EPA-650/4-74-034

August 1974
Environmental Monitoring Series
                                         xy
                                        £
                                        5
                                        x
                                        o
                                        \
                 ^
          SSE
                                          ^^ PRO^
                 Ax
 \
 LU
 O
 T
$

-------
Research reports of the Office of Research and Development, Environmental Protection
Agency, have been grouped into five series. These five broad categories were estab-
lished to facilitate further development and application of environmental technology.
Elimination of traditional  grouping was consciously planned to foster technology transfer
and a maximum interface  in related fields. The five series are:

          I.   Environmental Health Effects Research
          2.   Environmental Protection  Technology
          3.   Ecological Research
          4.   Environmental Monitoring
          5.   Socioeconomic Environmental  Studies

This report has been assigned to the ENVIRONMENTAL MONITORING series.  This series
describes  research conducted to develop new or improved methods and instrumentation
for the identification and  quantification of environmental pollutants at the lowest con-
ceivable significant concentrations.  It also includes studies to determine the ambient
concentrations of pollutants  in the environment and/or the variance of pollutants as
a function of  time or meteorological factors.
This report has been reviewed by the Office of Research and Development, EPA,
and approved for publication.  Approval does not signify that the contents
necessarily reflect the views and policies of the Environmental Protection
Agency, nor  does mention of trade names or commercial products constitute
endorsement  or recommendation  for use.

-------
                                     EPA-650/4-74-034
   PilSENCE  OF  SULFUR
                    by

              David N. McNeils

     National Environmental Research Center
           Las Vegas, Nevada 89114


          Program Element No. 21 AKB
                ROAPNo738


      EPA Project Officer:  William E. Wilson

       Chemistry and Physics Laboratory
     National Environmental Research Center
   Research Triangle Park,  North Carolina 27711
  NATIONAL ENVIRONMENTAL RESEARCH CENTER
    OFFICE OF RESEARCH AND DEVELOPMENT
   U.Si ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711

                August 1974

-------
                 DISTRIBUTION STATEMENT

This report is available to the public, for a nominal cost,  through the
National Technical Information Service, 5285 Port Royal Road,
Springfield, Virginia 22151.
This report is a dissertation submitted to the faculty of the University of
North Carolina in partial fulfillment of the requirements for the degree
of Doctor of Philosophy in the Department of Environmental Sciences
and Engineering.
                                ii

-------
                              ABSTRACT

The loss of visibility in atmospheres laden with photochemical smog is largely
attributed to the presence of sulfate aerosols.  This facet was thought to be the
most objectionable effect of the sulfate species and is reflected in the establish-
ment of the National Ambient Air Quality Standard for sulfur dioxide. Recent
reviews, however, have indicated that the adverse health effects appear  more
strongly associated with suspended particulate sulfate than with the sulfur
dioxide.

The dark-phase reaction of olef in-ozone-sulfur dioxide was studied in an
attempt to  elucidate the mechanism involved in the oxidative consumption of
the sulfur dioxide.  The effect of several variables on the reaction stoichio-
metry  and  on the aerosol production was also  investigated.   These variables
included the reactant concentrations,  the relative humidity,  molecular oxygen
concentration and the olef in species although propylene was the primary  olefin
studied.

The stoichiometry of the  propylene-ozone reaction was found to be a smooth
function of the  initial concentration ratio of  these species.  The olefin/ozone
consumption was  1 for a system in which the  olefin was initially in excess
and  1 for  a system having an initial ozone excess.  The consumption ratio
was unaffected by the addition of sulfur dioxide or by varying the relative
humidity over the range of 20 to 38 percent.

Molecular  oxygen had a significant effect on the reaction stoichiometry and
product formation in the propylene-ozone thermal reaction.   The propylene/
ozone consumption ratio was lower in a system in which the  molecular oxy-
gen concentration was reduced.  Oxygen apparently also contributed to the
regeneration of ozone and  the production of the hydroxyl radical species,  both
of which interacted with the propylene and with products of the reaction.

A tentative  mechanism has been proposed for  the oxidative consumption of
sulfur  dioxide in  the propylene-ozone-sulfur dioxide system  based on the
observations made during these  studies. Secondary reactions are a distinc-
tive feature of  this model which incorporated relatively few reactions to
explain the  major characteristics of the systems studied.

Aerosol formation occurred in the dark reaction of propylene and ozone
although the particulate formed did not grow to a size which  efficiently
scattered light.  The addition of sulfur dioxide markedly enhanced both the
formation and subsequent growth of the aerosol.  The same general patterns
of aerosol  formation, growth and decay were observed throughout the series
of experiments over a broad range of initial conditions.  These patterns  are
consistent  with a model which includes homogeneous nucleation,  condensation
and coagulation processes.

                                   iii

-------
An apparent equilibrium surface area was established for the aerosol when
the addition to the total surface area by condensation was equal to the  loss
by coagulation. A strong correlation was found to exist between this equi-
librium surface area and the volumetric conversion rate.  It was also
found that the correlation could be extended considerably in range to include
data from other experiments in which the aerosol was  photochemically
generated.  The establishment of the equilibrium surface area is due, there-
fore, to the physical processes rather than the chemical mechanism leading
to its formation.
                                  IV

-------
                              Table of Contents

                                                             Page



  List of Tables                                             ,vii

  List of Figures                                            viii

  I.    Introduction                                             1

 II.    Review of Previous Work                                  7

       A.  Studies of the Olefin-Ozone System                   7

       B.  Mechanism of the Olefin-Ozone Reaction               8

       C.  Reactions with Sulfur Dioxide                       14

       D.  Aerosol Production in Hydrocarbon Systems           15

III.    Theory of Aerosol Formation, Growth and Decay.           20

       A.  Nucleation                                          21

       B.  Condensation                                        26

       C.  Coagulation                                         29

 IV.    Experimental Arrangement and Procedures                 34

       A.  Experimental Design and Arrangement                 34

       B.  Experimental Apparatus                              41

           1.  Reactor                                         41
           2.  Ozone Analysis                                  44
           3.  Ambient and Dew Point Temperatures Measurements 46
           4.  Sulfur Dioxide Analysis                         47
           5.  Hydrocarbon Analysis                            49
           6.  Infrared Spectroscopic Product Analysis         50
           7.  Sulfur Balance Determination                    52
           8.  Light Scattering Measurements                   53
           9.  Condensation Nuclei Count Determination         55
          10.  Sub-microscopic Aerosol Analysis                56
          11.  Optical Particle Analysis                       61

       C.  Experimental Procedures                             63

                                      v

-------
 V.   Experimental Results and Discussion                          71

      A.  Reactant Consumption Data and Reaction Stoichiometry     71

      B.  Product Analysis                                         94

      C.  Tentative Reaction Mechanism                            111

      D.  Sulfate Analysis and Sulfur Mass Balance                114

      E.  Light Scattering Measurements                           119

      F.  Aerosol Development                                     123

      G.  Aerosol Measurements by Single Particle Light
          Scattering                                              158

VI.   Conclusions and Recommendations                             162

 References                                                       168

 Appendices                                                       175

      1.  Product Analysis by Fourier Transform Infrared
          Spectroscopy.                                           175

      2.  Observed and Predicted Reactant Concentrations for
          Gas Phase Experiments.                                  176

      3.  Aerosol Droplet Number, Surface Area, and Volume
          Concentrations Distributions for Aerosol Experiments.   196
                                    VI

-------
                             LIST OF TABLES


TABLE                                                                 PAGE

 2-1      Summary of Rate Constant Data	   10

 2-2      Relative Rate Constants	   17

 3-1      Coagulation Constants for Unequal Sized Particles 	   31

 4-1      Permeability of Tedlar and Scotchpak to Gases and Vapors  .   42

 4-2      Data Table for Number, Surface Area and Volume Distribution   62

 5-1      Summary of Experimental Conditions and Initial Reactant
          Concentrations  	   72

 5-2      Summary of Stoichiometric Data and Acetyaldehyde Production   75

 5-3      Product Analysis by Fourier Transform Infrared Spectroscopy  105

 5-4      Summary of Sulfuric Acid Analysis by X-ray Fluorescence
          Spectroscopy and Liquid Chromatography  	  116

 5-5      Volumetric Conversion Rate, Total Volume at 80 Minutes,
          Mean Volume Diameter Growth Rate, Coagulation Constant
          and Sulfur Dioxide Consumption at 80 Minutes for Aerosol
          Experiments	138

 5-6      Initial Sulfur Dioxide Reaction Rate and Oxidative Rate .  .  141

 5-7      Calculated Sulfur Dioxide Consumption based on the Observed
          Volume Concentration  	  141
                                vn

-------
                             LIST OF FIGURES


FIGURE                                                                PAGE

4-1       Schematic of Experimental Arrangement 	  36

4-2       Laboratory Arrangement for Olefin-Ozone-Sulfur Dioxide
          Studies 	 ......... 	  38

4-3       Aerosol Measurement Instrumentation 	  39

4-4       Fourier Transform Infrared Spectrometer .	  40

4-5       Variation of Relative Humidity with Time due to the
          Permeation of Water Vapor 	  43

4-6       Schematic of Reactor Port Arrangement ...........  45

4-7       Schematic of the Electrical Aerosol Size Analyzer 	  58

4-8       Relationship Between Electrical Mobility and Particle Size
          for Diffusion Charging where N  = 1 x 10  (ions/cm )(sec) .  60

4-9       Calibration Data for Optical Aerosol Analyzer 	  64

4-10      Variation of Reactant Concentration with Time Due to Wall
          Losses	  69

5-1       Variation of Concentration with Time for Propylene-
          Ozone-Sulfur Dioxide Reaction 	  78

5-2       Variation of Concentration with Time for Ethylene-Ozone-
          Sulfur Dioxide Reaction 	  79

5-3       Stoichiometry of Olefin-Ozone Reaction versus Initial
          Reactant Concentration Ratio  	  80

5-4       Concentration of Acetyaldehyde Produced versus Propylene
          Reacted for Nine Propylene-Ozone Experiments  	  86

5-5       Sulfur Dioxide Consumption versus Propylene Reacted for
          Different Initital Sulfur Dioxide Concentrations  	  88

5-6       Sulfur Dioxide Consumption versus Ozone Reacted for
          Different Initial Sulfur Dioxide Concentrations 	  89

5-7       Variation of the Apparent Rate Constant with Time Due to
          Secondary Reactions 	  92

                                   viii

-------
LIST OF FIGURES (Continued)
FIGURE                                                                 PAGE

5-8       Infrared Absorption Spectra of Reactor Contents - Runs 31
          and 21	   96

5-9       Infrared Spectra in Region of Carbon Monoxide Absorption -
          Runs 31 and 21	   97

5-10      Infrared Spectra in Region of Carbon Monoxide Absorption -
          Runs 78 and 79	   99

5-11      Infrared Spectra in Region of Carbon Monoxide Absorption -
          Runs 81 and 82	100

5-12      Infrared Spectra in Region of Carbon Monoxide Absorption -
          Runs 60 and 65	101

5-13      Infrared Absorption Spectra of Reactor Contents - Runs 28
          and 29	104

5-14      Comparison of Results for Sulfuric Acid Analysis 	  118

5-15      Variation of the Scattering Coefficient, visual Range,
          Surface Area Concentration and Sulfur Dioxide Concentration
          with Time for a Propylene-Ozone-Sulfur Dioxide Experiment  .  121

5-16      Variation of the Scattering Coefficient, Visual Range,
          Surface Area Concentration and Sulfur Dioxide Concentration
          with Time for an Ethylene-Ozone-Sulfur Dioxide Experiment  .  122

5-17      Aerosol Particle Number, Surface Area and Volume Development
          for a Propylene-Ozone Experiment 	  124

5-18      Total Volume, Number, Reciprocal of Number and Surface Area
          Concentration versus Time for the Developing Sulfuric Acid
          Aerosol - Run 101A	126

5-19      Total Volume, Number, Reciprocal of Number and Surface Area
          Concentration versus Time for the Developing Sulfuric Acid
          Aerosol - Run 101B	127

5-20   .   Total Volume, Number, Reciprocal of Number and Surface Area
          Concentration versus Time for the Developing Sulfuric Acid
          Aerosol - Run 102	128

5-21      Total Volume, Number, Reciprocal of Number and Surface Area
          Concentration versus Time for the Developing Sulfuric Acid
          Aerosol - Run 103	129
                                IX

-------
LIST OF FIGURES (Continued)


FIGURES                                                                PAGE

5-22      Total Volume, Number, Reciprocal of Number and Surface
          Area Concentration versus Time for the Developing Sulfuric
          Acid Aerosol - Run 104	130

5-23      Total Volume, Number, Reciprocal of Number and Surface Area
          Concentration versus Time for the Developing Sulfuric Acid
          Aerosol - Run 105	131

5-24      Total Volume versus Sulfur Dioxide Consumption at T + 80
          Minutes	143

5-25      Variation of the Equilibrium Surface Area with Volumetric
          Conversion Rate	145

5-26      Initial Volumetric Conversion Rate versus Product of Initial
          Propylene and Ozone Concentrations for Different Initial
          Sulfur Dioxide Concentrations  	  147

5-27 .     Development of the Volume Distribution with Time - Run 101A  148

5-28      Development of the Volume Distribution with Time - Run 101B  149

5-29      Development of the Volume Distribution with Time - Run 102 .  150

5-30      Development of the Volume Distribution with Time - Run 103 .  151

5-31      Development of the Volume Distribution with Time - Run 104 .  151

5-32      Development of the Volume Distribution with Time - Run 105 .  152

5-33      Variation of Volume Mean Diameter with Time for Six Aerosol
          Experiments	154

5-34      Rate of Change in Mean Volume Diameter versus Volumetric
          Conversion Rate	157

5-35      Variation of Aerosol Size Distribution with Time as Measured
          with Optical Particle Size Analyzer - Run 31	159

5-36      Variation of Aerosol Size Distribution with Time as Measured
          with Optical Particle Size Analyzer - Run 29	160
                                   x

-------
              AEROSOL FORMATION FROM GAS PHASE REACTIONS OF
            OZONE AND OLEFIN IN THE PRESENCE OF SULFUR DIOXIDE

                                CHAPTER I

                              INTRODUCTION


     A major source of the sulfur burden of our atmosphere results from

the combustion of solid and liquid fossil fuels.   These fuels contain

significant quantities of sulfur as inorganic sulfides, sulfur contain-

ing organic compounds and some small fraction as sulfates.  It is

estimated that sulfur dioxide accounts for upwards of 95% of the sulfur

released from the fuel combustion (Smith and Gruber, 1966).

     There are several undesirable effects attributed to this contaminant

which include the degradation of materials, retardation of plant growth

and production, and of most importance, the difect impact on man's

health.  In addition, literature reviews by Bufalini (1971) and Altshuller

and Bufalini (1971) provide many references citing the conversion of

sulfur dioxide to aqueous sulfuric acid droplets as one of the most basic

photochemical reactions.   The exact physical and chemical mechanisms

which enter into this conversion are not adequately understood but the

resulting aerosol is known to contribute significantly to visibility

reduction.  The scatter and absorption of visible radiation by gas

molecules and particulates are the causes of atmospheric visibility

reduction.  Small particles (0.1 to 1 pm diameter) suspended in the

atmosphere are particularly effective in scattering visible light.

     Until recently, this loss of visibility was thought to be the

most objectionable effect of the sulfate aerosol.  The sulfur dioxide

-------
                                                                        2





ambient air quality standards were based on judgemental decisions using




the best available data.  A review of the standard using the Community




Health Environmental Surveillance System (CHESS) data indicated that




this standard has a defensible base.  This same review, however, has




added a new facet to the problem, i.e., the fact that the adverse health




effects for certain health indicators, particularly asthma, appear




more strongly associated with suspended particulate sulfate than with




sulfur dioxide  (Environmental Protection Agency, 1974).  If new toxico-




logical data supports this finding then data collection and analysis




should be directed toward developing sulfate criteria.  Of paramount




concern at the present is the complete lack of knowledge concerning




the relationship of sulfur dioxide to suspended sulfate levels.  Addi-




tional information is also needed relative to the rate of conversion of




sulfur dioxide to sulfate.  The standard will require another review




when these data are available particularly if they indicate that the




simple control of sulfur dioxide is not the most efficient way of




protecting man's health from exposure to sulfur compounds.




     In the atmosphere sulfur dioxide is oxidized by either gas phase,




catalytic or photochemical processes.  In irradiated olefin-oxides of




nitrogen systems decay of sulfur dioxide and formation of sulfuric acid




do not occur to any appreciable extent until most of the nitric oxide is




converted to nitrogen dioxide.  Nitrogen trioxide is formed in the




reaction of ozone with nitrogen dioxide and the organic peroxides which




earlier reacted with the nitric oxide begin to accumulate.  Reaction




with one of these oxidizing species rather than the direct photooxidation




of the sulfur dioxide probably accounts for most of the sulfate aerosol




generation.  The oxidative interaction with the surfaces of aerosol




droplets or particles could also account in part for the decay of the

-------
sulfur dioxide.  The significance of this mode would depend on the




reactant concentrations, the physical/chemical nature of the aerosol and




its surface area.  The rate of this type of reaction may be limited by




the diffusion of sulfur dioxide to the active surfaces or by its diffu-




sion into and reactions throughout the aerosol material.  The third




process, i.e., the photochemical reaction of sulfur dioxide when irra-




diated by sunlight, has perhaps been the most thoroughly invesitgated.




In particle-laden air, the oxidation can proceed by a number of paths




with the aerosol surface serving as a catalyst for chemical reactions




involving the photochemical products.  In particle-free air, the oxida-




tion proceeds along fixed paths with components of the vapor phase.  The




primary photochemical reaction involves the absorption of solar radia-




tion by the sulfur dioxide followed by reactions of the excited sulfur




dioxide molecules with molecular oxygen to form an intermediate sulfate




species.  It is suggested that this intermediate then reacts with molec-




ular oxygen to the sulfur trioxide and ozone (Blacet, 1952).




     These processes are, for the most part, poorly understood and become




considerably more complex when in the presence of other air pollutants.




In recent years, however, major advances in the knowledge of homogeneous




gas phase reactions have been accomplished although the factors governing




the gas to particle conversions remain largely unresolved and are only




briefly discussed in published reviews of the chemistry of air pollution.




     Generally, aerosol production in gas phase reactions has been




monitored by condensation nuclei counters and integrating light scatter-




ing instruments. Both of these instruments provide evidence of the




presence of the aerosol but lack the ability to yield a quantitative




analysis of the mass or volumetric conversion rate or a size distribu-




tion of the aerosol produced.  The condensation nuclei counter provides

-------
a count of the number of nuclei present per unit volume.  The nuclei




are grown to a common size and their number concentration is sensed by




the attenuation of a light beam.  The light scattering devices measure the




integrated scattering which changes as the size distribution, refractive




index and the concentration of the aerosol change over the span of the




measurement period.




     Because of the inability of investigators to control or isolate




the various gaseous components or environmental factors leading to




aerosol production under field conditions, many of the studies designed




to investigate aerosol formation are conducted in chambers under con-




trolled environmental conditions.  The photochemical systems studied




usually involve a hydrocarbon, nitric oxide, and sulfur dioxide.  In




dark phase systems ozone is also introduced as a reactant.




     A review of the literature indicates that the thermal reaction of




olefins and ozone in the presence of sulfur dioxide yields a sulfate




aerosol (Groblicki and Nebel, 1971 and Cox and Penkett, 1972).  Recent




technological advances in both gas phase and aerosol measurement instru-




mentation now permits selected aspects of the olefin-ozone-sulfur dioxide




reaction with its attending aerosol production to be more thoroughly




investigated.  The principal advances relative to the studies undertaken




are the flame photometric detection of total sulfur, the application of




the chemiluminesence reaction of ethylene and ozone to ozone analysis,




the development of the Fourier transform infrared spectrometer for gas




phase product analysis and the electrical mobility analyzer for the




in situ measurement of submicroscopic aerosols.  Two new techniques




have also been applied to the sulfate analysis of filter samples.  X-ray




fluoresence is being used to measure the total sulfur as well as other

-------
elements and compounds, and liquid chromatography is being used to




analyze for water soluble sulfates.




     Data developed in these studies serve to develop a qualitative




understanding of the overall photochemical smog formation process as




well as to provide detailed quantitative input into newly developed




photochemical smog predictive models.  Finally, these data, particularly




those dealing with mass or volumetric conversion rates, mechanism and




gas phase product yields, serve to provide the information necessary




to establish criteria which ultimately are used as the basis for stand-




ards for control.




     Propylene was selected as the primary olefin to be studied because




it is a common atmospheric hydrocarbon, has moderate reactivity with




ozone and is a relatively simple molecule.  Ethylene was also studied




but, like other first members in homologous series, is atypical.  Neither




of these olefins produces organic aerosols in reaction with ozone.




     The general aims of the research reported herein were to elucidate




selected aspects of the oxidative mechanism of sulfur dioxide and to




measure its mass conversion rate from the gaseous to the condensed state.




     More specifically, the objectives of this research were:




     1.  to design a reaction vessel and sampling system which would




permit the study of selected dark phase reactions under controlled




conditions and reactant concentrations and which would yield reproducible




results.




     2.  to study the kinetics of the olefin-ozone-sulfur dioxide system




and to determine factors which influence the stoichiometry of the olefin-




ozone reaction and the formation of selected gas phase products.

-------
     3.  to assess and characterize the aerosol formed including measure-




ments of the size distribution, mass concentration, number concentration




and total light scattering both as a function of time and of system




component concentrations.




     4.  to infer from the aforementioned measurements the physical/




chemical mechanism leading to the formation and growth of the aerosol.




     5.  to advance a tentative mechanism for the oxidative consumption




of sulfur dioxide in the olefin-ozone-sulfur dioxide system based on




information contained in the literature and on the measurements reported




herein.

-------
                               CHAPTER II




                         REVIEW OF PREVIOUS WORK







A.  Studies of the Olefin-Ozone System





     The stoichiometry of the ethylene-ozone and the propylene-ozone




reactions were studied by Hanst et al. (1958).  Within experimental




error they found a 1:1 ratio of olefin to ozone consumption.   The initial




concentration ratios of the reactants in these studies were 1:1 for




ethylene and ozone and 3.36:1 for the propylene and ozone.   Stedman




et al.  (1973) reported similar results for various initial reactant con-




centrations where they show 1:1 stoichiometry by observation of equal




decrements of the reactants.  Vbraski and Cvetanovic (1960) and Wei and




Cvetanovic (1963) studied this reactions at much higher concentrations




(20,000 ppm olefin and 4,000 ppm ozone)  and found a quite different




result, i.e., the ratio of olefin to ozone consumed varied from 1.4 to




2.0 for the different olefins studied.  This difference was attributed




(Hanst et al., 1958) to the possible further reaction of the zwitterions




formed in the initial olefin-ozone reaction with olefins at the high




concentrations.  At lower concentrations, however, the zwitterion was




postulated to undergo unimolecular decomposition.  One other study, i.e.,




that of Bufalini and Altshuller (1965),  reports data for olefin-ozone




consumption ratios although propylene is not included.  This ratio was




found to increase with increasing olefin concentration and level off




when the olefin concentration was in excess.




     One other paramenter which has received cursory examination is the

-------
                                                                        8





initial molecular oxygen concentration in the reactor.  Oxygen can




participate in the reaction either as a scavenger of radicals or by




contributing to the formation of species which alter the consumption




ratio of the initial reactants.  Cox and Penkett (1972) and Stedman et




al.  (1973) report the independence of the olefin-ozone reaction from the




molecular oxygen concentration.  In both cases, nitrogen was employed as




the diluent gas and in the Cox and Penkett study, the oxygen mole




fraction was estimated at 0.2%.  Wei and Cvetanovic (1963), however,




found that in the absence of molecular oxygen the consumption ratio was




close to unity while in the presence of oxygen as previously reported,




this ratio increased to 1.4 - 2.0 depending on the olefin studied.  The




departure from 1:1 stoichiometry was believed to be due to a secondary




attack on the olefin.  An oxygen effect was also reported by Ripperton




et al. (1972) in their studies of the dark phase reaction of cyclohexene




and ozone.  A decrease in aerosol formation was noted concurrent with a




decrease in the molecular oxygen content.




     Cadle and Schadt (1952, 1953) studied the effect of varying the




reactor temperature and found no statistically significant change in the




bimolecular rate constant.  The temperature was varied from 8 to 27°C




in the ozone and 1-hexene reaction and from 20 to 50°C in the ozone-




ethylene reaction.






B.   Mechanism of the Olefin-Ozone Reaction





     The rate constants for the reaction of ethylene and propylene, both




terminally bonded olefins, with ozone are fairly well known.  The primary




step of the ozonalysis reaction most commonly cited is the Criegee




mechanism (1954, 1955) which leads to the formation of a stable carbonyl

-------
and a zwitterion as



                  0  + OL  +1  RC+HOO~ + RCHO                           (1)



where R can be a hydrogen atom, an alkyl group or an alkyl group contain-


ing an alcohol functional group.  This mechanism has been confirmed for


reactions in solution (Bailey, 1958 and Criegee, Blust and Zinke, 1954)


and as has been assumed by recent investigators to be operative in gas


phase reactions by analogy.  The zwitterion species has never been re-


ported as being observed in the vapor phase.  It was found to be a rela-


tively stable and long-lived species in the studies in solution.  However,


in the gas phase, if this species has an appreciable lifetime, then there


should exist an observable difference in the rate of reactant consumption


and product formation.  The observation of such a difference has only


been reported in one study, i.e., that of Cox and Penkett (1972) and was


interpreted as suggesting that an intermediate having an appreciable


lifetime is involved in at least one path to product formation.


     Representative rate constants for this ozonalysis reaction and many


of those to follow are listed in Table 2-1 along with the appropriate


references.


     For propylene, two modes of the mechanism are reported, one yield-


ing formaldehyde and an acetyl zwitterion and the other yielding acetyl-


aldehyde and the formyl zwitterion.  The zwitterions are unstable


intermediates which can, according to Leighton  (1961), react with an


aldehyde or ketone to form the ozonide (using the acetyl zwitterion as


the example)

                  R,            CH,         ,    R,     ,0    CH_,
                   XV            3\ +  -  k2   !\  /\ /  3
                       C  = O +     .C 00 •  -»•      ^C   C.             (2)


                                                 «

-------
                                                                        10
                               Table 2-1
                     Summary of Rate Constant Data
Rate Constant
      11
      14
      15
      16
      17
      18
      19
      20
      21
      22
      23
      24
      25
      26
                          -2
                          -3
                          4
  Value





1.82 x 10




1.47 x 10




2.13 x 10




2.93 .




2.35 x




2.35 x




2.05 x




2.58 x 10




4.40 x 10




1.47 x 10




8.36 x 103




2.27




8.37 x 101




(2.93 - 29.3)xl04 "




1.57 x 10~4
  Units
   -1 .  -1
ppm  mm
                          -17
                          -3
                          -4
   -2 .  -1
ppm  mm



ppm  min
    Reference





Garvin and Hampson, 1974




Stedman et al.,  1970



Stuhl, 1973




Niki et al., 1972



Niki et al., 1972




Niki et al., 1972



Garvin and Hampson, 1974




Garvin and Hampson, 1974



Heicklen, 1973




Garvin and Hampson, 1974




Washida et al.,  1973



Garvin and Hampson, 1974




Hampson et al.,  1973



Garvin and Hampson, 1974



Hendry, 1974

-------
                                                                          11




It can decompose,




                   CH3         k
                      \  +  -   3
                       X 00   -»•   CH4 + C02                          (3)


                    H          .

                                4
                               -»•   CH OH + CO                         (4)




                               k5
                               -»•   CH2=C=0 + H20                      (5)




                               k6
                               -*•   CH COOH                            (6)




                               k7
                               -»•   HCOOH (from formyl zwitterion)     (7)
it can rearrange
                   CH,          .           0

                     3\  +  -  k8       *
                        C 00   -»•   CH OC.                             (8)
H

           k
                                        .

                                         TI



                                                                      (9)
or it can dimerize




                                     CHK /°~°\ XCH3

                 2       C 00   ^          C           C              (10)

                     '        kio     /   \      /  \
                    H                  H       0 — O       H



Three reactions  with oxygen have also been proposed, the first two of



which are chain  initiating steps





                  CH^             kll
                        C 00  +0   •»•     'OH + CH C00;              (11)


                    H




                  CH3               k
                      \ +  -        12
                        C OO  +0   -»•      CH • + CO  +  -HO          (12)


                    H
                  CH3               k
                      X  +   -        13
                       C 00  4- 02  ->•       03  + CH3CHO            . (13)

-------
                                                                        12
These latter reactions can potentially account for the apparent excess



in olefin consumption observed in some of the studies previously dis-



cussed.  The ozone generated would simply react with additional olefin



while the formation of hydroxyl or hydroperoxy radicals is chain initi-



ating and can drive the system.  This role of the hydroxyl radical was



discussed by Leighton (1961) and Weinstock  (1971).  Niki et al.. (1972)



proposed a hydroxyl radical chain mechanism for the propylene system



which included the reactions



                                   k

                  •OH + CH CH=CH   -»•    CH CH - CH OH               (14)
                          J     ^         3       £


                                   k       00-

                CH CH - CH OH + 0  +    CH CH  - CH OH              (15)
                  J       £      £•        •    HO + CH (?•                 (16)




                                   K17         >P
                       •OH + HCHO  ->•    H20 + HC-                   (17)




Hydroxyl radicals react with hydrocarbons either by hydrogen atom abstrac-



tion or by addition.  The propylene reaction also proceeds via and



addition to the double bond or by hydrogen abstraction resulting in the



radical formation.  The aldehyde-hydroxyl radical reaction also yields



a free radical .  As the acetyl , f ormyl and hydroxyl radicals are



chain carriers, the aldehyde products of the olefin-ozone reaction can



sustain chain reactions .  The rate of the reaction of the aldehydes with



ozone, however, is negligible compared to propylene  (Bufalini, 1974).



     Carbon monoxide, another product of the ozone-olefin reaction will



also react with the hydroxyl radical  (Baulch et al . , 1968)




                                   k!8
                         •OH + CO  -*•    CO  + H-                    (18)



                                   k

                      H- + 0  + M  -»•    M + -HO                     (19)

-------
                                                                    13




The rate constant for the reaction of the hydroxyl radical with carbon



monoxide is about two orders of magnitude lower than for its reaction



with propylene or the two aldehydes.  The reaction of carbon monoxide



with ozone is considerably slower (Pressman et al., 1970) with a



reported rate constant of < 5.9 x 10    ppm   min



     Other reactions have been reported as follows which involve the



reactants or principal products of the olefin-ozone reaction:





                                                                    (20)





                                                                    (21)





                                                                    (22)





                                                                    (23)





                                                                    (24)





                                                                    (25)
                                       i      i



                                 k26          t?
                   •HO  + HCHO   -»•    HO  + HC-                    (26)




                                 k27            /
                  •H02 + CH3CHO  -»•    H202 + CH3C-                  (27)




     O'Neal and Blumstein (1973) proposed a new mechanism for the gas



phase olefin-ozone reaction involving biradical intermediates which react



by various modes.  One mode of their mechanism includes the Criegee



split already discussed.  The other major mode involves the ring opening



of the initial adduct into a species having two active sites, i.e.,



a biradical which can react in a variety of ways.  These involve three



different intramolecular hydrogen abstractions and rearrangements of



the biradical intermediates.
CH O- + CO

•HO + CO
HC- + 02

•HO + 0,
2 3


•OH + 0,
3

'OH + -H0_
2
~20
k
*21
!22
k
*23

k
24


k
*25

CH + CO

•OH + CO
•HO + CO

•OH + 20,,
2


•HO,, + 0,,
2 2

HO +0
2 2

-------
                                                                    14




C.   Reactions with Sulfur Dioxide




     A biradical species was also proposed by an unnamed referee for the



mechanism used by Cox and Penkett (1972) in their ozone-olefin studies



conducted in the presence of sulfur dioxide.  It was this biradical



or zwitterion species or an addition complex which was proposed by those



investigators as being the reactive intermediate responsible for the



oxidation of the sulfur dioxide.  Likewise, Groblicki and Nebel (1971)



suggested the oxidizing species as being a reactive intermediate but



included an ozonide, an epoxide or a free radical as the potential



identity of that species.



     Species which have been identified as capable of oxidizing the



sulfur dioxide are the hydroxyl radical, hydroperoxy radical and ozone.



The rate constants for the ozone and the hydroperoxy radical reactions are



quite slow relative to the rate constant with the hydroxyl radical and for



the three reactions are <2.93 x 10   (Davis et al., 1974), 1.32 (Payne


                           2                        -1   -I
et al., 1973) and 5.58 x 10  (Wood et al., 1974) ppm  min   respectively.



Other radical species which may also be considered for their potential



as the oxidizing agent are the alkoxyl, peroxyakyl and the acetyl



radicals.



     A three step mechanism has been proposed by Cox and Penkett  (1972)



to describe this olefin-ozone-sulfur dioxide system.  The scheme pre-



sented was independent of the nature of the intermediate species and of



other removal processes for this species in the system.  Their mechanism



postulates the following steps:




                            kl
            Olefin + Ozone  ->•   Intermediate (I) + Aldehyde





                     I      -*•   Wall, Decomposition, Reaction, etc.




                 I + SO     ->3  SO  + aldehyde

-------
                                                                        15
The time rate of change in concentration of the intermediate species

[R ] can be expressed as

            [RI]total = aV°l][°3] - k2 [I] - k3[l][S02]           II-l

where a is the fraction of the ozone molecules reacting which lead to

the formation of the short-lived intermediate.  They incorporated the

assumptions that the rate of formation of the intermediate species is

proportional to the initial rate of the olef in-ozone reaction and that a

pseudo steady state concentration of the intermediate species is

established, i.e., (dl/dt)      =0.  If follows then
                        •*! [01] [03]
                  [I] = k2 + k3 [so2]

and
                                        «k k [01] [0 ][SO ]

            -RS0  -      = k3[S°2][l] =    k  + V  [SO]
                             _          _
                     03      1   f. .   2 _ i                     __ .
                         "~                  }-
If the two assumptions are valid, then a plot of this latter relation-

ship, i.e., Rn,/Ror.0 vs [SO ]    would yield, via the intercept, the
fraction of ozone which reacts to form the oxidizing intermediate and

via the slope, the ratio of the second to third step rate constants.  Cox

and Penkett presented such a graph for the olef ins studied (hexene,

2-methyl pentene, trans-2-butene, cis-2-butene and cis-2-pentene)

although only the initial rates and reactant concentrations were used.


D.   Aerosol Production in Hydrocarbon Reactions


     Prager et al. (1960), Wei and Cvetanovic (1963), Rasmussen  (1964),

and Ripperton et al .  (1972) all observed aerosol formation in the

reaction of certain hydrocarbons and ozone alone.  Actually, only the

-------
                                                                        16





olefins out of the group of hydrocarbon categories consisting of




olefins, paraffins, aromatics and aldehydes, react readily with ozone




in the gas phase.  Table 2-2, developed by Niki et al. (1972), presents




relative rate constants for atomic oxygen and the hydroxyl radical as




well as for ozone in their reactions with the different hydrocarbon




classes.




     Groblicki and Nebel (1971) likewise observed aerosol formation in




the dark phase reaction of various olefinic hydrocarbons and ozone.  The




hydrocarbon concentration used in their studies was 4 ppm and the ozone




concentration was 0.6 ppm.  A condensation nuclei count during these




studies showed that a-pinene, 1-heptene, 3-heptene and styrene produced




approximately 10  particles/cm  when added to a chamber containing




ozone.  Cyclopentene, propylene and 2,4,4-trimethyl-l-pentene with




ozone resulted in the formation of about 4 x 10  particles/cm  while




the other hydrocarbons studied either produced less aerosol than the




propylene or none at all.




     Very few studies of the olefin-ozone-sulfur dioxide thermal reaction




have been conducted which included an investigation of aerosol formation




factors.  And, as previously indicated, aerosol measurement instrumenta-




tion was used to yield some integrated parameter on the aerosol such as




the light scattering coefficient, visibility reduction or total nuclei




count and not to provide a quantitative analysis of the aerosol produced.




Groblicki and Nebel (1971)  used a Sinclair-Phoenix Smoke Photometer to




monitor the formation of a light scattering aerosol in their studies




which included sulfur dioxide as a reactant.  They concluded that no




light scattering aerosol was formed unless propylene, ozone and sulfur




dioxide were all present and the amount of aerosol increases with the

-------
                                                                        17
                                 Table 2-2
                         Relative Rate Constants
       Compound

Olefins

   Propylene
   Ethylene
   Isobutene
   Trans-2-Butene
   2-Methyl-2-Butehe
   Tetramethylethylene

Aromatics
Relative Rate Constants*
   Benzene
   Xylene

Aldehydes

   Formaldehyde
   Acetaldehyde
   Proprionaldehyde
0

1.0
0.2
4.4
4.9
14
18
.007

0.05
.15
* .2
0,
3
1.0
0.3
1.7 *• 2
2.8 ^ 36
2.4
3.9 ^ 62

<0.2

	
	
OH

1.0
0.1
2.5
4.2
7.1
8.6
SO. 05
1,1
0.9
0.9
1.8
Alkanes
   n-Butane
  .008
0.24
*Absolute Rate Constants (ppm  min  )  for Propylene
                                  1.76 x 10'
              0           OH

          1.75 x 10"2  2.49 x 104
 Source of Table:  Niki et al. (1972).

-------
                                                                        18
amount of ozone added to the system.  They also concluded that the




sulfate aerosol formed was ammonium sulfate.




     Cox and Penkett (1972) used a condensation nuclei counter to test




for the presence of aerosol during some of their experiments.  The




major effort, however,  relative to the aerosol analysis, consisted of




determining the mass of sulfur collected in bubblers containing hydrogen




peroxide and on membrane filters.  Radiochemical techniques were used




to determine the mass of the   S collected based on the amount of




radioactivity present after correcting for counter efficiency and




radioactive decay.  A deionized water extraction of one filter yielded




a strongly acidic solution containing sulfate ions and it was concluded,




therefore, that the aerosol was sulfuric acid droplets.




     The aerosol formation during the Cox and Penkett studies was




observed to follow second order kinetics.  They also concluded that the




aerosol formation was not significantly affected by surface reactions.




This conclusion was based on the results of an experiment during which




an iron oxide aerosol was added to the reaction vessel increasing the




effective surface area by a factor of 2 to 3.  The rate of ozone con-




sumption and the rate of the sulfate aerosol formation were not signifi-




cantly affected by the additional surface.




     The role of water vapor concentration on aerosol formation in the




sulfur dioxide- ozone -cis-2-butene system was studied over a range of




relative humidities from about 10-78% and over a range of initial




sulfur dioxide concentrations  (0.2 - 3.61 ppm).  The aerosol formation




rate was observed to decrease with increasing relative humidity for all




of the sulfur dioxide concentrations studied.  Wall losses for all of




the system components also increased but the decrease in the aerosol




formation rate was too large to be accounted for by this type of loss.

-------
                                                                       19
Water vapor, it was concluded, inhibits the aerosol formation in the



olefin-ozone-sulfur dioxide reaction.



     Clark  (1972) conducted a comprehensive study of the formation and



growth of aerosols produced by the photooxidation of sulfur dioxide.



The advent of the recently developed Whitby Aerosol Analyzer permitted



for the first time, a study of the nucleation and growth of a forming



submicroscopic aerosol.  This instrument was an earlier version of the



electrical mobility analyzer used in the study reported on in this paper


                                                            •5

and has been described by Whitby and Clark  (1966).   A 625 ft" FEF



Teflon bag was used as the containment vessel and the ultraviolet



irradiation was provided by a cylindrical bank of fluorescent lamps.



The lamps used had about the same spectral energy distribution as sun-



light over the wavelength range of interest, i.e.,  2900 to 3400 £ for



SO., photooxidation.  Two important factors relative to the use of these
  t.


lights are that they essentially provide full ultraviolet energy output



instantaneously and that they do not introduce excessive heat.into the



system.



     The sulfur dioxide concentration ranged from 0.049 to 2.88 ppm



during these studies which yielded volumetric conversion rates of from



0.389 to 22.9 ym /cm -hr.  The volume concentration was observed! to



increase non-linearly with time during the initial  stages of the runs



which was attributed to a nucleation process.  The  increase in the



volume concentration was later characterized by a constant rate of con-



densation of material from the vapor to the condensed state.  The calcu-



lated total surface area of the aerosol formed during these studies was



found to achieve an apparent equilibrium value which appeared to be a



function of the volumetric conversion rate.

-------
                               CHAPTER III




              THEORY OF AEROSOL FORMATION, GROWTH AND DECAY






     Aerosols can be divided into two major classifications according




to the mechanism? involved in their formation.  Aerosols of large indi-




vidual particle size are primarily formed either by the communition or




nebulization of bulk materials into discrete particles small enough to




remain suspended in the gaseous medium.  The much smaller aerosols are




formed by chemical reactions leading to polymerized species or by the




physical agglomeration of molecules leading from the vapor to the con-




densed phase.  This formation mode is commonly referred to as "gas to




aerosol conversion1.




     This report is concerned with only the second classification, i.e.,




the transformation to particulate material resulting from the phase




transition from the gas to the liquid state.  The photooxidation of




sulfur dioxide and the oxidative consumption of sulfur dioxide in the




thermal reactions involving certain hydrocarbons and ozone are examples




of reactions that result in gas to aerosol conversions.  The initial




step in the agglomeration process is the nucleation of molecules from




a supersaturated vapor.  These nuclei may consist of clusters of the




condensable molecular species or foreign material (ions or particles)




of submicroscopic size.  The growth of the newly formed embryos continues




by the diffusion of condensable material from the saturated vapor phase




directly onto the existing particulate surface.  These processes




individually and in concert affect the time-dependent size distribution

-------
                                                                        21





spectra of the aerosol.  One other physical mechanism which influences




this distribution is coagulation.  This process is an agglomerative




interaction which occurs if the concentration of the aerosol droplets




is sufficiently high.






A.   Nucleation





     Nucleation is the first step in one type of gas to liquid or solid




phase transition process.  Homogeneous nucleation refers to the sponta-




neous formation of molecular aggregates or embryos resulting from the




random collisions of molecules in a supersaturated vapor.  The nucleation




is termed heterogeneous if the condensation process occurs on a pre-




existing solid or liquid phase, i.e., foreign particles or charged ions.




Prerequisite to any self-nucleation is the establishment of a super-




saturated vapor in the gas volume.  The supersaturated vapor develops




in chemical reactions between gaseous components only when the saturated




vapor pressure of the reaction products is less than the saturated vapor




pressure of the reactants.




     The classical treatment of nucleation is included in works by a




number of authors including Amelin (1967)  and Mason  (1971) .  The




experimental effort described in this paper involved a system which was




initially free of particles or ions.   The treatment to follow, therefore,




is devoted to spontaneous homogeneous nucleation.




     The embryos which rapidly form as a result of the statistical




fluctuations in the gaseous medium are also dissipated by evaporation




or breakup because of these same local thermal and density variations.




Only if the nuclei exceed some critical size (g*) will they survive




and continue to grow by condensation on the embryonic surfaces.  The




free energy of the system is at its maximum value when the size of the




nuclei is g*.

-------
                                                                        22
This variation of free energy in a supersaturated system with droplet



size can be plotted as
                   AG
                                     g*
where AG is the incremental change in free energy and g is the number of



molecules in the aggregate.  The supersaturated vapor is in a metastable



state with the chemical potential of the vapor being greater than that



of the condensed phase, i.e., y >y. .  The embryos will develop only if
                               a  b


they grow to a state where they are more stable than the vapor phase or



grow with a decrease in free energy to g>g*.  In the case of heteromole-



cular homogeneous nucleation, a three dimensional plot is required to



accurately represent the change in free energy.



     The total change in the free energy of a system upon the formation



of .an embryo of g molecules is described by the Gibbs free energy func-



tion for an isothermal, isobaric change as




                  AG =  (y,  - y )g + TT d2 0                        (ili-l)
                         b    a



where y   =  chemical potential/molecule in the vapor system
       cl



      y,   =  chemical potential/molecule in the condensed phase
       b



      d   =  diameter of the molecular aggregate



      o   =  surface tension



The second term on the right is the surface free energy which goes from



zero to this value upon nucleation and is significant because of the



large surface to volume ratio for these small clusters.

-------
                                                                        23
     In order to form nuclei of size g*,  dAG/dg must be  at  a  maximum,


i.e., dAG/dg = 0.  By differentiation of  equation  III-l  this  condition


is satisfied when
                       „           •
                   b    a     3g


The critical free energy change is,  therefore,



                  AG*  = 1^                                 IJI-3


where d* = the diameter of the critical nucleus containing  g*  molecules.


     An alternative expression relating the critical  diameter  for con-


tinued growth by condensation to the degree of  saturation can  be  developed


from equation III-2 by expressing g  in terms of the density and diameter


of the condensed molecular cluster and the difference in the chemical


potential of the vapor and the condensed phases in terms of the satura-


tion ratio (S) .   This relationship was first developed by Kelvin  (1870)


and can be expressed as

where  M  = molecular weight of the droplet


       p  = density of the droplet


       R  = universal gas constant


       T  = temperature .


S, the saturation ratio or degree of supersaturation is  expressed by.


the equality




                  s  m-£-m


where   p  = vapor pressure in the gas phase and


    p  (T) = saturated vapor pressure over a plane surface of the


             liquid at the temperature T.

-------
                                                                        24



     The rate at which vapor molecules are impinging per unit area on



the embryonic surface is given by




                  B  =  n p (2?r M R T)"15                         III-6




where n = Avogadro's number



and the rate of incorporation assuming an accommodation coefficient of



unity is simply the product of the impingement rate B and the surface



area.  And finally, the rate of nucleation in embryos/cm -sec can be



expressed as





                  |^- = it d*2 B exp (-AGVkTJ                      III-7
                  at



where  N  = number concentration of embryos cm   and



       k  = Boltzman's constant.



The preexponential term is a kinetic factor the expression of which



differs among various investigators (Becker and Doring, 1935; Reiss,



1950; Mason, 1971; and Mirabel and Katz, 1974).



     From the foregoing discussion, it follows that homogeneous nuclea-



tion occurs in a gas volume only if S > S*, i.e., at some critical super-



saturation condition.  The critical supersaturation condition is arbi-



trarily defined as that which will produce embryos capable of further



growth at the rate of one per unit volume per unit time.  The vapor



supersaturation ratio is included in the exponential term in equation



III-7 and the rate of nucleation, therefore,  increases sharply with the



degree of supersaturation.



     The critical supersaturation required for spontaneous homogeneous



nucleation can be calculated from that equation by setting dN/dt = 1.



As vapor supersaturation increases it can also be seen from equation



III-4 that the embryo diameter decreases and from equation III-7 that



the number of embryos increases.

-------
                                                                       25



     One of the earlier works dealing with the kinetics of the sulfur


trioxide and water reaction was conducted by Goodeve et al. (1934).


The resultant product of this reaction, molecular hydrogen sulphate, was


observed to be in a state of supersaturation and condensed to produce


sulfuric acid molecular aggregates.  They concluded from their reaction


velocity measurements that about one collision in one hundred results


in an association, and further, that the lack of a reaction in the


balance of the collisions was possibly due to an energy of activation,


a steric factor or the necessity of a third body collision to remove


the energy of association.


     The self-nucleation rate for the sulfuric acid-water system was


theoretically treated by Doyle (1961).  The rate of the transition was


characterized by a vector model developed by Reiss (1950) and adapted


for their study.  The existence of a saddle point was assumed on the free


energy surface, and the rate of flow of embryos through this pass in the


energy barrier was used to approximate the rate of nucleation.  The


nucleation rate was calculated and tabulated as a function of the partial


pressure of the sulfuric acid in the vapor phase for a system at 25°C and


50% relative humidity.  These data indicate that self-nucleation becomes

                                                          -9
appreciable at sulfuric acid partial pressures of about 10   mm although,


according to Doyle, the rate is uncertain by three to four orders of


magnitude at this pressure.  He attributed the source of the error to the


failure of the liquid droplet model at this supersaturation, uncertain-


ties in the acid partial pressure and uncertainties concerning the nature


of the dominant acid-bearing species at elevated humidities.


     Limited theoretical studies of the rate of formation of critical


sized molecular clusters at relative humidities have recently been


conducted by Kiang et al.  (1973)  for the sulfuric acid-water system.

-------
                                                                        26




Relative humidity is demonstrated to be a dominant factor in gas to


                                                                  9
particle conversions.  Their data indicates that only about 3 x 10



sulfuric acid molecules/cm  are required to achieve a unit nucleation



rate (cm~  sec" )  in a system at 10% relative humidity.  The errors



associated with this theoretical treatment relate to uncertainties in



the nucleation theory for small molecular aggregates.





B.   Condensation




     Growth of the molecular aggregates formed as a result of spontaneous



nucleation continues by two processes, condensation and thermal coagula-



tion.  The initial growth of the newly formed droplet due to the direct



condensation of material from the vapor phase is rapid according to



classical theories relative to the early growth due to coagulation.



Homogeneous nucleation followed by surface condensation occurs, according



to equation II1-4 when the supersaturation in the gas volume exceeds



some critical value.  Once an embryo exists, e.g., from the heteromole-



cular nucleation of sulfur trioxide and water, it becomes a soluble



nucleus upon which condensation can occur.  The equilibrium vapor pressure



over this newly formed surface is lowered according to its composition and



concentration.  Condensation will occur on this surface, therefore, at



a lower supersaturation level than required for the primary nucleation



or for condensation on an insoluble particle of the same size.  When the



solution droplet is in equilibrium with its surrounding air,, its vapor



pressure must equal the partial pressure of the water vapor, and the



saturation ratio equals the relative humidity of the air.



     An expression for the equilibrium water vapor pressure at the



surface of a droplet of solution is given by Mason (1971) as



                  .  p"    a1 M       P M                        TTT 0
                  In *•—  = —r——,_  -   ,  ^m                      III-8
                     p     p ' RdT     p '  RT

-------
                                                                        27





where  p1  =  equilibrium vapor pressure over the solution droplet




       a1  =  surface tension of the solution




       M   =  molecular weight of water




       p'  =  density of the solution




       P   =  osmotic pressure of the solution.




Mason also gives an expression for the osmotic pressure for solutions




whose densities vary linearly with concentration as




                                         * n
                           M      n       n2




where  p  =  density of water




       n  =  moles of solute




       n_ =  moles of water




       i  =  factor which depends on the composition and the concentra-




             tion of the solute.




The density of sulfuric acid solutions does not vary linearly with




concentration (Chemical Engineers Handbook, 1963) although over the




range of about 15 - 50% acid the relationship is nearly linear.  The




combination of equations III-8 and III-9 yields two predictions relative




to the percent acid for the equilibrium system of sulfuric acid and water.




First, the percent acid decreases with increasing relative humidity for




a fixed particle size and secondly, the percent acid increases with




particle size for a constant relative humidity.




     The rate of condensation on a droplet suspended in a gaseous medium




is expressed by the Maxwell equation (1890) corrected by Fuchs (1959)




for the influence of the droplet diameter as





                  dq     2ir D d M   ,
                  —•*•  =  ————   (D - D )
                  dt      R T C      p   PT




where q  =  quantity of vapor in the gas

-------
                                                                        28



      U  =  vapor diffusion coefficient



      c,  =  a coefficient the value of which depends on the Knudsen number



      p  =  equilibrium vapor pressure at the droplet temperature.




     The influence of the Knudsen number on the rate of condensation can



perhaps be better understood by defining the characteristics of the trans-



fer process regimes.  The Knudsen number (K ) itself is defined as the



ratio of the gas mean free path (1) to the particle radius, i.e.,



K  = 21/d.
 n


     In the free molecule regime where d $ .01 ym and K  -»• °° the particles
                                                       n


behave as giant molecules and do not disrupt the random collisions of



the molecules of the medium.  Transport properties for this regime may



be evaluated from the kinetic theory of gases.



     The continuum regime includes large particles, i.e., d ;> 1 ym and



K  -*• 0.  In this region, the transfer processes are adequately described



by the equations of continuum mechanics.  The molecular structure of



the medium is not included in the description.



     The region between these two extremes is referred to as the transi-



tion regime, i.e., where 0.01 < d < 1 lam and 13.2 > K  > 0.132.  Particles



in this size range disrupt the random motion of molecules in the medium,



and the molecular structure of the medium is still needed to describe



the system.



     A comprehensive treatment of the transport processes and the transfer



equations for an isolated sphere in these three regimes is presented by



Hidy and Brock  (1970).  The condensation rate can, therefore, be



approximated through the use of the equations for the mass flux on the



surfaces of a spherical droplet.



     As a droplet increases in size, the coefficient t decreases

-------
                                                                        29





approaching unity for d » 1 and the Maxwell equation applies.   Expressed



in terms of particle growth this equation becomes



                  dd   4 D M  ,      .                            TTT -11

                  dT = iTTd  (p - V '                           II1-11



For smaller droplets the coefficient increases and varies according to



the accommodation coefficient which is  sensitive to the molecular struc-



ture of the liquid.



     If the nuclei exceeds the critical diameter (equation III-4) in



a supersaturated system, then it will continue to grow with a decrease



in the free energy.  The transition would, according to theory, be rapid



and the droplet formed would grow without limit.  In practical systems



which can be conceived, however, the supersaturation does not remain



constant as the embryos compete for the condensable vapor which is being



consumed at an increasing rate.  The supersaturation, therefore, dimin-



ishes and the upper limit for the growth of the droplets is restricted.





C.   Coagulation




     Following the rapid growth of the  newly formed nuclei by condensa-



tion, the continued growth of the particles or droplets will be dominated



by coagulation assuming that the number concentration remains above about


  4   3
10 /cm .  Coagulation is a coalescence  of the nuclei due to the occurrence



of random collisions caused by the aerosol undergoing Brownian motion.



The theory predicting a second-order decay for the aerosol originated



with Smoluchowski  (1917) who applied the diffusion equations to the



Brownian motion of hard non-interacting spheres thermally agitated in a



continuum.  This aspect of aerosol stability has been extensively



treated by Fuchs (1964) and more recently by Hidy and Brock  (1970) and



Husar  (1971).  The basic coagulation equation follows from a consideration

-------
                                                                        30




of the number of collisions which occur per unit volume and time in an



aerosol due to diffusion and the number of particles is reduced by one



with each collision.  The equation describing the decay of the number



concentration of equally sized aerosol particles is given as




                  ^- =  - K N2                                       111-12
                  dt



where N  = the number concentration per unit volume cm   and



      K  = the coagulation constant (cm  sec  ) which is given by



           K  =  4n d D (D here being the particle diffusion coefficient).



     The value of K increases sharply as the system becomes more poly-



disperse and decreases as the particle size increases.  The coagulation



constant for unequal size particles is given by Fuchs (1964) as




                             dl + d2  °1 + °2
            K(d1,d2)  =  4 it (-±-^—-)  (-±-2—-) 0                 111-13




where d  + d   =  the diameter of the absorbing sphere in Fuchs1 model



      D  + D   =  the coefficient of relative diffusion for the two



                  particles



            B  =  a correction factor necessary to account for the



concentration discontinuity which exists at the surface of the absorbing



sphere and is a function of the aerosol transfer process regime.  Values



of K(d ,d ) are given in Table 3-1 from which it is evident that the



size distribution of the aerosol will rapidly become devoid of small



particles and the size of the larger particle is not significantly



affected by the interaction.  Several more elaborate approaches than



the limiting sphere calculations of Fuchs have been performed including



those of Sahni (1966), Smirnov (1969)  and Walter (1973).  The results



have not significantly altered the values obtained from the



original work.

-------
                                                   Table 3-1
                               Coagulation Constants for unequal sized particles




                                          K(dd) x 10~10        '
d]L(ym)
.002
:oo4
.01
.02
.04
.1
.2
.4
1
2
4
10
20
.002
4.5
7.5
30
90
300
1600
5000
13000
37000
77000
160000
400000
800000
.004

6
13
40
110
550
1400
4100
9500
20000
40000
100000
200000
.01


9
15
35
120
270
600
1600
3200
6600
16000
30000
.02



12
17
40
80
170
420
940
1700
4300
8500
.04




11
15
25
47
115
230
450
1100
2200
.1 .2 .4 1





7.2
8 5.2
11 5.3 4.0
24 9 4.7 3.4
45 16 7.1 3.7
90 30 12 5.2
220 72 28 10.3
430 140 54 19
2









3.2
3.4
5.6
9.6
                                                                                                   10
20
                                                                                              3.0




                                                                                              3.7    3.0




                                                                                              5.5    3.3    3.0
Adapted from Fuchs, 1964
                                                                                                                OJ

-------
                                                                        32



     One other phenomena affects the size distribution of the aging



aerosol, i.e., the sedimentation of the largest particles.  The net



tendency of the interactions which occur in the aerosol including the



nucleation, condensation, coagulation and sedimentation is for the size



distribution to become self-preserving in time.  A similarity theory to



describe the shape uniformity was originally presented by Friedlander



(1960) .   Subsequently Clark and Whitby (1967) presented data from a



number of investigations properly normalized to demonstrate this self-



preservation.  The data reported encompassed a wide range of experimental



conditions and suggest that a similarity spectrum forms in a matter



of hours for atmospheric aerosols originating from a number of different



sources and conditions.  The similarity spectrum fit, however, was not



that predicted by the theory and at present there exists no direct



verification of the theory (Hidy and Brock, 1970).



     Upon integration, equation 111-12 becomes




                  | - i.  = Kt                                   111-14

                       o





where N  = the initial particle concentration.  The straight line
       o


relationship between time and the reciprocal of the particle concentra-



tion dictated by this equation implies that K is invariant in time.  The



value of K derived from this relationship has been the main experimental



fact reported by previous investigators in treating gas to particle



conversions.  This limitation in quantitative aerosol analysis was pri*-



marily due to the lack of the sophisticated instrumentation necessary to



characterize the small particles generated by the binary homogeneous



nucleation process.



     A term to account for the deposition of aerosol particles on the



walls or floor of an experimental chamber was added to equation 111-12

-------
                                                                        33




by Langstroth and Gillespie (1947).   The decrease in the number concen-



tration due to this 'wall loss1 is proportional to N, the concentration



at any time t and the change in concentration can be expressed as




                  - ^T = B N + KN2                             111-15
                    dt



where 6 = the coefficient for wall loss.  In a chamber which is not



mechanically stirred, deposition on the walls must be diffusive.  The



small particles are transported near the wall by convective diffusion and



deposited on it by molecular diffusion through a stagnant air layer at



the containment surface.  The thickness of this layer was reported by



Langstroth and Gillespie to 20 ym although the value varies considerably



among investigators.  One of the more recent theoretical treatments was



performed by Van DeVate  (1972) in which a thickness of 0.85 mm was



calculated for the stagnant air layer.  The coefficient B, accounting for



the wall layer is given by Fuchs  (1964) as




                                    6 = H                      IU-16





where  S  =  the surface area of the chamber



       V  =  the chamber volume



       §  =  the thickness of the wall layer.  Implicit in this expres-



sion for the wall loss constant is the bias of the wall for small



particles and the dependence of this term on the surface to volume ratio



of the chamber.

-------
                               CHAPTER IV




                 EXPERIMENTAL ARRANGEMENT AND PROCEDURES






A.   Experimental Design and Arrangement





     The experimental series discussed in this paper was designed to




provide insight into the oxidative consumption of the reactant-sulfur




dioxide and into the subsequent growth of the resultant aerosol.  Initial




experiments were conducted on the propylene-ozone system for comparative




purposes and to establish a baseline for later studies conducted in the




presence of sulfur dioxide.  Nitrogen was used as the diluent gas in




certain of these studies to test the role of molecular oxygen on the




reaction kinetics and on the reaction stoichiometry.  Two levels of




relative humidity were also established in the reactor in an attempt




to determine the influence of that parameter on the gas-phase reaction.




     These same parameters as well as the initial reactant concentrations




were varied in the studies which included sulfur dioxide.  The propylene




and ozone were varied from conditions of olefin excess to ozone excess




and two levels of sulfur dioxide were used.  Target concentrations of the




primary reactants, propylene, ozone and sulfur dioxide, were generally




as close to representative atmospheric levels as possible consistent




with the limitations imposed by the instrumentation used.  This limita-




tion is most restrictive in the measurement of the time history for low




total reactant consumption.  A few runs were made at elevated concentra-




tions to obtain a higher product yield thereby increasing the sensitivity




for the measurement of minor reaction products.

-------
                                                                        35





     A high concentration of carbon monoxide was added to two of olefin-




ozone-sulfur dioxide runs to make use of its capacity as a hydroxyl




radical scavenger in determining the participation of that species in




the reaction.  Acetyaldehyde, which is reportedly formed as a result of




the propylene-ozone reaction, was investigated to determine its reactivity




in the presence of ozone or conversely its stability as a product of the




primary reaction.




     A series of low reactant consumption runs were made wherein the




volumetric conversion rate was limited in order to follow more closely




the aerosol formation and growth.




     Finally, a limited number of ethylene and ozone and ethylene-ozone-




sulfur dioxide experiments were conducted to determine similarities with




the propylene reactions and to observe the aerosol formation characteristics.




     A schematic outlining the complete laboratory arrangement is shown




in Figure 4-1.  The reaction containment vessel is shown at the top of




the figure while the reactant introduction system is shown on the right




side.  This system includes the gas cylinders, the pressure and flow




regulation, ozonization and humidification.  The diluent gas stream was




initially split twice with a metered fraction of the total flow passing




through the humidifier.  Upon recombination there were two air streams




of similar properties, i.e., composition, flow, humidity and temperature.




Ozone was added to one of these inlet lines and the olefin and the sulfur




dioxide to the other thereby limiting the reactions to the confines of




the reactor.




     The aerosol measuring instrumentation is shown directly below the




reactor where it was physically located.  The positioning of these devices




was dictated by the requirement to minimize aerosol line losses.

-------
                 OZONE
               GENERATOR
                                 VENT
                               DEW POINT
                              HYGROMETER
                 REACTOR
                                    AMBIENT,
                                      AIR
                        SULFUR DIOXIDE
                          ANALYZER
                           OZONE
                         ANALYZER
                            GAS
                       CHROMATOGRAPH
                       OPTICAL AEROSOL
                           COUNTER
                        CONDENSATION
                       NUCLEI COUNTER
       ROTAMETER

       ELECTRICALLY ACTUATED
       VALVE

       CHARCOAL COLUMN
FILTER HOLDER
   4
INJECTION
 SEPTUM
                                MOBILITY
                               ANALYZER
                                 FILTER
                                SAMPLES
                                INFRARED
                              SPECTROMETER
                               INTEGRATING
                              NEPHELOMETER
                               MAGNEHELIC
                                 GAUGE
                                                                           00
                                                                                              .
                           HUMIDIFIER
                          \	
                           A
                                       PROPYLENE

FIGURE 4-1 .  SCHEMATIC OF EXPERIMENTAL ARRANGEMENT.
                  SULFUR
                  DIOXIDE
 AIR OR
NITROGEN

-------
                                                                        37





     The gas-phase instrumentation is shown on the left along with the




dew point hygrometer.  The volume of the reactor would not permit con-




tinuous sampling by all of the instrumentation shown although to eliminate




dynamic instabilities and to minimize the response time of the instrumen-




tation it was necessary to have all measurement devices in continuous




operation.  The ambient air inlet shown at the upper left allows the




sulfur dioxide analyzer and the ozone analyzer to be switched to sample




the room air and the reactor intermittently.  Similar arrangements were




made for the remainder of the instruments.




     A photograph of the major portion of the instrumentation used




routinely during this study is shown as Figure 4-2.  The reactor is




shown at the top center and the aerosol instrumentation directly below




except for the integrating nephelometer.  This latter device, the




sampling section of which is oriented vertically, is located immediately




to the left of the reaction vessel.  The reactant introduction system




can be seen in the background near the center of the picture.  The gas




chromatograph is shown on the right and the sulfur dioxide and ozone




analyzers are to the left of the aerosol instrumentation table.  At the




extreme left is a computer and display scope used for the processing of




data from the optical particle analyzer.




     Figure 4-3 is a picture of the front panels of the condensation




nuclei counters, the optical particle analyzer and the electrical




mobility analyzer with its control module.  Figure 4-4 is a picture of




the Fourier transform infrared spectrometer.  The "folded path" cell




is in the background with the detectors, interferometer and recorder in




the foreground.

-------
FIGURE 4-2. LABORATORY ARRANGEMENT FOR OLEFIN-OZONE-SULFUR DIOXIDE STUDIES.
                                                                                         UJ
                                                                                         00

-------


FIGURE 4-3. AEROSOL MEASUREMENT INSTRUMENTATION.
                                                                           Co

-------

                                                          40
FIGURE 4-4. FOURIER TRANSFORM INFRARED SPECTROMETER

-------
                                                                        41



B.   Experimental Apparatus


     1.  Reactor


         The reaction bag was constructed of Dupont "Tedlar" PVF film


which was folded and then heat sealed along three sides.  When inflated


the bag took on a pillowcase configuration.  It was approximately 1.83m


in length and 0.94m in width and when fully inflated contained approxi-


mately 437 liters.  The surface to volume ratio for the fully inflated


bag was 7.87m  .


     "Tedlar" was selected for the reactor because of its low reactivity


with the type of compounds encountered in these studies.  The bag con-


figuration was employed because, in addition to its economy and sim-


plicity, the pressure in the system remains constant at ambient atmospheric


pressure as the volume is depleted by sampling.  A dilution correction is


required when rigid walled containers are used and makeup air is added


to replace the sampled volume.


     An aluminized polyester bag (3M Company 'Scotchpak1) was constructed


and placed over the "Tedlar" bag to eliminate the transmission of


ultraviolet radiation and also to reduce the rate of permeation of gases


and vapors into or out of the reactor.  A polyethylene tape was used to


seal the sections of this bag together.  Permeation rates for several


gases and vapors are shown in Table 4-1 for the two materials in cm  or

         2
g/(100 cm )(24 hrs) per mil thickness.  The permeation of water vapor


through the "Tedlar" was measured at a time when the relative humidity


in the laboratory was 54% and the ambient temperature was 27°C.  Figure


4-5 shows the increase in the relative humidity with time which is


estimated at 57.4 mg HO/(100 cm )(24 hrs) per mil.  The permeation of


water vapor was again measured but this time through both the "Tedlar"


and the 'Scotchpak1.  Over the anticipated lifetime of the experiments

-------
                                                                        42
                                Table 4-1


                             a              b
       Permeability of Tedlar  and Scotchpak  to Gases and Vapors
Tedlar
    Gases



        Oxygen




        Nitrogen




        Carbon Dioxide



        Hydrogen






    Vapors



        Water




        Ethyl Alcohol



        Benzene



        Acetic Acid
cm3/dOO cm2) (24 hrs) (mil) at 23.5°C
                .50




                .04



               1.7




               9.0
  g/(100 cm2)(24 hrs)(mil) at 23.5°C
                .50 (37.8 C, 90%RH)




                .08




                .22




                .11
Scotchpak
    Gases



        Oxygen




        Carbon Dioxide



        Air
  3        2
cm /(100 cm )(24 hrs)at 1 atmosphere
                .05



                .05



               <.02
Vapors
        Water
  g/(100 cm )(24 hrs)
                .003 (37.8°C, 95%RH)
 Adapted from DuPont Company



 Adapted from 3M Company

-------
                                                              43
50
40
30
20
10
           12345

                            TIME (hours)

FIGURE 4-5.  VARIATION OF RELATIVE HUMIDITY WITH TIME DUE TO THE
           PERMEATION OF WATER VAPOR.

-------
                                                                        44





to be conducted, one to two hours, no change in the relative humidity




could be detected within the resolution of the instrumentation used.




     The reactor was supported horizontally on a lattice frame suspended




from the laboratory ceiling and about two meters above the floor.  Air




was permitted to circulate freely about the outer bag.  Seven Swagelok




0-Seal, V O.D. stainless steel (316) adapters were fitted to the bag




with threaded teflon blocks in the pattern shown in Figure 4-6.  Addi-




tional thickness was provided at these points by the addition of a




layer of the polyethylene tape.  Two of the ports were used to introduce




the reactants into the system and the other five were reserved for




sampling.






     2.  Ozone Analysis




         A Meloy Laboratories, Incorporated, Model OA350 Ozone Analyzer




was used to monitor the ozone concentration in the reactor.  The gas




sample and ethylene are delivered simultaneously to a mixing zone where




they react and produce chemiluminescence in the 430 pm region.  This




light is photometrically detected and is directly proportional to the




ozone concentration.  The photocurrent thus produced from this flameless.




reaction is converted to a voltage, amplified and then is available as




an analog signal and/or is displayed on a panel meter.




     Oxidizing species other than ozone and reducing species which might




be expected to be encountered during this series of experiments do not




interfere with the analysis.




     The instrument has six .operating ranges to cover the span of from




0 to 10 ppm and its minimum detectable sensitivity is 0.001 ppm.




     Primary calibration of the analyzer was achieved by generating an




ozone atmosphere in the reactor, the concentration of which was determined

-------
                                                         45

.
I
~\
CD
5 ) 33 cm
  1 OZONE AND DILUENT GAS INLET
  2 MAGNEHLIC GAUGE, INTEGRATING NEPHELOMETER
  3 VACUUM, ELECTRICAL MOBILITY ANALYZER, IR SPECTROMETER,
      FILTER SAMPLES
  4 OZONE AND SULFUR DIOXIDE ANALYZERS
  5 DEW POINT HYGROMETER
  6 OPTICAL PARTICLE ANALYZER, CONDENSATION NUCLEI COUNTER,
      GAS CHROMATOGRAPH
    FIGURE 4-6. SCHEMATIC OF REACTOR PORT ARRANGEMENT.

-------
                                                                        46




using the Reference Method for the Measurement of Photochemical Oxidants




(Environmental Protection Agency, 1971) .  The instrument was spanned on




the concentration value thus obtained.  A three-way solenoid value was




installed, which when energized, closed the sample inlet port and allowed




ambient air to pass through a charcoal filter.  This provided ozone free




air to the reaction chamber for a zero calibration.




     The instrument also has an integral ozone generator (Pen-Ray ultra-




violet lamp) which provides a constant ozone source on demand to the




detector.  This secondary calibration was used to monitor the stability




of the analyzer and to determine primary calibration requirements.




     3.  Ambient and Dew Point Temperature Measurements




         A Cambridge Systems, Incorporated, Model 880 Dew Point Hygrometer




was used to sense the dew point temperature dynamically in the reactor and




was also used to monitor a remote Yellow Springs Instrument Company




Thermistor, Model 410X.  The thermistor had a 3.4 second time constant,




an active element 5.1 cm in length and was used to sense the temperature




in the bag.




     The hygrometer measures the dew point of a gas sample by presenting




a metal surface to the gas which will be at the temperature of the metal.




The surface is cooled to the dew point temperature of the gas being




sampled and condensation occurs.  At the dew point temperature, an




equilibrium condition exists and the rate at which molecules leave the




surface water is equal to the rate at which water molecules enter the




surface water.




     An optical system senses the presence of condensate on the metal




surface which is mirrored and an optical-thermal feedback system adjusts




and maintains the surface at the dew point temperature.

-------
                                                                        47




     The range of the hygrometer is -40°C to 50°C, its accuracy is ±1.1°C



and with this method being a primary measurement, no calibration is



required.



     The relative humidity of the gas sample defined as the ratio of the



actual vapor pressure in the gas to the saturation vapor pressure with



respect to water at the prevailing dry bulb temperature can be calculated



as follows:



_ ,  ..    ,   ....    Saturation Vapor Pressure over water at Dew Point Temp.
Relative humidity =_ .	—	—*•	.  . . .	——	*-—
                   Saturation Vapor Pressure over water at Ambient Temp.




The Smithsonian Meteorological Tables (list, 1958) were used to obtain the



appropriate saturation vapor pressure data.





     4.  Sulfur Dioxide Analysis



         A Meloy Laboratories, Incorporated, Model SA 185-2 Flame



Photometric Detector Sulfur Analyzer was used to monitor the sulfur



dioxide concentration in the reactor.  This instrument provides real



time analysis by monitoring the intensity of light emitted by sulfur as



it passes through a hydrogen rich flame.  High detection sensitivity



is achieved by the geometrical arrangement of the burner block and the



photomultiplier tube and specificity by a narrow band pass  (394 ym)



filter.  Sulfur containing molecules are converted to an S  species in



the hydrogen rich flame.  The hydrogen and hydroxyl radicals which are



also produced in the flame react with the S  to form activated S * as



follows:








                 •OH + -H + S  -»• S * + HO




Radiation with a maximum intensity at 394 ym is emitted when this excited



S * species reverts to a lower energy state.  The intensity of the light

-------
                                                                        48


produced is directly proportional to a power slightly less than the

square of the sulfur concentration.

     The range of the instrument is 0-1 ppm and its minimum detection

sensitivity is 0.005 ppm.

     Calibration of the sulfur analyzer was achieved through the use of

a 2 cm long, Metronics Associates, Incorporated, Dynacal Permeation tube

(#2140) and a constant temperature (30°C) Meloy SO  Gas Calibrator Model

CS-10.  The use of permeation tubes to generate low concentrations of

gases has been described by O'Keeffe and Ortman (1966).  The permeation

rate of the tube was calibrated gravimetrically over a period of 18 hours

using a Cahn Vacuum Electro Balance System Model R100 at 25°C as 558 pg/min.

     The concentration of SO  in ppm can be expressed as

                   /   N    R 22.4  T  760
                  c (ppm)  = ? — — —                        IV-1

where  R =  permeation rate (yg/min)

       F =  Gas flow rate (ml/min)

       M =  molar gas volume at operating conditions

       T =  Temperature of the gas (°K) and

       P =  Pressure of the gas (mm Hg).

The change in permeation rate with temperature is given by (Meloy, 1972) :

                                         T  - T
                  log R2 = log R  + 2950  2    l                 IV-2
                                           2 1

The permeation rate, therefore, for this series of experiments was 812.4

ug/min at the 30°C operating temperature and the output concentration used

for spanning the sulfur analyzer was 0.365 ppm.

     The instrument zero was achieved by passing air through an internal

charcoal column prior to its passage into the burner block.

-------
                                                                        49





     As this instrument is a total sulfur analyzer, a Millipore Mitex




(teflon) LSWP filter was added to the sample inlet to eliminate any




contribution from the sulfate aerosol.




     A Lear Siegler, Incorporated, Model III Second Derivative Air




Analyzer was used to measure the sulfur dioxide concentration in the




reactor when nitrogen was used as the diluent gas.  This instrument




measures the curvature or rate of change of the slope of intensity




with respect to wavelength for the absorption of ultraviolet radiation




at a prescribed wavelength.  Its minimum detectable concentration for




sulfur dioxide is 10 ppb using the 6.3)1, 1m long, multipass cell.  Its




relatively large volume and high sampling rate made extensive use of this




instrument impractical.  The output signal from the ultraviolet photo-




multiplier tube was recorded on an integral component Honeywell Electronik




193.






     5.  Hydrocarbon Analysis




         A Tracer, Incorporated, Model MT150 Gas Chromatograph was used




to monitor the concentration of the unreacted olefin and the product




acetyaldehyde in the reactor.  A flame ionization detector measured the




difference in the ion current due to the combustion of the sample.  Sample




was continuously drawn at 100 cm /min through a 5cm  loop the contents of




which could be injected via an eight position pneumatic valve into the




chromatographic column.  Poropak Q was used to pack the 2.75 m column




for the adsorption and desorption of the sample gases.  This material.




provides good separation based on the vapor pressure of gases.and allows




for complete resolution of the peaks of interest.  The column was operated




in an isothermal mode  (85°C) and regenerated for 30 minutes at 230°C




(Intersociety Committee, 1972).

-------
                                                                        50
     An analysis, using a 200 cm  sample injection and a silica gel




adsorption column, of the cylinder air used during the series of




experiments indicated the presence of methane, small amounts of ethane




and propane and trace levels of acetylene, n-butane, isobutane and




isopentane.  None of these gases interfere with the analysis or react




with ozone or sulfur dioxide under the conditions of the experiment.




     Calibration of the gas chromatograph was achieved by injecting




specified microliter quantities of pure olefin in the gas phase into a




51 I "Tedlar" bag along with carrier air.  Calibration samples from this




bag were periodically injected into the chromatographic column correcting




for wall loss in the bag.




     Operation of the column in the isothermal mode along with the semi-




continuous calibration allowed for the quantification of the gaseous




components by the measurement of peak heights which were recorded on a




Honeywell Incorporated Electronik - 193 chart recorder.






     6.  Infrared Spectroscopic Product Analysis




         A Fourier Transform Infrared Spectrometer was used to measure




the product inventory in the reactor at a terminal point in some of the




experimental runs.  This system consisted of two detectors operated at




liquid nitrogen temperature, a cell, a helium-neon laser for path align-




ment, a Nerst glower as a source of the infrared radiation, a Digilab




Incorporated, Model 496 scanning Michelson interferometer, a Data General




Nova 1200 dedicated computer for processing interferpgrams and spectra,




a teletype for user control and a Digilab Digital Plotter for plotting




the spectra.  One of the detectors, indium antimonide, was used to scan




the frequency region 2000 cm   to 3000 cm   and the other, mercury-




cadmium-telluride, to scan the region 700 cm   to 1200 cm  .  The glass

-------
                                                                        51





cell was 2.5m long and with the mirrored surfaces on the ends, path




lengths of 140 and 160 m were used for the measurements.   The operation




of a system of this type has been described in a recent article by Hanst,




et al., (1973).




     In operation the cell was evacuated and then allowed to fill to




ambient atmospheric pressure by opening a valve connecting the cell with




the reactor.  The infrared radiation from the Nernst glower entered the




interferometer where a scanning mirror projected a sine wave modulation




on each optical frequency passing through the cell to the detectors.




The superposition of the modulated frequencies reaching the detector




constitute the interferogram which is related to the spectrum through the




Fourier transform.  A reference spectrum was taken before each experimental




run which consisted of the diluent gas humidified to match the water con-




tent of the run to follow.  The digitized reference spectrum was stored




in the computer memory and was accessed after the sample was collected




in the cell.  The ratio of the sample spectrum to the reference was




computed and plotted.  The Lambert-Beer law was used to compute the con-




centrations from the equation
                  c ' TF-                                      IV-3





where I  =  the intensity of the radiation at the particular wave number



            for zero absorption



      I  =  the absorption diminished intensity



      k  =  the absorption coefficient



      £.  =  the path length used and



      C  =  the concentration to be determined.

-------
                                                                        52





     The absorption coefficients used in this study were obtained from




reference spectra of an extensive series of molecular species previously




studied at the National Environmental Research Center, Research Triangle




Park, North Carolina.






     7.  Sulfur Balance Determination




         Two techniques were used to determine the total sulfur content




of the sulfate aerosol formed in the olefin-ozone-sulfur dioxide reaction.




The first of these involved the use of Dupont Instruments Model 820 Liquid




Chromatograph and a Technicon Corporation Auto Analyzer to determine the




concentration of water soluble sulfates.  This method is the subject of




a technical report in preparation by Tejada (1974).




     The aerosol was collected on Millipore Corporation, Fluoropore FGLP




(0.2 urn pore size) 37 mm diameter filters.  The filter was extracted with




a 60/40 isopropyl alcohol (IPA)/water solution and the extract pumped




through a cation exchange resin (Dowex SOW - X2)  to remove interferences.




The extract is then pumped through a column of solid barium chloranilate




where barium sulfate precipitates out.  An equivalent amount of reddish




colored acid chloranilate ion is released and measured colorimetrically.




This method has a sensitivity better than 0.5 yg S0~/ml in 60% IPA.




     Calibration of this method was accomplished by preparing solutions




of sulfuric acid in IPA/water to span the concentration range of the




samples being analyzed.




     The second method made use of X-ray fluoresence techniques to




measure the total sulfur collected on the filters.  A Lawrence Berkeley




Laboratory X-ray Fluoresence Spectrometer which incorporated a dedicated




Texas Instruments computer Model T1960A and used three secondary targets,




i.e., Ti, Mo, and Tb to optimize the sensitivity over the periodic

-------
                                                                        53




table was used for this analysis.  A Northern Scientific Corporation



1024 Channel ADC together with the computer served as a pulse height



analyzer conducting an energy sort of the X-ray pulses.  The spectrometer



has been described in reports by Landis et al. (1970) and by Goulding



et al. (1972).  The composition of the samples was determined by irradia-



ting them for 400 seconds with monochromatic X-rays and observing the



characteristic K and L X-rays.  The secondary fluorescer, in this case,



Ti, was used to produce the near monochromatic X-ray source.  The fluoro-



pore filter was selected for this analysis because of its low mass and



very low level of impurities  (Dzubay and Stevens, 1973) in addition to



its high efficiency for collecting small particles.  A standard Millipore



Corporation Field Monitor was used as the filter holder with the addition



of one extra retainer incorporated as a spacer in an attempt to insure a



uniform deposition distribution on the filter surface.



     The analyzer was calibrated using evaporated foils of a number of



elements and compounds.  The smooth calibration curve drawn through these



points was used to determine elements for which foils were not available.



Individual spectra of up to forty elements are stored in the computer



along with a spectrum from a blank filter.  These spectra are compared



with the sample spectra using a stripping procedure in order to determine



the concentration of each element.





     8.  Light Scattering Measurements



         A Meteorology Research, Incorporated, Model 1500 Integrating



Nephelometer was used to measure the light scattering coefficient (b    )
                                                                    SCcl£


as a function of time during the experimental runs.  The design, operation



and calibration of this instrument have been described by Charlson et al.



(1969).

-------
                                                                        54


     The sampled gas is drawn into a cylindrical optical assembly with a

flash lamp mounted behind an opal glass screen on one side.  A photo-

multipler tube is mounted in one end of the assembly to detect the

scattered light from the flash illuminated volume.  A light trap and a

calibration mechanism are mounted on the opposite end.  A reference photo-

tube is located in the illuminated volume.  This volume is defined by

means of the physical arrangement which is such that the photomultiplier

tube is not visible from the flash lamp position.  Light received at the

photomultiplier tube is due, therefore, to scatter from particulate

material in the illuminated volume.  The reference phototube senses the

light directly from the .flash lamp and automatically eliminated the

effect of fluctuations in flash lamp output.  Ultraviolet cutoff filters,
       o
(-4100 A) are used with both phototubes to relate the instruments response

more nearly to the visible spectrum and to eliminate spurious radiation

in the ultraviolet region.

     The effective wavelength of the nephelometer is 500 run for either

aerosols or particle-free gases.  The human eye is most sensitive, however,

to radiation at 550 nm.  Horvath and Noll  (1968) have shown theoretically

and Alquist and Charlson  (1969) have shown experimentally that the

measured scattering can be related to light scattered at the other wave-

length as b^ = 0.84 b^.

     The normal sampling rate of this instrument is approximately 140 1pm

but because of the limited reactor volume the sampling rate was reduced

to approximately 300 cm /min.  The nephelometer output is independent of

flow rate and a range of time constants and flash lamp rates are select-

able.  For these studies a flash lamp rate of one per second and a ten

second time constant were used.

-------
                                                                        55




     Primary calibration of this instrument was achieved by sampling



Freon-12 and spanning the instrument on the tabulated light scattering


                     -4 -1
coefficient (3.6 x 10  m  ) which was derived from the Rayleigh theory



based on its measured refractive index (Charlson et al., 1969).  A



secondary calibration method is built into the instrument and consists



of a shutter arrangement on the end of the optical assembly opposite



the phototube.  A white surface behind the shutter is illuminated via



a fiber optic light pipe and the light reflected has been optically cali-



brated against gases of known scattering coefficients.



     There are two secondary meter scales with this instrument which do



not indicate the results of any measurement but rather relationships



that have extensive empirical support and some theoretical justification.



The visual range scale is based on the limiting contrast for daytime



visual determination of 0.02.  The resulting relationship is




                  L (m) * ill  = lil .                           IV-4

                   V      b550   b500



The other scale, i.e., mass concentration, relates the mass of material



in a box one meter in cross section and of length L  as
                                                   v


                                     2
                  L  x Mass =1.8 g/m  .                          IV-5




(Charlson et al., 1969).  The instrument panel meter displays the mass



concentration using the relationship between that parameter and the



light scattering coefficient which is indicated on the primary scale as




            mass (yg/m3) = 3.8 x 10  b   ^(m" ).                 IV-6
                                      scat



Values for both of these secondary -scales were calculated from the b
                                                                    SCcL u


measurements made during the experimental runs and are reported .herein.





     9.  Condensation Nuclei Count Determination



         An Environmental One Corporation, Model Rich 100 Condensation

-------
                                                                        56





Nuclei Counter was used to monitor the number concentration of particles




> 0.0025 pm diameter.  The instrument operates on the principle of a




cloud chamber in which submicroscopic droplets grow to micrometer size




as a result of condensation of water from a supersaturated environment.




The sample air is diverted, at an approximate frequency of once per




second, to a humidifier and then to a cloud chamber where the volume is




rapidly expanded.  The supersaturation achieved as a result of the




adiabatic expansion reaches in excess of 300%.  A light beam is focused




on a solid state light sensitive element across the expansion chamber.




The beam is attenuated by the growing aerosol the extent of which is




proportional to the number concentration of the aerosol particles.  At




the end of each cycle the chamber in pressurized and flushed out.




     The concentration range of the counter is from 300 to 10  particles/




cm  and has a linear response, as reported by the manufacturer, through




the 0-3x10  particles/cm  range scale and a non-linear response on the




0-10  particles/cm  scale.  The response time of the instrument is five




seconds at 50 cm /sec.




     Calibration by the manufacturer is based on the Pollack counter the




developmental history of which is given by Nolan (1972) .






    10.  Sub-microscopic Aerosol Analysis




         A Thermo-Systems, Incorporated, Model 3030 Electrical Aerosol




Size Analyzer was used to measure with time the size distribution of the




growing sulfate aerosol in the 0.003 to 1 pm size range.  The design,




operation and performance of a prototype of this instrument is described




in a paper by Liu et al.  (1973) and is based on the principle of




"diffusion charging-mobility analysis" described by Whitby and Clark  (1966)

-------
                                                                        57





     A flow schematic of the commercial model of the electrical mobility




analyzer is shown in Figure 4-7.  The sample aerosol is introduced into




a charging region where it is exposed to unipolar positive ions produced




in a corona discharge.  The charged aerosol then passes through the




mobility analyzer which consits of a precipitator, a high-efficiency




current collecting filter and an electrometer sensor.  The collection




rod, to which a variable negative voltage is applied, is immersed in a




sheath of clean air surrounded by an annular cylinder of the aerosol.




Small particles with high electrical mobility are drawn to the collection




rod under an applied voltage and larger particles are collected on the




current collecting filter.  The voltage on the rod is periodically




increased and particles of discrete size increments will be collected on




the rod with a concurrent decrease in the electrometer current.  Particles




with a sufficiently high electrical mobility are collected on the rod




for a discrete voltage while those of lower mobility are sensed by the




electrometers.  The cutoff mobility at which precipitation on the rod




occurs is given by:
                  ZP ' — 27UL -                              IV-7




where Q   =  the total air flow




      r   =  the inner electrode radius of the precipitator




      r   =  the outer electrode radius of the precipitator




      L  =  the length of the collecting rod and




      V  =  the applied voltage.




The mobility distribution of the aerosol is thus measured from which the




size distribution is inferred.  This inference is based on the electrical




mobility of the aerosol being a monotonically decreasing function of the




particle size.  Whitby and Clark (1966) report that for analyzers in

-------
                                                             58
                     AEROSOL
                    FLOWMETER
AEROSOL!
   IN   ^
       n
    SHEATH
     AIR IN
                AEROSOL
                 FLOW
                ADJUST
ABSOLUTE
 FILTER
               CHARGER
               SHEATH
                 AIR
                       ANALYZER
                        SHEATH
                          AIR
                                   J_
                                            /CHARGER
                                  TOTAL FLOW
                                    ADJUST
VACUUM
SOURCE
                                                  TOTAL FLOW
                                                  FLOWMETER
                                                ELECTROMETER
                                                ASSEMBLY WITH
                                               ABSOLUTE FILTER
       FIGURE 4-7. SCHEMATIC OF THE ELECTRICAL AEROSOL SIZE
                 ANALYZER.  [TAKEN FROM THERMO-SYSTEMS
                 INCORPORATED, 1974].

-------
                                                                        59
which the conditions approximate diffusion charging this relationship



holds for particles up to approximately 1 ym in diameter.



     Figure 4-8, adapted from the work of Bademosi (1971), shows the



relationship between the electrical mobility and the particle size for



diffusion charging for the conditions used in the design of this instru-



ment, i.e.,Nt=lx!0  (ions/cc)(sec).  N  is the positive ion con-
             o                              o


centration in the charging region and t is the residence time in that



region.  Up to approximately 0.02 ym, the mean charge on the particles



is one unit and, therefore, for this size and smaller particles, the



method provides an accurate sizing technique as the mobility Z  is
                                                              P


related to the particle radius by the equation,




                  Z  = 300 -^-      (cm2/volt-sec)               IV-8
                   p       6irrn



where e  =  the elementary charge unit is esu



      C  =  the Cunningham Correction factor



      n  =  the viscosity of the air and



      r  =  the particle radius.



     The mobility dependence decreases significantly with particle size



beginning at about 0.2 ym and is essentially zero for particles greater



than 1 ym in diameter.  The definition of the upper end of the distribu-



tion curve has, therefore, a higher degree of uncertainty than, for



example, the region between 0.02 and 0.2 ym.  Below 0.02 ym the distri-



bution of charges on the particles is either zero or one.  The determina-



tion of the fraction of particles charged and the.diffusion loss becomes



increasingly more tenuous as the particle size decreases.



     On this instrument the particle size interval from 0.0032 to 1.0 ym



has been divided into ten equal geometrical size intervals.  The reported

-------
                                                                  60
   0.1
  0.01
o
UJ
CM

 O
 0.001
 0.0001
    0.001
0.01
0.1
                                   Dp(^m)
     FIGURE 4-8. RELATIONSHIP BETWEEN ELECTRICAL MOBILITY AND
                PARTICLE SIZE FOR DIFFUSION CHARGING WHERE
                N0t = 1 x 10? (ions/cc)(sec).  [ADAPTED FROM THE WORK
                OFBADEMOSI, 1971].

-------
                                                                       61





diameter for each interval is the geometric mean size.   Table 4-2 shows



the computational sheet used to calculate the standard and the normalized



size distributions of count, surface area and volume.   The calibration



constant AN/AI is calculated from the expression






                  77- = Q  n  e f f                                  IV-9
                  AI    s . p    c p




where Q  = the sample air flow
       S



      n  = the mean particle charge
       P



      e  = the elementary unit charge in coulombs and



      f  = the fraction of charged particles and




      f  = the number collected on the current collecting filter.




     An absolute calibration of this instrument has not been accomplished



to date although an effort to determine the finite resolution of the



analyzer accounting for the charge distribution on the particles is



under study at the Particle Technology Laboratory, University of Minnesota.





    11.  Optical Particle Analysis



         A Royco Instruments, Incorporated, Model 220 optical aerosol



counter was used to measure the size distribution of aerosol particles



from approximately 0.3 to 10 ym in diameter.  The basic principle of



this instrument is that the light from a tungsten lamp is projected into



a viewing volume through which a dilute aerosol is passed, ideally with



only a single particle in the viewing volume at a time.  The light



scattered at 90° from the focused beam is sensed by a photomultiplier



tube and converted to a voltage signal.  The amount of light scattered



is related not only to the size of the particle but also to its shape,



absorbtivity, refractive index and orientation.  For this study the shape



of the aerosol droplets are spherical and the absorbtivity and refractive

-------
                        Table 4-2.  Data Table for Number, Surface Area and Volume Distribution'
                             Collector   IxlO~12   AIxlO~12
                                   AN
  D       Dpi    AN/AI
 .0032
 .0056
 .0100
 .0178
 .0316
 .0562
 .100
 .178
 .316
 .562
1.000
         .0042   1.102x10
         .0075   3.596x10
         .0133   1.069x10
         .0237   3.062x10'
         .0422   9.866x10
         .0750   4.775x10
         .133    2.356x10
         .237    1.356x10
         .422    7.781x10-
         .750    4.461x10'
Voltage     Amp.
   20
Amp.
   62
  196
  593
 1220
 2183
 3515
 5387
 7152
 8642
 9647
AN AS AS
AlogD AN AlogD
ANx4 P ird ASx4 P
5.54(-5)
1.77(-4)
5.56(-4)
1.76(-3)
5.59(-3)
1.77(-2)
5.56(-2)
1.76(-1)
5.59(-l)
1.77
AV
AN
l/6rrd
3.88(-8)
2.2K-7)
1.23(-6)
6.97(-6)
3.93(-5)
2.2K-4)
1.23(-3)
6.97(-3)
3.93(-2)
AV
AlogD
AVx4 P









  Taken from Thermo Systems Incorporated, 1974
                                                                         N=
                                                                                            S=
                                                                                                                  V=

-------
                                                                        63




index are fairly constant so that this counter can classify by size




as well as count the droplets.




     The signal from the Royco was processed by a Nuclear Data, Incorpo-




rated, system consisting of a Particle Sizing Amplifier, a Model ND 560




Analog to Digital Converter and a ND 812 Computer.  An ND 4410 Control




Function Module, an ND Tape Cassette, a teletypewriter and a Hewlett




Packard 1208 B Display Scope were used to manipulate, display and record




the distribution data.




     Calibration of this instrument was accomplished by a number of




techniques.  To compensate for degradation of the lamp, the counter is




equipped with a light pulse generator (mechanical chopper) to simulate




the passage of particles.  The amplification is adjusted until the signal




reaches the calibration value.  A zero check is made by placing a membrane




filter over the sample inlet.  The prime calibration procedure consisted




of generating an aerosol of monodisperse latex spheres  (refractive index




1.6) and sampling the suspended material with the optical aerosol analyzer.




The size of the spheres used were 0.3, 0.357, 0.481, 0.794 and 1.011 pm




and their corresponding peak channels were recorded with the multi-channel




analyzer.  The calibration data for this system is shown in Figure 4-9.






C.   Experimental Procedures




     The reactor was flushed with diluent gas three times before each




run to sweep out unreacted and product species from previous experiments.




The bag was evacuated via a vacuum line to yield a maximum deflection on




the Dwyer magnehelic gauge, < - 0.5 inches of water.  The reactor was




then filled with the diluent gas humidified to the same degree as for the




experiment which was to follow.  The diluent gas was passed through a

-------
                                                                    64
   2.0
   1.0

   0.9

   0.8

   0.7


_ 0.6


1 0.5

a.
o
   0.4




   0.3
   0.2
   0.1
              20
40
60       80


 CHANNEL
100
120
      FIGURE 4-9. CALIBRATION DATA FOR OPTICAL AEROSOL ANALYZER.

-------
                                                                        65





charcoal column followed by a 142mm Millipore type AAWP filter (0.8 ym




mean pore size) prior to its introduction into the reactor to remove




contaminant aerosols and undesirable reactive gaseous species.  The




diluent gases used in this experimental series were compressed air made




from a mixture of liquid oxygen and liquid nitrogen and compressed




nitrogen (High Purity Grade).




     A complete set of background measurements was made with the total




instrument complement to confirm the essentially zero concentrations of




the reactive species and particulate material.  The condensation nuclei




counter typically indicated a background nuclei count of about 400 con-




densation nuclei per cubic centimeter and its reported range is 300-10




en/cm .  If during these background measurements, positive concentration




values (other than the en/cm  level cited) were recorded, then the




flushing procedure was continued until a zero value was observed.




     The propylene used for the reactor was a calibration mixture  (Linde)




of 945 ppm in nitrogen while that used for the calibration of the gas




chromatograph was CP grade  (>99.5%) and was introduced into the reactor




by being injected into the diluent gas stream from a gas tight syringe.




The sulfur dioxide was also in a calibration mixture, in this case from




Scott Research Laboratories, Incorporated and contained 89 ppm in nitrogen.




An Orec Model 03V1 OZONATOR was used to provide the desired ozone concen-




tration.  Zero grade air was passed through a charcoal column and a Milli-




pore type AAWP filter before being exposed to the mercury vapor lamp in




the ozonizer.  A Millipore type FALP (teflon) filter (lym mean pore size)




was used on the output of the ozone generator to remove any condensation




nuclei which might be photochemically generated by the ultraviolet lamp.




The carbon monoxide which was used for two of the runs was CP grade and




was injected into the diluent stream in the same manner as was the ethylene.

-------
                                                                        66





     The propylene, sulfur dioxide and the diluent gas were delivered




to the reactor at a constant rate and mixture ratio over a five minute




period.  The humidification was provided as previously mentioned, byt




passing a metered portion of the diluent gas through bubblers containing




distilled water and immersed in a Forma Scientific Model 2800 Constant




Temperature Bath and Circulator.  The ozone was fed into one of the two




diluent gas streams for a portion of the five minute period depending




on the concentration desired.  Ozone was not, however, added during the




last thirty seconds of the filling time thereby permitting the other




reactants and diluent gas introduced during that terminal period to mix




thoroughly with the ozone.  Electrically actuated solenoid values were




used to initiate and terminate the introduction of the gas phase com-




ponents into the reactor and this flow was controlled through Brooks




Instrument Division flowmeters and Parker-Hannefin needle valves.  The




reactor was filled for each experiment to a positive pressure indication




on the magnehelic gauge, of 0.1" of water above ambient.




     The total gas flow into the bag was about 90 1pm which was considered




adequate to insure uniform mixing of the reactants.  The reactor was not




mechanically stirred during the experiments as the thermal gradients which




existed in and about the reactor were found to be adequate in maintaining




homogeneity among the gaseous species.  Wilson et al.  (1971) studied




the effect of stirring in a series of chamber experiments and noted a




decrease in the amount of aerosol with increased stirring rate.  If the




stirring was sufficiently fast, the observation of light scattering




aerosol could be completely eliminated.  The composition of the reactants




also influenced the effect noted with increased stirring.  Cox and Penkett




(1972) observed a similar result in that when forced circulation was used




during their experiments,  the effective rate of gaseous diffusion to

-------
                                                                        67





their reactor walls was increased by a factor of about five.   The effect




of stirring was not found to be a critical parameter on aerosol growth




in systems with initial low nuclei counts during studies conducted by




Kocmond et al. (1973).   Aerosol decay by coagulation in the dark, however,




was noted to be significantly increased by stirring.




     Sampling was initiated immediately after the reaction vessel was




full and was generally intermittent.  The electronics portion of the




instrumentation was kept operating while the pneumatic system was off in




some cases and cycled to sample room air in others.  These measures were




necessitated because of the limited total volume in the reactor and the




large sampling volume demands of the instruments.  The experiments were




designed to last until one of the principal reactants, i.e.,  the olefin




or ozone had been consumed by > 90%.




     A Westronics, Incorporated, Model MllE Multipoint Recorder was used




to record the ambient temperature and the output of the ozone and sulfur




dioxide analyzers, the condensation nuclei counter, the integrating




nephelometer, and the dew point hygrometer.  The output of the flame




ionization detector in the gas chromatograph was recorded on the Honeywell




Electronik 193 Strip Chart Recorder.  The Royco optical particle data




was stored in the memory of the Nuclear Data ND 812 computer system and




printed out on a teletypewriter.  The data output from the electrical




mobility analyzer was manually recorded.  The analog output from the




various instruments was digitized and recorded along with the other




digital data on computer cards for subsequent data manipulation.




     A series of measurements was performed to determine the loss of the




reactants in the containment vessel.  Propylene, ozone and sulfur dioxide,




were introduced into the reactor individually along with the diluent gas




and their concentration monitored over an extended period of time.  The

-------
                                                                        68





loss of the reactant species could be a function of molecular diffusion




to and reactions with the reactor walls and/or reactions with species




not purged from the bag after previous runs or introduced with the




species under study.




     The observed loss for these three species was apparently the result




of a first order reaction, with the species irreversibly lost to the walls




of the reactor.  Figure 4-10 shows the linear relationship resulting




from a plot of the logarithm of concentration versus time for the three




species for typical decay measurement runs.  The slopes of the lines




are the rate constants for the wall loss reaction as predicted from the




following equations :
                        = -k[A]                                     IV-10





where A =  some hypothetical species and




      k =  the velocity constant.




The equation simply states that the rate of the wall loss reaction is




directly proportional to the concentration of the reactant present.  By




integration the equation becomes :





                  In [A] = -kt + constant                            IV-11




from which the value of k can be obtained.  Reaction rates are also




often reported in terms of the half life of the species in the system.




For the first order reaction the half life is related to the rate constant




by




                  t>» = ^p..                                        iv-12





     The rate constants and half lifes measured for the three species in




the reaction containment vessel were as follows:

-------
                                                                   69
 10
  9
  8
  7
  6

  5
_   2
ex

O
I-
oc.

y 0.9
o 0.8
   0.7
   0.6

   0.5

   0.4

   0.3
0.2
0.1
      z        r
I      -
                                             kw-2.35x10-4 min-1  —
                                    kw = 9.7 x 10-4 min-1
                                                      S02
                                            kw = 1.52 x 10-4 min-1
                                    3

                                 t (hrs)
    FIGURE 4-10.  VARIATION OF REACTANT CONCENTRATION WITH
                TIME DUE TO WALL LOSSES.

-------
                                                                         70
                        Propylene           Ozone        Sulfur Dioxide



                             -4   -1          -4   -1           -4   -1
Rate Constant (k)     2.35x10  min      9.7x10  min      1.52x10  min



Half Life (t>5)          49.1 hrs.         11.9 hrs.        76.0 hrs.
     The data from several experiments were corrected for these different



rates of loss due to wall deposition although the net result was insigni-



ficant and is not included in this report.

-------
                                CHAPTER V




                   Experimental Results and Discussion






     The experimental results obtained during this study generally relate




to two separate yet interrelated scientific fields.  The first few sec-




tions of this chapter deal with the analysis of gas phase reactants and




products while the later sections are concerned with the gas to particle




conversion process and the growth of the newly-formed aerosol.  The inter-




relationships are made clear throughout the treatment to follow.  Extensive




use-has been made of tables and figures in supplementing the discussion




of the experimental results.




     A tabulation of the gas phase products observed by the Fourier




Transform Infrared Spectrometer for eleven runs is included in the




Appendix as is the listing of the observed and predicted values of the




reactant concentrations as a function of time for the basic series of




nineteen gas phase experiments.  The droplet number, surface area and




volume concentration distributions at several different times are also




tabulated in the Appendix for a series of six aerosol experiments.






A.   Reactant Consumption Data and Reaction Stoichiometry





     Several series of olefin-ozone-sulfur dioxide experiments were con-




ducted to obtain information relative to the Stoichiometry of the olefin-




ozone reaction, the influence of sulfur dioxide on the reaction, the




role of molecular oxygen and the effect of varying the relative humidity




on the system.  A summary of the initial conditions for one series of




nineteen experiments is shown in Table 5-1.  Sixteen of these runs

-------
Table 5-1.   Summary of Experimental Conditions and
            Initial Reactant Concentrations
             Propylene
ton
91
92
93
94
21
31
25
26
27
86
87
88
89
90
T(°C)
24
24
24
25
27
28
24
25
25
24
26
23
26
25
RH (%)
26
39
23
37
19
19
22
22
22
23
20
25
21
21
Olefin
(ppm)
2.580
3.012
2.753
2.930
8.125
8.555
1.878
3.547
3.461
.922
1.112
3.281
3.318
2.739
Ozone
(ppm)
1.660
1.651
2.573
2.683
2.882
2.714
2.377
1.080
.983
1.969
2.056
.274
.295
1.970
Sulfur Dioxide
(ppm)
0
0
0
0
0
.5250
.5827
.6206
.1939
.2010
.6507
.2075
.6373
.5803
Olefin/O:
1.554
1.824
1.070
1.092
2.819
3.153
.790
3.284
3.522
.468
.541
12.00
11.26
1.390

-------
 Table 5-1.  Continued
Propylene
:un
98
95
T(°C)
25
25
RH (%)
23
20
Olefin
(ppm)
2.825
3.025
Ozone
(ppm)
2.025
1.813
Sulfur Dioxide
(ppm)
.2039
.5582
Olef in/O
1.395
1.669
Ethylene
97
28
29
26
27
28
22
22
20
3.112
11.85
12.50
2.102
3.807
3.617
.5884
0
.5294
1.480
3.113
3.456
                                                                       U)

-------
                                                                        74
included propylene as the olefin species while the other three involved




ethylene.  The observed temperature with one standard deviation for the




propylene runs was 25.1 ± 1.9°C while the measured relative humidity




was 21.8 ± 2.0% for low water vapor concentration runs and 38.0 ± 1.9%




for the higher level runs.  The initial concentration of propylene ranged




from about 0.9 to 8.6 ppm for this series and for ozone the range was




from about 0.3 to 2.8 ppm.  The initial sulfur dioxide concentration was




in general, at one of three levels, 0.0, 0.2, or 0.6 ppm.  The ratio of




the initial concentrations of the two principal reactants typically




ranged from about 0.5 to 3.5 with this initial olefin-ozone concentration




ratio being about 12 for two of the runs.




     The amount of the reactants consumed in this series of experiments




is summarized in Table 5-2.  Also shown is the percentage of one of the




principal reactants which was consumed at the point at which the time




averaged stoichiometric change was measured.  This percentage generally




referred to the quantity of ozone reacted as the olefin was usually in




excess.  Ozone was in excess, however, for runs 86 and 87 and the per-




centage refers to the amount of the propylene consumed.  The amount of




sulfur dioxide reacted ranged from about 0.02 to 0.2 ppm in this series.




Only acetyaldehyde, of the stable aldehyde products formed in the olefin-




ozone reaction, could be measured with the gas chromatographic arrange-




ment used.  This species, which is formed in the reaction of ozone with




propylene, was followed in time for several of the experiments and the




concentration value attained near the end of the run is included in




Table 5-2.  Due to the longer time required for the acetyaldehyde to




elute from the gas chromatographic column and because of the lower




sensitivity for its measurement, it was not followed on every experiment.

-------
Table 5-2.  Summary of Stoichiometric Data and

            Acetyaldehyde Production
     Propylene
lun
91
92
93
94
21
31
25
26
27
86
87
88
89
90
AOL
(ppm)
1.811
1.957
1.852
1.975
3.917
3.934
1.630
1.498
1.468
.817
.954
.501
.539
1.999
A03
(ppm)
1.463
1.499
2.315
2.415
2.610
2.468
1.702
.978
.889
1.050
1.109
.249
.270
1.757
AS02
(ppm)
0
0
0
0
0
.174
.180
.141
.058
.078
.151
.022
.052
.205
AOL/AO
1.238
1.306
.800
.818
1.501
1.594
.958
1.532
1.651
.778
.860
2.011
1.996
1.138
Reacted
.88
.91
.90
.90
.90
.91
.72
.92
.90
.88*
.86*
.91
.92
.89
CH CHO
(ppm)
n.m.
n.m.
n.m.
n.m.
1.65
1.80
.73
.45
.58
n.m.
.48
n.m.
n .m.
.82
                                                                                   -J
                                                                                   en

-------
                                     Table  5-2.   Continued
                              Propylene
:un
98
95

97
28
29
AOL
(ppm)
2.084
1.721

1.265
4.557
4.964
A03
(ppm)
1.805
1.633
Ethylene
.980
3.039
2.876
ASO
(ppm)
.102
.165

.183
0
.299
AOL/AO
1.155
1.054

1.291
1.500
1.726
% 0
Reacted
.89
.90

.47
.80
.80
CH CHO
(ppm)
.84
.79




n.m. not measured




 *    % OL Reacted

-------
                                                                        77






Acetyaldehyde is not a product of the reaction of ozone with ethylene




and would not, therefore, be expected to be present in those runs.




     The variation of concentration with time for a typical propylene-




ozone-sulfur dioxide run is shown in Figure 5-1.  The near 1:1 stoichio-




metric consumption for the two principal reactants is readily apparent in




the near constant difference in their concentrations with time for this




experiment.  The consumption of the sulfur dioxide is markedly slower




than it is for the other two reactants.  The acetyaldehyde production is




also shown and as it does not appear to follow linearly the loss of the




propylene, the process leading to its appearance is more complex than




if it were formed in a single bimolecular reaction.




     The reaction involving propylene can be contrasted with a run in




which ethylene was the olefin reacted with the ozone.  The variation of




the concentration for the reactants in the ethylene experiment is shown




in Figure 5-2.  The initial concentrations for this run were about the




same as they were for the propylene experiment and although the same




trends in consumption are observed, the reaction proceeds much more slowly.




     The data listed in Tables 5-1 and 5-2 were used to plot the varia-




tion in the apparent stoichiometry of the olefin-ozone reaction with the




initial reactant concentration ratio (Figure 5-3).  There are a total of




thirteen data points plotted which represent ozone-propylene reactions




conducted in air.  The line drawn through the points appears to satisfy




these data and was fixed by eye.  Ten of these thirteen experiments were




conducted in the presence of sulfur dioxide.  The presence of the sulfur




dioxide had no apparent effect on the stoichiometry of the primary reaction




as all of the data closely follow a smooth curve.  This result indicates




that the oxidative consumption of the sulfur dioxide is due to a reactive




product of the propylene-ozone reaction rather than to a primary reactant

-------
10
            FIGURE 5-1.  VARIATION OF CONCENTRATION WITH TIME FOR
                       PROPYLENE-OZONE-SULFUR DIOXIDE REACTION.
03

-------
o
I—
<:
oc.
\-

LU
O

o
o
                                                                                                     VO
                  FIGURE 5-2. VARIATION OF CONCENTRATION WITH TIME FOR ETHYLENE-
                            OZONE-SULFUR DIOXIDE REACTION.

-------
  2.2
  2.0
  1.8
  1.6
  1.4
^•^
< 1.2
  1.0
  0.8
  0.6
CO
    -*L.
oPROPYLENE-OZONE                —
oETHYLENE-OZONE                 	
APROPYLENE-OZONE IN NITROGEN      	
• PROPYLENE-OZONE WITH CARBON MONOXIDE
         1
                      4    5
                 8
10    11   12
                               OL\
                               03/
   FIGURE 5-3. STOICHIOMETRY OF OLEFIN-OZONE REACTION VERSUS
            INITIAL REACTANT CONCENTRATION RATIO.
                                                                  03
                                                                  o

-------
                                                                        81




such as ozone.  This conclusion was fully substantiated through an




experiment in which ozone and sulfur dioxide were the only reactants.  The




consumption of those species in this experiment was slow enough so as




not to be distinguished from losses attributable to wall effects.




     The nature of the reactive product of the olefin-ozone reaction




which results in the consumption of the sulfur dioxide has been studied




by Cox and Penkett (1972).  Evidence that this product is a short-lived




intermediate of the primary reaction was derived from experiments in




which the sulfur dioxide was introduced at different times after the




initiation of the olefin-ozone reaction.  When the sulfur dioxide was




added after the ozone had sufficient time to react with the olefin, the




aerosol formation was barely perceptible, whereas when the sulfur dioxide




was initially present or added shortly after the run began, the aerosol




formation was significant.




     The data shown in Figure 5-3 indicate that the reaction is compli-




cated by secondary reactions.  The apparent stoichiometry of the reaction




is decidedly affected by the initial concentration ratio.  In general,




a system which initially has propylene in excess results in a consumption




ratio of the propylene to ozone greater than one while in a system




which initially has an ozone excess condition the value of this ratio




is less than one.  These results are in general agreement with those




reported by Wei and Cvetanovic (1963) for experiments conducted at much




higher reactant concentrations.  Several olefins including propylene




were studied by those investigators and the consumption ratio was found




to vary from 1.4-2.0 for the different olefins in the high olefin excess




system.  Bufalini and Altshuller (1965) also studied a series of olefins




but in concentrations much closer to values used in this study and




observed a similar trend.  Preliminary results from two current studies

-------
                                                                        82






in progress elsewhere are also in good qualitative agreement with those




reported herein.  One of these, directed by Williamson (1974) , makes use




of high initial reactant concentrations and is projected to include




propylene.  Initial results for the ethylene system indicate a consump-




tion ratio of less than one for the ozone rich system and equal to or




greater than one for the ethylene rich system.  The other investigation




is being conducted by Niki  (1974) and makes use of propylene concentra-




tions ranging from 2.0 to 5.8 ppm and ozone concentrations from about




1.4 to 8.3 ppm.  The stoichiometry which they have observed for these




propylene .studies ranges from 0.9:1 to 1.4:1.  For the study reported




herein, the consumption ratio varies over the range of experimental con-




ditions used from about 0.8 to 2.0 and it appears to approach this latter




value as a limit.  When a condition of initial ozone excess exists, these




data would support a postulate that either ozone is interacting with pro-




ducts of the primary olefin-ozone reaction or that propylene is being




produced.  Similarly for a system which is initially rich in propylene




these data could support a postulate that propylene interacts with a




reactive product of the initial reaction and/or that ozone is being pro-




duced as a result of a secondary reaction.  A reaction that could result




in the production of ozone is shown as reaction 13, Chapter II, as postu-




lated by Bufalini and Altshuller (1965) and involves an interaction of




the reactive intermediate with molecular oxygen.




     The three ethylene-ozone-sulfur dioxide experiments are also included




in Figure 5-3.  These data follow the same general trend as do the data




from the propylene system and the slightly increased scatter is believed




due to the method of olefin introduction into the reactor.  The propylene




was metered into the system at a constant rate for the entire reactor




filling period while the ethylene was injected via a septum into the

-------
                                                                        83




diluent gas stream.  It appears that the sulfur dioxide is oxidized by




a reactive product of the primary reaction in the ethylene system in




analogy to the oxidative process in the propylene system.




     The molecular oxygen concentration was sharply reduced for several




runs in an attempt to determine whether oxygen has a role in the reaction.




The results of four of these experiments in which nitrogen was used as




the diluent gas are recorded on Figure 5-3.  An obvious reduction in the




propylene to ozone consumption ratio occurred indicating that molecular




oxygen plays a significant role in modifying the consumption of one or




both of the primary reactants.  This result runs contrary to that reported




recently by Cox and Penkett (1972) and Stedman et al. (1973) .  Those




investigators concluded as a result of their studies that molecular




oxygen does not affect the stoichiometry of the primary olefin-ozone




reaction.




     Wei and Cvetanovic (1963) studied the reaction of high concentrations




of several olefins including propylene with ozone in the gas phase both




in the presence and in the absence of molecular oxygen and found a lower




consumption ratio in the absence of the oxygen.  Data collected in the




current study by Niki (1974) led him to the conclusion that oxygen has a




pronounced effect on the olefin-ozone system.




     The observation of this molecular oxygen effect has led to the




postulate that a secondary attack on the propylene is occurring in this




reaction.  As the consumption ratio changes in the presence or absence




of molecular oxygen, a secondary attack on the olefin by the reactive




intermediate can be eliminated as being responsible for the stoichiometric




change.  Other reactions which could account for the oxygen effect and




the deviation from 1:1 stoichiometry are those which involve an interaction




of the intermediate species and molecular oxygen, e.g., reactions 11-13

-------
                                                                        84





in Chapter II-B.  The products are the hydroxyl and the peroxyacetyl




radicals, the methyl radical, carbon dioxide and the hydroperoxy radical,




and ozone and acetylaldehyde for the three reactions respectively.  Each




of these reactions, therefore, results in the production of a species




capable of further interacting with the olefin in a chain propagating




reaction.  The three species formed also react with carbon monoxide




although the rate constant for the reaction of the hydroxyl radical with




carbon monoxide is approximately 10  times greater than the reaction




with the hydroperoxy radical and about 10   greater than the reaction of




carbon monoxide .with ozone.  The addition of an appropriate concentration




of carbon monoxide to serve as a scavenger in the system should, therefore,




provide a convenient method of testing for the formation and presence of




the hydroxyl radical.  As the hydroxyl radical would also react with the




olefin and sulfur dioxide in the system in addition to any aldehydes




formed and other minor products of the olefin reaction, a sufficient




carbon monoxide concentration was needed to affect the overall reaction




stoichiometry significantly.  In other words, it was desired to cause




the loss of the hydroxyl radical, if present, to be greater in its




reaction with carbon monoxide than its total loss in all of its other




possible reactions.  This can be expressed in functional form as
      k18 [-OH][CO] > [-OH] 
-------
                                                                        85.





     Two experiments in which 458 ppm of carbon monoxide was added to the




propylene-ozone-sulfur dioxide system were conducted and the stoichiometric




results are included in Figure 5-3.  These results substantiate the postu-




late relative to the formation of the hydroxyl radical in this reaction




and its subsequent secondary attack on the olefin species.




     One final result included graphically in Figure 5-3 is the lack of




an effect of a change in the relative humidity on the olefin-ozone




reaction over the range studied.  The propylene runs included one experi-




ment conducted in air and two in nitrogen for which the relative




humidity was elevated to about 38 percent.  Cox and Penkett (1972)




observed a water vapor concentration effect on the sulfur dioxide con-




sumption with a decrease in the sulfur dioxide oxidation rate occurring




with increasing relative humidity.  The data from two runs in the study




reported herein showed a slight but similar trend.  The consumption of




sulfur dioxide in run 60, which was conducted in an atmosphere at 20




percent relative humidity, was 0.071 ppm/ppm of ozone consumed while




for run 65 conducted at a relative humidity of 36 percent, this ratio




was 0.066 ppm/ppm of ozone consumed.  The products of these two experi-




ments as measured with the infrared spectrometer do not reflect an effect




due to the variation in the relative humidity.




     The acetyaldehyde concentration produced in several of the experiments




was plotted against the amount of propylene reacted (Figure 5-4).  Each




of the data points represents a separate run.  The measurements of both




the olefin and the aldehyde were made at the same time, late in the




experiments.  The line drawn was determined by fitting the data to a




linear function by the method of least squares.  The slope of the regres-




sion line forced through the origin is 0.44 which also indicates the




mean yield of acetyaldehyde in these reactions.  The data included are

-------
                                                               86
   1.8




   1.6




   1.4




   1.2
o 1.0
01
o
ce


^0.8
   0.6




   0.4




   0.2
                              AOL(ppm)


    FIGURE 5-4. CONCENTRATION OF ACETYALDEHYDE PRODUCED VERSUS

              PROPYLENE REACTED FOR NINE PROPYLENE-OZONE

              EXPERIMENTS.

-------
                                                                        87





from experiments conducted in the presence of sulfur dioxide, the oxida-




tive step of which purportedly produces acetyaldehyde.   The contribution




from that step amounted to approximately thirteen percent according to




data from runs 30 and 21 reported in the product analysis table in the




Appendix.  Making an adjustment for the contribution from the sulfur




dioxide oxidation reaction the modified production of 0.38 ppm of acety-




aldehyde per ppm of propylene reacted is in close agreement with the




results reported by Scott et al. (1957) and Wei and Cvetanovic (1963).




     A relationship between the consumption of the sulfur dioxide and




one or both of the primary reactants was also sought. .  Using the data




from runs in which the initial olefin to ozone concentration ratio was




equal to or greater than one, the plots shown in Figures 5-5 and 5-6 were




developed.  By including the initial sulfur dioxide concentration in the




plot, it can be seen from Figure 5-5 that an apparent linear relationship




can be established between the formation of the sulfur  species and the




olefin reacted.  These lines were constructed by fitting the data by the




method of least squares and forcing the regression line through the origin.




The sulfur dioxide concentrations employed in this analysis with one stand-




ard deviation were 0.605 ± 0.028 and 0.202 ± 0.007 ppm.  The data plotted




in Figure 5-6 demonstrate the non-linearity of the relationship between




the amounts of ozone consumed and the sulfur dioxide produced.  These




data further enforce the postulate concerning the identity of the oxidiz-




ing species for the sulfur dioxide as being a reactive  product of the




propylene reaction.




     The data plotted in Figure 5-3 and previously discussed show that




the stoichiometry of the propylene-ozone reaction for these experiments




was generally other than 1:1.  As the reactant consumption ratio dropped




when the partial pressure of the molecular oxygen was reduced, it was

-------
                                                                   88
  0.24
   0.20
o
<
   0.12
  0.08
  0.04
             0.4
0.6
                                                       =0.605 ±0.028 ppm
                                                      =0.202 ±0.007 ppm
0.8
1.2
1.6
2.0
2.4
                             AOL(ppm)
FIGURE 5-5. SULFUR DIOXIDE CONSUMPTION VERSUS PROPYLENE REACTED FOR
           DIFFERENT INITIAL SULFUR DIOXIDE CONCENTRATIONS.

-------
                                                               89
                                           [S0£]o= 0.605 ±0.028 ppn
                                                = 0.202+0.007 ppm
          0.4
1.2       1.6

 A 03 (ppm)
2.4
FIGURE 5-6.  SULFUR DIOXIDE CONSUMPTION VERSUS OZONE REACTED FOR
           DIFFERENT INITIAL SULFUR DIOXIDE CONCENTRATIONS.

-------
                                                                        90
speculated that ozone and/or the hydroxyl radical were being produced

in the reaction of a product of the ozone-propylene reaction with oxygen.

The reduction of the reactant consumption ratio when the carbon monoxide

was added enforced the speculation relative to the formation of the

hydroxyl radical which was subsequently scavenged by the carbon monoxide.

These measurements do not, however, demonostrate whether the change in

stoichiometry with the changing propylene/ozone initial condition is due

to the production of propylene, enhanced consumption of ozone or to an

interaction of these species with reaction products.  To investigate

further the reaction mechanism with regard to the stoichiometry, the

                              R             R
variation of the functions  ,  1 ,  ,  and ,  -i rn i  with time was noted.


If the only reactions involving these species were with each other then

R   = R   and the plots of the rate over reactant product versus time
 OL    03

would yield a straight line of zero slope, and the value of the function

would be equal to the second order rate constant.  If, however, a plot

of one of the functions initially displays a negative slope, then that

species is being produced in the reaction according to
                               .                                   _
                [OL] [03]     l    2  [OL] [03]


where [A] [B]  =  product of the concentration of reactants A and B lead-
                 ing to the ozone production and

          k   =  rate constant of the hypothetical bimolecular reaction.

Conversely, if a plot of one of the functions were to display a positive

slope initially then that species is being consumed by another process as
                                v                                 v_3
               [OL][03]  - *1   *2 [03]


     The variation of these two functions with time for run 98 is shown

-------
                                                                         91

in Figure 5-7.  The resultant curves are typical for the series of
experiments conducted in that the R   function, i.e., the function which
                                   OL
includes R  , increases with time to some maximum and then decreases to
          OL
a value that is approximately equal to that of the R   function.  The
function which includes R   also increases with time but at a much
slower rate and without the peak typical of the other curve.  The values
of the functions at the reported zero time are approximately equal and
represent the rate constant for the propylene ozone-reaction.
     Some mention is warranted, at this point, on how the derivatives of
the reactant concentrations with respect to time were estimated.  The
first five data points were fitted to a second order polynomial from
which the value of the third point and its derivative were estimated.
The first observed data point was then dropped and the sixth added to
establish a new set of five data points to be fitted.  This operation
was continued through the total set of data collected and was performed
using a computer program of the Statistical Analysis System  (SAS) (Service,
1972) .  Values required which were beyond the range of the observed data
were estimated by extrapolation.  This latter feature was implemented
repeatedly during this study to obtain the concentration and rate at
the zero time.
     One additional factor affecting the absolute accuracy of the initial
R-T and R   values is the fact that the primary reaction actually commenced
 ULi      (Jo
at a time prior to the zero time reported on these plots.  The reported
zero time, as previously indicated, was taken as the time at which the
reaction vessel became full.
     Data of the type reported in Figure 5-7 provide a measure of the
second order rate constant for the propylene-ozone reaction and yield
semi-quantitative information on the reaction stoichiometry.  The reaction

-------
                                                           92
FIGURE 5-7. VARIATION OF THE APPARENT RATE CONSTANT WITH TIME
          DUE TO SECONDARY REACTIONS.

-------
                                                                        93




which involves R   is observed in the figure to increase with time suggest-



ing that ozone is being consumed by a product of the principal reaction.



The postulated generation of ozone by a secondary reaction, e.g., the



reaction of the unstable intermediate with molecular oxygen, is either



nonexistent or is completely masked by the enhanced ozone consumption.



The product or products responsible for the additional loss of ozone are



unidentified at this time but are known not to include the products



measured and reported later in this paper.  The rate of reaction of



ozone with the aldehydes and carbon monoxide is too slow to be of signi-



ficance while the yield of the other products is too low to influence



the disappearance of the ozone.  The slowness of the acetyaldehyde-ozone



reaction relative to that of propylene with ozone was confirmed by con-



ducting one experiment with acetyaldehyde and ozone as the sole reactants.



The rate of loss of the ozone in the system with the aldehyde was such



that over the lifetime of a typical experiment in this series it could



not be distinguished from the first order loss of the ozone to the walls



of the reactor.  An equation of the form of V-3 may itself be symbolic



with the second term representing several reactions.



     The other curve in Figure 5-7 indicates that propylene is also



being subjected to a secondary reaction.  The postulate concerning the



formation of the hydroxyl radical in a reaction involving the interme-



diate species and molecular oxygen can be used to explain that curve.



As the hydroxyl radical is formed and reacts with propylene the function



plotted rises to its maximum value.  Formaldehyde and acetyaldehyde are



produced in the propylene-ozone reaction and eventually compete for the

                                        _j^


hydroxyl radical causing the function f  -, r  -.  to diminish to some
                                      IOLJ LO3J



equilibrium value.

-------
                                                                        94
     The value of k  based on the propylene rate data for run 98 was



1.46 x 10   ppm   min   and based on the ozone rate data was 1.39 x 10



ppm   min~ .  The mean and one standard deviation for k  computed from



the thirteen runs conducted in air and based on the rate of change in the



ozone concentration was 1.40 ± .38 (x 10   ppm   min  ).  For these runs,

               -RQL                                            -RQ3

the function ,  •,»  ,  was generally equal to or greater than ,  ,.-  -.





while for the runs conducted in the reduced oxygen atmosphere this



relationship was reversed.  This trend indicates that the relative rate



of olefin involvement in secondary reactions had diminished with the



oxygen depletion.



     The initial olefin/ozone concentration ratio was 12 for run 88 and,



therefore, the ozone would be expected to behave in a pseudo first order



rate.  The initial concentration of the ozone was 0.27 ppm and the R



function was essentially constant displaying only a slight increase over



the period of the experiment.  The first order rate constant for run 88


                            -2    -1    -1
based on ozone was 1.34 x 10   ppm   min
B.   Product Analysis



     An inventory of the products formed in the gas phase reactions of



propylene with ozone, propylene-ozone-sulfur dioxide, ethylene with



ozone and ethylene-ozone-sulfur dioxide was conducted by studying the



infrared absorption spectrum of the gaseous components in the reactor at



a time when approximately 90% of the ozone had been reacted (80% for



the ethylene runs).  The volume of the glass sample cell was about 60



liters and long infrared absorption path lengths were achieved by



'folding1 the beam with mirrors placed near the ends of the cell.  The



operation of the Fourier Transform Infrared Spectrometer is described in



Chapter IV-B,6.  Before each experimental run, a reference or background

-------
                                                                        95




spectrum of the properly humidified diluent gas was obtained and the




sample spectrum from the run was subsequently ratioed against this




background measurement.  The ratio recording technique eliminates many




of the interferences which may otherwise obscure the regions of interest




and increases the detectability limits for the system.




     Ratio recordings for three spectral regions, 700-1200 cm" , 2000-2300




cm   and 2600-3000 cm  , are shown in Figure 5-8 for runs 21 and 31.  The




infrared absorption path length for both of these measurements was 160 m




through air at one atmosphere which contained the products of the




propylene-ozone and the propylene-ozone-sulfur dioxide runs respectively.




Spectral regions which include strong water and carbon dioxide absorption




bands were excluded from these measurements because of the significant




interference which they cause on quantitative product analysis.  The con-




centrations of the identified product compounds calculated from the




spectrum which appears at the top of Figure 5-8 are 0.94 ppm carbon-




monoxide, 3.44 ppm formaldehyde, 1.59 ppm acetyaldehyde and a trace




amount of formic acid.  The regions of primary propylene, formaldehyde




and acetyaldehyde absorption are marked with a range bar and the prominent




absorption peak(s) used in the calculation are identified.  The lower




spectrum in Figure 5-8 is for the experiment (run 21) conducted without




any sulfur dioxide.  The product concentrations calculated from this




spectrum are 0.93 ppm for carbon monoxide, 3.21 ppm for formaldehyde,




1.49 ppm for acetyaldehyde, 0.2 ppm for ketene and again a trace of




formic acid was observed.




     A ratio plot with scale expansion of the region from 2000-2300 cm




for the two experimental runs is shown in Figure 5-9.  The scale expansion




technique tends to make the absorption lines more obvious without increas-




ing the background noise.  The upper spectrum shows the carbon monoxide

-------
                                      2200      2300 2600
                                                      2700
                                                                     2VD      9000
FIGURE 5-8. INFRARED ABSORPTION SPECTRA OF REACTOR CONTENTS.

-------
                                                                97
                                           RATIO WITH     rj
                                           SCALE EXPANSION'*
                              '••I-    I
                                KETENE
JRUN 21
-C3H6 + 03
                                           RAT 10 WITH
                                           SCALE EXPANSION;
      2000cm"1         2100              2200             2300
FIGURE 5-9. INFRARED SPECTRA IN REGION OF CARBON MONOXIDE ABSORPTION.

-------
                                                                        98





continuum for the run conducted in the presence of sulfur dioxide while




the lower plot shows a ketene spectrum superimposed on the carbon monoxide




continuum for the run which did not include sulfur dioxide.  The ketene




is identified by the double indentation on the carbon monoxide continuum




with the right lobe characteristically being slightly deeper and broader.




The appearance of ketene as a product of the propylene-ozone reaction




has been previously reported (Scott et al., 1957) and has been postulated




to result from the disassociation of the acetyl form of the zwitterion




species (Leighton 1961)  as shown by reaction 5 in Chapter II-B.  The




dramatic disappearance of the ketene spectrum in the presence of sulfur




dioxide indicates that the reaction rate of the oxidizing species with




the sulfur dioxide is much faster than is the rate of the decomposition




reaction.




     Ratio plots, with scale expansion of the spectral region which




includes ketene, for two additional propylene-ozone experiments conducted




in air at one atmosphere are shown in Figure 5-10.  The relative humidity




in the reactor was approximately 19% for run 78 and 33% for run 79.  The




presence of ketene in both of these runs and at approximately the same




concentration is readily apparent.  Nitrogen was used as the diluent gas




for the product spectra shown in Figure 5-11.  Run 81 was conducted in an




atmosphere at 36% relative humidity while the relative humidity for run




82 was about 20%.  The formation of ketene in those runs in an atmosphere




with a sharply depleted molecular oxygen content is also apparent.  Ratio




plots of two more runs conducted in the presence of sulfur dioxide are




shown in Figure 5-12 for which the relative humidity was 20 and 36%




respectively.  There is no ketene observable in these product absorption




spectra nor in that of run 31 previously discussed.




     Ratio recordings of the three spectral regions, 700-1200 cm  ,

-------
                                          SCALE EXPANS [ON
                                          RATIO WITH
                                          SCALE EXPANSION
                                                               99
     2000cm"1        2100               2200             2300
FIGURE 5-10. INFRARED SPECTRA IN REGION OF CARBON MONOXIDE ABSORPTION,

-------
                                                               100
                                           RAT 10 WITH
                                           SCALt EXPANSION
                                          RATIO WITH
                                          SCALE EXPANSION
      2000cm'1         2100              2200             2300
FIGURE 5-11. INFRARED SPECTRA IN REGION OF CARBON MONOXIDE ABSORPTION

-------
                                                              101
                                         I;: {RUN 60
                                          RATIO WITH
                                          SCALE EXPANSION i
                                             RUN 65
                                             C3H6 + 03 + S02
                                             LOWRH
                                          RATIO WITH
                                          SCALE EXPANSION
2000cm
2100
                                                       2300
FIGURE 5-12. INFRARED SPECTRA IN REGION OF CARBON MONOXIDE ABSORPTION

-------
                                                                       103





2000-2300 cm   and 2600-3000 cm   for two ethylene-ozone reactions are




shown in Figure 5-13.  Sulfur dioxide was included as a reactant in run




29 recorded in the lower spectrum.  The products observed from these




reactions are carbon monoxide, formyaldehyde,  and in .the case of Run 28




(without sulfur dioxide), a trace of formic acid.  For Run 29, a quanti-




fiable amount of formic acid was observed at approximately 1130 cm




     The results of the product identification and measurement for all




of the experiments analyzed by the Fourier transform spectroscopic method




are listed in the Appendix.  The concentration values tabulated are in




parts per million (ppm) and were calculated by applying the Lambert-Beer




equation to the absorption spectra as discussed in Chapter IV-B,6.  A




more detailed product evaluation is contained in Table 5-3 for eight of




these runs.  The run identification number, the relative humidity and




the product concentration expressed as ppm product per ppm of ozone




reacted are tabulated.  The error term noted below each product species




is based on the confidence with which the absorption coefficients used




in the measurements are known and does not relate to the measurement




itself.  The absorption coefficient for ketene was not measured due to




difficulties encountered in the quantitative isolation of pure ketene




for calibration purposes.  The value used for this species is, therefore,




a best estimate (Gay, 1974) and the error term is believed to be realistic.




     An evaluation of data in Table 5-3 shows that the production of




carbon monoxide is unaffected by the addition of sulfur dioxide to the




system.  Analogous to the formation of ketene from the decomposition of




the acetyl form of the intermediate species, carbon monoxide and water




were speculated as forming from the decomposition of the formyl form




of that species.  For this analogy to hold, a higher concentration of




carbon monoxide would be expected in the absence of sulfur dioxide.  This

-------
                               CO
                                      . RATIO PLOT
  1000
                   2000
                          2100
                                          2300 HOD
                                                            2100
                                                                    MOO
FIGURE 5-13. INFRARED ABSORPTION SPECTRA OF REACTOR CONTENTS.

-------
                                            Table 5-3



                  Product Analysis by Fourier Transform Infrared Spectroscopy
Propylene - Ozone Reaction
Run
60
(with S02)
65
(with S02)
31
(with S02)
21
(without SO2)
81
(without SO -in
82
(without SO -in
Ethylene - Ozone
29
(with S02)
RH CO* HCHO* CH CHO*
% (± 0.02) (± 0.03) (±0.1)
20 0.34 1.34 0.68
36 0.35 1.32 0.76
19 0.36 1.32 0.61
19 0.36 1.24 0.58
36 0.18 0.56 0.32
N2)
20 0.23 0.51 0.30
V
Reaction
20 0.38 1.42 0
CH CO* HCOOH*
(± 0.03) (± 0.07)
0 Trace
0 Trace
0 Trace
0 . 08 Trace
0.08 0
0.06 0
0 0.07
     28

(without SO )
22
0.28
1.38
Trace
*ppm/ppm ozone consumed
                                                                                    o
                                                                                    en

-------
                                                                       106





was not observed in these experiments as the production of carbon monoxide




appeared constant relative to the rate of ozone consumption.  Both forms




of the intermediate species are operative in oxidizing the sulfur dioxide




as evidenced by the formation of light scattering aerosols in both the




ethylene and propylene systems (data follows in Section E).   The formation




of the carbon monoxide observed may still be via that decomposition reac-




tion but direct evidence is lacking.  Reaction 4, Chapter II-B indicates




that carbon monoxide and methyl alcohol could be formed by another decom-




position reaction.  No methyl or other alcohol was observed in these




analyses and, therefore, reaction 4 is apparently not operational or




has a minor role in generating the carbon monoxide.




     Formic acid was observed in each of the experiments conducted in the




oxygen enriched system and is postulated to result from a recombination




of the hydroxyl and the peroxyformyl radicals.




     For the purpose of discussion, a mechanism is adopted which is




similar to the three step mechanism proposed by Cox and Penkett (1972)




in that the olefin-ozone reaction produces a reactive intermediate which




subsequently can decompose, rearrange, or react with other than the two




primary reactants and which is the oxidizing species for sulfur dioxide.




The new mechanism does not, however, exclude additional reactions parti-




cularly those which involve a secondary attack on the principal reactants.




Based on this mechanism, one should observe a higher production of




formaldehyde in the ozone-ethylene reaction conducted in the presence of




sulfur dioxide while in the absence of sulfur dioxide, production of formic




acid and carbon monoxide should be higher.  These features were not




observed in the reactions with ethylene, and in fact the tendency,




although slight, was in the opposite direction.




     An increase in both formaldehyde and acetyaldehyde production was

-------
                                                                       107





also expected for reactions of propylene and ozone occurring in the




presence of sulfur dioxide.  The data listed in Table 5-3 demonstrates




this feature in the case of formaldehyde but it is not well defined in




the case of acetyaldehyde.  A number of calculations were made to sub-




stantiate the increase in aldehyde yield consistent with the sulfur




dioxide oxidative step in the mechanism producing the products sulfur




trioxide and stable aldehyde.  The nine propylene runs listed in the




Appendix were used to determine the formation ratio of the two aldehyde




species.  The observed mean of the ratio of formaldehyde to acetyaldehyde




with one standard deviation for the nine experiments was 1.85 ± 0.20.




This finding tends to confirm the increase in the acetyaldehyde product-




tion concurrent with that of the formaldehyde.  The ratio of the formal-




dehyde concentration to that of carbon monoxide was also calculated for




the seven runs conducted in an oxygenated atmosphere.  Pour of the runs




were conducted without any sulfur dioxide in the system and for these




the mean ratio was calculated as 3.17 ± 0.26.  In the other three




experiments which included sulfur dioxide, the mean ratio was 3.80 ± 0.16.




A t-test shows these mean ratios to be significantly different at the 95%




level.  As the carbon monoxide concentration apparently remains constant,




the difference in the mean ratio values is due to a change in the form-




aldehyde concentration.  This result supports the finding of increased




formaldehyde, or more generally, aldehyde production in the propylene-




ozone thermal reaction when conducted in the presence of sulfur dioxide.




     The competition for the reactive intermediate was affirmed by noting




that a gain in carbon for run 21 was 0.40 ppm accounted for in the species




ketene while the corresponding loss relative to run 31 was 0.43 ppm carbon




accounted for in the two aldehyde species.  Although the initial reactant




concentrations were not precisely the same, it is also interesting to note

-------
                                                                       108


that 0.20 ppm sulfur dioxide was consumed in run 31 while 0.17 ppm ketene

was formed in run 21.  To determine whether this indication establishes

a real relationship between the two species would require further experi-

mentation.

     A product analysis of the two propylene-ozone experiments conducted

in a reduced oxygen atmosphere and listed in the table in the Appendix

indicates that no formic acid was formed.  This finding refutes the mecha-

nistic step shown as reaction 7 in Chapter II-B, in which the acid results

from the decomposition of the formyl form of the intermediate species.

An oxygen enriched atmosphere is required for the production of the formic

acid as it is for the formation of the hydroxyl radical.  The other

species which was formed with the hydroxyl radical from one of the pro-

posed interactions of the reactive intermediate with molecular oxygen is

the peroxyformyl radical.  It is speculated that a recombination of these

two radical species could result in the formation of the formic acid

according to


                  Hc'dO- + -OH  -*-  O  + HCOOH                      (28)


The species shown, however, is also known to be very unstable and rapidly

decomposes as

                         p
                        HCOO-  •»•  CO + 'HO                        (22)


A decomposition reaction of this form can also account for the essentially

constant production per ppm of ozone consumed of carbon monoxide in the

oxygen enriched system.  The occurrence of such a reaction in this system

is reinforced by the observation that the carbon monoxide was lower in

the oxygen deprived atmosphere.

     One further observation noted from these two experiments which made

use of nitrogen as the diluent gas is the marked reduction in the produc-

-------
                                                                       109





tion of both aldehyde species compared to the runs conducted in air.  The




production of formaldehyde for these two runs averaged about 0.54 ppm/ppm




of ozone reacted and for the acetyaldehyde was 0.31 ppm/ppm of ozone




consumed.  This result can be accounted for by another reaction of the




intermediate species with oxygen resulting in the formation of ozone and




aldehyde, reaction 13, Chapter II-B, for example.  It is also significant




to note that this reaction would contribute a major fraction of the




aldehyde formed.




     Ammonia was not observed as a product or a contaminant in these




experimental runs.




     Scott et al. (1957) also observed formaldehyde, acetyaldehyde, carbon




monoxide and ketene in the reaction of propylene with ozone.  In addition,




they reported values for carbon dioxide and water and an upper limit for




acid production.  The initial concentration for each of the two reactants




used in their experiments was 32 ppm.  The results described herein are in




good qualitative agreement with their findings.  Vrbaski and Cvetanovic




(1960) also conducted an extensive product analysis of the propylene-




ozone reaction although at considerably higher reactant concentrations.




Carbon monoxide and formaldehyde were not included in their analyses




performed by gas-liquid chromatography.  Products reported were acetyalde-




hyde, propionaldehyde, acetone, methyl alcohol, carbon dioxide and formic




acid.  The yield reported for formic acid (0.34 ppm/ppm of ozone consumed)




was considerably higher than observed in this study or in the study of




Scott et al. (1957).  The values reported for propionaldehyde and acetone




were very low being 0.003 and 0.008 ppm/ppm of ozone consumed respectively




while that for the alcohol was 0.033 ppm/ppm ozone reacted.




     Although there is some variation in product formation among the




different investigations, there is general agreement relative to the

-------
                                                                       110





formation of aldehydes as a major product of the olefin-ozone reactions.




These species result from the attack on and the rupture of the double




bond by ozone and the addition of an oxygen atom onto one of the fragments.




Through this mechanism and as observed in this study ethylene produces




formaldehyde and the propylene yields acetyaldehyde and formaldehyde.




There is also general agreement that carbon monoxide and carbon dioxide




are among the major products although carbon dioxide was not measured in




this study.  Alterations in the product inventory and yield due to the




presence of the sulfur dioxide are not reported elsewhere.




     It was not possible during these experiments to identify quantita-




tively all of the minor products formed during the reactions.  Some of




the products were below the detection limits of the instrumentation while




the characteristic absorption spectra of others appears in the region of




major products and were completely masked.  Carbon dioxide was not quanti-




tated as there existed a short free air path in the experimental arrange-




ment between the cell and the detector.  Despite these limitations, it




was still of interest to calculate a material balance on carbon, i.e., to




calculate the ratio of the concentration of carbon atoms found in the




identified products to the concentration of carbon atoms in the reacted




olefin.  This ratio was 62.5% for run 31 which included sulfur dioxide




and 66.6% for run 21 which included only the propylene and ozone.  The




ethylene reactions yielded a carbon balance of 56.3% for run 29 with




sulfur dioxide and 54.7% for run 28 which did not contain the sulfur




species.  Assurance that this difference was not due to equipment failure




was obtained by comparing the results of the olefin and acetyaldehyde




measurements made by the Fourier transform infrared spectrometer and the




gas chromatograph.  The four measurements of propylene and acetyaldehyde

-------
                                                                       Ill





agreed to approximately ten percent and the two measurements of ethylene




agreed to within three percent.




C.   Tentative Reaction Mechanism





     To summarize the evidence presented in the previous sections, the




following scheme provides the salient features of a mechanism for the




propylene-ozone-sulfur dioxide system which is consistent with the data




observed in these experiments.  The initial steps of the tentative mecha-




nism are those which directly influence the concentration of either the




sulfur dioxide or the species responsible for its oxidative consumption.




The final steps in the mechanism are presented to account for other pro-




ducts observed in the various experiments conducted.




     One scheme which is generally accepted for the ozonalysis reaction




is the Criegee mechanism in which the reactive intermediates are the




zwitterions.  The structure and reactivity of these species are considered




in popular practice to be biradical in character (Calvert, 1973).  It is




proposed that these reactive intermediates are responsible for the oxida-




tion of the sulfur dioxide in the system.  The products of the initial




reaction are stable aldehydes and the active biradical species according




to the equations





                  03 + CH3-CH=CH2  ->•  HCHO + CUCKOO-             (1)





                                   •*  CH CHO + HCHOO-             (la)





The intermediate can only interact with the olefin, molecular oxygen and




the sulfur dioxide.  The finding that the stoichiometry of the reaction




is affected by the presence of oxygen indicates that the intermediates




are not directly involved in a secondary attack on the olefin.  Inter-




actions of the intermediate with molecular oxygen occur as follows:

-------
                                                                       112

                                         0
                  RCHOO' + 0   -*•  -OH + RCOO-                     (11)

                  RCHOO- + 0   •*•  0  + RCHO                       (13)


A secondary attack on the olefin by species other than ozone generated

in the system has been demonstrated and through the calculated addition

of a prescribed amount of carbon monoxide, the agent responsible for the

secondary attack was tentatively identified as the hydroxyl radical.  A

reduction in the aldehyde yield observed in experiments conducted in a

reduced oxygen atmosphere indicates that reaction 13 is also active in

the system.  This ozone production is apparently offset by the increasing

participation of ozone in reactions involving unidentified products of the

primary reaction.

     An apparent decomposition reaction of the acetyl form of the biradical

intermediate was noted in the propylene-ozone interactions and resulted

in the production of ketene according to the reaction

                  CH CHOO-  -»•  CH =C=0 + HO .                    (5)

The rate of this reaction must be slow since the ketene was not observed

in systems which included sulfur dioxide as a reactant.  This result

along with a materials balance analysis enforced the identification of

the intermediate as the species which oxidized the sulfur dioxide parti-

cularly as none of the other species under consideration form ketene.

     The hydroxyl radical was involved in a secondary attack on the olefin

species.  The rate constant for these reactions is approximately one

hundred times greater than is the rate constant for the hydroxyl radical-

sulfur dioxide reaction thus eliminating that species from consideration

as the oxidizing agent.  The propylene-hydroxyl radical reaction is itself

chain propagating:

-------
                                                                       113


                  *OH + CH -CH=CH   -*•  CH CH-CH OH  .              (14)
                          J      *•       J     &

The final step in the proposed mechanism is the oxidative consumption of

the sulfur dioxide by.the reactive biradical species:


                  RCHOO- + SO   -»•  SO  + RCHO  .                   (29)


Other potentially oxidizing species for the sulfur dioxide have been

eliminated in the preceding sections.

     The mechanistic steps which follow next are advanced to account for

other products observed by the infrared spectroscopic analysis.

     An oxygen effect was observed in the production of formic acid thereby

eliminating the decomposition of the formyl form of the intermediate as

being responsible for its formation.  It is speculated, however, that a

reaction between the hydroxyl and the peroxyformyl radicals will produce

the acid observed:

                         0
                  •OH + HCOO-  ->•  0  + HCOOH   .                   (28)


     The formation of carbon monoxide from another decomposition of the

formyl form of the biradical species in analogy with the formation of

ketene from the decomposition of the acetyl form was not observed.  There

was an oxygen effect noted in the formation of the carbon monoxide and in

contrast to the ketene, the presence of sulfur dioxide did not affect

the observed concentration.  The peroxyformyl radical requires an enriched

oxygen atmosphere for its formation in this mechanism  (reaction 13), is

known to be highly unstable and on decomposition can account for the

carbon monoxide as follows:

                   O
                  HCOO-  ->•  CO + HO   .                            (22)

-------
                                                                       114


     The hydroxyl radical formed in the oxygen enriched system reacts


with the aldehyde species with about the same rate constant as with the


olefin.  This reaction leads to the formation of formyl and acetyl radicals


which in the presence of oxygen become an additional source of peroxy-


formyl and peroxyacetyl radicals:

                                           0
                  •OH  +  RCHO  -»•  HO  + RC-                     (16)


                   9.                ?•
                  RC-  +  °2    •*  RC'°°*  •                       <3°)


     The notable difference between the foregoing mechanism and the three-


step mechanism employed by Cox and Penkett (1972) is the relatively large


involvement of secondary reactions shown above.  The deviation from 1:1


stoichiometry results from a secondary consumption of the olefin and


secondary reactions between ozone and other products of the olefin-ozone-


sulfur dioxide system.  The observed oxygen effect in these studies which


affected the stoichiometry and the yield of the various products indicates


that in an oxygen deprived system, the consumption of ozone by product


species becomes more significant relative to the olefin consumption.  In


an oxygen enriched system, however, additional ozone is regenerated as


are the radical species.



D.   Sulfate Analysis and Sulfur Mass Balance


     Several samples, collected on Millipore Corporation Fluoropore


filters, were studied in an attempt to perform a sulfur balance on the


propylene-ozone-sulfur dioxide reaction and also to confirm the conver-


sion of the sulfur dioxide to a sulfate aerosol.  The filters were 37 mm


in diameter and had a mean pore size of 0.2 ym.  X-ray fluorescence spectro-


scopy, for a total sulfur measurement, and liquid chromatography, for


measurement of water soluble sulfates, were used to estimate the sulfuric

-------
                                                                       115




acid mist loading on tha filter.  Both techniques were employed as they



are still under development as methods for use on atuiospheric samples of



mixed components.  Also, a test for consistency was desired particularly



on this aerosol which is expected to contain only SO (HO) .
                                                    J  b  ii


     A summary of the analytical results from both methods is shown in



Table 5-4.  The total volume sampled, the calculated WEISS concentration



of the sulfuric acid aerosol in \ig/mJ. the equivalent concentration of



sulfur dioxide in ppm and the measured amount of sulfur dioxide consumed



are tabulated.  Using the liquid chromatographic results as a basis the



sulfur balance for the five samples analyzed by that method indicate



that recoveries of from 77 to 100 percent were achieved.  The two



samples for which the recovery was essentially 100S ware collected near



the end of the first hour of the experiments while tb.& other three samples



were collected approximately one half hour later into the run.  The



fraction of the sulfur unaccounted for by the filter analysis is believed



to have been lost to the system by diffusion to and obsorbance on the



reactor walls.  The extensive sampling requirements for runs 16, 24 and



32 coupled with the late period in the run for the filter sampling, i.e.,



after the ozone had been depleted by 90%, left the bag nearly empty and



the surface to volume ratio increased accordingly thereby enhancing the



potential for diffusive losses.



     As no ammonia was observed in the samples analyzed by the Fourier



transform infrared spectrometer it was assumed that the total sulfate



aerosol loading on the filters was due to sulfuric acid.  It also follows



from the high sulfur recoveries observed from the X-ray fluoreraence spectro-



scopic and liquid chromatographic analyses that the dominant if not the



only conversion for the sulfur dioxide is via an oxiclative process result-



ing in the formation of the acid aerosol.

-------
                                            Table 5-4
                     Summary of Sulfuric Acid Analysis by X-ray Fluoresence
                             Spectroscopy and Liquid Chromatography
Sample


  16

  24

  32

  60

  65
Sample


  16

  24

  32

  60

  65

... 2
ng (s)/cm
1348
596
1071
4149
3417

Vig(S)/ml

1.97
3.85
3.32
2.50

ug(s)
10.84
4.79
8.61
33.37
27.48

vg(so^)
29.3
9.85
19.3
99.6
75.0
X-ray Analysis
3
1147
853
1237
1289
1061
Liquid Chromatography
H2S04,M/*3,
1033
585
922
1282
966

SO (ppm)
.187
.139
.202
.210
.173

SO2 (ppm)
.168
.095
.150
.209
.57

Measured
ASO2 (ppm)
.202
.124
.187
.205
.158

% Recovery
83
77
80
100
99

-------
                                                                       117





     The results of the liquid chromatographic analysis were used as a




basis as the system was calibrated with sulfuric acid during the same




time period in which the samples were analyzed.  Calibration samples were




prepared by adding .1 N sulfuric acid aqueous stock solution to isopropy1




alchohol in the prescribed volumetric ratio and taking aliquots to yield




standards in the range of 0.5-5 yg (SO.)/ml (a span that encompassed the




samples collected).  The plot of peal; height for these standards vs con-




centration was non-linear in these low concentrations as was expected




from solubilities and kinetics considerations (Tejada, 1974) .




     Analysis of samples by X-ray fluorescence spectroscopy for elements




lighter than potassium (A=39) is hindered by the self-absorption (of the




low energy characteristic X-rays) within the filter mutrix or by the




aerosol loading  (Dzubay & Stevens, 1974).  The correction for self-absorp-




tion for particles < 2 ym, however, is small and the data to follow




demonstrates that aerosol generated in the gas phase reaction of propylene-




ozone sulfur dioxide consisted largely of sub-microscopic droplets.  A




ratio plot of the results of the total mass of sulfuric acid collected




on the filter as estimated by the X-ray fluorescence and the liquid




chromatographic technique is shown as Figure 5-14.  The unit ratio is




based on the acceptance of the results of the wet chemical analysis as




a reference.  Also, shown on this figure is the mass of the acid measured




by the two methods.




     Although the data are limited an initial evaluation of these results




led to a postulate explaining the high results obtained by the X-ray




fluorescence technique for low total mass loadings,  X-ray fluorescence




spectroscopic analysis under ideal conditions examines an area about 25 mm




in diameter on the filter.  The ratio of the examined to the total filter




area of the 37 mm diameter filters used was about 0.45, and the sample

-------
                                                                  118
SAMPLE
24
32
60
65
X-RAY
FLUORESCENCE
14.68
26.37
102.18
84.16
LIQUID
CHROMATOGRAPHY
10.06
19,65
101.68
76.56
RATIO
1.46
1.34
1.00
1.10
 >•


 Q_
 "*• i
 cr 2
 O
c
-------
                                                                       119



was assumed to be uniformly deposited over the total surface.  Precautions


were taken to facilitate the collection of a uniformly distributed sample


on the filter by adding an extra spacer ring to the in-line filter holder


to allow the captive aerosol more time to disperse laterally prior to


entrapment.  The data suggest, however, that the aerosol was preferentially


loaded on the central portion of the filter first.  For heavier loadings,


the deposition became more uniformly distributed as the increasing pres-


sure drop at the center made collection on the outer portion more favor-


able.  The data further suggest that for a total mass loading of >100 yg


of sulfuric acid and with the sampling arrangement and flow rate employed


the results were comparable.  This postulate was later substantiated by

                                        2
separately analyzing the central 1.98 cm  and the balance of filter sample


number 16 by the liquid chromatographic technique.  The mass loading per


unit area was found to be 1.7 times higher over the. central portion of the


filter than over the outer ring.




E.   Light Scattering Measurements.



     An integrating nephelometer was used to measure the light scattering


coefficient of the aerosol formed in the gas to particle conversion.


Scatter from particulate material over an angle from about 8.5 to 168.8°


was measured by this device.


     According to this instrument, no light scattering aerosol was formed


in reactions which did not include sulfur dioxide.' In addition, no light


scattering aerosol was observed over the lifetime 'of the experiment, in


the reaction of ozone with sulfur dioxide.


     The meteorological range or local visual distance was estimated from


the measurement of the scattering coefficient according to equation IV-4


adjusted for the wavelength used by the eye in sunlight.  These data are

-------
                                                                       120





reported for two representative experiments in Figures 5-15 and 5-16.




Included in these figures is a plot of sulfur dioxide consumption with




time and a plot of the variation in the total surface area with time as




measured with the Electrical Aerosol Size Analyzer.  The final stage of




the latter curve is affected by the rapid growth of the acid aerosol out




of the measurement range of this instrument.  The important feature for




the discussion and comparison to follow is related, however, to the




initial rate of growth to the achievement of some apparent stable total




surface area.




     Figure 5-15 contains plots of these data for a propylene-ozone-sulfur




dioxide reaction.  The initial rate of sulfur dioxide consumption is the




highest for the run and is responsible for the rapid establishment of a




large total surface area due almost exclusively to a nucleation process.




Although the total surface area is large, the newly formed embryos are




too small at early times to scatter light efficiently.  As the experiment




progressed, however, the aerosol particles grew by nucleation, condensa-




tion and coagulation becoming increasingly more efficient in scattering




light.  The mean surface area diameter for this run was estimated at 15,




30 and 60 minutes as 0.22, 0.26 and 0.30 urn respectively.  These mean




surface area diameter estimates are affected by the growth of the aerosol




out of the range of the analyzer and are, as a result, slightly smaller




than would be expected.  Concurrent with the increase in the scattering




coefficient is a decrease in the visual distance which, at eight minutes




into the run,  was almost 16 kilometers (km) and which diminished to about




2 km within forty-five minutes.




     Figure 5-16 contains plots for the ethylene-ozone-sulfur dioxide




system which are similar to those shown in Figure 5-15.  The ethylene-




ozone reaction has been shown previously to be slower than the propylene-

-------
                                                           121
       10
20
30
40
                            f (min)
FIGURE 5-15. VARIATION OF THE SCATTERING COEFFICIENT, VISUAL
           RANGE, SURFACE AREA CONCENTRATION AND SULFUR
           DIOXIDE CONCENTRATION WITH Till/IE FOR A PROPYLENE-
           OZONE-SULFUR DIOXIDE EXPERIMENT.

-------
                                                             122
co
o
                            t (min)

  FIGURE 5-16. VARIATION OF THE SCATTERING COEFFICIENT, VISUAL
             RANGE, SURFACE AREA CONCENTRATION AND SULFUR
             DIOXIDE CONCENTRATION WITH TIME FOR AN ETHYLENE-
             OZONE-SULFUR DIOXIDE EXPERIMENT.

-------
                                                                       123





ozone reaction and the sulfur dioxide consumption is likewise slower.  The




rate of establishment of surface area is reduced compared to the propylene




system and the stable total surface area level is also lower.  The scatter-




ing coefficient measurement shows that the rise of this function is appre-




ciably slower than in the propylene system although the mean surface




diameter of the particles is essentially identical for the two systems at




the times measured.  The estimated diametar at 15, 30, and 60 minutes was




0.22, 0.25 and 0.31 urn respectively.  This result indicates that although




the nucleation process is more limited, the growth of the acid aerosol




in the two systems follows the same pattern.  The degradation of the




visual distance with time liJcewise displays the relative slowness of the




ethylene system.  Visibility, which v/as reduced to about 16 km at eight




minutes in the propylene system was essentially unlimited and at the end.




of forty-five minutes was greater than 6 km compared to 2 kin in the pro-




pylene system.






F.   Aerosol Development





     The propylene-ozone reaction was monitored to determine if a nuclea-




tion process was active in this system and to follow the growth of any




aerosol formed.  A condensation nuclei counter was used to measure the




total number concentration as a function of tims while the electrical




mobility analyzer was used to obtain an estimate of the total surface




area and volume concentrations.  These data, for a typical olefin-ozone




reaction, are shown in Figure 5-17.  An induction period lasting almost




five minutes occurred prior to the initiation of the nucleation process.




Once started, however, the nucleation proceeded rapidly reaching a peak




concentration of approximately 95,000/cm  at fifteen minutes into the run.




Although the nucleation process continued, the peak and subsequent decay

-------
                                                                 124
     100
      80
      60
CSI
  co
      40
   o

  CO
      20
                                                          I
                                                          RUN 91
               10
20
30
40
50
60
70
                                    t (min)
     FIGURE 5-17.  AEROSOL PARTICLE NUMBER, SURFACE AREA AND VOLUME

                 DEVELOPMENT FOR A PROPYLENE-OZONE EXPERIMENT.

-------
                                                                       125





of the number concentration indicated that a transient equilibrium was




attained between the two processes and which was followed by a period of




increasing domination by the coagulation process.  The curves drawn




through these concentration data and those to follow were fitted by eye.




     The total volume concentration was observed to increase linearly for




almost thirty minutes and then remain essentially constant for the dura-




tion of the experiment.  The conservation of the total volume is consistent




with a system under the control of pure Brownian coagulation with no new




material being added to the aerosol volume.




     The total surface area concentration initially increased rapidly due




to the large surface to volume ratio of the newly formed embryos.  Conden-




sable material was limited in this system and at an early stage of the




aerosol's development, the coagulation process dominated causing the




total surface area concentration to diminish.




     There was no significant aerosol growth observed in the propylene-




ozone system and the count mean diameter of the aorosol at sixty minutes




was estimated by the electrical mobility analyzer as 0.018 ym.  There




were no particles > 0.042 pm observed nor was any light scattering aerosol




registered on the integrating nephelometer.




     Figures 5-18 through 5-23 show the total volume, surface area and




number concentrations for a series of six aerosol experiments which in-




cluded sulfur dioxide as a reactant.  The growth of the aerosol formed




in these reactions was limited by design to remain essentially within the




range of the electrical mobility analyzer.  Wall losses evident at higher




reactant concentrations were eliminated as were inaccuracies associated




with coupling of size distribution data from different measurement devices.




The initial reactant concentrations were also varied for these experiments




to obtain a range of sulfur dioxide consumption values.  Ona other variable

-------
                                                                126
    500
CO ICO
 E E
    400
    300
    200
  E
  o
 CO
    100
10     20     30     40     50    60

                   t (min)
                                                 RUN 101 A
                                               70
                                                            11
                                                             CM
                                                            10
                                                                CO
                                                              CO
                                                               X

                                                               CO
       7



       6



       5



       4
          CO
          o

       3  o
          t—I
          X
         Ha
       2



       1
80
     FIGURE 5-18. TOTAL VOLUME, NUMBER, RECIPROCAL OF NUMBER AND
                SURFACE AREA CONCENTRATION VERSUS TIME FOR THE
                DEVELOPING SULFURIC ACID AEROSOL.

-------
                                                           127
FIGURE 5-19. TOTAL VOLUME, NUMBER, RECIPROCAL OF NUMBER AND
           SURFACE AREA CONCENTRATION VERSUS TIME FOR THE
           DEVELOPING SULFURIC ACID AEROSOL.

-------
                                                         128
FIGURE 5-20. TOTAL VOLUME, NUMBER, RECIPROCAL OF NUMBER AND
           SURFACE AREA CONCENTRATION VERSUS TIME FOR THE
           DEVELOPING SULFURIC ACID AEROSOL.

-------
                                                            129
FIGURE 5-21.  TOTAL VOLUME, NUMBER, RECIPROCAL OF NUMBER AND
           SURFACE AREA CONCENTRATION VERSUS TIME FOR
           THE DEVELOPING SULFURIC ACID AEROSOL.

-------
                                                            130
        10     20    30     40     50    60    70    80     90

                            t (min)
FIGURE 5-22. TOTAL VOLUME, NUMBER, RECIPROCAL OF NUMBER AND
           SURFACE AREA CONCENTRATION VERSUS TIME FOR THE
           DEVELOPING SULFURIC ACID AEROSOL.

-------
                                                          131
      10
20     30
40     50    60

 t (min)
70
80
FIGURE 5-23. TOTAL VOLUME, NUMBER, RECIPROCAL OF NUMBER AND
           SURFACE AREA CONCENTRATION VERSUS TIME FOR
           THE DEVELOPING SULFURIC ACID AEROSOL.

-------
                                                                       132


is displayed on these plots, that being the reciprocal of the number con-


centration which is related to the coagulation constant by equation 111-14.


The general patterns in the formation and growth of the aerosol observed


in this series of runs appeared similar for each run over the range of


experimental conditions employed.


     The process by which the acid aerosol was formed and developed in
           i

this system is postulated to include the three mechanisms of nucleation,


condensation and coagulation.  Initially, for a brief instant, the vapor


phase mixture was free of all nuclei.  Homogeneous nucleation then occurred


spontaneously due to the diffusional or effusional aggregation of mole-


cules in the supersaturated system.  Supersaturation in this system is


postulated to have been primarily achieved as a result of the gas phase


reaction leading to the production of species of sufficiently low


volatility which allowed the critical saturation to be exceeded.  The


surface free energy was critical to the stability of these newly formed


embryos because of their small size.


     The solution droplets formed persisted and grew by condensation


because the supersaturation exceeded that required by equation III-8.  The


size of the nuclei apparently exceeded the critical diameter required by


the Kelvin equation (III-4) and the supersaturation was maintained result-


ing in the continued growth of the droplets with a decrease in the free


energy.  Eventually the supersaturation diminished due to a depletion of


the reactants in the gas phase and the subsequent reduction in the pro-


duction of acid vapor.  The vapor was extracted faster than it was being


produced and a practical limit was reached.


     The coagulation process in this system immediately followed the


commencement of the nucleation phase because of the rapid buildup in the


number concentration and the Brownian collisions which followed.  The

-------
                                                                       133





rate of coagulation eventually became equal to the nucleation rate and




the maximum number of nuclei was observed.  The nucleation rate was




finally surpassed and eventually the decay typical to a pure coagulation




process was noted.




     The number concentration was observed to be decaying at the start of




the experiments indicating that the build up of the aerosol by nucleation




was very rapid and was near completion at the recorded zero time or was




at least dominated completely by the coagulation process.  Ozone was




added to the reactor late in the charging period of five minutes and was




not added in the final thirty seconds for any of these aerosol runs.  In




an attempt to monitor the nucleation process, the condensation nuclei




count was recorded through the reactor charging period for runs 104 and




105.  In these two cases the ozone was added to the reactor, already




partially charged with sulfur dioxide, propylene and air, for thirty




seconds commencing at 3.75 minutes into the filling period.  Prior to




the addition of the ozone, the observed nuclei count was approximately




400/cm .  The initiation of the homogeneous nucleation process was




essentially instantaneous with the maximum concentration of 940,000/cm




attained approximately thirty seconds before the recorded zero time.  Here




also, the maximum concentration of nuclei corresponded to a transient




equilibrium condition during which the rate of addition of particles to




the system by nucleation was equal to the loss by coagulation.




     The relationship between the number concentration data and the coagu-




lation constant predicts that the reciprocal of the number concentration




should increase linearly with time for a system in which coagulation is




the only removal process.  The slope of the line formed as a result of




this relationship is numerically equal to the average coagulation constant




of the polydisperse aerosol system.  The plots of the reciprocal number

-------
                                                                       134





concentration in Figures 5-18 through 5-23 show that after some initial




period, a linear relationship was established.  Wall or sedimentation




losses would have been demonstrated by an increasing slope of this curve




commencing at the time of occurrence.  It was expected that should this




phenomena occur, as it did in some of the higher concentration experiments,




it would happen late in the run when the surface to volume ratio of the




reactor has increased significantly.




     The higher value of the coagulation constant noted for the first




several minutes of the runs can be accounted for by the relatively large




difference between the diameter of the newly formed embryos and the mean




diameter of the growing aerosol.  According to the theory of coagulation,




for particles of unequal sizes, the coagulation constant increases rapidly




with an increasing diameter ratio (Table III-l and equation 111-13).




     The total surface area concentration for the six runs was observed




to rise rapidly commencing prior to the recorded zero time.  The large




number of nuclei formed as a result of the gas to particle conversion




coupled with the large surface to volume ratio for the small particles




was responsible for this initial increase in surface area.  As the




aerosol growth continued, condensation tended to increase while coagula-




tion reduced the total surface area observed until an apparent equilibrium




condition between these processes was established. . The surface area




function remained essentially constant until the condensable material in




the system diminished and the growth process was dominated by coagulation.




     The increase in the total volume concentration proceeded at an




appreciably slower rate than did the surface area concentration, again




because of the small volume and relatively large surface area of the




nuclei.  The initial rise in the total volume was due to nucleation and at




some later time, when sufficient surface area had been established, to

-------
                                                                       135




condensation.  When the condensation process ceased due to the depletion



of the vapor phase the total volume concentration was conserved as there



were no system losses such as to the reactor walls.  The slight decline



noted in the total volume concentration curve for run 104 was due to the



growth of the aerosol, or more specifically to some of the aerosol particles



exceeding the range of the electrical mobility analyzer.  Losses to the



walls were contraindicated as the reciprocal number concentration curve



was linear through this period consistent with a pure coagulation process.



     At any time in the developmental history of the aerosol produced in



the thermal reaction of propylene, ozone and sulfur dioxide, the relative



contribution of each of the three processes varys.  The expression for



the rate of generation of particles in a particular class of the size



distribution per unit volume for the simultaneous occurrence of nuclea-



tion, condensation and coagulation takes the form  (according to Hidy and



Brock, 1970 and Brock 1974):





—„ '    = h I  b(v ,v')  n(v )n(v') dv1 - n(v,t)  /  bn(v')dv'
  3t         b     C        C                     b
                                   V
                                coagulation
                                                                  V-4
                         condensation    nucleation




where  n  =  concentration of species in number per unit volume



       b  -  collision parameter for Brownian motion in the free molecule

             regime



       v  =  volume of particular class of molecular aggregate



       Y  =  rate of condensation



       YN = rate of generation of particles in the class by nucleation.





The rate of change in the mass concentration of the sulfuric acid (S) ,



which is coupled to the equation for the rate of generation, takes the

-------
                                                                       136




functional form




9[H SO ]               .     »                   ~

  3*      = k3(J[l][S02] -  | Y(v,s)n(v,t)dv -  £  v?N(v,s)dv      V-5





assuming that k   [l][SO ] equals the rate of formation of the acid species



due to the thermal reaction.  And, if a pseudo-steady state is proposed



for the acid concentration then



                              00        00

            k3Q [I][S02]   =  I fndv + { v?Ndv                     V-6





Owing to the form of the nucleation term, nucleation will take place only



at high acid concentrations but condensation, once an aerosol phase has



been established, will occur at a much lower supersaturation.



     Once an aerosol has been formed it is difficult to distinguish



between continued nucleation and condensation relative to their respective



contribution to the aerosol mass.  Small nuclei attach to the established



aerosol surface in much the same manner and according to laws of the same



form as do the condensing molecules.  For example, one variable measured



during this study was the change in the mean volume diameter with time.




The governing rate equation based on condensation is   . °  = «'S  ,   ,
    '       '       n                                 dt        molecules



while for the attachment of nuclei the expression is —, °  = «"S   , ..
                                                      dt        nuclei



The equations are of the same form differing in the constants «' or «".



For a constant mass of acid in molecular form and a constant acid mass as


        d Dpv
nuclei,    c   would also be constant as was observed in the initial period



of growth in these experiments.  This could occur when the rate of produc-



tion of the acid is constant which, according to the initial linearity



of the volumetric conversion rate, was the case for this series of



experiments.



     An empirical result of computer simulation studies performed by

-------
                                                                       137





Brock  (1974) indicates that for a system involving only the nucleation




and coagulation processes, the reciprocal number concentration curve is




markedly curved during the early stages of aerosol development in agreement




with observations made during the studies reported herein.  In contrast,




the results of a simulation involving only the condensation and coagulation




processes indicate that the reciprocal number concentration increases




linearly with time.  These empirically determined results can be used for




the six aerosol experiments to distinguish between the nucleation and con-




densation processes from the reciprocal number concentration curves.  The




nucleation process appears to persist for a longer period in systems with




a lower sulfur dioxide consumption and a lower total surface area (run




103 for example) while for runs in which the loss of sulfur dioxide is




greater, the surface and volume increase more rapidly at the expense of




the supersaturation level.  As the acid vapor drops below the critical




saturation in these high sulfur dioxide consumption runs  (runs 104 and




105 for example) condensation commences earlier and at a lower super-




saturation state than required by nucleation.




     The volumetric conversion rate, total volume at 80 minutes, volume




mean diameter growth rate, coagulation constant and the sulfur dioxide




consumption through the first 80 minutes, all calculated from the data




shown in Figures 5-18 through 5-23, are listed in Table 5-5.




     The relative humidity was approximately the same during each of




these aerosol experiments.  The volumetric conversion rate was calculated




for the period during which this function was constant, i.e., for at




least the first twenty minutes of the run.  The values listed for the




volume and the sulfur dioxide consumption reflect measurements made when




the experiments were eighty minutes old and when the volume accumulation




was becoming stabilized.  The coagulation constant was calculated from the

-------
                            Table 5-5.

Volumetric Conversion Rate, Total Volume at 80 minutes, Mean Volume
   Diameter Growth Rate, Coagulation Constant and Sulfur Dioxide
         Consumption at 80 minutes for Aerosol Experiments

                                    d5
Run
101A
101B
102
103
104
105
RH (%)
19
19
18
20
20
18
dv . urn
cm hr
217
182
139
89
437
483
-^
iirn
cm
158
150
86
61
235
290
\M*k*
pv ,ym
~dtT (hr)
.391
.340
.295
.258
.655
.671
KxlO 9
1
1
1
1
1
1
/ 3 -1*
(cm sec )
.26
.44
.49
.61
.27
.28
ASO (p
11
10
8
4.
17.
21.
Pi



5
5
5
                                                                                             LJ
                                                                                             CO

-------
                                                                       139




linear portion of the curve of the reciprocal of the number concentration



versus time.  The rate of change in the mean volume diameter related to



the initial period of each run for which this function was apparently



constant.



     The mean coagulation constant with one standard deviation calculated



from the reciprocal number concentration curves for these runs was 1.39 ±


         -9   3
0.14 (x 10  cm /sec).  The mean, count mean diameter calculated from the



data for the six runs which appears in the Appendix with one standard



deviation was 0.064 ± 0.014 ym.  According to Fuchs (1964), the coagula-



tion constant for equal sized particles having diameters between 0.04 and


                                         -9            -9   3     /
 0.1 um would be in the range of 0.72 x 10  to 1.10 x 10  cm /sec (Table



IIT1).   The agreement between the theoretically predicted and the meas-



ured values is very high particularly when the polydispersity is considered.



     The count mean diameters of the distributions indicate that the



particles are larger than the molecules in the suspending gas.  In this



system in which the Knudsen number is greater than 13.2, i.e., the free



molecule regime, the kinetic theory of gases may be applied to aerosol



coagulation.  This implies that the particles are behaving like giant



molecules mixed in the gas in such a way that they merely constitute



just another component of the multicomponent gas phase system.



     For these low level runs, the total amount of sulfur dioxide reacted



was small and the reaction rate was slow.  The volumetric conversion rate



was essentially constant during at least the first twenty minutes of these



runs as evidenced by the total volume concentration curves on Figures 5-18



through 5-23.  Based on the initial, near linear consumption of the sulfur



dioxide, an estimate of the consumption rate of that species during the



initial period where the volumetric conversion rate was constant could



be calculated.  In the performance of this calculation it was assumed that

-------
                                                                       140


the sulfur dioxide consumption was directly proportional to the formation


of the sulfur trioxide during this period, that the sulfuric acid to


water ratio in the particles was constant and that equilibrium existed


between the vapor and the condensed phase of the aqueous sulfuric acid.


Bray  (1970) generated tabulations of the relative humidity and the water


vapor pressure over aqueous solutions of sulfuric acid as a function of


the temperature and density at the equilibrium condition.  The tabulation


also included a correlation of the percent acid and the density of the


aqueous sulfuric acid solution as reported in the Chemical Engineers


Handbook  (1963) .


     The following expression relates the oxidation rate of the sulfur


dioxide with the volumetric conversion rate of the aqueous sulfuric acid


solution through the percent acid and the solution density data:




                  dfS02]     dtVSO,(H,0) 1      MSO,
                  	  =  	i—t—i_   fp 	£              V-7

                    dt



where f  =  the mole fraction of sulfuric acid in the aqueous solution
            at the temperature and relative humidity of the system


      p  =  density of the solution and


      M  =  molecular weight of the species.


Using the average temperature of 27°C, average relative humidity of 19


percent for the six aerosol runs and data from Bray's tabulation (density =


1.477 g/cm ; percent acid = 58.43), the initial reaction rate of the sulfur


dioxide was calculated.  Dividing this result by the initial sulfur dioxide


concentration yields the oxidation rate by the reactive species in this


dark reaction.  The results of these computations are listed in Table 5-6.


     The mean oxidation rate with one standard deviation calculated from


these data is 0.360 ± .096 percent/hour which is in good agreement with

-------
                                                                        141
nun
101A
101B
102
103
104
105
dt IFF"1 HI
7.80 x 10"4
6.54
4.97
3.17
. 15.67
17.36
Table 5-6.  Initial Sulfur Dioxide Reaction Rate and Oxidative Rate.






                             "X J "1)       Oxidative




                                                  0.414



                                                  0.299



                                                  0.229



                                                  0.476



                                                  0.435



                                                  0.307
Table 5-7.  Calculated Sulfur Dioxide Consumption based on the Observed



            Volume Concentration.
                                            ASO  Calculated

      Run        ASO_ Calculated  (ppm)      7———	•:—
                    2             ee        ASO  Observed




                                                  2.0



                                                  2.1



                                                  1.5



                                                  1.8




                                                  I.-9



                                                  1.9
101A
101B
102
103
104
105
0.022
0.021
0.012
0.008
0.033
0.041

-------
                                                                       142






the observations of Cox and Penkett (1971) who reported an oxidation




rate of 0.4 percent per hour for sulfur dioxide in their studies.




     It was also of interest to determine whether the total volume con-




centration of aerosol measured at eight minutes into the run correlated




with the sulfur dioxide consumption using the density, percent acid, and




equilibrium conditions assumed in the previous calculations.  The sulfur




dioxide concentration was calculated, which according to the assumptions




would be responsible for the volume observed.  The calculated sulfur




dioxide concentration was then compared to the amount reacted.  The




results of this analysis are reported in Table 5-7.  The mean ratio of




the calculated to the observed sulfur dioxide consumption based on these




data is 1.9 ± 0.2 indicating that the volume concentration measured was




somewhat greater than can be accounted for by this type of computation.




There are a number of factors which could possibly account for this dif-




ference including an inaccuracy in the original assumptions of constant




acid to water ratio in the volume and equilibrium between the vapor and




the condensed phase of the aqueous sulfuric acid.  It is believed, however,




that this error more probably stems from the lack of an absolute calibra-




tion on the aerosol measuring instrument.




     The observed volume and the sulfur dioxide consumption data, both




at eighty minutes into the run, were plotted for the six aerosol experi-




ments (Figure 5-24).  The line was fit through the origin by a least




squares regression analysis and a slope of approximately 13.5 was obtained.




These data demonstrate the linearity of the relationship existing between




these two parameters and as predicted by equation V-7.




     The data displayed in Figures 5-18 through 5-23 show that the total




surface area concentration grows quite rapidly to some apparent equilibrium




level where it remains essentially constant for a significant period of

-------
                                                                143
  300
  200
CO
4^
ro
  100
                         8         12        16

                                AS02(ppb)
20
24
     FIGURE 5-24. TOTAL VOLUME VERSUS SULFUR DIOXIDE CONSUMPTION
                ATT+80 MINUTES,

-------
                                                                 144





time.  The total volume concentration of the aerosol can be seen to be



increasing linearly to and in some cases beyond the time at which the



equilibrium surface area is established.  This indicates that the equili-



brium condition is established as a result of the gain of surface area



due to condensation and nucleation being offset by the loss due .to



coagulation.  The equilibrium condition is destroyed when the reactants



have largely been consumed, the degree of supersaturation diminished and



new surface area is not being formed.  The coagulation process continues



and the total surface area is thereby reduced.



     Clark  (1972) noted, for his experiments dealing with the photooxida-



tion of sulfur dioxide, a strong correlation between the equilibrium sur-



face area and the volumetric conversion rate.  A linear relationship was



demonstrated in a log-log plot of the two parameters.  This correlation



was tested for the series of six aerosol experiments under discussion



the results of which are shown in Figure 5-25.  The line shown is fit



to the six data points by the method of least squares regression analysis.



The equation which best fits these data is given as



                                        dv
                  log S0  =  0.583 log  (^-) + 0.239              (V-8)
                       E                at



for which a correlation coefficient of 0.9933 was obtained.



     Also plotted on this figure are the data from Clark's experiments



as well as some data from sulfur dioxide photooxidation studies conducted



by the University of Minnesota and by Calspan Corporation.  Although it



is generally considered a tenuous operation to extend a regression line



beyond the limits of the data it was successfully done through more than



two log decrements with these data.  It is immediately apparent that all



of the data plotted could be fit to a single regression line with a high



correlation coefficient.  Of note in this observation is the fact that

-------
   104
CO
 o
   103
   10?
                         I    I  I  I  Mill
OMcNELIS EXPERIMENTS
D CLARK (1972)
A CALSPAN (KOCMOND ET AL, 1973)
• UNIV. OF MINNESOTA LARGE BAG (KOCMOND ET AL., 1973)
• UNIV. OF MINNESOTA SMALL BAG (KOCMOND ET AL., 1973)
                                                           LILIJI       II  II
     0.1     0.2  0.3 0.4   0.6 0.8 1
                         2    34   6  8 10      20   30 40   60 80100    200    300500
                                dV    ,   ,
                                — (unr/cnH-hr)
                                at
                                                                                                    tn
         FIGURE 5-25. VARIATION OF THE EQUILIBRIUM SURFACE AREA WITH VOLUMETRIC CONVERSION RATE.

-------
                                                                       146




the data plotted for which the volumetric conversion rate < 30 ym /cm -hr



are from experiments involving the photooxidation of sulfur dioxide while



those with the volumetric conversion rate > 80 ym /cm -hr are from a thermal



reaction involving olefin-ozone-sulfur dioxide.  The tendency toward the



equilibrium surface area is controlled, therefore, by the physical processes



of nucleation, condensation and coagulation rather than the particular



chemical reaction leading to the gas to particle conversion.



     The volumetric conversion rate measured for the initial period of the



aerosol growth is plotted in Figure 5-26 against the product of the initial



ozone and propylene concentrations for two different initial concentrations



of sulfur dioxide.  Included in this plot are eight of the experiments



discussed earlier in this report for which both the reactant concentra-



tion data and the volumetric conversion rates were available and which



involved higher reactant concentrations.  The data point closest to the



origin represents run 104 of the aerosol series.  The curves shown are the



least squares regression lines fit to the data and forced through the



origin.  The volumetric conversion rate for the higher concentration runs


                   AV
was estimated from —  over the first few minutes of the reactions only.



These data indicate that for a given sulfur dioxide concentration, the



aerosol formation follows second order kinetics with respect to the olefin



and ozone concentrations.  These data also indicate a predictive capability



for the volumetric conversion rate which, for the low concentration runs,



was observed to remain constant for a prolonged period.



     The volume distribution of the aerosol was followed in time for the



six runs.  The distributions plotted at selected time intervals for these



runs are shown in Figures 5-27 through 5-32.  Depending on the time at



which the distributions were measured they were influenced to a different



degree by one or more of the three processes:  nucleation, condensation

-------
                                                                147
   300
CO
   200
   CO

   o.
   100
                                                    0.204 ±0.010 ppm
              123456

                              ffOLj [03])0 (ppm2)

  FIGURE 5-26.  INITIAL VOLUMETRIC CONVERSION RATE VERSUS PRODUCT OF
              INITIAL PROPYLENE AND OZONE CONCENTRATIONS FOR
              DIFFERENT INITIAL SULFUR DIOXIDE CONCENTRATIONS.

-------
                                                                      148
   300
   200
00
o
   100
        0.013    0.024      0.042    0.075     0.133     0.237     0.422    0.750

                                      Dp (yLtm)
   FIGURE 5-27. DEVELOPMENT OF THE VOLUME DISTRIBUTION WITH TIME.

-------
                                                                149
    0.013    0.024     0.042     0.075    0.133     0.237    0.422    0.750
                              Dp(/im)
FIGURE 5-28. DEVELOPMENT OF THE VOLUME DISTRIBUTION WITH TIME.

-------
                                                            150
                               Dp
FIGURE 5-29. DEVELOPMENT OF THE VOLUME DISTRIBUTION WITH TIME.

-------
   200
a
oo
                                                                    151
                                                            RUN 103
                                45MIN
                                                  MIN
        T+ 5 WIN
        0.013     0.024    0.042    0.075     0.133    0.237     0.422    0.750

                                     Dp
     FIGURE 5-30. DEVELOPMENT OF THE VOLUME DISTRIBUTION WITH TIME
00
o
   100
        0.013     0.024     0.042     0.075    0.133     0.237    0.422     0.750

                                   Dp (/.tm)

   FIGURE 5-31.  DEVELOPMENT OF THE VOLUME DISTRIBUTION WITH TIME.

-------
                                                                  152
      0.013    0.024    0.042     0.075    0.133     0.237    0.422     0.750
FIGURE 5-32.  DEVELOPMENT OF THE VOLUME DISTRIBUTION WITH TIME.

-------
                                                                       153



                                                               dV
and coagulation.  The frequency parameter used in these plots, ———,

                                                                 °9 P



presents the distribution in such a manner that the area under the curve



in a given size range is proportional to the volume in that range.



     There are several notable features in these developmental volume



distributions which were similar for each of the experiments.  Most



obvious is the shift in the distributions to larger mean diameters with



time.  The rate of change appears, however, to decrease at later times



and the mean volume diameter appears to be approaching some limit.  The



increase in the total volume i's also readily apparent, particularly in



runs in which the sulfur dioxide consumption was highest.



     Although in the logrithimic display of these data, it appeared



that the distributions were becoming narrower, in fact the spread increased



over a larger range of particle diameters.  The count distribution data



(included in the Appendix) also reflect the increasing polydispersity



with time.  From theoretical considerations, the tendency for the size



distribution, due to droplet growth by diffusion of vapor, i.e., conden-



sation, is to become narrower.  The increasing polydispersity in this



system, therefore, must have been due to the formation of.larger particles



by collisions and coalesence of smaller ones, also theoretically predicted.



The tail of the distribution curves for measurements made late in the runs



show that for the runs in which the amount of sulfur dioxide reacted was



high, there was some few particles which outgrew the range of the electrical



mobility analyzer.



     The volume mean diameters of the growing aerosol were calculated for



each of the runs at several different times out to beyond eighty minutes



in most cases.  These data are plotted in Figure 5-33 and show that the



rate of change in the volume mean diameter is apparently constant for

-------
                                                                154
I O
                                                           RUN 104
                                                RUN 101B        —
         FIGURE 5-33. VARIATION OF VOLUME MEAN DIAMETER WITH
                    TIME FOR SIX AEROSOL EXPERIMENTS.

-------
                                                                       155





the initial period of at least twenty minutes.  The slopes of the linear




portion of these curves are listed in Table 5-5 and range from 0.258 pm/hr




for run 103 to 0.671 ym/hr for run 105.  The higher rates correspond to the




runs in which the largest amount of sulfur dioxide was reacted.  The mean




volume diameters for each of the runs tended to reach some upper limit




of development which for the conditions of these experiments appeared to




be about 0.25 to 0.30 ym.




     The curves plotted in Figure 5-33 also indicate two distinct growth




regimes which are more obvious on the curves having the higher initial




rate of change in the mean volume diameter.  The knee of these curves




corresponds in time to the point at which the reciprocal number concentra-




tion curves became linear (Figures 5-18 through 5-23).  The coagulation




process essentially commences in these systems concurrently with the




initiation of the gas phase reaction because of the very rapid increase




in the number concentration of nucleated droplets and continues through-




out the experimental period.  The nucleation process, however, occurs




initially when the supersaturation exceeds some critical value and recedes




when either the supersaturation drops too low to support the process or




when sufficient surface area is formed and the condensation process com-




petes for the material in the vapor phase.  It is suggested for these




curves, that the initial linear increase in the volume mean diameter is




due to a coupling of the nucleation and the coagulation processes while




the later and slower increase is due to condensation and coagulation.




The relative sharpness of the knee depends on the accumulative effect of




the speed with which the critical surface area in the system develops and




with which the acid in the vapor phase is depleted by condensation.  The




net result of this additive effect is a sharp decline in.and termination




of the homogeneous nucleation process.

-------
                                                                       156





     The constant rate of change observed initially in the volume mean




diameter was found to be linearly related to the volumetric conversion rate




measured for each of the aerosol experiments through the logarithms of the




two functions.  These data are plotted in Figure 5-34 and the line shown




is fit by the method of least squares regression analysis.  The equation




which best fits this data is given by




                       dD

                  Io9 (~^)  = 0.6102 log (^)  - 1.8172
                        at                  at




for which a correlation coefficient of 0.9880 was obtained.




     The results of these aerosol characterization studies could have an




immediate application in the generation of sulfuric acid mist of known




physical properties for controlled laboratory animal exposures.  The data




displayed in Figure 5-26 indicate that the volumetric conversion rate of




condensable material from the vapor to the aerosol phase is predictable




from a knowledge of the initial reactant concentrations.  And, for a




given initial concentration of sulfur dioxide, the aerosol formation




follows a second order rate law with respect to the two primary reactants,




propylene and ozone.  The curve drawn in Figure 5-34 shows that the rate




of change in the mean volume diameter is also predictable based on the




volumetric conversion rate.  Finally, the curves previously discussed and




displayed in Figure 5-33 show that some limit on the development of the




mean volume diameter occurs in practical systems and is likewise predict-




able.  For the conditions of these experiments a limit of 0.25 to 0.30 urn




was imposed on the growth of the mean volume diameter.




     Although it is believed that aerosols having other mean volume dia-




meters could be generated through this mechanism, the mean volume diameter




observed in these studies is of particular interest.  Previous studies,




summarized by Lewis et al.  (1972), made use of aerosols of 0.8, 2.5 and

-------
                                                                   157
ft
     1.0
     0.9
     0.8
     0.7
     0.6

     0.5

     .0.4
     i
     0.3
     0.2
     0.1
      5x101
                                    0.6102    —
102
5x102
103
                                dv
                                dt
       FIGURE 5-34. RATE OF CHANGE IN MEAN VOLUME DIAMETER VERSUS
                  VOLUMETRIC CONVERSION RATE.

-------
                                                                       158





7.0 vim mass median diameters (mmd) to measure the effect of sulfuric acid




mist on the pulmonary function of guinea pigs.  Few particles of the




largest size penetrate beyond the nasal-pharangeal region where they can




cause an obstruction in the larger airways.  The two smaller sizes can




penetrate the lower regions of the respiratory system and interact with




the smooth muscle lining around the terminal and other small bronchioles.




It is this smooth muscle which contracts to produce in humans the symptoms




of asthma, the physiological response most strongly associated with




suspended particulate sulfate (Environmental Protection Agency, 1974).




An acid aerosol with a mass mean diameter of the size generated during




the aerosol studies discussed in this dissertation (approximately 0.3 ym)




has not been used in exposure studies.  It is suggested that these smaller




particles would penetrate the alveoli, the division of the pulmonary




region of the lung which has no smooth muscle and as a result, a different




physiological or histopathological response may be observed.






G.   Aerosol Measurements by Single Particle Light Scattering.





     The aerosol generated in the series of the experiments just discussed




was limited in growth by controlling the initial reactant concentrations.




That series led to the development of an internally consistent mechanism




to describe the relative contribution of the different processes leading




to the production and growth of the acid aerosol.  The aerosol produced




in the higher reactant concentration experiments was also monitored with




an optical aerosol analyzer.  The results of these measurements for typical




experimental runs are shown in Figures 5-35 and 5-36 for a propylene-ozone-




sulfur dioxide system and for a ethylene-ozone-sulfur dioxide system




respectively.  The general trends displayed for the two different olefins




are similar with the production of the light scattering aerosol occurring

-------
                                                              159
                                                  RUN 31
                                                      • 03 - S02
  20   40    60    80   100   120   140   160   180   200   220  240

                             CHANNEL

FIGURE 5-35.  VARIATION OF AEROSOL SIZE DISTRIBUTION WITH TIME AS
            MEASURED WITH OPTICAL PARTICLE SIZE ANALYZER,

-------
                                                               160
3000
                                                  RUN 29
                                                  C2H4 • 03 - S02
   20    40    60    80   100   120  140   160   180  200  220   240

                               CHANNEL

  FIGURE 5-36. VARIATION OF AEROSOL SIZE DISTRIBUTION WITH TIME AS
             MEASURED WITH OPTICAL PARTICLE SIZE ANALYZER.

-------
                                                                       161





earlier and at a faster rate in the propylene system.  The fact that




ethylene produces a similar aerosol is itself of note supporting the basic




features of the gas phase mechanism postulated earlier.  Principal features




of these aerosol size distributions are the first peak which occurs at




about 0.38 pm, the saddle at approximately 0.76 ym and the second peak at




channel 140 for which the diameter is 1.35 ym.  Although the aerosol




outgrew the range of the electrical mobility analyzer, the measurement




by that instrument indicated a somewhat lower mean size for the first




peak.  More important, however, is the existence of the second peak




which occurred in each of the higher concentration runs.  It is speculated




that this secondary peak and the saddle point occur as a result of the




enhanced coagulation among particles of unequal sizes.  This mechanism




would account for the peak becoming more pronounced at the expense of




the smaller particle sizes through the saddle point of the combined




distribution for these systems of higher reactant consumption.  More




significant in the observation of the second peak is the fact that




although only a relative few droplets grow to this larger size they can




account for an appreciable fraction of the total aerosol mass.

-------
                                CHAPTER VI




                     CONCLUSIONS AND RECOMMENDATIONS





     As a result of the measurements made during these experiments and




reported on in this dissertation, the mechanism leading to the oxidative




consumption of sulfur dioxide is better understood.  In addition, the




mechanism leading to the production, growth and decay of a sulfuric acid




aerosol from the thermal gas phase reaction has been explained in terms




of the basic processes of nucleation, condensation and coagulation.




Finally, the coupling of the gas phase and the aerosol information pro-




vides a unique set of data with which to evaluate atmospheric and photo-




chemical smog chamber results.  The principal conclusions reached as a




result of these studies follow in a summarized format.




     1.  The reactor vessel and sampling system designed and assembled




for these experiments were found to perform satisfactorily in the study




of selected dark phase reactions and in the measurement of specific gas




and condensed phase products.  A reactor having a larger volume is sug-




gested for use in any additional studies of the type conducted to allow




for more nearly continuous sampling with all of the instrumentation.




     2.  Wall effects did not affect the results or the conclusions of




this study significantly.  The losses of the individual reactants were




measured and found not to affect the stoichiometric results.  The sulfur




mass balance and the aerosol reciprocal number concentration measurements




indicate that a measurable loss of sulfur to the walls occurred only at




the terminus of the runs when the surface to volume ratio of the reactor




became excessively large.  The lower recovery of total carbon evaluated

-------
                                                                       163





near the end of the runs was believed due to the carbon content of




unmeasured products.




     3.  The stoichiometry of the propylene-ozone reaction was found to




be a smooth function of the initial concentration ratio of these species.




The olefin/ozone consumption was > 1 for a system in which the olefin was




initially in excess and < 1 for a system having an initial ozone excess.




The consumption ratio was unaffected by the addition of sulfur dioxide




to the reactor or by varying the relative humidity over the range of 20




to 38 percent.




     4.  Molecular oxygen had a significant effect on the reaction stoichi-




ometry and product formation in the propylene-ozone thermal reaction.  The




propylene/ozone consumption ratio was lower in a system in which the molec-




ular oxygen concentration was reduced.  Oxygen also contributed to the




regeneration of ozone and the production of the hydroxyl radical species,




both of which interacted with the propylene and with products of the reac-




tion.  Although the formation of the hydroxyl radical had been postulated,




these experiments provided the first direct evidence for its participation




and its role.




     5.  Acetyaldehyde, formaldehyde, carbon monoxide, ketene and formic




acid were observed products of the propylene-ozone reaction.  The concen-




trations of acetyaldehyde, formaldehyde and carbon monoxide were affected




by the molecular oxygen concentration.  Ketene was apparently unaffected




by the availability of oxygen while the production of formic acid depended




on its presence.  Ketene was not observed in any of the reactions involving




sulfur dioxide as a reactant indicating that its rate of formation is




significantly slower than the rate of the bimolecular reaction between




sulfur dioxide and the intermediate species.




     6.  The mechanism which follows has been advanced in this dissertation

-------
                                                                       164





for the oxidative consumption of sulfur dioxide in the propylene-ozone-




sulfur dioxide system and which is consistent both internally and with the




data observed in this study.  Secondary reactions are a distinctive




feature of this model which incorporates relatively few reactions to




explain the major characteristics of the system studied.  The rate con-




stant for the propylene ozonalysis step of the mechanism was estimated




to be 1.40 ± 0.39 (x 10   ppm   min  ).  The steps of the proposed




mechanism are as follows:





                   0  + CH -CH=CH   •*  HCHO + CH CHOO-            (1)
                    «j     
-------
                                                                       165




patterns are consistent with a model which includes homogeneous nucleation,




condensation and coagulation.  Nucleation commenced in the system which




included sulfur dioxide almost immediately upon initiation of the gas




phase reaction.  Coagulation of the newly formed embryos and the growing




droplets commenced within seconds and persisted throughout the experi-




mental period.  Condensation began with the establishment of a stable




aerosol phase and soon replaced nucleation as the primary gas to particle




conversion process.




     9.  An apparent equilibrium surface area was established for the




aerosol when the addition to the total surface area by condensation was




equal to the loss by coagulation.  A strong correlation was found to exist




between this equilibrium surface area and the volumetric conversion rate.




It was also found that the correlation could be extended considerably in




range to include data from experiments conducted elsewhere in which the




aerosol was photochemically generated.  The establishment of the equili-




brium surface area is due, therefore, to the physical processes rather




than the chemical mechanism leading to its formation.




    10.  The volume distributions were observed to shift to larger sized




droplets with time.  An upper limit for droplet growth was noted and




was attributed to the depletion of the reactant concentrations, of the




concentration of the sulfuric acid in the vapor phase and of the number




concentration of the aerosol droplets.  The initial rate of change in the




mean volume diameter was found to be highly correlated with the volumetric




conversion rate according to a power function relationship.




    11.  Finally, the apparent dependence of the volumetric conversion




rate on the initial reactant concentrations and of the rate of change of




the mean volume diameter on the volumetric conversion rate provide a pre-




dictive capability potentially useful in the generation of a sulfuric

-------
                                                                       166





acid aerosol of known physical properties for controlled laboratory




animal experimentation.




     It is recommended that additional research be conducted to measure




or to obtain through appropriate mathematical models the sulfur dioxide




oxidation rate.  The reaction scheme advanced in this dissertation with




its mechanism for ozone regeneration and hydroxyl radical production could




be tested in existing kinetic models comparing the results with the con-




centration profiles observed in this study.




     Experimentation along the lines conducted in this study should also




be continued to include other olefinic compounds.  The structure and




reactivity of the intermediate species and the reaction products would




vary depending on the unsaturated hydrocarbon investigated.  Tetramethyl-




ethylene, for example, would yield acetone as an initial product in lieu




of the aldehydes.  The hydroxyl radical reaction with the acetone is




expected to be significantly slower than its reaction with the aldehydes




and the product inventory and primary reaction stoichiometry would be




affected.  The products of the secondary reaction between the olefin and




the hydroxyl radical require identification as do the products of the




decomposition reactions of the hydroxyl radical-olefin adducts.  The




products of these composite reactions could be followed in time with a




system like the Fourier Transform Infrared Spectrometer and the concen-




tration profiles evaluated to determine the relative potential of the




various olefins in photochemical smog development.




     The hydroxyl radical is, perhaps, the most important oxidizing species




in photochemical smog particularly in the early stages before the ozone




concentration develops.  Although the initial reaction of the hydroxyl




radical with saturated and unsaturated hydrocarbons is established, its




mode of reaction with aromatics is speculative.  Experimental evidence is

-------
                                                                       167





needed to fix the exact mechanism of that reaction.   Specification of the




oxidation mechanism for sulfur dioxide in reactions  involving the hydroxyl




radical with olefins, parraffins and aromatic hydrocarbons is also needed.




These reactions potentially contribute to the development of secondary




aerosols in the ambient atmosphere which are largely responsible for the




aggravation of respiratory ailments and reduced visibility associated with




photochemical smog.

-------
                               REFERENCES
Ahlquist, N.C. and Charlson, R.J., "Measurement of the Wavelength
Dependence of Atmospheric Extinction Due to Scatter," Atmospheric
Environment, Vol. 3, p. 551 (1969).

Altshuller, A.P. and Bufalini, J.J., "Photochemical Aspects of Air
Pollution:  A Review," Environmental Science and Technology, Vol. 5, p.  39
(1971).

Amelin, A.G., Theory of Fog Condensation, S. Monson Wiener Bindery Ltd.,
Jerusalem, 1967, pp. 1-39.

Bademosi, F., "Diffusion Charging and Related Transfer Processes in
Knudsen Aerosols," Ph.D. Thesis, University of Minnesota, Minneapolis, 1971.

Bailey, P.S., "The Reactions of Ozone with Organic Compounds," Chemical
Reviews, Vol. 58, p. 925  (1958).

Baulch, D.L., Drysdale, D..D and Lloyd, A.C., "High Temperature Reaction
Rate Data," Department of Physical Chemistry, The University, Leeds,
England, 1968.

Becker, R. and Doring, W., "Kinetic Treatment of Embryo Formation in
Supersaturated Vapours," Annalen der Physik, Vol. 24, p. 719 (1935).

Blacet, F.E., "Photochemistry in the Lower Atmosphere," Industrial and
Engineering Chemistry, Vol. 44, p. 1339  (1952).

Bray, W.H., "Water Vapor Pressure Control with Aqueous Solutions of
Sulfuric Acid," Journal of Materials, Vol. 5, p. 239 (1970).

Brock, J.R., Private Communication, University of Texas, Austin, 1974.

Bufalini, M., "Oxidation of Sulfur Dioxide in Polluted Atmospheres,"
Environmental Science and Technology, Vol. 5, p. 685 (1971).

Bufalini, J.J., Private Communication, Environmental Protection Agency,
Research Triangle Park, North Carolina, 1974.

Bufalini, J.J. and Altshuller, A.P., "Kinetics of Vapor-Phase Hydrocarbon-
Ozone Reactions," Canadian Journal of Chemistry, Vol. 43, p. 2243 (1965).

Cadle, R.D. and Schadt, C., "Kinetics of the Gas Phase Reaction of Olefins
with Ozone," Journal of the American Chemical Society, Vol. 74, p. 6002
(1952).

-------
                                                                         169


  Cadle, R.D. and Schadt, C., "Kinetics of the Gas Phase Reaction between
  Acetylene and Ozone," The Journal of Chemical Physics, Vol.  21, p.  163
   (1953).

  Calvert, J.G., "Interactions of Air Pollutants," Presented at the National
  Academy of Sciences, Conference on Health Effects of Air Pollutants,
  Washington, D.C. (1973).

  Charlson, R.J., Ahlquist, N.C., Selvidge, H. and MacCready,  P.B. Jr.,
  "Monitoring of Atmospheric Aerosol Parameters with the Integrating Nephlo-
  meter," Journal of the Air Pollution Control Association, Vol. 19,  p. 937
   (1969).

  Clark, W.E., "Measurements of Aerosols Produced by the Photochemical
  Oxidation of SO  in Air," Ph.D. Thesis, University of Minnesota, Minneapolis,
  1972.

  Clark, W.E. and Whitby, K.T., "Concentration and Size Distribution Measure-
  ments of Atmospheric Aerosols and a Test of the Theory of Self-Preserving
  Size Distributions," Journal of Atmospheric Sciences, Vol. 24, p. 677 (1967).

  Chemical Engineers Handbook, 4th Edition, (J.H. Perry, ed.), McGraw-Hill,
  New York, 1963, p. 3-79.

  Cox, R.A. and Penkett, S.A., "Photooxidation of Atmospheric SO ," Nature,
  Vol. 229, p. 486 (1971).

\Cox, R.A. and Penkett, S.A., "Aerosol Formation from Sulfur Dioxide in the
  Presence of Ozone and Olefinic Hydrocarbons," Journal of the Chemical Society,
  Faraday Transactions I, Part 9, Vol. 68, p. 1735 (1972).

  Criegee, R., Blust, G. and Zinke, H., "Eine Neuartige Synthese von Ozoniden,"
  Chemische Berichte, Vol. 87, p. 766 (1954).

  Criegee, R., Kerchav, A. and Zinke, H., "Der Bildungsbereich Monomerer
  Ozonide," Chemische Berichte, Vol. 88, p. 1878 (1955).

  Doyle, G.J., "Self Nucleation in the Sulfuric Acid-Water System," The
  Journal of Chemical Physics, Vol. 35, p. 795 (1961).

  Dzubay, T.G. and Stevens, R.K., "Applications of X-Ray Fluorescence to
  Particulate Measurements," Second Joint Conference on Sensing of Environ-
  mental Pollutants, Instrument Society of America, Pittsburgh, Pennsylvania,
  1973, pp. 211-218.

  Dupont Company, "Tedlar" PVF Film, Bulletin TD-3,. Wilmington, Delaware.

  Environmental Protection Agency, National Primary and Secondary Ambient
  Air Quality Standards, Part 410, Chapter IV, Title 42, Code of Federal
  Regulations, Federal Register, Vol. 36, p. 8195, 1971.

  Environmental Protection Agency, Health Consequences of Sulfur Oxides;  A
  Review from CHESS, 1970-1971, Environmental Health Effects Research Series,
  EPA-650/1-74-004, Research Triangle Park, North Carolina, 1974.

-------
                                                                       170


Friedlander, S.K., "Similarity Considerations for the Particle-Size
Spectrum of Stratospheric Aerosols," The Journal of Meteorology, Vol.  17,
p. 479 (1960).

Fuchs, N.A., Droplet Evaporation and Growth in a Gaseous Medium, Pergamon
Press, London, 1959.

Fuchs, N.A., The Mechanics of Aerosols, MacMillan Company, New York,  1964.

Garvin, D. and Hampson, R.F. (editors), Chemical Kinetics Data Survey VII.
Tables of Rate and Photochemical Data for Modelling of the Stratosphere.
National Bureau of Standards, NBSIR 74-430, Washington, D.C.,  1974, pp.  10-56.

Gay, B.W. Jr., Private Communication, Environmental Protection Agency,
Research Triangle Park, North Carolina, 1974.

Goodeve, C.F., Eastman, A.S. and Dooley, A., "The Reaction between Sulfur
Trioxide and Water Vapours and a New Periodic Phenomenon," Faraday Society,
London.  Transactions, Vol. 30, p. 1127 (1934).

Goulding, F.S., Jaklevic, J.M., Jarrett, B.V. and Landis, D.A., Advances
in X-Ray Analysis, Plenum Press, London, 1972, p. 470.

Groblicki, P.J. and Nebel, G.J., "The Photochemical Formation  of Aerosols
in Urban Atmospheres," Chemical Reactions in Urban Atmospheres, (C.S.
Tuesday, ed.), American Elsevier, New York, 1971, pp. 241-267.

Hampson, R.D.  (editor), Chemical Kinetics Data Survey VI.  Photochemical
and Rate Data for Twelve Gas Phase Reactions of Interest for Atmospheric
Chemistry, National Bureau of Standards, NBSIR 73-207, Washington, D.C.,
1973.

Hanst, P.L., Lefohn, A.S. and Gay, B.W. Jr., "Detection of Atmospheric
Pollutants at Parts-Per-Billion Levels by Infrared Spectroscopy," Applied
Spectroscopy, Vol. 27, p. 188  (1973).

Hanst, P.L., Stephens, E.R., Scott, E.R. and Doerr, R.C., "Atmospheric
Ozone-Olefin Reactions," The Franklin Institute, Philadelphia, Pennsylvania
(1958) .

Heicklen, J., "Photochemistry of CH ONO, Reactions of CH 0 Radicals,  Reactions
of CH O  Radicals," Reviews, Pennsylvania State University, University Park
(19731.

Hendry, D.G., Private Communication, Stanford Research Institute, Menlo
Park, California, 1974.

Hidy, G.M. and Brock, J.R., The Dynamics of Aerocolloidal Systems, Vol.  1,
Pergamon Press, New York, 1970, pp. 47-358.

Horvath, H. and Noll, K.E., "The Relationship Between Atmospheric Light
Scattering Coefficient and Visibility,"  Atmospheric Environment, Vol. 3,
p. 543 (1969).

-------
                                                                       171


Husar, R.B., "Coagulation of Knudsen Aerosols," Ph.D.  Thesis,  University
of Minnesota, Minneapolis, 1971.

Intersociety Committee, Methods of Air Sampling and Analysis,  American
Public Health Association, Washington, D.C., 1972,  p.  103.

Kelvin, Lord, "On the Equilibrium of Vapour at a Curved Surface of a
Liquid," Proceedings of the Royal Society, Edinburgh,  Vol.  7,  p. 63 (1870).

Kiang, C.S., Stauffer, D., Mohnen, V.A., Bricard, J. and Vigla, D.,
"Heteromolecular Nucleation Theory Applied to Gas-To-Particle  Conversion,"
Atmospheric Environment, Vol. 7, p. 1 (1973).

Kocmond, W.C., Kittelson, D.B., Yang, Y.H. and Demerjian, K.L., "Determi-
nation of the Formation Mechanisms and Composition of Photochemical Aerosol,"
NC5365-M-1, Calspan Corporation, Buffalo (1973).

Landis, D.A., Goulding, F.S. and Jaklevic, J.M., "Performance  of a Pulsed-
Light Feedback Preamplifier for Semiconductor Detector X-Ray Spectrometers,"
Nuclear Instruments and Methods, Vol. 87, p. 211 (1970).

Langstroth, C.O. and Gillespie, T., "Coagulation and Surface Losses in
Disperse Systems in Still and Turbulent Air," Canadian Journal of Research,
Vol.. 25B, p. 455 (1947).

Leighton, P.A., Photochemistry of Air Pollution, Academic Press, New York,
1961, pp. 158-183.

Lewis, T.R., Amdur, M.O., Fritzhand, M.D. and Campbell, K.I.,  Toxicology
of Atmospheric Sulfur Dioxide Decay Products, Environmental Protection
Agency, Research Triangle Park, North Carolina, 1972,  pp. 1-17.

List, R.J., Smithsonian Meteorological Tables, Smithsonian Institution
Press, Publication Number 4014, Washington, D.C., 1949.

Liu, B.Y.H., Whitby, K.T. and Pui, D.Y.H., "A Portable Electrical Aerosol
Analyzer for Size Distribution Measurement of Submicron Aerosols," Presented
at the 66th Annual Meeting of the Air Pollution Control Association, Chicago
(1973).

Mason, B.J., The Physics of Clouds, Clarendon Press, Oxford, 1971, pp. 1-30.

Maxwell, J.C., "Theory of the Wet-Bulb Thermometer," Collected Scientific
Papers, Cambridge, Vol. 11, p. 625  (1890).

Meloy Laboratories Incorporated, "Operation, Maintenance and Parts Manual
for Portable S02 Gas Calibrator, Model CS-10," Springfield, Virginia, 1972.

Mirabel, P. and Katz, J.L., "Binary Homogeneous Nucleation as  a Mechanism
for the Formation of Aerosols," The Journal of Chemical Physics, Vol. 60,
p. 1138  (1974).

Niki, H., Private Communication, Ford Motor Company, 1974.

-------
                                                                       172


Niki, H., Daby, E.E. and Weinstock, B., "Mechanism of Smog Reactions,"
Photochemical Smog and Ozone Reactions, American Chemical Society, Wash-
ington, D.C., 1972, pp. 27-34.

Nolan, P.J., "The Photoelectric Nucleus Counter," Proceedings of the
Royal Society, Dublin, Series A, Vol. 4, p. 161 (1972).

O'Keeffe, A.E. and Ortman, G.C., "Primary Standards for Trace Gas Analysis,"
Analytical Chemistry, Vol. 38, p. 760  (1966).

O'Neal, H.E. and Blumstein, C., "A New Mechanism for Gas Phase Ozone-
Olefin Reactions," International Journal of Chemical Kinetics, Vol. V,
p. 397 (1973).
Payne, W.A., Stief, L.J. and Davis, D.D., "A Kinetics Study of the Reaction
of HO  with SO  and NO," Journal of the American Chemical Society, Vol. 95,
p. 7614  (1973).
Prager, M.J., Stephens, E.R. and Scott, W.E., "Aerosol Formation from
Gaseous Air Pollutants," Industrial and Engineering Chemistry, Vol. 52,
p. 521  (1960).

Pressman, J., Arin, L.M. and Warneck, P., "Mechanisms for Removal of CO
from the Atmosphere," GCA-TR-70-6-G, GCA Corporation, Bedford, Massachusetts
(1970).

Rasmussen, R.A., "Terpenes:  Their Analysis and Fate in the Atmosphere,"
Ph.D. Thesis, Washington University, Saint Louis, Missouri, 1964.

Reiss, H., "The Kinetics of Phase Transitions in Binary Systems," The
Journal of Chemical Physics, Vol. 18, p. 840 (1950).

Ripperton, L.A., Jeffries, H.E. and White, 0.,  "Formation of Aerosols by
Reaction of Ozone with Selected Hydrocarbons,"  Photochemical Smog and
Ozone Reactions, American Chemical Society, Washington, D.C., 1972,
pp. 219-231.

Sahni, D.C., "The Effect of a Black Sphere on the Flux Distribution in an
Infinite Moderator," Journal of Nuclear Engineering, Vol. 20, p. 915 (1966).

Scott, W.E., Stephens, E.R., Hanst, P.L. and Doerr, R.C., "Further Develop-
ments in the Chemistry of the Atmosphere," Proceedings of the American
Petroleum Institute, Vol. 37, p. 171 (1957).

Service, J., "A Users Guide to the Statistical  Analysis System," Student
Supply Stores, North Carolina State University, Raleigh  (1972).

Shirnov, V.I., The Rate of Aerosol Particle Growth by Coagulation and
Condensation, Gidrometeorozdat, Moscow, 1969.

Smith, W.S. and Gruber, C.W., Atmospheric Emissions from Coal Combustion -
An Inventory Guide, Publication Number 999-AP-24, U.S. Department of Health,
Education and Welfare, Washington (1966).

-------
                                                                       173

Smoluchowski, M. von, "Versuch einer Mathematischen Theorie der
Koagulationskinetik Kolloider Lo'sungen," Zeitschrift fuer Physikalische,
Chemie, Vol. 92, p. 129 (1917).

Stedman, D.H., Morris, E.D. Jr., Daby, E.E.,  Niki,  H.  and Weinstock,  B.,
"The Role of OH Radicals in Photochemical Smog," Presented at the 160th
Annual American Chemical Society National Meeting,  Chicago (1970).

Stedman, D.H., Wu, C.H. and Niki, H., "Kinetics of  Gas-Phase Reactions of
Ozone with Some Olefins," The Journal of Physical Chemistry, Vol. 77,
p. 2511 (1973).                      '

Stuhl, F., "Determination of Rate Constants for the Reactions of OH with
Propylene and Ethylene by a Pulsed Photolysis - Resonance Fluoresence
Method," Berichte der Bunsengesellschaft fuer Physikalische Chemie, Vol.  77,
p. 674  (1973).

Tejada, S., "Determination of Soluble Sulfates from CVS Diluted Exhausts:
An Automated Method," Environmental Protection Agency, Research Triangle
Park, North Carolina, in preparation.

Thermo Systems Incorporated, "Operating and Service Manual - Model  3030
Electrical Aerosol Size Analyzer," Saint Paul, Minnesota, 1974.

Van DeVate, J.F., "The Thickness of the Stagnant Air Layer in Aerosol
Containments and the Aerodynamic Diameter of Aggregates of Small Spheres,"
Journal of Colloid and Interface Science, Vol. 41,  p.  194 (1972).

Vbraski, T. and Cvetanovi6, R.J., "Relative Rates of Reaction of Ozone
with Olefins in the Vapor Phase," and "A Study of the Products of the
Reactions of Ozone with Olefins in the Vapor Phase  as Determined by Gas-
Liquid Chromatography," Canadian Journal of Chemistry, Vol. 38, pp. 1053
and 1063  (1960).

Walter, H., "Coagulation and Size Distribution of Condensation Aerosols,"
Aerosol Science, Vol. 4, p. 1  (1973).

Washida, N., Martinez, R.I. and Bayes, K.D.,  "The Oxidation of Formyl
Radicals," Reviews, University of California, Los Angeles (1973).

Wei, Y.K. and Cvetanovic, R.J., "A Study of the Vapor Phase Reaction of
Ozone with Olefins in the Presence and Absence of Molecular Oxygen,"
Canadian Journal of Chemistry, Vol. 41, p. 913  (1963).

Weinstock, B., Chemical Reactions in Urban Atmospheres,  (C.S. Tuesday, ed.),
American Elsevier, New York, 1971, p. 54.

Whitby, K.T. and Clark, W.E.,  "Electrical Aerosol Particle Counting and
Size Distribution Measuring System for the 0.015 to 1 u Size Range,"
Tellus, Vol. 18, p. 573 (1966).

Williamson, D.G., Private Communication, California Polytechnic State
University, San Luis Obispo, California, 1974.

-------
                                                                       174
Wilson, W.E. Jr., Merryman, E.L., Levy, A. and Taliaferro,  H.R.,  "Aerosol
Formation in Photochemical Smog I.  Effect of Stirring," Journal  of the
Air Pollution Control Association, Vol. 21, p. 128 (1971).

Wood, W.P., Castleman, A.W. Jr. and Tang, I.N., "Mechanisms of Aerosol
Formation from Sulfur Dioxide," Presented at the 67th Annual Meeting of
the Air Pollution Control Association, Denver (1974) .

3M Company, "Scotchpak" Data Sheet, St. Paul, Minnesota.

-------
            Appendix 1:  Product Analysis by Fourier Transform Infrared Spectroscopy







Run                    RH  (%)   CO  (ppm)    HCHO (ppm)    CH CHO (ppm)   CH2CO  (ppm)    HCOOH (ppm)




 28   C2H4 + 03          22        .84          4.14          0              0             Trace





 29   C«,H,, + O., + SO.,    20      1.09          4.06          0              0             .19
 21
 79
 81





 82
:3»6 + °3
:3H6 + °3
:3H6 + o3 + so2
:3H6 + °3
;3H6 + °3
:3H6 +. 03 + so2
:,H, + o, + so.,
36 3 2
3H6 + °3 in N2
:3H6 + 03 in N2
19
19
19
19
33
20
36

36
20
175
.93
.94
.56
.70
.71
.79

.73
.87
2
3
3
1
2
2
2

2
1
.12
.21
.44
.77
.27
.82
.97

.23
.94
1
1
1

1
1
1

1
1
.25
.49
.59
.99
.36
.43
.70

.30
.12
. 18 Trace
.20 Trace
0 Trace
.32 Trace
. 34 Trace
0 Trace
0 Trace

.32 0
.22 0
                                                                                                            01

-------
                                                                 176
Appendix 2:  Observed and Smoothed Reactant Concentrations
             for Gas Phase Experiments.
             Smoothed data are listed through the first half
             of the 'OBS1 sequence followed by the observed
             concentration values.  The time derivatives of
             the concentration values are indicated by the
             1DOT'  notation.

-------

OBS
1
2
3
4
5
6
7
8
9
10
11
12
13
15
16
17
IB
19
20
21
22
23
24

TIME
0.0
1.5
3.0
4.5
6.0
9.5
14.5
20.5
26.5
35.0
43.5
52.0
0.0
1.5
3.0
4.5
6.0
9.5
14.5
20.5
26.5
35.0
43.5
52.0

HC
8.12507
7.662u5
7.23351
6.83947
6.47991
5.77509
5.16662
4.73827
4.46861
4.20761
4.06259
4.04746
7.70000
6.77300
5.79800
5.10900
4.77300
4.45400
4.20200
4.09200
4.03400

HCOOT
-.320182
-.297188
-.274194
-.251199
-.228205
-.174552
-.111191
-.067231
-.040845
-.024701
-.009421
.005861






RUN It
03
2.88215
2.54341
2.23673
1.96211
1.72000
1.32i55
0.95053
0.66120
0.45569
0.27253
0.15747
0.12115
2.55000
2.21000
2.COOOO
1.70000
1.28000
0.95500
0.66200
0.44700
0.27300
O.X7200
0.11400
21.
03DOT
-.236510
-.215139
-.193769
-.172398
-.149031
-.iOSiBO
-.066193
-.043283
-.030780
-.018168
-.008905
.000358







S02
0
0
0
0
0
0
0
0
0
0
0
0







S02DOT
0
0
0
0
0
0
0
0
0
0
0
0







FLAG
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0

OLX03
23.4177
19.4878
16.1794
13.4198
11.1454
7.6320
4.9110
3.1329
2.0363
1.1467
0.6397
0.4903







-------
_.RUN_»_25.__

OBS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
3<»
35
36

TIME
0.0
1.5
3.0
4.5
6.0
7.5
n.o
20.0
2R.O
3P.O
47.0
57.0
66.0
75.0.
«3.0
91.0
100.0
109.0
0.0
1.5
3.0
4.5
6.0
7.5
13.0
20.0
28.0
38.0
47.0
57.0
66.0
75.0
83.0
91.0
100.0
109.0

HC
.87786
.78501
.69609
.61109
.53001
.45285
1.20348
0.96989
0.79430
0.63495
0.53740
0.45650
0.40226
0.35848
0.33503
0.31216
0.28179
0.24754
1.87600

1.53900
1.18700
0.97500
0.78800
0.64000
0.53*00
0.45900
0.40300
0.35700
0.33300
0.31800
0.27700
0.24900

HCOOT
-.0632034
-.0605892
-.0579751
-.0553610
-.0527469
-.0501327
-.0405476
-.0301656
-.0203730
-.0139678
-.0100405
-.0073793
-.0054750
-.0039788
-.0034759
-.0031598
-.0035897
-.0040197










03
2.37699
2.26547
2.16267
?. 06858
1.99261
1.92047
1.70577
1.50335
1.3?309
1.15703
1.06069
0.98858
0.92370
0.85641
0.80746
0.76169
0.71554
0.67506
2.27000
2.15000
2.08000
1.98000
1.70000
1.50000
1.33000
1.15000
1.06000
0.99600
0.92000
0.85700
. 0.80700
0.76300
0.71500

03DOT
-.0772514
-.0714405
-.0656295
-.0598186
-.0496536
-.0465367
-.0360289
-.0272264
-.0201304
-.0139203 .
-.0102874
-.0077714
-.0071519
-.0067102
-.0060014
-.0054422
-.0048130
-.0041839










SO 2
.582743
.574805
.567062
.559513
.552157
.544996
.518966
.494454
.475969
„ ,463097_
.454454
. .446463
.439819
.43?438
.424536
.417090
.409335
..402307

.566000
.561000
.552000
.520000
.493000
.476000
.461000
.455000
.446000
.440000
.433000
.424000
.417000 _
.410000
.4Q?000

S0200T FLAC
-.00535633
-.00522694
-.00509755
-.00496815
-.00483876
-.00470937
-.00418712
-.00306539
-.00204479
-.00124745
-.00093915
-.00080307
-.00083868
-.00088?37
-.00090368
-.00090212
-.00082131
-.00074049
(
(
(
<
(
(
(
(
(
(
(
(
(
(
(
(
(
(

> OLX03
4.46365
4.04389
3.66808
3.33?67
3.04871
2.79015
2.05286
1.45809
1.05093
0.73466
0.57002
0.45129
0.37157
0.30700
0.27053
0.23777
0.20163
0.16711
3
J
)
J
)
)
)
)
)
J
)
)
)
)
)
) •-
) w
)

-------


DBS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
IB
19
20
21
22
23
24
25
26
27
28
29
30
31
32


TIME
0.0
1.5
3.0
6.0
10.0
17.0
24.0
32.0
39.0
47.0
55.0
64.0
71.0
80.0
91.0
0.0
1.5
3.0
4.5
6.0
10.0
17.0
24.0
32.0
39.0
47.0
55.0
64.0
71.0
80.0
91.0


HC
3.54659
3.42474
3.30908
3.19961
3.39634
2.85124
2.52829
2.39025
2.28256
2.20528
2.13821
2.08635
2.34329
2.01263
1.99044
1.9U425
3.46000

2.76800
2.54200
2.40900
2.25600
2.23200
2.12800
2.07900
2.06VOO
2.00000
1.98000
1.99000


HCDOT
-.0833005
-.0791711
-.0750416
-.0709122
-.0667827
-.0557708
-.0365000
-.0191743
-.0133804
-.0104328
-.0068144
-.0364363
-.0045340
-.0033217
-.0016095
.0004831









R
03
1.07984
0.98602
0.89997
0.82167
0.75113
0.61791
0.47074
0.36625
0.28003
0.22113
0.16859
0.13032
0.10151
0.08298
0.06345
0.04652
0.98000
0.91800
0.72900
0.61300
0.46900
0.36700
0.28000
0.22200
0.16800
0. 13000
0.13200
0.08400
0.06200
0.04700

UN * 26.
03DOT
-.0651314
-.0599585
-.0547357
-.0496128
-.0444400
-.0320097
-.0185722
-.3135722
-.0099257
-.0077414
-.0057763 .
-.0042370
-.0030956
-.0024547
-.0018860
-.0011908










502
.620152
.610093
.600614
.591718
.583402
.565275
.539787
.521412
.508221
.500380
.491768
.484739
.479418
.475815
.471807
.467689
.602000
.593000
.583000
.565000
.540000
.521000
.503000
.501000
.492000
.484000
.480000
.476000
.471000
.468000


S02D3T
-.00690340
-.00651270
-.00612500
-.03573731
-.00534961
-.33411949
-.00315637
-.00225174
-.00153991
-.00118088
-.00093265
-.00076722
-.00060213
-.00047723
-.00041340
-.00033539










FLAG
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
3
0
0
0
0
3
0
0
0
0
0
0
3
0
0
0


OLX03
3.82976
3.37638
2.97806
2.62902
2.32575
1.76182
1.19018
0.87544
3.63919
0.48765
0.36047
0.27189
0.20792
0.16700
0.12S28
0.09231









-------
RUN # 27.
DBS TIME
1 O.C
2 1.5
3 3.0
4 4. 5
5 6.0
6 7.5
7 13.0
0 19. U
9 25. C
10 3i.O
U 39.0
12 44.0
13 49.0
14 58. u
15 68. C
16 76.0
17 84.0
18 93.0
19 101.0
20 C.O
21 1.5
22 3.0
23 4. 5
24 6.0
25 7.5
26 1J.O
27 19.0
28 25.0
29 21.C
30 39.0
21 44.0
32 4S.O
33 58.0
34 68.0
35 76.0
36 84.0
37 7j 7}









Co
O

-------

1
OBS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
2'J
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40

TIKE
0.0
1.5
3.0
6.0
11.5
17. C
22.5
27.5
32.5
37.5
42.5
47.5
52.5
57.5
65.0
7C.O
75. C
80.0
85.0
95.0
J.C
1.5
3iO
6.0
11.5
17.0
22.5
27.5
32.5
37.5
42.5
47.5
52.5
57.5
65.0
70.0
75.0
80. C
85. C
95.0

HC
11.8498
11.6368
11.4344
11.0613
10.5001
1C.C644
9.6343
9.2907
9.0227
8.8315
8.5778
8.3489
8. 1824
8.C6S7
7.8996
7.7715
7.6382
7.5139
7.4188
7.2931
11 .8500
11.4500
11.0300
1C.51CO
10.0500
9.6710
9.251U
9.0340
fi. 857C
8.572C
8.3280
8.2C6C
8.0430
7.9210
7.772C
7 .6090
7.5550
7.3920
7.2970

HCCOT
-.14557C
-.133494
-.131418
- . 117266
-.093963
-. 081193
-.070649
-.058653
-. C51840
-.046160
-.043700
-.039880
-.030425
-.024483
-.025875
-.023407
-. 025500
-.021179
-.016873
-.008261










PUN *
03
3.80672
3.67754
3.55317
3.31887
2.93931
2.63336
2.37533
2.18218
1.99914
1 .82686
1.66971
1. 54457
1.42567
1.31398
1.16381
1.09662
1.03477
0.95780
0.89021
0.76726
3.57000
3.29000
2.95000
2. 63COO
2.37000
2.19000
2.00COO
1.82000
1.68COO
1.53000
1.44000
1. 31000
1.16000
1.10000
1.04000
0.95600
0.88500
0.7690C
28.
03DOT
-.0877223
-.0845148
-.0813074
-.0748925
-.0631319
-.0511641
-.C43S306
-.0389075
-.0350000
-.0328000
-.0282000
-. 0252COO
-.0232458
-.0204060
-.0164605
-.0147936
-.0138800
-.0139278
-.0131113
-.0114784











SO 2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0











SO 200 T
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0











FLAG
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

OLX03
45.1090
42.7947
40.6282
36.7111
30. 8629
26.5032
22.8845
20.2740
18.0577
16.1340
14.3225
12.8955
11. 6654
10.6035
9.1937
8.5224
7.9038
7.1968
6.6042
5.5957








1— »
oo


-------

DBS
1
2
3
4
5
6
7
8
9
13
11
12
13
14
15
16
17
13
19
20
21
22
23
24
25
26
27
28
29
30
31
32

TIME

2.5
5. C
10.0
15.0
20.0
25.0
30.0
35. C
40.0
45.0
53.5
58.5
£4. C
77.5
87.0
C.O
• 2.5
5.0
10. 0
15.0
20.0
25.0
30.0
35.0
40.0
45. C
S3. 5
53.5
64. C
77.5
87.0

hC
12.5020
11.8036
11 .2398
10.7103
10.2431
9. 8461
9.5693
9. J521
9. 0576
8.8549
8.6362
8.3576
8.1707
8. 04 30
7.7354
7.5384
12. 68CC
11.5400
11.2000
10.72CO
10.2500
9.8200
S.5760
9 .3590
9.0880
8. /620
8.7350
8.274C
8 .2200
8.0160
7.7310
7.5420

HCOOT
-.306241
-.252446
-. 198650
-. C99454
-.082960
-.067920
-. C55700
-.05208U
-. 045580
-.045447
-.038772
-.033168
-.032937
-.023632
-.021487
-.019979








R
03
3.61723
3.42806
3.24861
2.91891
2.62600
2.37486
2.1500C
1.94086
1. 76000
1.61748
1.47818
1.25973
1.14786
1.06217
0.36160
0.74162
3.42000
3.26000
2.92000
2. 620CO
2.38000
2- 15000
1.94000
1 .76000
1. 61000
1.49000
1.25000
1. IbOOO
1 .05000
0. 87700
C.7350C
UN * 29.
Q3DOT
-.0776162
-.0737240
-.0698319
-.0620476
-.0552000
-.0486000
-. 0432000
-.03860^0
-.033UOOO
-.C295763
-.0267066
-.0230118
-.0205612
-.0161656
-. 0135495
- .0117086









SQ2
.529393
.500223
.473445
.427065
.391029
.362486
.339829
. 320343
.303571
.289401
.276559
.259623
.251110 _
.243495
.231950
.229914
.501000
.472000
.428000
..390000 .
.363000
.3400J9
.320000
.304000
.289000
.277000 __
.259000
.252000
.243000
.232000


S 02 DOT
-.0121465
-.0111896
-.0102328
-.0083190
-.0065800
-.0053200
-.0043000
-.0036800
-.0031400
-.0027146
-.0023525
-.0018238
-.U015378
-.0012313
-.. 0004 791.
.0000503









FLAG





1
1
1
1
1
1
0
0
0
0
0
0
o
0
0
0
0
0
0
0
0
0

ni *m
45-???*,
40.4635
^A. 5136
31.2625
76.R904
23.3832
70.5740
18.1512
15.9413
14.3226
12.7659
10.5284
9.37B9
8.5398
6.6648
5.5907









-------
RUM * 31.
OBS
1
2
3
4
5
6
7
8
9
10
11
12
13
15
16
17
18
19
20
21
22
23
24
TIME
0.0
1.5
3.0
4.5
5.5
10.5
18.5
24. 0
32.0
37.5
46.0
55.0
0.0
1.5
3.0
4.5
5.5
10.5
18.5
24. 0
32.0
37.5
46.0
55.0
HC
8.55502
8.05260
7.58708
7.15846
6.89321
5.81298
5.08467
4.83663
4.62060
4.52596
4.43847
4.45258
8.62200

6.77000
5.81200
5.06200
4.87700
4.6230U
4. 502 JO
4.47800
4.43800
HCOOT
-.347250
-.322649
-.298048
-.273448
-.257047
-. 175045
-.070146
-.0463U6
-.022449
-.016054
-.004532
.007667






03
2.71353
2.37742
2.07558
1.8080O
1.66658
1.11775
0.60891
0.41815
0.24538
0.17306
0.09287
0.06894
2.38000
2.07000
1.81000
1.65000
1.08000
0.61900
0.42UOO
0.24500
0.17300
0.10200
0.06500
0300T
-.235492
-.212649
-. 189807
-.166964
-. 143902
-.092676
-.051840
-.032987
-.018640
-.012724
-.006143
.000825







•
•
•
•
•
•
•
•
*
•
•
•
•
•
•
•
•
•
•
S02
524980
482373
452138
430607
429385
397888
368151
358121
351134
347638
341452
335170
526000
480000
453000
436000
428000
397000
368000
358000
352000
347000
342000
335000
S02DOT
-.0325282
-.0242803
-.0160334
-.0114937
-.0080189
-.0052269
-.0028558
-.0017553
-.0009143
-.0007421
-.0007133
-.0006826






FLAG
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
OLX03
23.
19.
15.
12.
11.
6.
3.
2.
1.
0.
0.
3.






2143
1445
7476
9425
4881
4974
0961
0224
1338
7833
4122
3070







-------
                                                RUN  ft 86.
UbS
        TIKE
                                HCCJT
                                                        U3DUT
SO 2
SG2DOT
FLAG
CLXC3
1
2
j
4
c
e
1C
11
12
13
Is
It
lo
17
le

2C
22
23
24
26
2b
3u
31
33
3s
Ji
Jo
27
Jd
0.0
1.5
s . 5
6 .0
11. C
l7 .u
23.0
2S.C
37. C
47.0
Sf.C
o5 .c
73. C
81. o
S7.C

C'.C
l.L
1.5
3.C
c.G
1 l.L
17. c
23. C
2S. C
37.o
•J7.C
65.0
73. C
o 1 .C
69. C
S7.C
11S.L
.921666
.901666
.462203
.63311.6
.71 19D4
ibuiJSc
• 4s J t' 13
. _>c t i;ic
. Jo looo
IloIuJJ
.15SC.1S


.9CSCCC

.77tuCO
.7iCCCC
.5150CC
.43SCCC
• 3 / 1 u 0 0
• 
-------
RUN « 87.
CBS TIME
1 0.0
2 1.5
3 3.0
4 4.5
5 8.0
6 16.0
7 24.0
8 32.0
9 41.0
10 50.0
11 59.0
12 67.0
13 76.0
14 65. 0
15 93.0
16 0.0
17 i.5
18 3.0
19 4.5
20 8.0
21 16. U
22 24.0
23 32.0
24 41.0
25 50.0
26 59.0
27 67.0
28 76.0
29 35. J
30 S3.0
HC
1.1116C
1.07261
1.03446
0.99716
0.9134J
C. 7392C
0.57677
0.47589
0. 39184
0.32611
0.27222
0.23874
0.20769
0. 17911
0.15748
1.101CO

0.937CO
0.732CO
0.57500
0.471CO
0.40100
0. 319CG
0.275CO
0.240CO
0.2C4CO
0.185l'0
0. 1550C
HCCOT
-.0262750
-.0257125
-.0251500
-.0245875
-.0232750
-.0202750
- .0165553
-.0121465
-.CC86349
-.0066120
-.0053507
-.0039449
-.0034257
-.0029255
-.0024809








03
2.05569
1.98461
1.91883
1.85834
1.75180
1.56696
1.44498
1.35889
1.27473
1.19481
1.13102
1.G8507
1 .03339
0.98602
0. 94687
1.99000
1.91000
1.86000
1. 74000
1.57000
1.44000
1.36000
1.28000
1.19000
1. 13000
1.09000
1.03000
0.98500
0. 94800
03DOT
-.0491465
-.0456182
-.0423900
-. 0385618
-.0282022
-.0197362
-.0140006
-.0110185
-.OJ9 1)150
-.0077733
-.0068353
-.0058903
-.0054581
-.0050675 •
-.0047203








SO 2
.650706
.644627
.638706
.632941
.620101
.593566
.573175
.556781
.542474
.530812
.521174
.514488
.507768
.502908
.500089

.638000
.635000
.618000
.595000
.572000
.557000
.543000
.530000
.522000
.514000
.508000
.503000
.500000
S02DOT
-.00410461
-.00399995
-.00389530
-.00379064
-.00354644
-.00298760
-.00231829
-.00191739
-.00145676
-.00120446
-.00097612
-.00078915
-.00063919
-.00044070
-.00026427








FLAG
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
OLX03
2.28510
2.12B72
1.98496
1.85307
1.60010
1. 15830
0.83342
0.64668
0.49950
0.38964
0.30788
0.25905
0.21462
0. 17660
0.14911









.
CD
cn

-------

K
OBS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32


TIME
0.0
1.5
3.0
4.5
6.0
7.0
13.0
22.0
29.0
36.0
45.0
53.0
61.0
69.0
77.0
83.0
0.0
1.5
3.0
4.5
6.0
7.0
13.0
22.0
29.0
36.0
45.0
53.0
61.0
69.0
77.0
83.0


HC
3.28134
3.23994
3.20084
3.16405
3.12957
3.10786
2.99912
2.93428
2.87909
2.84561
2.81450
2.79193
2.78071
2.78130
2.78685
2.79382
3.30600

3.05500
3.01100
2.94200
2.86900
2.84400
2.82500
2.78000
2.78500
2.78500
2.77500
2.80000


HCDOT
-.0283691
-.0268321
-.0252952
-.0237582
-.0222213
-.0211966
-.0150488
-.007S727
-.0060033
-.0049590
-.0028560
-.0018513
-.0011875
.0004272
.0009615
.0013623








R

03
.273524
.256070
.239875
.224940
.211383
.202266
.158285
.111033
.085744
.065424
.045584
.033364
.024829
.018804
.014230
.011904
.256000
.240000
.225000
.211000
.203000
.156000
.112000
.086000
.065000
.046000
.033000
.025000
.019000
.014000
.012000
UN * 88.

0300T
-.0120562
-.0112163
-.0103764
-.0095365
-.0090037
-.0087549
-.0065635
-.0045476
-.0034246
-.0026655
-.0018921
-.0013545
-.0009750
-.0006769
-.0004666
-.000308M










S02
.207480
.206?20
.205001
.203P23
.202686
.201950
.197918
.193106
.190705
.188952
.187600
.186*13
.185400
.184195
.183066
.181085


.203000
.202000
.197000
.194000
.190000
.189000
.188000
.186000
.186000
.184000
.183000
.182000


S0200T
-.000853374
-.000826182
-.000793991
-.000771800
-.000744609
-.000726481
-.000617717
-.000432187
-.000293443
-.000245*21
-.000135639
-.000146571
-.000150000
-.000148579
-.000133674
-.000209996










FLAG
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0


OLX03
.897525
.879650
.767803
.711724
.661539
.628613
.474718
.325802
.246864
.1*6172
.12829S
.091149
.069041
.052300
.039657
.033258









-------
RUN # 89.
OBS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
IS
16
17
18
19
20
21
22
23
24
25
26
TIME
0.0
1.5
3.0
4.5
7.0
13.0
19.0
26.0
34.0
44.0
52.0
63.0
74.0
0.0
1.5
3.0
4.5
7.0
13.0
19.0
26.0
34.0
44.0
52.0
63.0
74.0
HC
3.31782
3.26072
3.20828
3.16049
3.09116
2.98867
2.94850
2.90455
2.85921
2.80017
2.79586
2.77920
2.76977
3.33000
3.19000
3.09000
2.99000
2.93300
2.92800
2.84300
2.81000
2.78100
2.80000
2.76200
HCOOT
-.0396135
-.0365132
-.0334128
-.0303124
-.0251452
-.0133228
-.0087489
-.0059588
-.0051314
-.0035170
-.0018419
-.001 1860
-.0005301







03
.294753
.275311
.257096
.240105
.215418
.167647
.131815
. 100686
.073722
.050283
.036865
.023264
.016354
.276000
.255000
.242000
.214000
.167000
.132000
.100000
.074000
.050000
.037000
.024000
.016000
03DOT
-.0133694
-.0125524
-.0117354
-.0109183
-.0093294
-.0071878
-.0054665
-.0040996
-.0030091
-.0020703
-.3015407
-.0009323
-.0003240







S02
.637251
.634696
.632234
.629864
.626121
.618188
.611975
.605966
.600173
.593636
.589276
.585133
.582906

.630000
.626000
.618000
.612030
.606000
.600000
.594000
.589000
.585000
.583000
S0200T
-.00173412
-.00167233
-.00161053
-.0015'4'B'7T"
-.00144575
-.00119857
-.OOC39351
-.00081444
-.00070402
-.00057796
-.00046381
-.00(128957
-.00011534







FLAG
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
OLX03
.977936
.897714
.824834
.758849
.665892
.501041
.388657
.292448
.210785
.140800
.103070
.964655
.045297








-------
J?U N_JL 9 0.	
OBS
1
2
3
4
5
6
7
8
9
10
11
12
13
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
35
36
TIME
o.o
1.5
3.0
4.5
15.0
23.0
27.0
31.0
39.0
47.0
55.0
69.0
7«.0
Sfl.O
95.0
103.0
111.0
0.0
1.5
3.0
4.5
3.0
15.0
23.0
27.0
31.0
39.0
47.0
55.0
69.0
78.0
88.0
95.0
103.0
111.0
HC
2.73851
2.62523
2.51674
2.41301
2.18956
1.R2065
1.54386
1.42899
1.35060
1.21466
1.10789
1.02832
0.91759
0.86769
0.81524
0.78448
0.75779
0.73939
2.74300

2.18200
1.81600
1.54000
1.36200
1.20300
1.11700
1.02200
0.92000
0.86800
0.81800
0.78000
0.76000
0.739QO
HCOOT
-.0771067
-.0739234
-.0707401
-.0675569
-.0601292
-.0452739
-.0314440
-.0259903
-.0216875
-.0160125
-.0125021
-.0093736
-.0068246
-.0057272
-.0046008
-.0038541
-.0028179
-.0017816









03
1.97049
1.87248
1.78022
1.69372
1.52317
1.247P9
1.01499
0.91?48
0.82547
0.69851
0.61114
0.53590
0.42505
0.36562
0.31411
0.27848
0.24373
0.21338
1.87000
1.78000
1.70000
1.51000
1.25000
1.02000
0.91400
0.81700
0.69900
0.60900
0.53800
0.42400
0.36600
0.31300
0.28300
0.23700
0.21600^
0300T
-.0672579
-.0634219
-.0595859
-.0557499
-.0*64426
-.0351251
-.0273047
-.0232000
-.0204154
-.0141097
-.0108605
-.0089872
-.0070730
-.0059651
-.0052988
-.0046182
-.0040686
-.0035191









S02
.580315
.570500
.561140
.552235
.533223
.50P065
.479342
.471RA6
.465237
.454214
.444872
.436136
.420877
.411129
.399639
.391147
.382920
.375522

.559000
.555000
.533000
.50?000
.480000
.47?000
.465000
.454000
.445000
.436000
.421000
.411000
.400000
.391000
.3fl?000
..376000
S02DOT
-.00669454
-.00639150
-.00608846
-.00578542
-.00507832
-.00379175
-.00234598
-.00198750
-.00163486
-.00131166
-.00117401
-.00110370
-.00109513
-.00113158
-.00115905
-.00108024
-.00097653
-.00087282









FLAG
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
OLX03
5.39619
4.91569
4.48035
4.08697
3.33507
2.27197
1.56700
1.30392
1.11488
0.84845
0.67708
0.55108
0.39002
0.31725
0.25607
0.21846
0.18470
0.15777








oo
oo

-------
RUN « 91.

DBS
1
2
3
i*
1
6
7
8
9
10
11
12
13
l
-------
RUN #92.

UBS
1
2
3
5
6
7
b
9
10
11
12
13
14
15
lo
17
18
19
20
21
22
23
24
25
2c
27
20
29
30
31
32
33
34
TIME
0.0
1.5
3.0
6.0
11.0
17. u
19.9
24.0
30.0
39.0
53.0
60.0
6U.C
77.0
86.0
94. C
108. C
0.0
1.5
3.0
6.0
11.0
17.0
19.9
2
-------
RUN » 93.

OBS TIPE
1 U.O
2 1.5
3 3.0
4 6.0
5 b.C
6 14.2
7 16 .0
e 2^.0
9 32.0
1C 41. t
11 49.0
12 58.0
13 67. C
14 76.0
15 . 86.0
It 94. U
17 112.0
Id C.C
IS 1.5
20 3.0
21 6.C
22 8.C
23 14.2
24 It.C
25 24.0
26 32.0
27 il.O
28 49.0
29 sa.o
30 o7.(.
31 76. C
32 66. C
33 St.C
3t 112.0
HC
2.753CO
2.663Q5
2. g 7 5-59
2.vl051
2.3066J
2.C1704
1.94220
1.65715
1.-6415
1.31642
1.2C939
1.14/73
I.C7674
1.C1874
0.9676U
0.93a97
0.90064
2.7*600

. 2.329CO
1.917CO
l.t71CO
1 .45800
1.3C200
1.16700
l.L70tO
l.COiCC
C.942CO
HCDOT
-.C6C925C
-.0590031
-.0570S13
-.0532375
-.050o75C
-.C427313
-. 0^.04250
-.0311619
-.0219354
-.0154697
-.0112881
-.0091785
-.0070786
-.CC58107
-.OC44J20
-.0032751
-.C007394








03
2.57259
2.42836
2.29375
2.05338
1.93120
1. 53751
1.54181
1.25198
1.03362
0.84093
0.72247
0.61362
0.51900
0.44773
C. 37565
0.32065
0.25754
2.42000
2.31000
2.04000
1 .92000
1.53000
1.26000
1.C300C
O.U4200
0.717CC
0.61500
0.44400
0.33300
C.2560C
03DOT S02 S02DOT
-.C993578
-.0929463
-.0865347
-.C737115
-.0642510
-.0466190
-.0424233
-.0328215
-.0250017
-.C13791C
-.0143150
-.C114303
-.0095944
-.0079461
-.CC64676
-.0052832
-.0026133








0
0
0
0
0
0
C
0
0
C
0
0
0
0
0
0
0








0
0
0
0
0
0
0
0
0
0
0
0 •
0
0
- o""
0
0








FLAG
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
CLXC3
7.08235
6.46686
5.90869
4.94971
4.45451
3.20206
2.99451
2.07471
1.51338
I. 10702
0.87374
0.70427
0.55883
0.45612
0.36351
0.30794
0.23200







H^

-------
FUN
      94.
OBS 1 I^t
1 0.0
2 1.5
3 3.0
4 5.0
5 11. C
o 12 .0
7 17.0
8 24. C
9 32 .u
1C 41. C
11 -,'J.U
12 57.0
13 71.0
Is 79.0
15 87. C
lo 94.0
17 " 102. u
10 C.C
19 1 .t
20 3.C
21 5.C
22 11. 0
23 12. C
2-. 17. o
25 24.0
2t> 32.0
27 41.0
28 -,9.0
29 57.C
30 71.0
31 79.0
32 ti/.C
33 94.0
34 IC2.C
HC
2.S2957
<:. 61015
2.55138
2.11/32
1 .b75t3
1.3J7JO
1 .2 Jc.,03
i . i -j 7 1 1,
1 . 01.028
1 .00705
C.97^79
0.9550;-

2.570CC
2.12500
1.9C800
1 .'to 6 JO
1 .24500
1 .15000
C.9UCOO
0.95SOO
HCDJT 03
-.0813174 2.08281
-.0779099 2.52o69
-.L 7<. 5025 2.38047
-.U69-jf.92 2.20094
-.0^6329^ 1.79184
-.0^«-J57U 1.732^-1
-.'02^234 U21V52
-.021J206 0.99821
-.Ul4jft70 0.82002
-.0119761 C. 09432
-.GJ90542 0.59111
-.0062152 O.s5375
-.0045929 0.39237
-.0015511 0.30230
.OOC0713 0.2o799
2.5400C
2.37000
2.19000
1.78000
1.49000
1.2100C
1.00000
O.U2300
O.C92CO
0.59100
0.45500
0.39200
0.34100
0.30200
U3D'JT
-.107338
-.1007UO
-.094171
-.0853oO
-.OoJo9o
-. 058176
-.046-590
-.C25269
-.018362
-. 014026
-.012051
-.007217
-.UO5950
-. 004893
-.003o84







•
SO 2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0








SO 2 DOT
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0








FLAG
I
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
J
0
0
r.'LX03
7.U5949
7.10037
t.-1738
5.e.l544
3.H9280
3.*6tlJ6
?'.o5's5°
I. J9636
O.Uf 620
0.68401
(i. *7.^157
0. 39r-13
0.33235
0.29050
0.25394







i — •
— 	 . _ — is 	

-------
KUN # 95.
08S
1
2
3
5
6
7
a
10
11
12
14
15
lo
17
18
20
21
22
23
24
25
26
27
20
29
30
TIME
0.0
1.5
J.O
b.d
9.0
16.0
40.0
4b.O
56.0
64.0
72.0
80.0
86.2
0.0
1.5
3.0
4.5
6.U
9.0
lo .O
24.0
40.0
48.0
56.0
64.0
72.0
dO.O
86.2

3,
2.
2.
2.
2.
2.
2.
1.
1.
1.
1.
1.
1.
1.
1.
3.

2.
2.
1.
1.
1.
1.
1.
	 U.
HC
02490
93108
84039
75284
50901
18532
92o78
59480
46080
41877
37562
33550
31134
30361
02«tOO

52400
154QO
92100
59300
42000
36100
33oOO
31000

HCDOT
-.0635928
-.0615022
-.0594115
-.0573209
-.0541152
-.0510489
-.0412V25
-.0304795
-.0159087
-.0110915
-.0061819
-.OOoJl37
-.0040172
-.0020207
-.0004734









1
1
1
1
1
1
1
0
0
o
0
0
0
0
0
1
1
1
1
1
0
0
0
0
0
0
0
c

•
•
•
•
•
•
•
•
•
•
•
•
*
*
•
•
*

•
•
*
*
•
•
•
•
•
03
81263
71052
61396
52294
39415
20979
01tS2
79480
51492
42433
35126
29291
245H6
20522
18002
71000
61000
53000
28000
01000
79000
52100
42oOO
35100
29300
24oOO
20500
leooo
P3DOT
-.0699221
-.0662244
-.0625267
-.0588291
-.0531593
-.0477266
-.0357746
-.0252097
-.0140306
-.0109594
-.0085375
-.0068375
-.0055 3<: 1
-.0044754
-.0036550








S02
.558233
.550147
.542290
.534661
.523406
.512349
.485605
.460639
.429429
.420580
.416343
.410486
.404343
.397914
.392736
.549000
.543000
.536000
.512000
.484000
.462000
.428000
.422000
.416000
.411000
.404000
.398000

S0203T FLAG OLXOS
-.00546656 1
-.00531432 1
-.00516208
-.00500984
-.00477640
-.00467407
-.00364176
-.00271643
-.00140179
-.00099844
-.00073750
-. 0007?000
-.00078571
-.0008 21 '+3
-.00084911 1
5.48302
I 5.01367
4.58427
4.19241
3.65922
3.23609
2.22696
1.53140
0.82120
0.63171
0.49835
0.40294
0.32754
0.26911
0.23467
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

-------
RUN # 97.

UdS
1
<:
3
4
5
6
/
0
9
10
11
12
1 j
1*
16
17
Id
19
20
21
22
23
24
25
67
2.t>99f2
2.52573
2.4122S
2.32294
 C 0 C
2.1 /UOO
2. 144CC

HCDOT
-.0750C11
-.0626476
-.0502141
-.C377806
-.C253471
-.0142708
-.013^027
-.0143035
-.0145299
-.0137691
-.0134352
-.0126737
-.ClC80<-9
-.G1UJ725
-.CU97140
-.00'J9002
-.0074063
-.OC69B79
-.0066059
-.00otf2<:3
-.CC63659
-.0073328
-.CC59027
-.0084550
-.O04'j?l«
-.0042000
-.0110000
-. Cl 8cOO(,
-.0262030










C3
2.152"26
2.06753
2.03436
2.00274
1.97268
l.V2.t02
I.db3t8
1.82963
1.79353
1. 76975
1. 74839
1. 70574
1.63103
1.55542
1.52U75
1.5Cc60
1.44159
1.41457
1. 39098
1.3C327
i. 31336
1.2B-.98
1.24979
1.22762
1.21360
1. 194b6
1.17343
I. 14914
1.122CO

2.00000
1.98000
1.92000
1.U2JCO

1.71000
1.63000
~l.r£OCO~~
1.50000
1.42000

03 DOT
-.0236669
-.0226331
-.C215973
-.0205615
-.0195257
-.0177994
-.0160731
-.0128996
-.0119756
-.0110515
-.0103123
-.0091372
-.0084096
-.0077=07
-.0074919
-.0074949
-.0069521
-.0068182
-.0066568
-.0071580
-.OU80991
-,OOV729tt
-.0075222
-.OC62198
-.0072200
-.0060000
-.0091429
-.0102857
-.0114206










S02
' ".568442
.581254
.574567
.568379
.562691
.554321
.548496
.537719
.531948
.5P6491
. 521997
.508731
.489047
.473652
.469393
.465277
.453992
.44975*
.446410
.441629
.434289
.429909
.424899
.420206
.417C36
.413929
.410882
.407897
.404974

.570000
.560000
.555000
.548000
.538000
.532000
.526000
.510000
.488000
.474000
.47CCCC
.465000
.454000
.450000
. itf.ono

S02DOT
-".00^95912"
-.00462497
-.00429182
-.00395867
-.00362552
-.00307027
-.00235090
-.00236143
-.00226723
-.00224564
-.00224862
-.00219786
-.00183737
-.001^0323
-.00138295
-.00132240
-.001 14125
-.00104998
-.00104764
-.00109156
-.00118283
-.00125961
-.00128811
-.00128005
-.00125544
-.00123084
-.00120623
-.00110162
-.00115702










FLAG
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
0
-Q-
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

OLX03
6.54282
6.22117
5.94915
5.72' 51
5.54518
5.39344
5.21136
4. 93666
4.78903
4.6^855
4.54*78
4.30824
3.93452
3.M315
3.49777
3.4005^
3. 13232
3.U3869
2.9*663
2.B?907
2.6B751
2.60290
2.49474
2.^2883
2.36757
2.32741
2.276*5
2. 16682
2.07233




i— i
vo
£*.




-------

LES lift
I C.C
2 1.3
3 3.C
4 <«.3
5 o.C
o 13. C
7 21.0

-------
                                                                 196
Appendix 3:  Aerosol Droplet Number, Surface Area, and Volume

             Concentrations Distributions for Aerosol Experiments.



             DP             Geometric Mean Diameter

             AMP            Electrometer Current Sensed

             DIFFERENCE     Current Difference

             DX                dx
                            d log D
                                   P

             DX                ,
             r                dx
                            x d log D

-------
101A
OP
.0042
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4220
.7500



AMP
0.0
9.2800
9.1700
8.5900
4.6900
0.8140
0.0360
0.0
0.0
0.0
0.0


5
DIFFERENCE
0.0
0.396E 06
0.620E 06
0.119E 07
0.3B2E 06
0.371E 05
0.84BE 03
0.0
0.0
0.0




ON
0.
1582248.
2480079.
4776716.
1529624.
148598.
3393.
0.
0.
0.

N
2630164.00

OS
0.0
280.06
1378.92
8407.02
8550.60
2630.18
188.63
0.0
0.0
0.0

s
5358.85

0V
0.0
0.3SO
3.050
33.294
60.114
32.840
4.173
0.0
0.0
0.0

V
33.46 •

~HDN7Nf
0.0
a. 6516
0.9429
1.8161
0.5816
0.0565
0.0313
0.0
' 0.0
0.0




OS/5-
0.0
"0.0523
0.2573
1.5688
1.5956
0.4908
0.0352
0.0
0.0
0.0




OV/V
0.0
0.0105
0.0912
0.9952
1.7969
0.9816
0.1247
O.D
0.0
0.0




-------
10
DP
.0042
.0075
.0133
.0237"
.0*22
.0750
.1330
.2370
.4220
.7500



I1A
' AMP
0.0
0.0
6.9500
6.8500
6.1900
2.8600
0.6510
0.0700
0.0020
0.0
0.0


15
DIFFERENCE
"0.0
0.0
0.107E 06
0.202E 06 '-
0.329E 06
0.105E 06
0.137E 05
0.922E 03
0.156E 02
0.0 ' " "



ON "
0.
0.
427602.
808368.
1314151.
421919.
54753.
3688.
62. "
0.

	 "N " 	
757635.25

"•• " os
0.0
O.D
	 237.75
1422.73
7346.10
7467.96
3344.29
.... .... 649>14 	
34.80
0.0

5
5050.69
— ov --
0.0
o.o
- 0.526
5.634
51.646
	 93.244
67.347
~*- 25.708
2.446
0.0

V
61.64
DN/N
0.0
o.o -
0.5644
1.0670
1.7345
0.5569
-" 0.0723
0.0049
0.0001
0.0


OS/S
0.0
0.0
0.0471
0.2817
1.4545
1.4786
0.6027
0.1285
0.0069 '
0.0



DV/V
0.0
0. 3
0.0085
0.3914
0.8379 " '
1.5128
1.0926
0.4171
0.0397
0.0



VO
00

-------
101A
OP
.0042
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4220
.7500
AMP
0.0
0.0
5.5800
5.5200
5.4700
4.6400
1.9390
0.3240
0.0720
0.0210
45
DIFFERENCE
0.0
0.0
0.641E 05
0.1 53E 05
0.819E 05
0.129E 06
0.380E 05
0.342E 04
0.397E 03
0.669E 02

ON
0.
0.
256562.
61240.
327551.
515891.
152197.
13668.
1587.
268.

OS
0.0
0.0
142.65
107.78
1831.01
9131.27
8*62.18
2405.65
887.31
473.76

DV
0.0
0.0
0.316
0.427
12.873
~114.012 "~"
187.203
95.269
62.382
58.885

-OWN" * -
0.0
0.0
0.7722
0.1843
' 0.9859
1.5528
0.4581
0.0411
0.0048
0.0308

DS/S
0.0
0.0
0.0243
0.0184
0.3124
1.5581
1.4440
0.4105
0.1514
0.0808

DV/V
0.0
0.0
Q.3024
0.0032
0.0969
0.8583
1.4092
0.7172
0.4696
0.4433
0.0060






N
332240.50
S
5860.40
V
132.84






ID

-------
101A 75
DP AMP DIFFERENCE
.0042 0.0 0.0
.0075 4.7500 0.180E 06
.0133 4.7000 0. 107E 05
.0237 4.6900 0.612E 04
.0422 4.6700 0.138E 05
.0750 4.5300 0.965E 05
.1330 2.5100 0.47BE 05
.2370 0.4800 0.525E 04
.4220 0.0930 0.521E 03
.7500 0.0260 O.S03E 02
0.0080









ON OS DV DN/N DS/S OV/V
0. 0.0 0.0 0.0 0.0 0.0
719203. 127.30 0.159 1.9947 0.0216 3.0010
42761. 23.78 0.053 0.1186 0.0041 0.0003
24497. 43.11 0.171 0.0679 0.0074 0.0011
55249. ~ 308.84 2.171 0.1532 0.0528 0.0140
385820. " 6829.01 " 85.266 1.0701 1.1675 0.5483
191307. 10636.67 235.308 0.5306 1.8184 1.5131
20991. ~ 3694.39 146.306 0.0582 0.6316 0.9408
2085. 1165.69 81.953 0.0358 0.1993 0.5270
321. 568.51 70.662 0.0009 0.0972 0.4544
N S V
360558.00 5849.32 155. 51




t\j
0
0

-------
1018
DP
.0042
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4220
.7500



AMP
0.0
7.0300
6.9100
5.9000
1.B700
0.1550
0.0
0.0
0.0
0.0
0.0


5
DIFFERENCE
0.0
0.432E 06
0.108E 07
0.123E 07
0.169E 06
0.740E 04
0.0
0.0
0.0
0.0



DN
o.
1726078.
4318760.
4935940.
676807.
29605.
0.
0.
0.
0.

N
2921797.00

OS
0.0
305.52
2401.23
8687.25
3783.35
524.01
0.0
0.0
0.0
0.0

s
3925.34

0V
0.0
0.381
5.312
34.403
26.599
6.543
0.0
0.0
0.0
0.0

V
18.31

DN/N
0.0
0.5908
1.4781
1.6894
0.2316
0.0101
0.0
0.0
D.O
0.0




" "DS/S"
0.0
0.0778
0.6117
2.2131
0.9638
0.1335
0.0
0.0
0.0
0.0




DV/V
O.O
0.0208
0.2901
1.8790
1.4527
0.3573
O.D
0.0
0.0
0.0




-------
1018
DP
.0042
.0075
.0133
.0237
-.0422
.0750
.1330
.2370
.4220
.7500




' " ftHP
0.0
6.3100
6.2900 ~
6.2500
5.5200
2.4600
0.5220
0.0520
0.0
0.0
0.0


16
DIFFERENCE
0.0
0.719E 05
0.428E 05
0.224E 06
0.302E 06
" 0.925E 05
0.1 HE 05
0.705E 03
0.0
0.0




ON
0.
287673;
171040.'
894105.
~ 1207598. 	
370158.
44293.
2820. "
0. '" "
0.

N
744421.00

DS
0.0
-50.92 - -
95.10
15/3.62
	 6750.47 	
6551.79
Z4&Z.68
496.40
0.0
O.D

S
4495.24

- 'DV
O.D
0.064
0.210
6.Z3Z
47.459
81.805
!>4.<>BU
	 19.659
0.0
0.0

"v 	
52.48

'DN/N 	
0.0
0.3864
0.2296
1.2011
1.6222
0.4972
0.0595
"0.0038
~o.o -
0.0




DS/S
0.0
0.0113
0.0212
0.3501
1.5017 ~
1.4575
0.5478
0.1104
0.0
0.0




DV/V
0.0
0.0012
0.0040
0.1188
0.9044
1.5589 '
1.0382
0.3746
O.D
0.0




-------


•
•
•
•
•
•
•
•
•
•



101B
OP
0042
0075
0133
0237
0422
0750
1330
2370
4220
7500





40

AMP DIFFERENCE
0
5
5
5
5
4
1
0
0
0
0


.0
.3900
.3700
.3400
.3200
.3000
.5840
.2700
.0590.
.0150
.0030


0.0
0.719E
0.321E
0.612E
0.101E
0.130E
0.310E
0.286E
0. 342E
0.535E



05 ""
05
04
06
06
05
04
03
02




ON
0.
267673.
128283.
24495.
402533.
518756.
123831.
11445.
1369.
214.

N
374649.25

OS
0.0
50.92
71.33
43.11
2250.16
9181.97
6885.02
2014.26
765.53
379.01

5
5410.32

0V
0.0
0.064 '"•
0.158
0.171
15.820
114.645
152.312
79.769
53.820
47.108

V
115.97

ON/N
0.0
0.7678
0.3424
0.0654
1.0744
1.3846
0.3305
0.0305
0.0037
0.0306




DS/S
0.0
0.0094
0.0132
O.OOBO
0.4159
1.6971
1.2726
0.3723
0.1415
0.3701




OV7V"
0.0
o.ooo*
0.0014
0.0015
0.1364
0.9886
1.3134
0.6879
9.4641
0.4062




-------
1018
OP
.0042
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4223
.7500



AMP
0.0
4.5600
4.5400
4.5100
" 4.4900
4.3300
2.4700
0.4920
0.0910
0.0240
0.0090


75

" DIFFERENCE
0.0
0.719E
0.321E
0.612E
0.158E
" 0.888E
0.466E
0.544E
0.521E
0.669E



05
05
04
05
05
05
04
03
02




"W
0.
287673.
128283.
24495.
63142.
355260.
186407.
21750. "-
; ' 2085.
268.

N
267J4U.50

OS
0.
50.
71.
43.
352.
6288.


0
92
33
11
97 ~
10
10364. 2O
3828.
1165.
473.

s
5659
04
68
76


.52

DV
0.0
0.064
0.158
0.171
2.481
78.512
229.280
* 151. 5 99
81.952
SB.BBb

V
150.78
" DNXN
0.0
1.0761
0.4798
0.0916
0.2362
1.3289
0.6973
" 0.0814
0.0078'
O.oOlO



OS/S
0.
0.
' 0.
"0.
0.
1.
1.
0.
0.
0.

0
0090
0126
0076
0624
1111
8313
6764
2060
0837



DV/V
0.0
0.0034
0.0010
0.0011
0.0165
0.5207
1.52D7
1.0055
0.5435
0.3905




-------
102
OP
.0042
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4220
.7500










" AMP
0.0
6.9900
6.8400
5.3000
1.5500
0.0100
0.0
0.0
0.0
0.0
0.0








5
DIFFERENCE ON DS DV DN/N DS/S DV/V
0.0 0. 0.0 0.0 0.0 0.0 0.0
0.539E 06 2157608. 381.90 0.477 0.6189 0.0982 0.0294
0.165E 07 6585036. 3661.28 8.100 1.8388 0.9413 0.4992
0.115E 07 4593000. 8083.68 32.013 1.3174 2.0783 1.9732
0.152E 06 607745. 3397.29 23.884 0.1743 0.8735 1.4722
0.477E 03 1910. 33.81 0.422 0.0005 0.0087 0.025.0
0.0 0. 0.0 0.0 0.0 0.0 0.0
0.0 0. 0.0 0.0 0.0 0.0 0.0
0.0 0. 0.0 0.0 0.0 0.0 O.O
0.0 0. 0.0 0.0 0.0 O.O 0.0

N S V
3486324.00 3889.49 16.22




to
o
en

-------
102
OP
.0042
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4220
.7500




" AMP"
~0~.0~
5.9700
5.9400
5.9200 "
5.1500
2.1100
D.3600"
0.0
0.0
0.0
0.0 "


16
DIFFERENCE
"' 0.0
0.108E 06
0.214E 05
0.236E 06
0.300E 06
0.836E 05
0.848E 04
0.0
0.0
0.0




DN " " *
0.
431516. '
"" 85522.
943095.
" 1199705.
334250;
33926.
0.
0.
0.

N
757003.25

OS
0.3
76.38
" 47.55 	
1659.85
6706.35 "•
5916.22
18B6. 31
0.0
0.0
O.O

s
4073.16

- ov-
0.0
0.095
0.105
6.573
47.148 '
73.869
41.729
0.0
0.0
0.0

•V" 	 ' '
42.38

ON/N
0.0
0.5703
0.1130
1.2458
1.5848
0.4415
0.0448
0.0
0.0"
0.0




DS/S
0.0
0.0188
0.0117
0.4075
1.6465
1.4525
0.4631
0.0
- o.o "
0.0




OV/V
0.0
0.0023
0.0025 "
0.1551
1.1125
1.7430
0.9846
0.0
0.0
0.0




-------
102
OP AMP
.0042 0.0
.0075 4.6800
.0133 4.6500
.0237 4.6300
.0422 4.5900
.0750 3.7200
.1330 1.3400
.2370 0.1400
.4220 0.0
.7500 0.0
0.0







45
DIFFERENCE ON OS
0.0 0. 0.0
0.108E 06 431516. 76.38
0.214E 05 85522. 47.55
0.122E OS 48992. 86.23
0.858E 05 343337. 1919.25
0.114E 06 454580. 8046.06
0.283E 05 113088. 6287.68
0.190E 04 7594. 1336.47
0.0 0. 0.0
0.0 0. 0.0

N S
371156.50 4449.91






DV
0.0
0.095
0.105
0.341
13.493
100.462
139.098
52.927
0.0
0.0

V
76.63






DN/N
0.0
1. 1626"
0.2304
0.1320
0.9250 "
" 1.2248
0.3047
0.0205
0.0
0.0









DS'/S"
0.0
0.0172
0.0107 "
0.0194
0.4313
1.8081
1.4130
0.3003
0.0
0.0









OV/V
0.0
0.0012
0.0014
0.0045
0.1761
1.3110
1.8152
0.6907
0.0
0.0







NJ
	 r-»

-------
102
OP AMP
.00*2 0.0
.0075 3.9000
.0133 3.8100
.0237 3.0000
.0422 3.7900
.0750 3.4900
.1330 1.7400
.2370 0.2400
.4220 0.0
.7500 0.0
0.0








75
DIFFERENCE ON OS DV ON/N OS/S DV/V
0.0 0. 0.0 0.0 0.0 0.0 0.0
0.324E 06 1294562. 229.14 0.286 2.6466 0.0539 0.0033
0.107E 05 42761. 23.78 0.053 0.0874 0.0056 0.0006
0.306E 04 12247. 21.55 0.085 0.0250 0.0051 O.OOlb
0.296E 05 118392. "661.81' 4.653 0.2420 0.1557 0.0542
0.836E 05 334250. 5916.22 73.869 0.6833 1.3918 0.8601
0.353E 05 141360. 7859.61 173.873 0.2890 1.8490 2.0244
0.325E 04 13018. 2291.10 90.733 0.0266 0.5390 1.0564
0.0 0. 0.0 0.0 0.0 0.0 0.0 " "
0.0 0. O.O 0.0 0.0 0.0 0.0

N 5 V
489146.75 4250.80 85.89





o
oo

-------
103
DP
.0042
.0075
.Oi33
.0237
.0422
.0750
.1330
.2370
.4220
.7500



AMP 1
0.0
5.4500
5.3900
3.5400
0.7070
0.0080
0.0
0.0
0.0
0.0
" 0.0'

5
HFFERENCE
0.0
0.216E 06
0.198E 07
0.867E 06
0.690E 05
0.382E 03
"0.0
0.0
0.0
0.0



DN
0.
863046.
7910596.
3469857.
275853.
1528.
0.
0.
0.
0.

" N

	 os
0.0
152.76
" 4398.29
6106.95
1542.02
27.05
0.0
0.0
0.0
0.0

5

ov
0.0
0.191
9.730
24.185
10.841
0.338
0.0
0.0
0.0
0.0

V

~tm/N '
0.0
T>.275T~
2.527T
1.1085
0.0881
"0.0005
0.0
0.0
o; o" "
0.0



DS/S
0.0
"" 0.0500
1.4389
1.9978
0.5045
0.0088
0.0
0.0
0.0
0.0



OV/V
0.0
0.0168
0.8595
2.13*3
0.9576
0.0298
O.D
0.0
0.0 '
0.0


3056.76
                  11.32
                                                                       K)
                                                                       to

-------
103
OP
.0042
.0075
.0133
.0237
" .0422
.0750
.1330
.2370
.4220
.7500




"AMP 	
0.0
0.0
4.4100
4.3700
3.2000
1.0750
0.1470
0.0
0.0
0.0" "•
0.0


15
DIFFERENCE
0.0
0.0
0.428E 05
0.358E 06
0.210E 06
0.443E 05
0.346E 04 "•"""
0.0
0.0
o.o 	




• oft' 	
0.
o. 	
171040.
1433016.
838610.
177248.
13853.
0.
o. "•••
o.

*
658441. 25

DS
0.0
0.0
95.10 "
2522.11
	 4687.83 "'•
•--•- 3137.29
770.24
0.0
0.0
0.0

s - -- 	
ZB03.14

DV
0.0
0.0
0.210
9.988
32.957
" 39.172
17.040
0.0
0.0
0.0

-y -
24.84

" ON/N
0.0
o.o 	
0.2598
1.2736
0.2692
0.0213
0.0
0.0
0.0




os/s
0.0
0.0 "
0.0339
0.8997
1.6723
1.1192
0.2749
0.0
0.0
0.0




DV/V
0.0
0.0
0.0085
0.4021
1.3267
1.5769
0.6859
0.0
0.0
0.0




-------
103
OP
.00*2
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4220
.7500




45
AMP DIFFERENCE
0.0
3.5000
3.4400
3.4200
3.3000
2.2800"
0.7000
0.1100
0.0170
0.0
0.0

0.0
0.216E 06
0.214E 05
0.3676 05
0.101E 06
0.754E 05
0.139E 05
0.1 26E 04
0.132E 03
0.0


ON
0.
863046.
85522.
146976.
402533.
301780.
55602.
5044.
529.
0.

N
465257.00

OS
0.
152.
47.
25B.
2250.
5341.
3091.
887.
295.
0.

S
3081


0
76
55
68
16
50
45
80
77
0


.41

DV
0.0
0.191
0.105
1.024
15.820
66.693
68.390
35.159
20.794
0.0

V
52.04


-OKI7S- -
0.
l'
0.
0.
0.
0.
0.
0.
0.
0.


0
8550 "
1838
3159
8652 "
6486
1195
0108
0011 -
0




	 DS/S
0.0
0.0496
0.0154
0.0839
0.7302
1.7335
1.0033
0.2881
0.0960
0.0





DV/V
0.
o ;
0.
t).
0.
1.
1.
0.
0.
0.



0
0037
0020
0197
3040
2815
3141
6756
3995
0




-------
103
OP
.0042
.0075
.0133
.0237
'.0422'
.0750
.1330
.2370
.4220
.7500



AMP
0.0
0.0
2.7000
2.6900
"" 2.6800
2.2400
0.8970
0.1500
' 0.0270
0.0030
0.0


75
" OIFFEREF
0.0
0.0
0.107E
0.306E
0.434E
'" 0.641E
0.1 76E
0.167E
0.187E
0.134E


JCE

05
04
05
05
05
04
03
02


ON
0.
o-
42761.
12248.
173641;-
256513. 	
70397.
6672.
747.
54.

N
140758.13

OS
0.



DV
0
' 0.0
23.
21.
970.
— 4540.
3914.
1174.
417.
94.

S
Z789
78
56
66
28 	
09
19
56
75


.21
0
0
" ~0
0
..... 6
	 56
86
46
29
11

V

.0
.0
.053
.085
.824
.689
.589
.500
.356
.777


59.47

rON/N '
o.o
0.0
0.3038 "•
0.0870
1.2336
1.8224
0.5001
0.0474
0.0053
0.0004




OS/S
0.0
0.0
0.0085
0.0077
0.3480
1.6278
1.4033
0.4210
0.1497
0.0340




OV/V
0.0
0.0
0.0009
0.0014
0.1148
0.9533 ~ ~ 	
1.4560
0.7819
0.4936
0.1980




-------
104
OP
.0042
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4220
.7500



AMP
0.0
10.1700
10.1000
9.9200
6.7500
1.9490
0.2410
0.0020
0.0
0.0
0.0


5

DIFFERENCE ON
0.0
0.252E 06
0.192E 06
0.971E 06
0.474E 06
0.816E 05
0.563E 04
0.271E 02
"0.0 " ""
0.0



0.
1006876.
769661.
3882614.
1894666.
326228.
22523.
108.
0.
0.

N
1975673.00

DS
0.0
178.22
427.94
6833.40
10591.18
5774.23
1252.30
19.09
0.0
0.0

S
6269.09

DV
0.0
0.223
0.947
27.062
74.460
72.096
27.704
0.756
0.0
0.0

V
50.81

DN/N
0.0
0.5096
0.3896
1.9652
0.9590
0.1651
0.0114
0.0001
0.0
0.0




DS/S
0.0
0.0284
0.0683 "
1.0900
1.6B94
0.9211
0.1998
0.0030
--Q-.0
0.0




DV/V
0.0
0.3044
0.0186
0.5326
1.4654
1.4189
0.5452
0.0149
0.0
0.0




-------
10'
OP
.0042
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4220
.7500



it
AMP
0.0
0.0
9.0800
8. 8900
8.6200
5.5600
1.5400
" 0.2570"
0.0460
0.0050
o.o


15
"DIFFERENCE
0.0
0.0
0.203E 06
0.827E 05
0.302E 06
0.192E 06
0.302E Ob
0.286E 04
"' 0.319E 03 "
0.223E 02




ON
0.
0.
812442.
330695.
1207598.
767820.
120910.
11445.
1276.
89.

N
813068.00

DS
0.0
0.0
"•*51.7Z"
582.02
6750.47
13590.41
6722.59
2014.26
713.33
157.92

S
7745.68

- - ovr 	
0.0
0.0
0.999
2.305
47.459
169.688
148.719
" 79.769
50,150
19.628

V
129.68

ON/N
0.0
0.0
0.9992
0.4067
1.4852
0.9443
0. 1487
0.0141
0.0016
0.0001




DS/S
0.0
0.0
0.0583
0.0751
0.8715
1.7546
0.8679
~ " ' 0.2600
0.0921"
0.0204




OV/V
0.0
0.0
0.0077
0.0178
0.3660
1.3085
1.1468
0.6151
0.3867
0.1514




-------
104
OP
.0042
.0075
.0133
.0237
.0*22
.0750
.1330
.2370
.*220
.7500




AMP
0.0
0.0
6.9200
6.9100
6.9000
6.5*00
3.6700
0.7650
0.1520
0.0*80
0.0200


45
DIFFERENCE
0.0
0.0
0.107E 05
0.306E 04
0.355E 05
0. 137E 06
0.684E 05
0.831E 0*
0.809E 03
0.125E 03




ON
0.
0.
42757.
12248.
142070.
548170.
273767.
33249.
3237.
500.

N
263999.00

OS
0.0
0.0
23.77
21.56
794.17
9702.61
15221.44
5851.84
1809.42
884.35

S
8577.29

DV
0.0
0.0
0.053
0.085
5.583
121.146
336.733
231.746
127.210
109.919

V
233.12

DN7tf -
0.0
0.0 ~
0.1620
0.0464
0.5381
2.0764 '
1.0370
0.1259
0.0123
0.0019




"DS/S
0.0
"0.0 "
0.0028" "
0.0025
0.0926
1.1312 "
1.7746
0.6822
0.2110
0.1031




DV/V ~
0.0
0.0
0.0002
0.0004
0.0240
0.5197
1.4445
0.99*1
0.5*57
O.*715




-------
104
OP
.0042
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4220
.7500


AMP "
0.0
5.8200
"' 5.7600
5.7500
5.7300
5.6200
3.8500
0.8820
0.1550
0.0400
0.0150

75
OIFFEREI
0.0
0.216E
0.107E
0.612E
0. 109E
0.845E
0.699E
0.986E
0.895E
0.112E

YCE

06
05
04 -
05
05
05 '
04
03
03

- ON '
o.
863046.
42757.
24497.
43410.
338070.
279704.
39432.
3579i"
446.

N

OS
0
_...- — -152
23
43
	 242
5983
15551
6940
2000
789

5


.0
.76
.77
.66
.84 	
.55
.11
.80
.60



DV
0.0
0.191
0.053
0.171
1.706
74.713
344.036
"274.844
140.665
98.142

v

01
0
2
0
0
0
o
0
o
0

N/N
.0
.1115
.1046 "
.0599
.1062 "
.8271
.6843
.0965
.0088
.0011
	 - • 	
os/s
0.0
0.0193
0.0030
0.0054
0.0306
0.7544
1.9606
0.8749
0.2522
0.0995

OV/V
0.0
0.0008
0.0002
0.3097
0.0073
0.3198
1.4726
1.1764
"0.6021
0.4201



-------
105
DP
.0042
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4220
.7500

" AMP ' "
0.0
10.4800
10.4700
10.0700
6.3600
1.6150
0.1570
0.0
0.0
0.0
5
DIFFERENCE
0.0
0.360E 05
0.428E 06
0.114E 07
0.468E 06
0.696E 05
0.370E 04
0.0
" 0.0
0.0

ON
0.
143843.
1710398.
4544008.
1872566.
278478.
14796.
0.
0.
0.

OS
0.0
25.46
950.98
7997.45
10467.64
4929.05
822.64
0.0
0.0
0.0

0V
0.0
0.032
2.104
31.672
73.592
61.544
18.199
0.0
O.D
0.0

DN/N
0.0
0.0672
0.7989
2.1224
0.8746
0.1301
0.0069
D.O
D.O
0.0

DS/S
0.0
0.0040
0.1510
1.2698
" " 1.6620
0.7826
0.1306
0.0
0.0
0.0

OV/V
0.0
0.0007
0.0450
0.6770
1.5730
1.3154
0.3890
0.0
0.0
0.0
0.0






N
2141021.00
s
6298.30
V
46.79







-------
105
DP
.0042
.0075
.0133
.0237
.0422
.0750
.1330
.2370
.4220
.7500



AMP
"(T.O
0.0
8.7800
8.5800
8.4200
5.4600
"1.6300
0.2790
0.05*0
~ 0.0080
0.0

15
DIFFERENCE
0.0
0.0
0.214E
0.490E
0.292E
0.183E
0.318E
0.305E
0.358E
0.357E

06
05
06
06
05
04
03
02

ON
0.
0.
855199.
195969.
1168134.
731530.
127318.
12204."
1432.
143.

N

DS
0.0
	 ovo" •
475.49
344.91
" 6529.87
12948.07
7378.89
2147.90
800.32
252.67

S

DV
0.0
0.0
1.052
1.366
	 45.908
'" 161.668
156.601
85.062
56.266
31.405

V


'" ON/N
0
	 o
1
0
1
" 0
0
— o
0
0


.0
.0
.1064
.2535
.5112
.9464
.1647
.0158
.0019
.0002



OS/S
0.0
• o.o
0.0622
0.0451
0.8542
1.6938
0.9260
0.2810
0.1047
0.0331



OV/V
0.
0.
0.
0
0
0078
0.0131
0.3405
1.1990 ~
1.
0.
0.
1615
6309
4173
0.2329




772981.00
                 7644.53
                                  134.83

-------
105 45
OP AMP DIFFERENCE DN OS
.0042 0.0 0.0 0. 0.0
.0075 0.0 0.0 0. 0.0
.0133 7.2000 0.118E 06 470363. 261.52
.0237 7.0900 0.612E 04 24495. 43.11
.0422 7.U700 0.345E 05 138124. 772.11
.0750 6.7200 0.146E 06 582550. 10311.13
.1330 3.6700 0.696E 05 278479. 15483.43
.2370 0.7150 0.782E 04 31296. 5508.18
.4220 0.1380 0.786E 03 3144. 1757.23
.7500 0.0370 0.937E 02 375. 663.26
0.0160
N 5
382206.00 8699.99







0V DN/N DS/S OV/V
0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0
0.579 1.2307 0.0301 0.0026
O.ltl 0.0641 0.0050 0.0008
5.428 0.3614 0.0887 0.0241
128.744 1.5242 1.1852 0.5712
342.529 0.7286 1.7797 1.5197
218.136 0.0819 0.6331 0.9678
123.540 0.0082 0.2020 0.5481
82.439 0.0010 0.0762 0.3658

V
225o39




to
t—>
O

-------
105
OP AMP
.0042 0.0
.0075 5.6700
.Oi33 5.6400
.0237 5.6000
.0422 5.5900
.0750 5.4700
.1330 4.3400
.2370 1.2930
.4220 0.2610
.7500 0.0690
0.0250


75
DIFFERENCE ON OS 0V DN/N DS/S OV/V " "" "
0.0 0. 0.0 0.0 0.0 0.0 0.9
0.108E 06 431516. '" 76.38 0.095 1.4057 0.0088 0.0003 »
0.428E 05 171040." " ~ 95.10'"" 0.210 0.5572 0.0109 0.0007
0.306E 04 12248. 21.56 O.085 0.0399 0.0025 0.0003
0.118E 05 47357. 264.72 ~ "' ~ 1.861 0.1543 0.0304 0.0062 '"
0.540E 05 215830. 3820.19 "" "47.698 0.7031 0.4388 0.1589
0.718E 05 287149. 15965.50 353.193 0.9354 1.8338 1.1769
0.140E 05 55976. 9851.70 390.150 0.1823 1.1316 1.2996
0.149E 04 5976. ~ 3340.48 " 234.849 0.0195 0.3637 0.7823
0.196E 03 785. 1389.69 172.730 0.0026 0.1596 O.S753

N S V - -
306968.75 8706.32 300.22
. ' ro
NJ
o-

-------
                                                       INSTRUCTIONS

   1.   REPORT NUMBER
        Insert the EPA report number as it appears on the cover of the publication.

   2.   LEAVE BLANK

   3.   RECIPIENTS ACCESSION NUMBER
        Reserved for use by each report recipient.

   4.   TITLE AND SUBTITLE
        Title should indicate  clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, if used, in smaller
        type or otherwise subordinate it to main title, when a report is prepared in more than one volume, repeat the primary title, add volume
        number and include subtitle for the specific title.

   5.   REPORT DATE
        Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of issue, date of
        approval, date of preparation, etc.).

   6.   PERFORMING ORGANIZATION CODE
        Leave blank.                                  .                                                    .            .    .

   7.   AUTHOR(S)
        Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.).  List author's affiliation if it differs from the performing organi-
        zation.                         •

   8.   PERFORMING ORGANIZATION REPORT NUMBER
        Insert if performing organization wishes to assign this number.

   9.   PERFORMING ORGANIZATION NAME AND ADDRESS
        Give name, street, city, state, and  ZIP code. List no more than two levels of an organizational hirearchy.

   10.  PROGRAM ELEMENT NUMBER
        Use the program element number  under which the report was prepared. Subordinate numbers may be included in parentheses.

   11.  CONTRACT/GRANT NUMBER
        Insert contract or grant number under which report was prepared.

   12.  SPONSORING AGENCY NAME AND ADDRESS
        Include ZIP code.

   13.  TYPE OF REPORT AND PERIOD COVERED
        Indicate interim final, etc., and if applicable, dates covered.

   14.  SPONSORING AGENCY CODE
        Leave blank.

   15.  SUPPLEMENTARY  NOTES                                                                                        .
        Enter information not included elsewhere but useful, such as:  Prepared in cooperation with, Translation  of, Presented at conference of.
        To be published in, Supersedes, Supplements, etc.                                '                       • •

   16.  ABSTRACT
        Include a brief (200 words or less) factual summary of the most significant information contained in the report. If the report contains a
        significant bibliography or literature survey, mention it here.    .                    .                         •

   17.  KEY WORDS AND DOCUMENT ANALYSIS
        (a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
        concept of the research and are sufficiently specific and precise to be used as index entries for cataloging.

        (b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc.  Use open-
        ended terms written in descriptor form for those subjects for which no descriptor exists.

        (c) COSATI FIELD GROUP - Field and group assignments are to be taken from the 1965 COSATI Subject Category List. Since the ma-
        jority of documents are multidisciplinary in nature, the Primary Field/Group assignment(s) will be specific discipline, area of human
        endeavor, or type of  physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
        the primary posting(s).

   18.  DISTRIBUTION STATEMENT
        Denote reusability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
        the public, with address ana price. /

   19. & 20. SECURITY CLASSIFICATION
        DO NOT submit classified reports to the  National Technical Information service.

   21.  NUMBER OF PAGES
        Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list,  if any.

   22.  PRICE
        Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Form 2220-1 (9-73) (Reverie)

-------
                                                        INSTRUCTIONS

    1.   REPORT NUMBER
        Insert the EPA report number as it appears on the cover of the publication.

    2.   LEAVE BLANK

    3.   RECIPIENTS ACCESSION NUMBER
        Reserved for use by each report recipient.

    4.   TITLE AND SUBTITLE
        Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, if used, in smaller
        type or otherwise subordinate it to main title. When a report is prepared in more than one volume, repeat the primary title, add volume
        number and include subtitle for the specific title.

    5.   REPORT DATE
        Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of issue, date of
        approval, date of preparation, etc.).

    6.   PERFORMING ORGANIZATION CODE
        Leave blank.                                                                                       ...

    7.   AUTHOR(S)
        Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.).  List author's affiliation if it differs from the performing organi-
        zation.     .                   •

    8.   PERFORMING ORGANIZATION REPORT NUMBER
        Insert if performing organization wishes to assign this number.

    9.   PERFORMING ORGANIZATION NAME AND ADDRESS
        Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hirearchy.

    10.  PROGRAM ELEMENT NUMBER
        Use the program element number under which the report was prepared. Subordinate numbers may be included in parentheses.

    11.  CONTRACT/GRANT NUMBER
        Insert contract or grant number under which report was prepared.

    12.  SPONSORING AGENCY NAME AND ADDRESS
        Include ZIP code.                                                                                          .

    13.  TYPE OF REPORT AND PERIOD COVERED
        Indicate interim final, etc., and if applicable, dates covered.

    14.  SPONSORING AGENCY CODE
        Leave blank.

    15.  SUPPLEMENTARY NOTES
        Enter information not included elsewhere but useful, such as: Prepared in cooperation with, Translation of, Presented at conference of,
        To be published in, Supersedes, Supplements, etc.         .                              '             .  •'

    16.  ABSTRACT                                                                                     ...
        Include a brief (200 words or less) factual summary of the most significant information contained in the report. If .the report contains a
        significant bibliography or literature survey, mention it here.    .                            •      .

    17.  KEY WORDS AND DOCUMENT ANALYSIS                                               ,   '
        (a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
        concept of the research and are sufficiently specific and precise to be used as index entries for cataloging.
        (b) IDENTIFIERS AND OPEN-'
        ended terms written
                      ND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc. Use open-
                      in descriptor form for those subjects for which no descriptor exists.

    (c) COSATI FIELD GROUP - Field and group assignments are to be taken from the 1965 COSATI Subject Category List. Since the ma-
    jority of documents are multidisciplinary in nature, the Primary Field/Group assignment(s) will be specific discipline, area of human
    endeavor, or type of physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
    the primary posting(s).

18.  DISTRIBUTION STATEMENT
    Denote reusability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
    the public, with address and price.  /'

19. & 20.  SECURITY CLASSIFICATION
    DO NOT submit classified reports to the National Technical Information service.

21.  NUMBER OF PAGES
    Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any.

22.  PRICE
    Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Form 2220-1 (9-73) (Reverse)

-------