 12.5 Lpm 15 Lpm Figure C-17. Average mode size as a function of flow and impinger concentration for MS2 viral particles.

PAGE 88

88 APPENDIX D EFFECT OF SURFACE TENSION AND VISCOSITY DATA SETS Effect of Surface Tension Test 1: 03/21/2008a 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 Baseline 12.5 Lpm -0 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 Soap Added 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 Soap Added 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 Soap Added 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 Soap Added 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 2.0E+03 4.0E+03 6.0E+03 8.0E+03 1.0E+04 1.2E+04 1.4E+04 0.0 0.1 1.0 10.0 100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure D-1. Size distri bution and adjusted avg concentratio n of PSL at 12.5 Lpm (03/21/08a).

PAGE 89

89 Test 2: 03/21/2008b 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 Baseline 12.5 Lpm -0 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 Soap Added 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 Soap Added 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 Soap Added 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 Soap Added 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 5.0E+03 1.0E+04 1.5E+04 2.0E+04 2.5E+04 3.0E+04 3.5E+04 4.0E+04 0.0 0.1 1.0 10.0 100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure D-2. Size distri bution and adjusted avg concentratio n of PSL at 12.5 Lpm (03/21/08b).

PAGE 90

90 Test 3: 03/24/2008 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 5.0e+4 6.0e+4 Baseline 12.5 Lpm -0 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 5.0e+4 6.0e+4 Soap Added 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 5.0e+4 6.0e+4 Soap Added 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 5.0e+4 6.0e+4 Soap Added 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 5.0e+4 6.0e+4 Soap Added 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 2.0E+03 4.0E+03 6.0E+03 8.0E+03 1.0E+04 1.2E+04 1.4E+04 1.6E+04 0.00.11.010.0100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure D-3. Size distri bution and adjusted avg concentratio n of PSL at 12.5 Lpm (03/21/08b).

PAGE 91

91 Summary of Surface Tension Impinger Concentration (ppm) 0.01 0.1 1 10 100 Adjusted Final Reaerosolized Concentration (#/cm^3) 1e+4 2e+4 3e+4 4e+4 5e+4 Figure D-4. Summary of adjusted average concen tration of PSL at 12.5 Lpm. The lower end of the box represents the 25th percentile, the middle line represents the median, and the upper end of the box represents the 75th percentile.

PAGE 92

92 Effect of Viscosity Test 1: 03/31/2008 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 With Oil 12.5 Lpm -0 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 With Oil 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 With Oil 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 With Oil 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 With Oil 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 2.0E+03 4.0E+03 6.0E+03 8.0E+03 1.0E+04 1.2E+04 1.4E+04 1.6E+04 0.0 0.1 1.0 10.0 100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure D-5. Size distri bution and adjusted avg concentratio n of PSL at 12.5 Lpm (03/31/08).

PAGE 93

93 Test 2: 04/02/2008a 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 5.0e+4 6.0e+4 With Oil 12.5 Lpm -0 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 5.0e+4 6.0e+4 With Oil 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 5.0e+4 6.0e+4 With Oil 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 5.0e+4 6.0e+4 With Oil 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 1.0e+4 2.0e+4 3.0e+4 4.0e+4 5.0e+4 6.0e+4 With Oil 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 5.0E+03 1.0E+04 1.5E+04 2.0E+04 2.5E+04 0.0 0.1 1.0 10.0 100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure D-6. Size distri bution and adjusted avg concentratio n of PSL at 12.5 Lpm (04/02/08a).

PAGE 94

94 Test 3: 04/02/2008b 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 1.6e+5 With Oil 12.5 Lpm -0 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 1.6e+5 With Oil 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 1.6e+5 With Oil 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 1.6e+5 With Oil 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+4 4.0e+4 6.0e+4 8.0e+4 1.0e+5 1.2e+5 1.4e+5 1.6e+5 With Oil 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0.0E+00 1.0E+04 2.0E+04 3.0E+04 4.0E+04 5.0E+04 6.0E+04 0.00.11.010.0100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure D-7. Size distri bution and adjusted avg concentratio n of PSL at 12.5 Lpm (04/02/08b).

PAGE 95

95 Summary of Viscosity Impinger Concentration (ppm) 0.01 0.1 1 10 100 Adjusted Final Reaerosolized Concentration (#/cm^3) 1e+4 2e+4 3e+4 4e+4 5e+4 6e+4 Figure D-8. Summary of adjusted average concen tration of PSL at 12.5 Lpm. The lower end of the box represents the 25th percentile, the middle line represents the median, and the upper end of the box represents the 75th percentile.

PAGE 96

96 APPENDIX E REAEROSOLIZATION WITH BIOSAMPLER DATA SETS Test 1: 03/15/2008 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 8.0e+2 1.0e+3 BioSampler 12.5 Lpm -0 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 8.0e+2 1.0e+3 BioSampler 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 8.0e+2 1.0e+3 BioSampler 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 8.0e+2 1.0e+3 BioSampler 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 8.0e+2 1.0e+3 BioSampler 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0 20 40 60 80 100 120 140 160 180 0.0 0.1 1.0 10.0 100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure E-1. Size distri bution and adjusted avg concentrati on of PSL at 12.5 Lpm (03/15/08).

PAGE 97

97 Test 2: 03/19/2008a 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 8.0e+2 1.0e+3 1.2e+3 1.4e+3 1.6e+3 BioSampler 12.5 Lpm -0 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 8.0e+2 1.0e+3 1.2e+3 1.4e+3 1.6e+3 BioSampler 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 8.0e+2 1.0e+3 1.2e+3 1.4e+3 1.6e+3 BioSampler 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 8.0e+2 1.0e+3 1.2e+3 1.4e+3 1.6e+3 BioSampler 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 8.0e+2 1.0e+3 1.2e+3 1.4e+3 1.6e+3 BioSampler 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0 50 100 150 200 250 0.0 0.1 1.0 10.0 100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure E-2. Size distri bution and adjusted avg concentrati on of PSL at 12.5 Lpm (03/19/08a).

PAGE 98

98 Test 3: 03/19/2008b 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 BioSampler 12.5 Lpm -0 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 BioSampler 12.5 Lpm 0.1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 BioSampler 12.5 Lpm 1 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 BioSampler 12.5 Lpm 10 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 10 100 1000 0.0 2.0e+2 4.0e+2 6.0e+2 BioSampler 12.5 Lpm 100 ppmParticle Diameter, dp (nm)dN/dlog(d p ) (#/cm 3 ) 0 20 40 60 80 100 120 0.0 0.1 1.0 10.0 100.0 Impinger Collection Liquid Concentration (ppm)Adjusted Average Reaerosolized Concentration (#/cm^3) Figure E-3. Size distri bution and adjusted avg concentrati on of PSL at 12.5 Lpm (03/19/08b).

PAGE 99

99 Summary of BioSampler Impinger Concentration (ppm) 0.01 0.1 1 10 100 Adjusted Final Reaerosolized Concentration (#/cm^3) 5.0e+1 1.0e+2 1.5e+2 2.0e+2 2.5e+2 3.0e+2 Figure E-4. Summary of adjusted average concen tration of PSL at 12.5 Lpm. The lower end of the box represents the 25th percentile, the middle line represents the median, and the upper end of the box represents the 75th percentile.

PAGE 100

100 LIST OF REFERENCES Ace Glass Inc. (2008). "7540 Im pinger." Ace Glass Catalog, 212. Aranha-Creado, H., and Brandwein, H. (1999). "A pplication of Bacteriophages as Surrogates for Mammalian Viruses: A Case for Use in Filter Validation Based on Precedents and Current Practices in Medical and Environmental Virology." PDA J. Pharm. Sci. Tech., 53(2), 75-82. Demokritou, P., Gupta, T., and and Koutrakis, P. (2002). "A High Volume Apparatus for the Condensational Growth of U ltrafine Particles for Inhala tion Toxicological Studies." Aerosol Sci. Tech., 36(11), 1061-1072. Floyd, R., and Sharp, D. G. (1978). "Viral Aggreg ation: Effects of Salts on the Aggregation of Poliovirus and Reovirus at Low pH." Appl. Environ. Microb., 35(6), 1084-1094. Floyd, R., and Sharp, D. G. (1977). "Aggregation of Poliovirus and Reovirus by Diltuion in Water." Appl. Environ. Microb., 33(1), 159-167. Friedlander, S. K. (2003). "Gas-to-Particle Conversion." Smoke Dust and Haze, Oxford University, 275-305. Fuchs, N. A. (1964). "Absorp tion of Aerosols by Bubbling." The Mechanics of Aerosols, Pergamon, New York, 240-245. Ghiaasiaan, S. M., and Yao, G. F. (1997). "A Theoretical Model for Deposition of Aerosols in Rising Spherical Bubbles due to Diffusion, Convection, and Inertia." Aerosol Sci. Tech., 26(2), 141-153. Golmohammadi, R., Valegard, K., Fridborg, K., and Liljas, L. (1993). "The Refined Structure of Bacteriophage-MS2 at 2. 8 Angstrom Resolution." J. Mol. Biol., 234(3), 620-639. Grant, S. (1994). "Virus Coagula tion in Aqueous Environments." Environ. Sci. Technol., 28(5), 928-933. Grinshpun, S. A., Willeke, K., Ulevicius, V., Juo zaitis, A., Terzieva, S., Donnelly, J., Stelma, G. N., and Brenner, K. P. (1997). "Effect of Impaction, Bounce and Reaerosolization on the Collection Efficiency of Impingers." Aerosol Sci. Tech., 26(4), 326-342. Hering, S., and Stolzenburg, M. (2005). "A Method for Particle Size Amplification by WaterCondensation in a Laminar, Thermally Diffusive Flow." Aerosol Sci. Tech., 39(5), 428-436. Hinds, W. C. (1999). "Condensation and Evaporation." Aerosol Technology, John Wiley and Sons, Inc., New York, 278-303. Hinds, W. C. (1999). "Produc tion of Test Aerosols." Aerosol Technology, John Wiley and Sons, Inc., New York, 428-432.

PAGE 101

101 Hogan, C. J., Kettleson, E. M., Lee, M. H., Ra maswami, B., Angenent, L. T., and Biswas, P. (2005). "Sampling Methodologies and Dosage Assessment Techniques for Submicrometre and Ultrafine Virus Aerosol Particles." J. Appl. Microbiol., 99(6), 14221434. Hogan, C. J., Lee, M. H., and Biswas, P. (2004). "Capture of Viral Particles in Soft X-RayEnhanced Corona Systems: Charge Dist ribution and Transport Characteristics." Aerosol Sci. Tech., 38(5), 475-486. Lin, X., Reponen, T., Willeke, K., Grinshpun, S., Foarde, K. K., and Ensor, D. S. (1999). "Longterm Sampling of Airborne Bacteria a nd Fungi into a Non-evaporating Liquid." Atmos. Environ., 33(26), 4291-4298. Lin, X., Reponen, T., Willeke, K., Wang, Z., Gr inshpun, S., and Trunov, M. (2000). "Survival of Airborne Microorganisms During Swirling Aerosol Collection." Aerosol Sci. Tech., 32(3), 184-196. Lin, X., Willeke, K., Ulevicius, V., and Grinshpun, S. (1997). "Effect of Sampling Time on the Collection Efficiency of All-Glass Impingers." Am. Ind. Hyg. Assoc. J., 58(7), 480-488. Madigan, M. T., Martinko, J. M., and Parker, J. (2003). Brock Biology of Microorganisms, Prentice Hall, New Jersey, 231-260, 520-546. Okuyama, K., Kousaka, Y., and Motouchi, T. (1984). "Condensational Growth of Ultrafine Aerosol Particles in a New Particle Size Magnifier." Aerosol Sci. Tech., 3(4), 353-366. Pich, J., and Schutz, W. (1991). "On the Theory of Particle Deposition in Rising Gas Bubbles: The Absorption Minimum." J. Aerosol Sci., 22(3), 267-272. Reponen, T., Willeke, K., Grinshpun, S., and Ne valainen, A. (2001). "Biological Particle Sampling." Aerosol Measurement: Principl es, Techniques and Applications, K. Willeke, and P. A. Baron, eds., John Wiley and Sons, Inc., New York, 471-492. Russell, L., and Singh, E. (2006). "Submi cron Salt Production in Bubble Bursting." Aerosol Sci. Tech., 40(9), 664-671. Shields, P. A., and Farrah, S. R. (2002). "Cha racterization of Virus Adsorption by Using DEAESepharose and Octyl-Sepharose." Appl. Environ. Microb., 68(8), 3965-3968. Sioutas, C., Kim, S., and Chang, M. (1999). "Development and Evaluation of a Prototype Ultrafine Particle Concentrator." J. Aerosol Sci., 30(8), 1001-1017. Sioutas, C., and Koutrakis, P. (1996). "Inertial Separation of Ultrafine Particles Using a Condensational Growth/ Virt ual Impaction System." Aerosol Sci. Tech., 25(4), 424-436. SKC Inc. (2008). "BioSampler Op erating Instructions." 1.

PAGE 102

102 Terzieva, S., Donnelly, J., Uleviciu s, V., Grinshpun, S. A., Willeke, K., Stelma, G., and Brenner, K. P. (1996). "Comparison of Methods for Detection and Enumeration of Airborne Microorganisms Collected by Liquid Impingement." Appl. Environ. Microb., 62(7), 2264-2272. Thomas, J. J., Falk, B., and Fenselau, C. (1998). "Viral Characterizatoin by Direct Analysis of Capsid Proteins." Anal. Chem., 70(18), 3863-3867. Tseng, C. C., and Li, C. S. (2005). "Collecti on Efficiencies of Aerosol Samplers for VirusContaining Aerosols." J. Aerosol Sci., 36(5-6), 593-607. van Voorthuizen, E. M., Ashbolt, N. J., and Scha fer, A. I. (2001). "Role of Hydrophobic and Electrostatic Interactions for Initial En teric Virus Retention by MF Membranes." J. Membrane Sci., 194(1), 69-79. Vidaver, A. K., Koski, R. K., and Van Etten, J. L. (1973). "Bacteriophage f6: a Lipid-Containing Virus of Pseudomonas phaseolicola ." J. Virol., 11(5), 799-805. Vonnegut, B. (1954). "Method and Apparatus for M easuring the Concentration of Condensation Nuclei." (2,684,008). Weissenborn, P. K. (2006). "Surface Tension of Aqueous Electrolytes." Encyclopedia of Surface and Colloid Science, P. Somasundaran, ed., CRC Press, Boca Raton, FL, 6029-6032. White, F. (2002). Fluid Mechanics. McGraw-Hill, New York, NY. Willeke, K., Grinshpun, S. A., Ulevicius, V., Terzieva, S., Donnelly, J., Steward, S., and Juozaitis, A. (1995). "Microbial Stress, B ounce and Re-aerosolization in Bioaerosol Samplers." J. Aerosol Sci., 26(1), S883-S884. Willeke, K., Lin, X., and Grinshpun, S. (1998) "Improved Aerosol Collection by Combined Impaction and Centrifugal Motion." Aerosol Sci. Tech., 28(5), 439-456. Wu, C. Y., and Biswas, P. (1998) "Particle Growth by Condensation in a System with Limited Vapor." Aerosol Sci. Tech., 28(1), 1-20.

PAGE 103

103 BIOGRAPHICAL SKETCH Lindsey Ann Riem enschneider was born in Augusta, GA, to David Riemenschneider and Sandra Applegate on June 30, 1983. She moved to Ukiah, CA, shortly thereafter, where she lived with her parents and two brothers, Michael and Paul, until she graduated from Ukiah High School in June 2001. She attended the Universi ty of California at Davis from 2001, where she received a B.S. degree in Civil Engineering with an emphasis in environmental engineering and a minor in geology. Following graduation from UC Davi s, Lindsey enrolled in the masters degree program in the Environmental Engineering Sciences Department at the University of Florida, where she was a member of the Dr. C.Y. Wu Air Resources Res earch Group. There, she served as a Research Assistant and a Teaching Assistant, as well as Tr easurer and President of the UF Student Chapter of the Air and Waste Management Association during the 2006 and 2007 academic years, respectively. She received her Mast er of Engineering degree in environmental engineering in August 2008 and entered the enviro nmental engineering consulting industry with Camp Dresser McKee as an Engineer II in the Water/Wastewater Services Group.