United States
            Environmental Protection
            Agency
            Environmental Research
            Laboratory
            Duluth MN 55804
EPA-600/3-80-094
December 1980
            Research and Development
v>EPA
Environmental Effects of
Western Coal Combustion:

Part  IV - Chemical and
Physical Characteristics
of Coal Fly Ash

-------
                RESEARCH REPORTING SERIES

Research reports of the Ctfice o< Researcn anc 'Development  U S  Environmental
Protection Agency have been grouped into nine series  "hese nine broad cate-
gories were established to fac  litate further de\/elopmen' and application of en-
vironmental technology  Elimination  ot traditional grouping was  consciously
planned to foster Technology transfei and a n'aximum interface in related fields
The nine series are

      1   Environmental  Health Effects Research
      2   Environmental  Protection Technology
      3   Ecological Research
      4   Environmental  Monitoring
      5   Soc'oeconomic Environmental Stud es
      6   Scientific and Technical Assessment Reports (STAR)
      7   Interagency  Energy-Environment Research and  Development
      8   "Special" Reports
      9   Miscellaneous Reports

This report has been assigned to the ECOLOGICAL RESEARCH series This series
describes research on  the effects of pollution on humans, plant and animal spe-
cies, and materials Problems  are assessed  for their long- and short-term influ-
ences Investigations include formation, transport,  and oathway studies to deter-
mine the fate of pollutants and their effects This work prov des the technical basis
for setting standards to minimize undesirable changes m living organisms in the
aquatic, terrestrial, and atmospheric environments
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                            EPA  600/3-80-094
                                            November
                                                     /
       ENVIRONMENTAL EFFECTS OF WESTERN COAL COMBUSTION:
PART IV - CHEMICAL AND PHYSICAL CHARACTERISTICS OF COAL  FLY  ASH
                              by

            David F. S. Natusch and David R. Taylor
                    Department of Chemistry
                   Colorado State University
                 Fort Collins, Colorado  80523
                       Grant No. R803950
                        Project Officer

                        Donald I. Mount
               Environmental Research Laboratory
                   Duluth, Minnesota  55804
               Environmental Research Laboratory
              Office of Research and Development
             U.S. Environmental Protection Agency
                   Duluth, Minnesota  55804
              F::--. i -., L/'--ry
              2CO v. jr.-, O.v born Stiost
              Cliicago,  Illinois   60604

-------
                                   DISCLAIMER

     This report has been reviewed by the Environmental  Research Laboratory-
Duluth, U.S. Environmental Protection Agency,  and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection  Agency, nor does mention of
trade names or commercial products constitute  endorsement or recommendation
for use.
                   U,S. environments! Protection  ^nc
                                      ii

-------
                                  FOREWORD

     This report, part IV of a series, provides considerable information on
the physical and chemical properties of coal fly ash.  National reliance on
coal for more of our energy will result in large increases in fly ash
production.  To properly dispose of this material and more importantly to
find effective ways to use it as a resource rather than as a waste requires
such knowledge.


                                      Norbert A. Jaworski, Ph.D.
                                      Director
                                      Environmental Research Laboratory-Duluth
                                     ttt

-------
                                  ABSTRACT

     Coal fly ashes from a number of different sources have been extensively
studied.  Morphologically, fly ashes can be classified in terms of as many
as twelve different particle types, which include,  primarily, solid, hollow,
encapsulating, and agglomerated particles.  Both non-magnetic particles,
which consist maninly of silicon, aluminum, calcium, potassium, and sodium,
and magnetic particles, which consist mainly of hematite and magnetite, are
found in all the ashes studied.   The relative proportion of non-magnetic to
magnetic particles varied considerably among coal  types.  The electrical
resistivity of the coal fly ash decreases with increasing alkali metal
content.

     Potentially volatile trace elements, namely.  As, Se, Ga, In, Pb, and
Cd, are almost alsys preferentially associated with the smaller, more
respirable fly ash particles.  Many other minor and trace elements may also
show an inverse particle size dependence, depending on the coal fly ash.
Elements that do not exhibit any dependence of concentration on particle
size include Al, Ba, Ca, Co, Fe, K, Mg, Rb, Sc, Si, Sr, Ta, Ti, and the rare
earths.  Surface analytical techniques indicate that many of the more volatile
and potentially environmentally hazardous trace elements are associated with
the surface layer of coal fly ash.  Thus, the concentration levels of Pb,
Tl, Cr, and Mn on the surface were much higher than in the bulk material.
These elements are readily soluble in aqueous solution in contrst to the
bulk of coal fly ash, which is very insoluble.  It is believed that enrichment
of trace elements results from the volatilization of the elements in the
combustion zone and subsequent condensation and/or adsorption as the
temperature falls in power plant plumes.

     Most potentially hazardous inorganic species associated with coal fly
ash can be leached into aqueous solution but are unlikely to give rise to
solution concentrations of great concern.  On the other hand, localized
concentrations of soluble inorganic species could cause damage to a few lung
cells in the microregion of particle deposition following inhalation.

     The physical and chemical characteristics and behavior of polycyclic
organic matter (POM) associated with coal fly ash has also been studied.
Theoretical models, which have been generally verified by experiment, show
that POMs associated with coal fly ash by adsorption from the vapor phase.
This is a highly temperature-dependent process, although such factors as the
energetics of adsorption, particle surface area, and vapor phase POM concen-
trations all influence the vapor-to-particle adsorption process.  The models
show that the adsorption process will occur within fractions of a second
under the conditions found in a typical power plant emission stream and
within a few seconds in the emitted plume.

                                     iv

-------
     Surface-adsorbed polycyclic organic compounds (mainly pyrene and
benzo[a]pyrene) exposed to different nitrogen and sulfur oxides are shown to
be highly reactive.  Surfaces studied included coal fly ash, alumina, and
silica gel.   Reactions with nitrogen dioxide and sulfur trioxide appear to
proceed most completely, with both reactions yielding several oxidation and
substitution products.  It is concluded that reactions of particulate POM
with plume gases can result in major changes in both the chemical and
toxicological characteristices of this species.  On the other hand, adsorption
of POM onto coal fly ashes result in significant stabilization against
photochemical decomposition, therefy preserving the integrity of the original
compounds against photo-oxidation.  A new non-photochemical oxidation sequence
has, however, been discovered for those hydrocarbons containing a benzylic
carbon atom.

     Overall, it can be stated that coal fly ash collected by electrostatic
precipitators or other high-temperature control devices have essentially no
associated POM.   Fly ash emitted to the atmosphere, however, has much more
POM but not enough to make a discernible impact on any terrestrial aquatic
system.

-------
                                  CONTENTS

                                                                        Page

Foreword	   iii

Abstract	    iv

Figures	   vii

Tables	    xi

Acknowledgment 	  xiii

  1.  Introduction 	     1

  2.  Conclusions  	     3

  3.  Recommendations  	     6

  4.  Physical Characteristics of Coal Fly Ash	     7

          Particle Morphology  	     7
          Mass, Density, and Ferromagnetism Distributions  	    18
          Electrical Resistivity Distribution  	    21
          Surface Area Distribution  	    27

  5.  Chemical Characteristics of Coal Fly Ash	    29

          Matrix Element Composition and Distribution  	    29
          Mineral Composition  	    39
          Trace Element Composition and Distribution 	    43
          Inorganic Chemical Speciation and Leaching Characteristics  .    66
          Surface Characterization 	    77
          Health Effects 	    90
          Particulate Associated Organic Compounds 	    91

Summary	   115

References	   116
                                    vii

-------
Appendices	129

      A.  Experimental Procedures and Instrumental Techniques
          Used to Analyze Coal Fly Ash	130
      B.  Matrix Element Distributions for a Size, Density,
          and Magnetically Fractionated Coal Fly Ash	141
      C.  Trace Element Distributions for a Size, Density,
          and Magnetically Fractionated Coal Fly Ash	150
      D.  Results of Different Coal Fly Ash Leaching Experiments .... 175
      E.  Glossary of Abbreviations  	 200
                                     viii

-------
                               LIST OF FIGURES

Number                                                                  Page

   1    Scanning electron micrographs of representative
        coal fly ash particles	    9

   2    Scanning electron micrographs of coal fly ash particles,
        showing representative encapsulated and encapsulated
        particles and cenosphere shells  	   13

   3    Calculated surface tension at 1400°C as a function
        of the percentage of encapsulating particles 	   17

   4    Particle size distributions resulting from coal
        combustion in different boiler types 	   20

   5    Particle size distribution of emissions from similar
        production units at the same power plant equipped
        with either an electrostatic precipitator (ESP)
        or a Venturi wet scrubber	20

   6    Mass distribution of non-magnetic and magnetic coal
        fly ash fractions as a function of particle size	22

   7    Mass distribution of size- and density-classified
        coal fly ash with and without crushing	25

   8    Resistivity as a function of reciprocal absolute
        temperature for two different coal fly ashes	26

   9    Concentration of silicon in non-magnetic and
        magnetic coal fly ash fractions as a function
        of particle size and density	31

  10    Concentration of aluminum in non-magnetic and magnetic
        coal fly ash fractions as a function of particle
        size and density	32

  11    Concentration of iron in non-magnetic and magnetic
        coal fly ash fractions as a function of particle
        size and density	33
                                    lx

-------
Number                                                                  Page

  12    Concentration of sulfur in non-magnetic and magnetic
        coal fly ash fractions as a function of particle
        size and density	34

  13    Concentration of potassium in non-magnetic and magnetic
        coal fly ash fractions as a function of particle
        size and density	36

  14    Concentration of sodium in non-magnetic and magnetic
        coal fly ash fractions as a function of particle
        size and density	37

  15    Concentration of calcium in non-magnetic and magnetic
        coal fly ash fractions as a function of particle
        size and density	38

  16    X-ray powder diffraction patterns of non-magnetic
        and magnetic coal fly ash particles	41

  17    X-ray powder diffraction patterns of
        density-separated coal fly ash particles	42

  18    Concentration of manganese in non-magnetic and magnetic
        coal fly ash fractions as a function of particle
        size and density	45

  19    Concentration of zinc in non-magnetic and magnetic
        coal fly ash fractions as a function of particle;
        size and density 	  47

  20    Concentration of europium in non-magnetic and magnetic
        coal fly ash fractions as a function of particle
        size and density	48

  21    Concentration of gallium in non-magnetic and magnetic
        coal fly ash fractions as a function of particle
        size and density	49

  22    Concentration of antimony in non-magnetic and magnetic
        coal fly ash fractions as a function of particle
        size and density	,	50

  23    Concentration of thallium in coal fly ash fractions
        as a function of particle size	52

  24    Concentration of antimony in coal fly ash fractions
        as a function of particle size	52
                                      x

-------
Number                                                                  Page

  25    Concentration of beryllium in coal fly ash fractions
        as a function of particle size	53

  26    Concentration of cobalt in coal fly ash fractions
        as a function of particle size	53

  27    Concentration of arsenic in coal  fly ash fractions
        as a function of particle size	54

  28    Concentration of lead in coal fly ash fractions
        as a function of particle size	54

  29    Concentration of nickel in coal fly ash fractions
        as a function of particle size	55

  30    Concentration of chromium in coal fly ash fractions
        as a function of particle size	55

  31    Concentration of manganese in coal fly ash fractions
        as a function of particle size	56

  32    Concentration of vanadium in coal fly ash fractions
        as a function of particle size	56

  33    Dependence of As, Ni, and Cd concentrations
        in coal fly ash on particle size	63

  34    Apparatus for ambient temperature soxhelt leaching 	   76

  35    Ion microprobe depth profiles for the Group A elements
        (Ti, Al,  Si) in unleached and leached fly ash	79

  36    Ion microprobe depth profiles for the Group B elements
        (Fe, S) in unleached and leached fly ash	80

  37    Ion microprobe depth profiles for the Group B elements
        (K, Na, Li) in unleached and leached fly ash	81

  38    Ion microprobe depth profiles for the Group C elements
        (Pb, Tl)  in unleached and leached fly ash	82

  39    Ion microprobe depth profiles for the Group C elements
        (Cr, Mn,  V) in unleached and leached fly ash	83

  40    Ion microprobe depth profiles for the Group D elements
        (Ca, Mg)  in unleached and leached fly ash	84

  41    Auger Electron Spectrometry (AES) elemental
        depth profiles for unleached fly ash	88


                                     xi

-------
Number                                                                  Page

  42    Dependence of mole fraction (x) of PAH
        adsorbed on temperature  	  95

  43    Dependence of half-time of adsorption on temperature 	  97

  44    Percent conversion of irradiated benzo[a]pyrenecompared
        with a non-irradiated sample as a function of time	102

  45    Percent conversion of fluorene adsorbed on
        coal fly ash as a function of time	104

  46    Chromatograms of extracts from pyrene adsorbed on
        coal fly ash before and after exposure to N02 and S03  	 108

  47    Chromatograms of extract from benzo[a]pyrene adsorbed
        on coal fly ash before and after exposure to N02 and S03 .... 109

  48    Chromatograms of extracts from pyrene adsorbed on
        silica gel exposed to N02, showing changes with aging  	 110

  49    Low-resolution ESCA spectra of a (A) coal fly ash
        surface following pyrene adsorption and (B) the
        same coal fly ash surface following adsorption of
        pyrene and exposure to N02	Ill

  50    High-resolution ESCA spectrum of a coal fly ash
        surface following adsorption of pyrene and
        exposure to N02	112
                                     xii

-------
                               LIST OF TABLES

Number                                                                  Page

   1     Relative abundance (wt %) of morphologic particle types
         in four coal size-classified fly ash fractions	    1

   2     Encapsulating particle bulk mass contribution (wt %)
         for the non-magnetic fraction of coal fly ash	14

   3     Encapsulating particle bulk mass contribution (wt %)
         for the magnetic fraction of coal fly ash	15

   4     Mass distribution (wt %) of non-magnetic fractions
         of fly ashes from midwestern bituminous and
         western sub-bituminous coals	23

   5     Mass distribution (wt %) of magnetic fractions of
         fly ashes from midwestern bituminous and
         western sub-bituminous coals  	   24

   6     Specific surface areas (m2/g) of coal fly ash fractions
         from a midwestern bituminous coal	28

   7     Typical matrix element composition ranges of some
         American and British coal fly ashes	30

   8     Summary of elemental concentration dependence
         on particle size	58

   9     Boiling points of possible inorganic species
         evolved during coal combustion  	   60

  10     Elemental surface concentrations and layer thicknesses
         calculated using the surface deposition model 	   62

  11     Enrichment factors for in-stack total
         suspended particulate material  	   66

  12     Upper-limit concentrations of carbonate
         compounds in coal fly ash	68

  13     Carbonate species contribution to leachable
         elemental concentrations 	   69
                                    xiii

-------
Number                                                                  Page

  14     Changes in measured pH resulting from column leaching
         of a bulk a western sub-bituminous coal  fly ash
         with solutions of different pH	74

  15     Elemental concentrations in coal fly ash leachates	86

  16     Summary of analytical results-surface characterization
         of coal fly ash	89

  17     Polycyclic aromatic compounds identified in emitted
         coal fly ash	91

  18     Polycyclic aromatic hydrocarbon emission factors
         for coal-fired furnaces 	  92

  19     Measurement of POM emitted from a coal-fired
         power plant stack	98

  20     Decomposition of irradiated polycyclic aromatic
         hydrocarbons adsorbed on coal fly ash	103

  21     Decomposition of non-irradiated polycyclic aromatic
         hydrocarbons adsorbed on coal fly ash	105
                                     xiv

-------
                               ACKNOWLEDGMENTS

     We thank Dr. R. K. Skogerboe for his editorial advice and assistance in
the preparation of this report.

     This research was primarily funded by the U.S. Environmental Protection
Agency, Environmental Research Laboratory-Duluth, Research Grant No. R803950,
awarded to Natural Resource Ecology Laboratory, Colorado State University,
and Fisheries Bioassay Laboratory, Montana State University.  Additional
support was provided by the U.S.  Department of Energy under Reseach Grant
Nos. EE-77-S-02-4347 and DE-AS 02-78 EV 04960 and by the National Science
Foundation under Research Grant No. ERT-74-24276.
                                    XV

-------
                                  SECTION 1

                                INTRODUCTION

     Production of electric power from the combustion and conversion of
fossil fuels represents a ubiquitous and increasing means of obtaining energy
in most countries throughout the world.  It is now apparent that, for the
next several decades at least, deficits in the energy budget of the world
will be made up primarily from increased coal usage.   As a result, consider-
able emphasis is currently being placed on assessing the environmental impact
of coal combustion and conversion operations.  It is to this end that the
present work, dealing with the characterization and assessment of the poten-
tial environmental impact of coal fly ash, is directed.

     Particulate matter in the form of airborne fly ash constitutes a major
pollutant emission from both the combustion and conversion of coal.  Indeed
some estimates indicate that coal fly ash, consisting of material retained by
collection devices within a coal fired power plant, may soon constitute one
of the largest tonnages of solid waste materials produced in the United
States.  It is appropriate therefore to seek information about both the
nature and potential environmental impact of this material as emitted to the
atmosphere and as retained by control devices in a power plant.  To this end,
a number of samples of coal fly ash have been studied.  These have been
derived from a variety of power plants in which combustion of either mid-
western bituminous coal or western sub-bituminous coal occurs.  For the
purposes of the present project, however, one fly ash of each type has been
studied extensively.  Thus, fly ash representative of the burning of a
western sub-bituminous coal was obtained from the Corrette Plant located in
Billings, Montana.  That representative of a midwestern bituminous coal was
obtained from the Iowa/Illinois Gas and Electric Plant in Davenport, Iowa.
The coal burned in the first plant was obtained from the Rosebud Mine located
at Col strip, Montana; that burned in the second plant was obtained from the
Captain mine in southern Indiana.

     In all cases two distinct types of fly ash are considered.  These are:

      1.  Fly ash collected in bulk from mechanical or electrostatic precipi-
          tating devices and representative of material retained in a power
          plant.

      2.  Fly ash collected from flue gas in the stack downstream from the
          precipitator and thus representative of the material emitted to the
          atmosphere.

     In all the studies of the physical and chemical characteristics of coal
fly ash, emphasis was placed on the dependence of fly ash properties on

-------
particle size since this parameter,  perhaps more than any other,  is of
primary importance in determining the behavioral characteristics  of partic-
ulate material.   Thus, it is the aerodynamic particle size which  determines
the ability of a control device to collect a particle; it is the  aerodynamic
particle size which determines the atmospheric lifetime; it is the physical
particle size which determines the amount of potentially undesirable material
which can be leached from the fly ash; it is the aerodynamic particle size
which determines the potential inhalation toxicological  impact of airborne
fly ash, and it is both the physical and aerodynamic particle size which have
a profound effect upon the methodology employed for sampling and  analysis of
coal fly ash.

     The text of this report is devoted to a description of the studies which
have been performed for the specific purpose of obtaining a better under-
standing of the nature and behavior of coal fly ash.  However, in order to
perform many of these studies, it has been necessary to develop and/or apply
a large number of specific analytical techniques.  These techniques have
involved sample handling, sample characterization and the interpretation of
the results from the analyses.  A number of such methodology — related
studies are presented in appendix A of this report.

-------
                                  SECTION 2

                                 CONCLUSIONS

     The following conclusions can be drawn from the results presented
herein.

      1.  Coal fly ash as obtained from a conventional coal fired power plant
          exhibits a number of morphological forms.   For most fly ashes, the
          most predominant forms include small, solid, spherical particles;
          hollow particles or cenospheres; and hollow spheres containing
          small solid particles encapsulated inside of them (plerospheres).
          The occurrence of quite large numbers of small particles inside the
          plerespheres is not thought to be of much significance from the
          standpoint of environmental impact.

      2.  The mass of material of a given aerodynamic size range of coal fly
          ash particles emitted to the atmosphere is a function of both the
          nature of the combustion system and the type of control device.
          Cyclone coal-fired systems generally emit the smallest particles.

      3.  The relative mass proportions of non-magnetic versus magnetic
          particles in coal fly ash are mainly dependent on the coal that is
          burned.  Density is determined more by the morphology of the
          particles than by their composition.

      4.  The electrical resistivity of coal fly ash is a function of both
          the concentration of alkali metals in the particles and the
          temperature.  High sodium and potassium concentrations and low
          temperatures lower the resistivity of the particles and promote
          collection by electrostatic precipitators.

      5.  The specific surface areas of size fractionated fly ash particles
          above 15 urn in physical diameter are generally less than 1 m2/g.
          However, values of up to 5 m2/g have been found for small (<5 urn
          diameter) respirable particles.  Specific surface area appears to
          be highest in the smallest (<20 urn) and largest size (>74 urn) size
          fractions.  The latter observation is believed due to the porous,
          irregular shapes and high internal surface areas of the larger
          particles.

      6.  Coal fly ash particles are largely, made up of an impure alumino-
          silicate glass.  Major elements are Al, Fe, Si, Ca, K, Na and Ti.
          As such, the particles are essentially insoluble.

-------
 7.   A considerable number of potentially toxic trace elements;  As,  Br,
     Cd, Cu, Ga,  Mo, Pb,  Sb,  Se and Zn are preferentially associated
     with smaller fly ash particles.   The implications of this behavior
     are that such trace  elements may be preferentially emitted  from a
     coal fired power plant even under conditions where highly efficient
     particle collection  devices are being utilized.   Since small
     particles have longer atmospheric residence times, these toxic
     trace elements will  have corresponding longer residence times in
     the atmosphere.  Finally, such particles are capable of penetrating
     to the innermost (pulmonary) regions of the human lung when
     inhaled.

 8.   It has been  established that those trace elements which are prefer-
     entially associated  with small particles are in  fact highly
     enriched in  a thin layer at or near the particle surface.

 9.   Material present in  the particle surface layer has been shown to be
     easily capable of mobilization in aqueous solution and can  be
     readily extracted into body fluids.

10.   It is postulated that tl)e enrichment of certain  trace elements  on
     fly ash particle surfaces results from their being volatilized  at
     the temperatures encountered in the coal combustion process.   As
     the temperature falls downstream from the combustion zone,  these
     vapors condense onto the external surface of co-entrained fly ash
     particles moving toward the stack exit.   Considerable evidence  is
     available in support of this volatilization-condensation model
     although it is not considered to be either fully validated nor  to
     be the only possible mechanism which can result  in surface asso-
     ciation of trace elements.

11.   The distribution of  elements in coal fly ash is  considered to be
     due to three primary factors.   These include the particle size, the
     matrix composition of the particle, and the geochemical nature  of
     the element in question.  Thus, elements belonging to the sidero-
     phile, chalcophile and lithophile classifications behave quite
     distinctly in terms  of their partitioning behavior during and
     following combustion in a coal fired power plant.  The principal
     minerals in coal fly ash are mullite, quartz, magnetite and
     hematite.

12.   The chemical forms in which trace and minor elements exist in the
     surface layer of coal fly ash particles are primarily sulfates,
     carbonates,  silicates and oxides with smaller amounts of several
     chlorides and possibly mixed or complex salts.

13.   The composition of aqueous leachates of coal fly ash is determined
     by the extent to which equilibrium is achieved between the solid
     fly ash and the leaching solution.  The four most important param-
     eters which commonly influence this equilibrium are the composition
     of the fly ash, the  ratio of fly ash to solution volume, the method
     by which the fly ash and solution are brought into contact, and the
     chemical characteristics of the leaching solution.

-------
14.   Leaching of coal fly ash generally proceeds quite slowly, but
     relatively high amounts of certain trace elements may dissolve
     during this time due to their presence on the particle surfaces.
     The use of leaching solutions of low pH results in greater disso-
     lution of metals than does leaching at higher pH.

15.   A number of polycyclic organic species (POM) have been shown to be
     associated with coal fly ash which is actually emitted from the
     stack of a coal fired power plant.   On the other hand, fly ash
     collected and retained inside the power plant stack system rarely
     contains detectable amounts of polycyclic organic matter.

16.   Polycyclic organic species are strongly associated with the
     smallest particles of fly ash emitted from a coal fired power
     plant.

17.   Polycyclic organic matter associated with fly ash is predicted to
     be present in the surface layer of the fly ash.

18.   Strong evidence exists to support the idea that polycyclic organic
     matter is initially formed in the gas phase at elevated tempera-
     tures within the power plant stack system.  As the temperature
     falls, however, POM vapor rapidly adsorbs onto the surface of
     coentrained coal fly ash particles.  This process apparently takes
     place at or close to the stack exit and is probably complete within
     a short distance of the stack exit in the power plant plume.

19.   Particulate polycyclic organic matter emitted from a coal fired
     power plant is stabilized against photochemical decomposition as  a
     direct result of being adsorbed onto the surface of the coal fly
     ash.  A number of polycyclic organic species are shown to oxidize
     in the absence of light as a direct result of being adsorbed onto
     the surface of coal fly ash thereby significantly changing their
     potentially carcinogenic properties.

20.   A number of potentially carcinogenic polycyclic organic species
     have been shown to react with stack gases such as nitrogen dioxide
     forming a number of derivatives the most prominent of which are
     nitration derivatives.  This process happens fairly rapidly and is
     likely to have profound effects on the toxicological potential of
     coal fly ash particles.

-------
                                 SECTION 3

                              RECOMMENDATIONS

1.   It is recommended that consideration be given to establishing particle
    emission standards in the future in terms of particle mass distribution
    as a function of size rather than simply in terms of total mass emitted
    as at present.

2.   Emphasis in further research should be placed on identifying the actual
    chemical forms in which potentially toxic trace metals are present.

3.   Considerable emphasis should be placed on elucidating the role(s) of
    coal fly ash in the conversion of sulfur dioxide and sulfur trioxide to
    sulfuric acid and metal sulfates both within a coal  fired plant stack
    system and in the plume emitted therefrom.

4.   One or more methods for conducting laboratbry leaching studies of coal
    fly ash should be standardized and the present simple mathematical
    models describing the leaching process extended to have predictive
    capability for field situations.

5.   Studies should be conducted to determine the extent(s) to which differ-
    ent soil types are capable of adsorbing or otherwise fixing trace metals
    present in leachates percolating through soil (e.g., in the vicinity of
    a coal fly ash pond).

6.   More precise emission factors should be obtained for polycyclic organic
    matter emitted from various types of coal fired power plants.

7.   Considerable study should be directed towards understanding the rate and
    extent of transformations of polycyclic organic matter which occur in
    the emitted plume of coal fired power plants.

8.   Emphasis should be placed on determining how the toxicological charac-
    teristics (e.g., mutagenic and carcinogenic properties) of emitted coal
    fly ash varies as a function of time and distance from the point of
    emission.

-------
                                  SECTION 4

                  PHYSICAL CHARACTERISTICS OF COAL FLY ASH
PARTICLE MORPHOLOGY

     Two general classification schemes have been reported describing dif-
ferent morphological classes of coal  fly ash.   The first,  (Natusch et al.,
1978a) based mainly on optical  and scanning electron microscopy studies,
indicates five classes of fly ash can be distinguished:

 1.   Large irregular particles  often  exhibiting extensive  spherical  pitting
     (Figure 1A).   These particles predominate in the physical  size fraction
     greater than 74 micrometers (urn) and are frequently found in fly ashes
     derived from coal burned in power plants utilizing a  chain grate stoker
     and in industrial boilers.

 2.   Solid spherical particles  (Figure IB) whose solidity  is confirmed by
     etching deep into the particles  with a stream of positive argon ions
     derived from an ion milling apparatus.  These solid particles predomi-
     nate in the physical size  range  below about 10 urn.

 3.   Hollow spherical particles (Figure 1C) whose interior voids are con-
     firmed by breaking the particles or etching through their shells prior
     to examination by scanning electron microscopy (SEM).  These hollow
     spheres, which are referred to as cenospheres (Raask, 1968), predominate
     in the 20-74 urn physical size range of fly ash.  Their densities are
     generally less than 1 g/cm3 so that they float on effluent settling
     ponds.  They commonly comprise about 5 percent of the mass of fly ash
     retained in power plant precipitators.

 4.   Hollow spherical particles containing a number (5-100) of small solid
     particles encapsulated inside them (Figures ID and IE).  The encapsu-
     lated particles called plerospheres (Raask, 1968) are observed by
     breaking the outer shell of the  host particle, or etching through it
     with argon ion bombardment, prior to SEM examination.  In some cases the
     encapsulated particles can also  be observed by optical microscopy by
     utilizing an objective immersion oil whose refractive index is close to
     that of the shell material of the host particle.

 5.   Agglomerates of many small (<10  pm) spherical particles forming large
     nonspherical  particles which predominate in the physical size range
     greater than 74 urn (Figure IF).

-------
     Optical microscopic observation of fly ash shows that, most of the parti-
cles are colored, semi-transparent spheres suggesting an impure glass.
Magnetic particles are, however, opaque.   In most cases the particles have a
slightly roughened surface (Figure 1) to which adhere numerous, apparently
crystalline particles of submicrometer size.   This adhering material  is
readily removed by washing with either aqueous or organic solutions (Bolton
et al., 1973).

     In the second scheme, described by Fisher et al.  (1978), a total of
eleven particle types were differentiated.  These particle types differ from
those of Natusch et al. (1975) principally in that optical microscopy has
been incorporated into the classification scheme.  In addition, Fisher et al.
(1978) collected and fractionated an ash to determine the mass distributions
of the various particle types.  Table 1 gives the results of Fisher et al.
(1978) and contains both a general description of the particle types  and
their mass size distributions.  The fly ash was collected downstream  from the
electrostatic precipitator in the stack breeching of a large western  U.S.
power plant burning a low sulfur (0.5%), high ash (23%) coal.  Although the
mass distribution by class cannot be considered necessarily indicative of the
distributions resulting from the combustion of other coals in other power
plants; it is probably representative of a typical plant system.  The rela-
tive abundance of all particle types, except non-opaque, solid spheres was
found to increase with the increasing particle sizes of the fly ash frac-
tions.  Non-opaque solid spheres were found to be inversely dependent on
particle size.   Particle density was found to be inversely related to
particle size indicating that the smaller particles contained fewer hollow or
encapsulating particles.

     The different particle shapes appear to be primarily a function  of three
factors, the chemical composition, combustion zone temperature and the parti-
cles residence time in the combustion zone.  The irregular particles  or those
with only slightly rounded edges were the result of limited high temperature
exposure.  The bulk of fly ash, especially the smallest size fractions,
consist almost exclusively of smooth spherical particles.  This suggests that
they had been exposed to higher temperatures or longer residence times,
although their appearance is partly attributable to the lower melting point
expected because of their size.  Fly ash particles, especially the spherical
ones, tend to consist of magnetic and non-magnetic fractions.  While  the
latter reflect a diversity of particle types such as the cenospheres, solid
spheres, and plerospheres previously described, the former consist of mainly
black solid spherical particles, the particle colors being caused mainly by
the presence of magnetite, Fe304.
Figure 1.  Scanning electron micrographs of representative coal fly ash
           particles.  Approximate diameters are in parenthesis.  A. clinker
           (85 pm), B. solid (20 pm), C. hollow (40 urn), ID. encapsulating
           (70 pm), E. encapsulating (60 pm), F. agglomerate (250 pm).
           (Reprinted by permission of D.F.S. Natusch from Natusch et al. ,
           1975 and by A. Loh from Loh, 1975.)

                                      8

-------

-------
  TABLE 1.   RELATIVE  ABUNDANCE  (WT %)  OF  MORPHOLOGIC  PARTICLE  TYPES IN
      FOUR  SIZE CLASSIFIED  FLY  ASH FRACTIONS*  (FISHER ET AL.,  1978)

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.


Particle type
Amorphous, non-opaque
Amorphous, opaque
Amorphous, mixed opaque and
non-opaque
Rounded, vesicular, non-opaque
Rounded, vesicular, mixed opaque
and non-opaque
Angular, lacy, opaque
Non-opaque, cenosphere
Non-opaque, plerosphere
Non-opaque, solid sphere
Opaque sphere
Non-opaque sphere with crystals
CombinedH particle types 2 and 6
VMD (urn)
Density§ (g/cm3)
Volume mean diameter (VMD) (um)t
20
7.25
0.42
0.77
12.39
2.27
1.34
41.11
0.51
25.58
1,56
6.80
—
20
1.85
6.3
2.13
0.18
0.09
6.67
0.24
0.57
26.22
0.21
56.01
0.90
6.79
--
6.3
2.19
3.2
0.79
--
--
2.91
--
0.27
13.20
—
79.16
0.33
3.18
0.15
3.2
2.36
2.2
0.33
--
--
2.99
0.03
0.33
7.91
--
86.99
0.24
0.95
0.24
2.2
2.45
* A total of approximately 3000 particles for each size fraction were
  classified from random fields using three microscope slides per
  fraction.

t Samples aerodynamically sized by collection with an in-stack fraction-
  ator (McFarland et al., 1977).

H Combined due to inability to distinguish between these classes for the
  finer particles.

§ Apparent density determined by gravimetric displacement of 1-propanol.


                                     10

-------
     Probably the most interesting aspect of fly ash morphology is the phe-
nomenon of particle encapsulation; a phenomenon that has hereto been regarded
mainly as a scientific curiosity (Matthews and Kemp, 1971; Fisher et al.,
1976).   However, the occurence of encapsulation in a substantial fraction of
fly ash particles could have a significant influence on the overall envi-
ronmental impacts of fly ash.  This is because the encapsulated particles are
almost always smaller than 5 micrometers (urn) in physical diameter (Figure
2).  Such particles, if released from their encapsulating hosts, are diffi-
cult to control (Jones, 1972), have long atmospheric lifetimes (Butcher and
Char!son, 1972; Heindrycdx and Dams, 1974), provide substantial surface area
for heterogeneous reactions (Urone and Schroeder, 1969), produce significant
light scattering (Butcher and Charlson, 1972), and may penetrate the inner-
most regions of the human lung when inhaled (Natusch and Wallace, 1974).
Thus, reduced environmental impacts may result from encapsulation of large
numbers of small particles emitted by enclosing them within larger, easily
collected, host particles; this should be recognized as a desirable "natural"
control process.

     In order to establish whether or not encapsulation does influence the
environmental impact of fly ash, several ashes were studied to determine
specifically:

      1.   The physical and chemical characteristics of the encapsulated and
          the encapsulating particles.

      2.   The extent of encapsulation.

      3.   The probable encapsulation mechanism.

Fly ashes derived from five representative midwestern bituminous and western
sub-bituminous coals collected from electrostatic precipitators of modern
coal fired power plants were chosen for study.  All plants utilized pul-
verized fuel feeds.

     Fly ash particles were studied either individually or in groups by means
of scanning electron microscopy (SEM) using carbon or gold coatings to
increase particle conductivity.  In order to identify a hollow or encapsu-
lating particle it was necessary to rupture the outer shell.  This was
achieved by etching with a stream of positive argon ions, by cutting the tops
off particles embedded in a plastic matrix using a irncrotome, by crushing the
particles between two microscope slides, or by rolling a cylindrical metal
bar over them.  The latter was found to be most suitable for rupturing
particle shells with minimal loss of their contents.

     Evidence indicating that encapsulated particles are actually contained
inside a host particle, as shown in Figures 2A, C, D, and E, and do not
simply move there during particle rupture or SEM observation was obtained in
several ways.  First, encapsulated particles could occasionally be observed
through holes in the encapsulating shells which were too small to permit
their entry (Figure 2C).  Secondly, where shell rupturing took place after
particles had been carbon or gold coated, encapsulated particles exhibited
considerable electrostatic charging indicating that they were not exposed at
'the time of coating.  This effect is illustrated by the relative brightness

                                      11

-------
of the encapsulated particles compared to the host shells in Figures 2A and
2E.   Thirdly, within a given size fraction,  the crushed portion contained a
significant number of small  spherical  particles which would have been removed
by the size separation had they not been encapsulated (Figures 2D and E).   In
addition, encapsulated particles could sometimes be observed by optical
microscopy using an oil immersion objective  to render the intact encapsula-
ting shells transparent (Matthews and Kemp,  1971).   It seems probable,
therefore, that encapsulation is not an artifact of sample preparation and
observation.

     In order to study the dependence of the encapsulation phenomenon on
particle size, density, and magnetic character; the fly ashes were subdivided
sequentially with respect to these parameters and examined by Scanning
Electron Microscopy.  Bulk fly ash was first separated into physical size
fractions by mechanical sieving.  Each of these fractions was divided into
magnetic and non-magnetic fractions using a  bar magnet.  It is recognized
that this separation is highly technique dependent, but it does serve to
differentiate crudely between particles containing high and low concentra-
tions of magnetic (iron) compounds.  These fractions were then separated into
several density fractions by the float - sink method (Ruch et al. , 1974)
using appropriate mixtures of either chloroform and diiodomethane or water
and thallium formate/malonate (Clerici's solution) (Natusch et al., 1975).
The contributions of encapsulating particles to the bulk mass for non-
magnetic and magnetic fractions are presented in Tables 2 and 3.

     Both hollow and encapsulating particles were observed in all fly ashes
studied.  In all cases, encapsulating particles occurred primarily in the
size range 20 - 90 urn and in the density ranges <1.6-2.7 g/cm3 and 2.1 -
3.4 g/cm3 for non-magnetic and magnetic fractions, respectively, so only
these fractions were considered in the following numerical estimates of
encapsulation.  Cenospheres were found to occur exclusively in the lowest
density fractions and exhibited no preference for either magnetic or non-
magnetic particles although their incidence  has been correlated with high
iron content by others (Raask, 1968).

     The number of particles encapsulated inside a single host varied from
about ten to several hundred with 30-100 particles being commonly observed
(Figures 2A, D, E).  Almost all encapsulated particles were found to be
solid and to have physical diameters in the  range 0.1 to 5 urn.  Encapsu-
lating shell thicknesses were typically about 5-10 percent of the diameter
of the host particle (Figure 2).  Elemental  analyses of associated encapsu-
lated and host particles by energy dispersive X-ray emission spectrometry
excited by the SEM beam showed that the matrices were closely similar,
indicating that both types were derived from essentially the same local
matrix.
Figure 2.  Scanning electron micrographs of coal fly ash particles showing
           representative encapsulated and encapsulating particles (A, C,
           D, E) and cenosphere shells (B, F).  (Reprinted by permission
           of D. F. S. Natusch.)

                                      12

-------
13

-------
          TABLE 2.   ENCAPSULATING PARTICLE  BULK MASS  CONTRIBUTION
           (WT %)  FOR THE NON-MAGNETIC FRACTION OF  COAL FLY ASH*


                                       Particle size  (|jm)
ueiia i i,_y v.y/ >-»i )
<1.6
1.6-2.3
>2.3
20-45
0.8-1.3
0.04-1.48
NDt-7.01
45-60
0.2-0.4
0.05-0.11
0.04-0.97
60-90
0.2-0.5
0.01-0.23
0.03-0.55
          * Data represent ranges for five different coal  fly ashes.

          t ND = Not detectable.
     As expected, encapsulated particles have essentially no surface growths
since it has been shown, both here and by other workers (Fisher et al.,
1976), that microcrystal growth (Fig.  2C) usually occurs by exposure to
moisture during collection and storage of fly ash.   It has been suggested
(Fisher et al.  1976) that S02 is hydrated and oxidized on the surface to
H2S04 which subsequently reacts with either CaO or NH3 to form crystalline
salts.  It aslo has been suggested (Sarofim et al. ,  1977) that the crystal-
line material may result from deposition of fine silica particles.  It has
been reported (Raask, 1968; Fisher et al., 1976) that the gas contained
inside cenospheres and encapsulating particles consists mainly of carbon
dioxide with traces of nitrogen and water vapor.  The internal pressure  has
been calculated to be in the range 0.3 to 0.5 atm.  (Raask, 1968).   On heat-
ing, all particles decrepitate (lose water) at about 300°C.  The softening
and slow collapsing of individual particles was observed when they were
heated to 1150-1300°C on a high temperature optical  or electron microscopic
stage, but true melting does not take place until approximately 1350°C.
Exact fusion temperatures were not determined.

     The extent to which encapsulation occurs was estimated as follows.
First, the relative number of encapsulating particles present in each sub-
sample was determined by counting the number of spherical and non-spherical
particles by optical microscopy and then counting the relative numbers of
spherical particles shown to be encapsulating after each sub-fraction was
crushed.  This latter observation was made by means of scanning electron
microscopy for which a typical field of view is shown in Figure 2D.  Since
encapsulating particles were always found to be spherical, this approach
established the number fraction of encapsulating particles in each sub-
sample.  Secondly, number-fractions were generally equivalent to mass-
fractions because of the narrow size and density of each sub-sample.  The


                                      14

-------
           TABLE 3.   ENCAPSULATING PARTICLE BULK MASS CONTRIBUTION
              (WT %) FOR THE MAGNETIC FRACTION OF COAL FLY ASH*


                                            Particle size (urn)
UGIIO i ^y ^y/ i-ni j
<2.l
2.1-2.9
>2.9
20-60
NDt
ND-0.5
0.39-1.4
60-90
ND
0.01-0.12
0.05-1.00
          * Data represent ranges for five different coal fly ashes.

          t ND = Not detectable.
resulting percentages of the total sample mass shows that the mass percent-
ages of the five different fly ashes involved in encapsulation were, 2.4%,
9.3%, 8.7%, 10.2%, and 1.8%.

     Since a substantial fraction of the mass involved in encapsulation
consists of respirable particles (Natusch and Wallace, 1974), it is appro-
priate to estimate just what mass percentage of all respirable particles
produced are encapsulated.   Based on the foregoing microscopic observations
and X-ray analyses, it can be assumed that both encapsulated and host parti-
cles have the same density, that shell  thicknesses are 5 to 10 percent of the
host particle diameters, that host interiors are 30 to 70 percent occupied,
and that all encapsulated particles are within the respirable range.  These
assumptions enable calculation of the percentage mass of each sample present
as encapsulated, respirable particles.   The ranges obtained were, 1 to 2%, 4
to 6%, 4 to 6%, 5 to 7%, and 0.8 to 1.2% for the five ashes studied.  The
mass percentage of free particles in the respirable range (^5 urn) can be
determined, for comparison, by plotting the cumulative fly ash mass percent-
age versus logarithmic physical diameter.  This indicated that a closely
Gaussian relationship existed and that the mass percentages of free respir-
able particles in each of the five fly ashes was 20%, 37%, 29%, 10%, and 25%.

     Comparison of the mass percentages of both encapsulated and free parti-
cles of respirable size shows that, of all respirable particles produced by
each power plant, the following mass percentages are encapsulated:  5 to 9%,
10 to 14%, 12 to 17%, 33 to 41%, and 3 to 5%.  Thus, if quantitatively
released to the atmosphere, encapsulated particles could significantly
increase respirable particle emissions from a coal fired power plant.
                                      15

-------
     Based on present information, it is not possible to state definitively
whether the phenomenon should be regarded as being potentially beneficial or
potentially hazardous.  However, the evidence obtained in this study and by
others (Raask, 1968; Matthews and Kemp, 1971; Fisher et a "I., 1976), suggests
that encapsulated particles are actually formed inside their host particles
in, or close to, the coal combustion zone.   This further suggests that the
mechanism of formation involves the following steps:

      1.   Molten spherical droplets of fly ash matrix material are formed at
          temperatures above 1350°C.  The size distribution and viscosity of
          these droplets is determined by their matrix composition (and thus
          the original coal), the temperature, and the configuration of the
          combustion chamber.

      2.   Due to the production of gases within the molten droplets, they may
          expand like a balloon to an extent which is determined by their
          surface tension and the internal  pressure.   It has been suggested
          by Raask (1968) that the expanding gas is produced by reaction of
          iron oxides with residual carbon:

                         2 Fe203.+ C ? 4 FeO + C02

Alternatively, decomposition of inorganic carbonates present in the coal
could be involved.   Either suggestion would account for the substantial
amount of carbon dioxide found in particle voids and for the observation of
small localized voids (Figure 2A, 2D, 2E).   Expansion to equilibrium size in
this way has been estimated to occur in less than a millisecond (Raask,
1968).

      3.   As the hollow spheres cool, their exterior surfaces solidify first
          leaving a molten interior whose viscosity and surface tension are
          sufficiently low to enable detachment of material from the concave
          inner surface.  This, together with the reduction of internal
          pressure within the sphere on cooling, promotes "budding off" of
          molten droplets whose surface tension is now great enough to pre-
          vent extensive coalescence.  Support for this process is provided
          by the fact that, for a given fly ash, the surface tension of
          molten matrix material, as calculated from the known composition of
          each sub-fraction (Morey, 1954), exhibits a significant correlation
          with the percentage of encapsulating particles (Figure 3).  Further
          support comes from the close matrix identity observed for encapsu-
          lated and host particles.

     In conclusion, it would appear that encapsulation does not result in the
containment of small particles which might otherwise be emitted to the atmo-
sphere.  On the other hand, it is unlikely that it gives rise to additional
particulate emission.  If, however, encapsulated particles were subjected to
sufficient mechanical shock to cause their rupture, a substantial increase in
respirable particle emission should be anticipated.
                                      16

-------
                       10      20     30     40      50
                       ENCAPSULATING PARTICLES  (%)
Figure 3.   Calculated surface tension at 1400°C as a function of the
          percentage of encapsulating particles.
                                  17

-------
MASS, DENSITY, AND FERROMAGNETISM DISTRIBUTIONS

     In addition to particle morphology, a number of other physical proper-
ties have been investigated in an attempt to elucidate the formation and
behavioral characteristics of fly ash.   Such properties include the mass
distributions according to particle size, particle density, and ferromagnetic
properties.   The available data are incomplete and apply only to fly ash
collected after a control device or from the device itself,,  Consequently, it
is not presently possible to relate physical properties to parameters such as
plant operating conditions or the type of coal from which the fly ash was
derived.   Nevertheless, it is qualitatively apparent that the physical prop-
erties of fly ash depend upon both of these parameters.

     Measurements of the distribution of fly ash particle mass with size are
of two distinct types.   The first involves determination of the aerodynamic
particle size distribution, which normally involves isokinetic collection of
fly ash directly from a stack gas stream.  Several sampling devices are
available, but the most common involve the principle of inertial impaction
and enable collection and size classification of fly ash in situ.  Alterna-
tively, bulk fly ash, such as might be collected from an electrostatic
precipitator or bag house, can be differentiated into physical size fractions
by mechanical sieving.   In either case it is necessary to separate the
various size fractions prior to weighing or microscopic counting for size
estimation.

     These two types of measurement are quite distinct.  Aerodynamic size
determination enables prediction of the atmospheric (air-stream) behavior of
each size fraction, such as is important for establishing parti cul ate collec-
tion efficiency, atmospheric residence, and inhalation characteristics
(White, 1963; Butcher and Charlson, 1972; Natusch and Wallace, 1974; Yeh et
al., 1976).   On the other hand, physical size determination provides a
straightforward measurement of the physical dimensions of the particles and
can be directly related to particle number.  Interconversion between aero-
dynamic and physical sizes can be accomplished using the relationship
(Koltrappa and Light, 1972):

                                              PfL,                     (1)
where D   is the aerodynamic particle diameter, D  is the physical particle
       ac                                        i
diameter, p,,, is the particle density, p   is 1 g/cm3 by definition, and
           r                           ae
C(D  ) and C(D ) are the Cunningham slip correction factors for the diameters
   36         r
D   and D .  Raabe (1976) has compared the commonly used conventions for
 ae      r
description of the aerodynamic size of respirable aerosols including a de-
tailed description of the slip correction and dynamic shape factors.  Aero-
dynamic size distributions are often presented as lognormal probability
functions of the grouped mass or number data derived from inertial impaction
instrumentation.
                                      18

-------
     Fisher et al.  (1978) compared fly ash fractions collected isokinetically
(Ondov et al., 1976, Ondov et al.  1979) with electrostatic precipitator
samples from the same plant fractionated in the laboratory.   The results show
that, for the larger and generally less uniform particle sizes, the agreement
between laboratory and field separations is not particularly good.   However,
for the smaller size fractions (physical diameter below about 2 urn) agreement
was considerably better.

     Despite the comparative simplicity of determining both the aerodynamic
and physical size distributions of fly ash mass, the number of available
measurements of fly ash prior to collection by control equipment is rela-
tively sparse.  It has been established, however, (Southern Research Insti-
tute, 1975) that the size and morphology of fly ash depend not only on the
nature of the mineral inclusions in coal, as discussed in the previous sec-
tion, but also on the manner in which the coal is burned.  This latter
dependence is illustrated in Figure 4 for fly ash derived from coal burned in
a chain grate stoking unit, a pulverized coal fed unit, and a cyclone fired
unit (Southern Research Institute, 1975).  In each case the fly ash was
sampled upstream from control equipment so it is representative of that
generated by combustion.

     It is apparent from Figure 4 that the fly ash mass approximates a log-
normal distribution over the aerodynamic size range considered.  Furthermore,
although the geometric standard deviations of the three distributions are
similar, their mass median diameters differ considerably (i.e., Cyclone ~
6 urn; Pulverized = 18 urn; Stoker = 42 urn).

     From a practical standpoint, one is primarily interested in the aero-
dynamic size distribution of the fly ash which is actually emitted from a
coal fired power plant.  This is, of course, largely determined by the
collection efficiency of the particle control equipment.  Specifically, the
size distribution of the emitted fly ash is determined by the product of the
function describing the size dependence of fly ash mass entering a control
device and that function describing the dependence of collection efficiency
of this device on particle size.  Examples of the physical size distributions
of fly ash mass emitted from a coal fired power plant equipped with different
control devices are presented in Figure 5 (Ondov et al. , 1975).  The results
indicate that the size of the particles emitted by the two control  devices
were roughly comparable.  However, the mass of material emitted in the
respirable range was greater for the scrubber system than for the electro-
static precipitator (Ondov et al., 1979).  Thus, determination of the densi-
ties of different fly ashes, and sub-fractions thereof, provides a means of
interconverting aerodynamic and physical sizes according to Equation 1.  In
addition, some differentiation between distinct morphological and composi-
tional characteristics can be achieved.  For example, cenospheres can readily
be distinguished from solid particles on the basis of density as can predomi-
nantly carbonaceous particles from aluminosilicates (Fisher et al., 1978).

     Determination of fly ash particle density is achieved most simply by
means of the traditional "float-sink" method which employs a series of
liquids of different densities to separate the particles (Ruch et al., 1974;
Olson and Skogerboe, 1975).  Alternatively, separation can be achieved by


                                      19

-------
                        1000
                           ; III II I I I I  M II I I I  I I I  II
                                            PULVERIZED
                                            COAL-FIRED
                         100
                        E
                        a.
                        1 10
                        d

                        fe
                          01
                         001
                                                   J_L
                          001 01 051  5 10   50   9095 99  99.9 SI999
                               WEIGHT % LESS THAN STATED SIZE
Figure 4.   Particle size distributions resulting  from coal combustion in
            different boiler types.   (Reprinted by permission of A.  E.
            Vandegrift and the Midwestern Research Institute from Vandegrift
            and Shannon, 1971.)
                10
              cc 10
              LU
              CD
              LU
              LU
                10
                             ESP
0.01  O.I     I

   DIAMETER
                                   10
                         ioa
                       LU
                       m

                       |,o

                       LU
                       > ,_
                       F 10
                       LU
                       K 10
                                                          SCRUBBER
0.02     O.I  0.2 0.5 I

   DIAMETER (jim)
Figure 5.   Particle size distribution of emissions  from similar production
            units at the same power plant equipped with either an  electro-
            static precipitator  (ESP) or a Venturi wet scrubber.   (Reprinted
            by permission of J.  Ondov from Ondov  et  al., 1976.)
                                       20

-------
placing the particles in a liquid in which a density gradient has been
established (Loh, 1975).

     While determination of particle density is of considerable interest in
its own right, more definitive insights are obtained if density separations
are carried out in conjunction with sequential size separations and with
differentiation between ferromagnetic and non-ferromagnetic particles.  This
three dimensional fractionation scheme has been described in Section IV.

     Typical mass distributions are presented in Figure 6 and Tables 4 and 5.
Figure 6A shows the mass distribution in a typical midwestern U.S.  bituminous
coal fly ash while Figure 6B shows how the bulk ash is distributed between
non-magnetic and magnetic fractions.  Tables 4 and 5 enable comparisons to be
made between mass distributions of midwestern bituminous and western sub-
bituminous coal fly ashes for non-magnetic and magnetic fractions,
respectively.

     The data in Figure 6A and 6B illustrate two points.  First, at least for
electrostatic precipitator ash, the mass of fly ash in a given fraction
varies inversely with the size of particles.  Thus, the number of particles
in the smallest (respirable) size fraction greatly exceeds those in the
larger size fractions.  Second, the distribution of mass between non-magnetic
and magnetic fractions of the fly ash are reasonably constant as shown in
Figure 6B.  This suggests that no single fraction can be isolated solely by
magnetic properties and also suggests that iron is fairly uniformly distri-
buted in the fly ash.

     A number of characteristics of coal fly ash can be distinguished from
the data presented in Tables 4 and 5.  It is apparent that both fly ashes are
compositionally heterogeneous, since there are considerable differences
between their mass distributions.  As discussed above, much of the variation
in densities observed is attributable to morphological rather than composi-
tional characteristics.  This is rather well illustrated by the data in
Figure 7, where density distributions are presented as a function of particle
size before and after crushing the fly ash.  The observed shift to higher
density on crushing indicates the presence of vesicular particles and ceno-
spheres in the larger size fractions, as discussed previously.  Interest-
ingly, the X-ray powder diffraction patterns of each of the subfractions
presented in Tables 4 and 5 reveal no convincing differences in matrix compo-
sition which depend upon either size or density (Natusch et al., 1975).  This
finding further supports the contention that the density distributions in fly
ash are largely determined by morphology rather than by composition.  There
are, however, very distinct differences between the amount of magnetite
(Fe304) present in the magnetic and nonmagnetic fractions.  This suggests
that magnetite is primarily responsible for the magnetic properties of coal
fly ash.

ELECTRICAL RESISTIVITY DISTRIBUTION

     The electrical resistivity of coal fly ash is an important physical
property from the standpoint of control.  Thus, it has been established
(Bickelhaupt, 1974; 1975a, 1975b) that the collection efficiencies of


                                      21

-------
     O


O   Z
P   <

O   Z
s   i

Q  D
OQ
                                                               00
                                                               o
                                                               OJ
                                                                   E
                                                                   d.
                                                                   UJ
                                                                   N
                                                                   CO

                                                                   UJ
                                                                   _l
                                                                   o
              d       o

                NOIiOVdd
                                                 o       o

                                                1H9I3M
Id


1



A c
^ N
^ CO
6 ^
OJ
V
10 to  ID
      OJ f-
 C  C C 
 O  O D5iH
•r- ••- fO
•P -(-> E   -
 O  O    -C
 C  (O in O
 3  S- 3 —I
4- 4- r—
      Q. E
 it)  U    O
   •r- U S-
 (fl +J .r- 4-
 n3  0) 4J
    C CD -C
 to  O> C O
 C  (0 O)—I
 o  E m
•r-    E   •
-P TJ  I  r- tO
4-  I      S-
    C i—  Q)
i—  O  (0  O
 03  c -P
 o     o  >,
 O   • -P -Q
   CQ

•r-     O  Ol
-!->   •    4->
 O)  C  to  C
 C  O  C -r-

 CO -t-> -r-  O-
 e  3 +j  oj
   -O.  O Q£

 C  il  $-
 (0 4-> 4-
    to
 CJ •!-•)-> 4->
•i- "O JT  C
^ *     C^l QJ
 OJ -C •,-  E
 C  tO  0)  Q)
 O^ fO  ?  S-
 m        u
 E -^  co  c
 I  r—  H3 .r-
 C  3
 o oa -a
 C     0)
     .  tO •!-
4- 
       s- _c
 c   .  o. u
 o  ai  x  to
•r-  M  a>  a>

 3  tO  tO 4—
^3     CO
•i-  a>  o
                                                                                  a>
                                                                         to
                                                                         • r"
                                                                         •a
                                                                            U -(->
                                                                                  to
                                                                                  01
                    NOIlOVdd 1H9I3M
                                                                            s_ •,- u
                                                                            ^3 ^* *^
                                                                         to Q.-P -P
                                                                         to    to i-
                                                                         ro 4- ••- m
                                                                         S O T3 O.
                                                                                         a>
                                                                                         S-

                                                                                         O5
                           22

-------
  TABLE 4.   MASS DISTRIBUTION (WT %) OF NON-MAGNETIC FRACTIONS OF FLY ASHES
            FROM MIDWESTERN BITUMINOUS AND WESTERN SUB-BITUMINOUS
                 COALS (D.F.S.  NATUSCH, UNPUBLISHED RESULTS)
Type of coal
Midwestern coal



Western coal



Particle
size
(um)
<20
20-60
60-90
>90
<20
20-44
44-74
>74
Percent of
i
<1.6 1.6-2.0
*
1.4 1.3
0.7 1.0
0.1 0.1
*
0.2 0.4
0.5 0.8
0.2 0.3
total mass in density ranges
ndicated (g/cm3)
2.0-2.3
0.2
12.1
0.6
0.1
*
0.5
1.0
0.4
2.3-2.7
28.0
12.9
1.1
0.6
0.7
21.3
45.6
16.1
2.7-3.0
*
0.1
0.6
0.5
*
0.3
0.6
0.2
>3.0
*
*
0.1
0.2
*
0.2
0.5
0.2
   Less than 0.05%.
electrostatic precipitators increase with decreasing fly ash resistivity.
Bickelhaupt (1974; 1975a, 1975b) has further shown that both the surface and
volume resistivities of fly ash, at precipitator operating temperatures, are
inversely proportional to the specific concentrations of alkali metals that
are thought to act as charge carriers.   These studies have shown that
considerable differences in electrical  resistivity occur between different
fly ashes, and correlations are observed between fly ash resistivity and
alkali metal content.

     In addition, resistivity changes with temperature.   This is illustrated
in Figure 8 (Kropp et al.,  1979) where a maximum is observed at about 180°C.
It was found (Kropp et al., 1979) that the use of a wet scrubber, of either
Venturi or centrifugal type, helped reduce resistivity.   As a result, the
collections efficiencies of in-line electrostatic precipitators was improved,
probably due to reduction in flue gas temperatures.

     No studies relating particle size to either surface resistivity (Bickel-
haupt 1975a) or volume resistivity (Bickelhaupt, 1975b) have been reported.
                                      23

-------
    TABLE 5.   MASS DISTRIBUTION (WT %)  OF  MAGNETIC  FRACTIONS OF  FLY ASHES
            FROM MIDWESTERN BITUMINOUS  AND WESTERN  SUB-BITUMINOUS
                 COALS (D.F.S.  NATUSCH,  UNPUBLISHED RESULTS)
Type of coal
Midwestern coal



Particle
size
(Mm)
<20
20-60
60-90
>90
Percent of
i
<2.1 2.1-2.5
* 0.6
0.2 0.6
0.5 0.8
0.1 0.2
total mass in density ranges
ndicated (g/cm3)
2.5-2.9
0.4
1.8
1.0
0.3
2.9-3.4
1.3
11.5
0.2
0.2
3.4-3.6
14.9
3.1
*
0.1
>3.6
0.5
0.1
*
*
Western coal



<20 * * *
20-60 * * *
60-90 0.1 0.1 0.2
>90 * * *
*
1.0
6.8
1.5
* *
* *
0.1 0.1
* *
 * Less than 0.05%.
However, a number of authors have reported changes in sodium and potassium
concentrations as a function of particle size (Loh,  1975;  Davison et al.,
1974; Coles et al. , 1979; Klein et al.,  1975; Cambell et al., 1978; Block and
Dams, 1976).   These results indicate that some, but not all, coal fly ashes
show increasing sodium and potassium concentrations in smaller particles.
Sodium and potassium concentrations also have been reported to decrease  with
decreasing density (Loh, 1975).  Similar, though less pronounced, density
dependencies are observed in the magnetic fractions, but size dependencies,
if any, are obscure (Loh, 1975).  Since  both decreasing density and decreas-
ing physical  size contribute to decreasing aerodynamic size, it is apparent
that the efficiency of electrostatic precipitation per unit mass of size-
classified fly ashes showing concentration trends increases with decreasing
aerodynamic particle size.  This is an extremely desirable characteristic.
It should be pointed out, however, that  since not all fly ashes show such
concentration trends, improved collection efficiencies for small particles
will not occur in all cases.
                                      24

-------
            .On
        I-
        o
           0.5-
        e>
        u
           1.0
        h-
        o
        <
        oc.
        o
        UJ
           0.5
                          —  Crushed
                          -  Uncrushed
                 2.1  2.5  2.9
                 2J  2.5 2.9
                                             B
2.1  2.5 2.9
2.1  2.5 2.9
   v
                           DENSITY(g/cm

              PARTICLE DIAMETER: A -  <20^m, B-20'44jjm
                                    C - 44-74>im,D->74>jm
Figure 7.  Mass distribution of size and density classified coal fly ash
          with and without crushing.  The shift to higher densities after
          crushing is indicative  of hollow or vesicular  particles.
          (Reprinted by permission of A. Loh from Loh, 1975.)
                                   25

-------
        10
          14
        10
          IS
     E
     o

     E
        10
     en
     to
     LU
        10'
        10'
O  MEASURED VALUES
   AT ~9 %H20
                          • COMPUTER  VALUES
                   3.0


                   60
                   140
  2.6
2.2
1.8
   112      182      282
  223     360      540

      TEMPERATURE
 1.4


442
827
Figure 8.   Resistivity as  a  function of reciprocal absolute temperature
           for two different coal  fly ashes.  Upper curve Na20 0.25 weight
           percent of ash.   Lower  curve Na20 2.31 weight percent of ash.
           (Reprinted by permission of the Air Pollution Control Association
           and R.  E.  Bickelhaupt from Bickelhaupt, 1975a.)
                                     26

-------
SURFACE AREA DISTRIBUTION

     The specific surface area of fly ash particles is an important parameter
in determining a number of behavioral characteristics.  It is the surface
area of a particle which determines the electrostatic charge which can be
placed on that particle in an electrostatic precipitator (White,  1963;
Bickelhaupt, 1974; 1975a, 1975b); it is the surface area of a particle which
determines the extent of condensation or adsorption of species from the gas
phase (Davison et al., 1974; Natusch and Tomkins, 1978); and it is the sur-
face area of fly ash which significantly affects the rate and extent of its
aqueous leaching (Natusch et al., 1975; Matusiewicz and Natusch,  1979).

     To a reasonable approximation, one would expect the specific surface
areas (m2/g) of fly ash to increase linearly with decreasing particle diam-
eter since the particles are predominantly spherical.   Similar trends would
also be expected for non-spherical  particles having similar shape factors
(Butcher and Charlson, 1972).  In general, the specific surface area, S , is
given (Exner and Hausner, 1974) by:

                                  S   =  —                            (2)
                                  \     Dp                            U;

where a = shape factor, D = mean particle diameter (pm), and p = particle
density (g/m3).

     The shape factor varies from 1 for spherical particles to 20 for irregu-
lar, elongated particles.  The specific surface area is directly proportional
to the shape factor and shows great variability of S  with the morphology of
the particles.                                      w

     Specific surface areas of a size and density fractionated midwestern
bituminous coal fly ash are given for illustrative purposes in Table 6.   All
measurements were made by adsorbing and desorbing the inert gases nitrogen
and helium.

     In general, the expected trend of increasing area with decreasing parti-
cle size was observed.  However, there are exceptions, especially in the
largest size fractions (>74 pm).   It would be expected that these fractions
would have a smaller number of particles per unit mass and consequently would
have a lower specific surface area, but the opposite was true.  The particles
in these fractions are more irregular in shape and apparently exhibit a
significant internal surface areas as well.  Such behavior would result from
a particularly porous surface, surface crystal formation or cracks that
permitted the adsorbate gas to penetrate to the interior.  The morphological
studies reported in section IV support this.  The possible influences of the
small encapsulated and/or the hollow particles on surface area are unknown
but could provide an explanation for the discrepency.

     Several of the fractions show no significant dependence of surface area
on particle diameter.  These data suggest that the internal surface area is
effectively proportional to particle volume rather than external  surface
area.  In this regard, it has recently been suggested (Natusch, 1978b) that
                                      27

-------
collisionally efficient condensation processes may result in deposition of
material from the gas phase predominantly onto the external  particle surface,
whereas much less efficient adsorption processes (Natusch and Tomkins, 1978)
can deposit gases and vapors on both the internal and external surfaces of a
particle.

     In general, the specific surface areas can be seen to increase as the
densities of the particles decrease.  This is attributed to the fact that
there are more particles per unit weight of sample and more hollow particles
in the less dense fractions.

     The magnetic and non-magnetic fly ash fractions, although not identical,
show similar trends.  However, the large specific surface areas of the
larger, heavier magnetic fraction compared to their non-magnetic counterparts
do stand out as an exception.  Possibly these particles have a more porous
structure or a greater internal surface area.

     The results in Table 6 appear to be lower than those reported by some
authors (Kaakinen et a!., 1975; Miguel, 1976) but about the same as those of
Watt and Thorne (1965).  It should be noted, however, that only Miguel (1976)
reports data for fractionated fly ash, so direct comparison between the data
of these authors is not possible.  It does appear that an upper specific
surface area limit of about 4-5 m2/g is reasonable for some bulk fly ashes,
especially ashes having a high proportion of non-spherical particles.
        TABLE 6.  SPECIFIC SURFACE AREAS (m2/g) OF COAL FLY ASH FRACTIONS
                        FROM A MIDWESTERN BITUMINOUS COAL
                 Particle size (pm)
                                       S  density ranges (g/cm3) indicated
                                        w
                                         <2.1   2.1-2.5   2.5-2.9   >2.9
Non-magnetic



<20
20-44
44-74
>74
0.35
0.31
0.22
0.96
0.35
0.29
0.26
0.25
0.33
0.27
0.24
0.13
0.30
0.26
0.24
0.29
Magnetic



<20
20-44
44-74
>74
__*
0.31
0.37
1.16
0.30
0.28
0.22
0.68
0.24
0.14
0.15
0.18
    No meaningful data.
                                      28

-------
                                  SECTION 5

                  CHEMICAL CHARACTERISTICS OF COAL FLY ASH
MATRIX ELEMENT COMPOSITION AND DISTRIBUTION

     Elemental analyses of fly ash shows that the major (matrix) elements are
Al,  Fe and Si together with up to a few percent of Ca, K, Na, S, and Ti.
These elements are generally present as oxides, carbonates, sulfates, or
silicates.  Fly ashes from western sub-bituminous coals usually exhibit
higher levels of calcium than do midwestern bituminous coals.  Numerous
authors (Klein et al., 1975; Ondov et al., 1975; Coles et a!., 1979; Cambell
et al., 1978; Bickelhaupt, 1975a) have reported analyses for the matrix
elements in different fly ashes.  The reported compositions vary over con-
siderable ranges.  The range of values, as well as the values for the
National Bureau of Standards (NBS) Standard Reference Material, SRM 1633, are
reported in Table 7.

     To further characterize the distributions of fly ash matrix elements, a
midwestern sub-bituminous coal was analysed with respect to particle size,
density, and ferromagnetism.  The results of these analyses are given in the
tables in Appendix B.  However the dependence of concentration on these
variables can be found by examination of Figures 9 to 16.  Reference should
be made to Figure 6 for the mass distribution of the fractions.

Silicon and Aluminum

     The highest concentrations of aluminum and silicon were found in the low
density, non-magnetic fractions as shown in Figures 9 and 10.  This is
because of the abundance of alumino-silicates (from optical microscopy data)
and increasing amounts of mullite (Si02) in the low density fractions (from
X-ray diffraction data).  It is noteworthy that both silicon and aluminum are
still present at significant concentrations in the high density, magnetic
fractions.  This is probably because of the presence of encapsulating parti-
cles (see Section IV) or particles which contain mixed phases.  It can also
be seen that aluminum and silicon concentrations do not vary greatly with
particle size, confirming data from optical and electron microscopy.

Iron

     The iron concentrations in the separated fly ash samples are presented
in Figure 11.  The highest concentrations of iron are found in the most
dense, magnetic particles.  Data from optical microscopy and X-ray diffrac-
tion have shown the presence of magnetite and hematite in the magnetic


                                     29

-------
    TABLE 7.   TYPICAL MATRIX  ELEMENT  COMPOSITION  RANGES  OF  SOME  AMERICAN
                    AND  BRITISH  COAL  FLY  ASHES  (WT %  OF  ASH)
Element
Al
Ca
Fe
K
Na
S
Si
Ti
American*
7-16
0.4-22
2-15
0.2-2.9
0.2-1.7
0.2-1§
10-28
0.4-1.6
Britisht
12-18
1.2-5.5
4-10
1.5-3.5
0.2-1.4
0.3-0.8(1)
1.9-24
0.4-0.7
NBS-SRM 1633H
12.7
4.7
6.2
1.6
0.3
NR£
21
0.7
   * Data adapted from Bickelhaupt (1975a)  for several  American ashes.

   t Data adapted from Watt and Thorne (1965)  for 14 British ashes.

   H Data of Ondov et al.  (1975) for the NBS standard reference material
     (SRM) 1633.

   § Soluble S042-.

   <|> Reported as  S03.

   4 NR = Not reported.
fractions which explains the high concentrations of iron that are present.
The iron concentrations in this fly ash sample are high because of the high
pyrite content of Illinois coals used in the power plant (Rao and Gluskoter,
1973).  The iron is still present at relatively high concentrations in the
least dense, non-magnetic fractions.

Sulfur

     Sulfur concentrations can be seen to be higher in the non-magnetic
fraction but do not show any significant dependence on particle size or
density (Figure 12).   Sulfur was not found to be associated with iron,
indicating that most of the pyrite in the coal had decomposed during coal
combustion.

                                     30

-------
*o
36

32
g2B

Z 24
0
^20
o:
•- 16
^^ 1 w
III
UJ
O 12
§ a
4
ft
NON-MAGNETIC




*
DENSITY(g/cm3)
<2 1
2. -2.5
«-, 2.5-2.9


•


•

\

\
\
S
\
\
s
s
\
s
__ >2.9
• ,"B
:';/
•X
•.•.
'•:'•'.
;'•''.

—









r\
^
\
\
\
\
\
\
\
\
\
•^
•:|:
•:•:
::>'
I;.'-
.•»•

1




••)




PC
A
s
\
\
s
s
\

;^.-
;•;•
:•'.'•
:';/
:-::

—









-s
\_
_!ii

i







.

rr~m






\
N
\
\
\
\
\

*.*,


X;
;':';:
!••'
__





-

.
•

          <20
             <20
20-44          44-74
PARTICLE SIZE  (/xm)
28

Co 24
Z 20
O
^ 16
a:
z l2
LJ
0 8
Z
0
0 4

O
•
MAGNETIC
•
m

•

.

,













—
















MB


^^
\
\
N
\
\
\
\
\
\
\
\
\
•


t • t *
• . *,
•x
X;

f * t •

)',*













•ft





V?
•.' ,
1 * , *
*, '.
;.'•!
'•:]:
•'.•'.
,• ,'
:•:•

~"











-

.

-
.


  20-44         44-74
PARTICLE  SIZE (/xm)
>74
Figure 9.   Concentration of silicon in non-magnetic and magnetic coal fly
          ash fractions as a function of particle size and density.  (Data
          of Loh, 1975.)
                                31

-------
oc
rt A
28
?24
O 20
1-
< ifi
o: |e
CONCENT
O *. 00 fo

•
•
.

•
,




^ 24
Z 20
O
2:.9|I]B 	
•^
•*•
__r


\
\
\
\
^^^


l ^
:;j:
^^
U| -

<20 20-44 44-74 >74











PARTICLE SIZE (/xm)
MAGNETIC















—







H
1
N
\
^
\
\
\
\
•

j:|:
I;!;
•':•:
•:•':
— i




TH



•':•:
.*.•
•v
•.*•':
•!•;






•

•

             <20
 20-44         44-74
PARTICLE  SIZE
>74
Figure  10.  Concentration of aluminum in  non-magnetic and magnetic coal
           fly ash fractions as a function of particle size and density
           (Data of Loh, 1975.)
                                  32

-------

>0 44
1* 40
Z 36
g 32
tr 28
QC 24
h- 20
UJ 16
O 12
0 8
0 4
n
NON-MAGNETIC

; DENSITY (g/cm3) ;
f ^ i

r
/
-
" r-IH
__J!v| 1 l^n
* ^^H
1. 1-2.5 :::;:
1.5-2.3
>2.9 m



1"
•

         <20
                        20-44          44-74
                        PARTICLE SIZE (/im)
^ 48
O^ 44
*"" 40
^ 36
 or
 l-
  -
 o
   28
   24
   20
   16
   12
    8
    4
   0
                             MAGNETIC
                              I
            <20
                          20-44         44-74
                         PARTICLE SIZE (pun)
>74
Figure 11.   Concentration of iron in non-magnetic  and magnetic coal  fly
           ash fractions as a function of particle size and density
           (Data of Loh, 1975.)
                                33

-------

1.4
§? 1.2
_
O '-O
P
< .8
h-
Z .6
LJ
O
Z A
0
0
.2


-
_

-

-

_


.


- j^
|\











^MM
*.",
;•>
^i-
;!;':
NON- MAGNETIC













~|
1 F
—
*•

\
*•
•*
-••

















n;
Ic












~
;.";*.
.•.'
* •

-
r-i DE:NSITY(g/cm3) _
^^rt













<2
J.I
m 2.1-2.5
2.5-2.9
>2.9









>
*.'.'



-








\[
\
s


:•:•
—






_


_


-

         <20
20-44          44-74
PARTICLE SIZE
>74
    1.2
 g
 s
 o:
 a. 4
 8.2
   MAGNETIC
            <20
  20-44        44-74
 PARTICLE SIZE  (/im)
>74
Figure 12.  Concentration of sulfur in non-magnetic and magnetic coal fly
          ash fractions as a function of particle size and density.
          (Data of Loh, 1975.)
                               34

-------
Potassium

     Figure 13 shows the concentration of potassium in coal  fly ash fractions
separated by size, density and magnetic properties.

     Potassium concentrates in the low density particles in both the magnetic
and non-magnetic fractions.  Potassium may be associated with the glassy,
alumino-silicates which predominate in the less dense fractions.   The
presence of encapsulating particles or particles with different phases could
explain the significant quantities of potassium found in the magnetic
fractions.   Potassium also appears to be more concentrated in the smaller
particles indicating that the element may exhibit vaporization-condensation
partitioning behavior.

Sodium

     Figure 14 shows that sodium concentrations are also higher in the small
size, low density fractions, following trends similar to those for potassium.
The non-magnetic fractions show higher concentrations than the magnetic
fractions as for potassium, but the differences between the two appear to be
more pronounced.  It is uncertain whether this difference is significant.

Calcium

     Figure 15 shows that calcium concentrations increase with increasing
density and particle size for the non-magnetic particles but appear to be
fairly evenly distributed in the magnetic particles.  The reasons for this
are not clear but could be due to the presence of amorphous calcium compounds
since X-ray powder diffraction did not show any calcium containing crystals.
It should be noted that, even though the high density, non-magnetic fractions
contain a high concentration of calcium, the total mass of these fractions is
small.  Consequently, the high density, non-magnetic fractions contribute
only a small percentage of the total bulk sample calcium.

Titanium

     Titanium appears to be quite homogeneously distributed in the separated
fly ash fractions.  The data were not plotted but are presented in Appendix
B.  Titanium would not be expected to volatilize in appreciable quantities
during coal combustion and would tend to remain with the same elements with
which it was associated in the original coal.

     The elements discussed above comprise the major and minor constituents
of the fly ash matrix.   They are also the elements which tend to be present
at higher concentrations (~0.05 to 3% by weight) in the coals burned; their
presence as minor to major constituents in the ash is not surprising.  Their
combined presence in the glassy composite of the ash, primarily as mixed
oxides, carbonates, and ferro- or alumino-silicates strongly influences the
physical, chemical, and solubility properties of the ash (see discussion
below).
                                      35

-------

-2.8
~ 2.4
0
g2.0
<
(V _
H '-6
Z
1 1.2
8 -s
.4
n
NON-MAGNETIC


DENSITY (g/cm3
-

*•

•
-
•

\
\

\
\
\
\
\
\
\
^

I*',*
<:[
'.•'.•
'.•'.•
*« •
•'.•;
:•*:•
^•M








•k




0
\
\
S
\
\
s
\
\
^
<2.l
_, 2.1-2.5
*.*•
',•'.'
:•:•
*"•*
•*•*
"«*•
• •*
'•!•
2.5-2.9
-. >2.9





••


\
N
\
\
\
\
\

,*•*
, •*
•*•"
.*.*
,*•*

—


i
•••
HI
hU




-








fN
\
S
\
\
\
^
\

-
:•:•
• •*
*•*•
— i
—
m
          <20
             <20
20-44          44-74
PARTICLE SIZE  (/im)
 20-44         44-74
PARTICLE SIZE (ftm)
 >74
— 2.4
^ 2.0
O
L 1 »^5
^T

h- L2
LU o
0 '8
z
0 .4
O
n
MAGNETIC
-



~


-

rim
1
II



T^T
*• •
'. .
.*.•
•.*,
;•;•
••.';
•'.''.
•'.•;
•.*,
-

—










••


N
^
V
s^1
^
\
s
^


•,".
'\\,
;-j.
M-
;•>
,',•

^







rm
|

II
rs;
S
\
S
S
s
::::

;!\
•;;':
X;
'.•',•
;.';^
—










1

||
>74
Figure 13.   Concentration of potassium in non-magnetic and magnetic coal
           fly ash fractions as a function  of particle size and  density.
           (Data of Loh, 1975.)
                                36

-------
«HJUU
3800
36OO
3400
-*• 3200
*o>3000
^2800
*-" 2600
2 2400
2 2200
< 2000
CC 1800
t^ 1600
LU 1400
^ 1200
O 1000
O 800
600
400
200
; NON-MAGNETIC

-
_
•
_
-
-
-
-
-
-
T
\
S
,
\
,
>
S
\
\
\
s
s
s




^m
•':•;
t* t*
•:•'.
•/*
yl
:>;
•.*.

"•








-
-
DENSITY (g/cm3) ^
<2.l








Ps
\
\
\
N
V
^
^
\
S
'. *,

;!;•:
'•!]
•'.•;
••/•:
•:•:
:'•:•
2. -2.5
2.5-2.9







H


13
\
$

^
TT
'.•;•

''.''.

•:'•:

>2.9



__




J
ii:
01






.








^
s^
V
>
^
N
",* ,


.• *
•i '.
\. .
•: ;

•
—
-
-
-
—


—
••
-
           <20
             <20
 20-44           44-74
 PARTICLE  SIZE
    >74
3ZUU
3000
— * 2800
\ 2600
O>
^2400
"^ 2200
Z 2000
o
^ 1800
 ;•
;:•:
•;.';
•:•:
•.'•'.

—
-
—
^"^













V|



••^
\
\
\
\
\
\
\
s
B« t»
.*.*
•';•:
•i-'.
'.•'.;
;:•*.
".;'.•.
t+t*
:'•''•
















Pvl
\

V
V
v
\
^S

—
-
-
-
-
_
~
 20-44         44-74
PARTICLE SIZE (/im)
>74
Figure 14.  Concentration of sodium in non-magnetic and magnetic coal  fly
           ash  fractions as a function of particle size and density.
           (Data of Loh, 1975.)
                                 37

-------
16
14
S|2
z
O 10
F
< p
or 8
[-
Zj 6
0
0 4
0
2
NON- MAGNETIC
DENSITY (g/cm3)

- _,

-

•
_
•
S
s
s
s
s
\
s
s
\



r~i K~I
••••
•X
•*• *
1 • '



1
N
\
^
\
•Mi
;X
•X
I
—





<2
.1
2.1-2.5
2.5-2.9



..—
^
i •*
'•*,

>2.9




J
•I-'.-
~
in

•



-

r\
\
'•:[:
rl-
-
         <20
20-44          44-74
PARTICLE  SIZE (/im)
>74
10
14
z
O 10
\-
§ 8
\-
^^™*
O
0 4
o
2


MAGNETIC
~
_

-
-
-

-

—









— i
•. •.
•".*
*. .
•*»
•y.
•'.•i
* •
*,*.
,*•*

—























• •
.•»"
:•:'
.*,•
•:•':
•X
• •*
\\'
v'.
^~

















-^
\
\
\
N
\
•X
•.'•*
X*

*•"•
—









-
—
-

-
1 '
-

-

            <20
 20-44         44-74
 PARTICLE SIZE  (p,m)
>74
Figure 15.  Concentration  of calcium  in non-magnetic and magnetic coal fly
           ash  fractions  as a function of particle size and density.
           (Data of Loh,  1975.)
                                 38

-------
MINERAL COMPOSITION

Mineral Composition of Coal
     Identification
more readily with a
                    of the chemical compounds present in fly ash can be made
                    knowledge of the inorganic minerals present in coal.   The
mineral composition of coal varies widely with the origin of the coal but a
summary of the most common types of minerals which occur in coal is presented
below (Rao and Gluskoter, 1973; Pringle and Bradburn, 1958; 0'Gorman and
Walker, 1971; Gluskoter, 1967; Littlejohn and Watt, 1963; Reid, 1971).

Clay Minerals--
     These are the most commonly occurring inorganic minerals in coal and
consist primarily of kaolinite [Al203*2Si02«2H20], and montmorillonite
[(Mg,Ca)-Al203'5Si02-4H20].  The general formula for these minerals is
MAl2(Si205)(OH)4.

Sulfide Minerals--
     These consist of pyrite [FeS2, cubic], marcasite [FeS2, rhombic], and
sphalerite [Zn, Fe S].  General formula MS or MS2.

Carbonate Minerals—
     These consist of calcite [CaC03], dolomite [Ca(Mg,Fe), (C03)3] and
siderite [FeC03].  General formula (Ca, Mg, Mn, Fe) C03.
Halide Minerals--
     These consist mainly of sylvite
formula MX (X = F, Cl, Br, I).
                                     (KC1) and halite (NaCl).  General
Silicate Minerals—
     These consist mainly of quartz [Si02].

Iron Minerals—
     Among the most common in this group are magnetite [Fe304], hematite
[Fe203] and lepidocrocite [Fe203-H20].

Shale Minerals--
     These minerals, closely related to the clay minerals, consist of illite
[(K,H20)(Al2[AlSi]Si3010)(OH)2], muscovite [K20-3Al203-6Si02-2H20] and
biotite [k(MgFe)3(A!Si3010(OH)2].

Accessory Minerals—
     There are a large number of minerals whose concentrations are heavily
dependent on the particular coal source.  Among these minerals are feldspars
[K,Na)20-Al203-6Si02], garnet [3CaO-Al203-3Si02], hornblende [CaO-3FeO-
4Si02], gypsum [CaS04-2H20], apatite [9CaO-3P205-CaF2], kyanite [Al203-Si02],
tourmaline [HgAl3(BOH)2Si4019], aluminum phosphate A12(P04)3, and zircon
[ZrSi04].  Many more minerals may be present in even smaller quantities.
                                     39

-------
Mineral Composition of Coal Fly Ash

     X-ray powder diffraction was used to identify those minerals present in
coal fly ash.   Samples of a separated midwestern bituminous coal  derived fly
ash were first ground to a particle size of less than 44 urn then a thin film
of particles was placed in the diffractometer.

     Figure 16 shows X-ray diffraction patterns for magnetic and non-magnetic
fly ash particles.   The magnetic particles in the < 20 urn size fraction
exhibited diffraction peaks which can be identified to originate from
magnetite (FesO^) and hematite (Fe203).   The bluish-black,  opaque spheres (as
observed by optical microscopy) are thus definitively classified as magnetite
and the reddish-brown spheres as hematite.  The non-magnetic particles showed
diffraction peaks from two minerals; quartz (Si02) and mullite (3Al203'2Si02).

     Since only crystalline material with regularly repetitive atomic planes
result in sharp diffraction peaks, it is this material which is easiest to
identify.  The broad bands observed in the diffraction patterns indicate the
presence of amorphous compounds.   These bands are attributed to the glassy
aluminosilicates observed by optical microscopy.  The presence of quartz and
amorphous materials in the magnetic fraction show the overlap of the ferro-
magnetic separation of fly ash particles.  No magnetite diffraction peaks are
observed in the diffraction pattern of the non-magnetic particles, indicating
that ferromagnetic separation is not only efficient in removing most of the
magnetic particles, but also removes some non-magnetic particles that are
agglomerated with the magnetic particles.

     Figure 17 shows X-ray diffraction patterns for the density fractions of
non-magnetic fly ash particles in the 20-44 urn size fraction.  Small quanti-
ties of hematite are present in the density fraction between 2.5 g/cm3 and
2.9 g/cm3.  This is because hematite is only weakly magnetic and is not
efficiently separated by the magnet.  Both quartz and mullite are present in
significant quantities in the least dense fraction even though their densi-
ties are greater than 2.1 g/cm3.   This may be attributed to encapsulating
particles (see Section IV) or particles which contain mixed phases (a single
particle containing two or more compounds of varying density).  The ratio of
the areas under the broad band to the quartz peak (20 = 2.67°) indicates that
the glassy, aluminosilicate particles also tended to be more concentrated in
the least dense fraction.

     The four main minerals in fly ash were found to be magnetite, hematite,
quartz and mullite.  Magnetite and hematite are probably formed from the
oxidation of iron containing minerals (such as pyrite, marcasite and
sphalerite) present in the coal.   Quartz particles can be from the original
quartz in the coal or formed from the breakdown of clay minerals in the coal.
On heating, clay minerals (such as kaolinite) first dehydrate and then break
down to form mullite at 1200-1400°C (Brindley and Nakahira, 1957).

     These results are consistent with other work (Simons and Jeffrey, 1960;
Watt and Thome, 1965; Hulett and Weinberger, 1979) where it also was found
that the principal mineralogical components of fly ash were quartz, mullite,


                                     40

-------
     z
     UJ
     H-
     Z
     UJ
     >
     UJ
     o:
                   NON-MAGNETIC
                    H-hematite
                    Q-quartz

                    Mu-mullite
          M
                          M
          MAGNETIC
        H-hematite
        Q-quartz
          -magnetite
      60*
       50*
   40°        30°
26 ANGLE
20°
Figure 16.
X-ray powder diffraction patterns of non-magnetic and magnetic
fly ash particles.  (Reprinted by permission of A. Loh from
Loh, 1975.)
                               41

-------
         z
         UJ
         I-
         z
         UJ
         UJ
         or
                    DENSITY  2.lg/cm*
                       Q-quortz
                       Mu-mullite
DENSITY-2.1-2.5 g/cm
  Q-quortz
  Mu-mullite
                     DENSITY = 2.5-2.9 g/cm
                        Q-quartz
                        Mu-mullite
                        H- hematite
                                  HH,MU
                   Mu ..    MuH
           SO5""    505      40°       30°
                           20  ANGLE
                           20°
Figure  17.  X-ray powder  diffraction patterns of density separated coal
           fly ash particles.  (Reprinted by permission of A. Loh from
           Loh, 1975.)
                                  42

-------
magnetite and hematite and a large amorphous glass fraction.   The non-
crystalline fraction ranged between 52 and 87% for a number of British fly
ashes (Simons and Jeffrey, 1960; Watt and Thome 1965) and between 28 and 80%
for the American fly ashes that were studied (Simons and Jeffrey, 1960;
Hulett and Weinberger, 1979).

TRACE ELEMENT COMPOSITION AND DISTRIBUTION

Dependence of elemental concentration on particle size, density and
ferromagnetism

     It is now well established that many high temperature combustion and
smelting operations emit particles containing toxic elements such as Be, Cd,
As, Se, Pb, Sb, Hg, Tl and V into the atmosphere (Schroeder,  1971).  Many of
these elements are enriched in ambient urban aerosols by as much as 100-1000
fold over their natural crustal abundance (Gordon and Zoller, 1973).  In
addition, it has been reported that a number of toxic elements (e.g., Sb, Cd,
Ni, V, Sn and Zn) in urban areas have mass median equivalent diameters (MMED)
of the order of one micrometer or less.   This diameter is considerably
smaller than those reported for such common matrix elements as Al, Fe and Si
having MMED values in the range of 2.5-7.0 urn (Rahn et al., 1971; Lee et al.,
1972; Colovos et al., 1974; Gladney et al., 1974).  It is therefore meaning-
ful to determine whether certain toxic trace elements predominate in the
smallest particles emitted from participate sources or whether the mass
median diameter differences in urban aerosols are simply due to mixing of
particles characteristic of individual source emissions.

     The work reported in this section was conducted in two parts.  In the
first section, the midwestern fly ash that had been previously fractionated
was examined to establish what trends in trace element composition might
occur as a function of particle size, density and ferromagnetism.  The
results are given in Appendix C.  Representative figures based on the tables
in Appendix C are given in the discussion that follows.

     The second section involved a much more detailed study in which the
concentrations of a large number of elements were determined strictly as a
function of particle size.  In this study the fly ash was highly fractionated
and samples from both within the plant as well as airborne material were
examined.  A large number of analytical  techniques were employed in order to
examine the relative merits of each for fly ash analyses and to establish the
data firmly.

     Both studies were designed to establish whether elements present in fly
ash particles emitted from coal fired power plants exhibit a dependence of
element concentration on particle size.   A summary follows.

Magnesium—
     Magnesium appears to be homogeneously distributed in the separated fly
ash fractions.  Since magnesium is not expected to volatilize in appreciable
quantities during coal combustion, it would be expected to remain with the
same elements with which it was associated in the original coal.


                                      43

-------
Manganese--
     This element is more concentrated in the high density,  large particle
size, non-magnetic fractions as shown in Figure 18.   Manganese is expected to
be only partially volatilized during coal combustion.   The trends in manga-
nese are similar to those of several other elements, especially barium and
arsenic and, to some extent, those of the matrix element,  calcium (see
Figure 15).

Arsenic--
     This element shows trends very similar to those of manganese (see Figure
18), with the highest concentrations generally found in high density, large
particle size, non-magnetic fractions.  Arsenic and its compounds can be
volatilized in the combustion of coal and deposit on the surface of fly ash
particles.   This condensation or adsorption process may be more favorable for
non-magnetic particles.  The higher concentrations in the larger particle
fractions may be due to associations with refractory minerals in coal.  These
minerals are often not completely volatilized and tend to form larger irreg-
ular particles rather than small, spherical particles.  The results are
somewhat analogous to those of Lamb (1975) who found arsenic to be more
concentrated in the low density, large particle size, non-magnetic fractions
of roadside dust.  He attributed this to selective adsorption of As by a soil
component.   A similar type of adsorption mechanism may be occurring in fly
ash.  However, it has been pointed out (Kaakinen et al., 1975) that As is
mostly retained as arsenate/arsenite in high-calcium coals.   In fact, compar-
ison with the results for calcium (Figure 15) indicates that As correlates
well with calcium revealing the probable existence of Ca-As compounds.

Barium—
     The data show that barium is concentrated in the high density, non-
magnetic fractions.  This distribution is similar to that of manganese and
arsenic (see Figure 18) indicative of a similar partitioning mechanism.  In
contrast to calcium, the element does not concentrate in the magnetic
particles.

Strontium--
     Strontium may be slightly more concentrated in the non-magnetic  frac-
tions, but otherwise it is relatively evenly distributed.

Cesium—
     Cesium is only present in trace quantities in fly ash.  Being an alkali
element, cesium would be expected to follow the concentration trends  of
potassium and sodium.  However, the data show that cesium is more concen-
trated in the magnetic particles.  Cesium and its compounds are expected to
volatilize in the combustion of coal thus the predominance of cesium  in
association with the magnetic particles may be due to selective deposition.

Rubidium--
     Rubidium appears to behave as an alkali metal but no definitive  conclu-
sions can be deduced from the data.
                                      44

-------
     3200
   •
     2800
     2000
     leoo
LU
O

O
O
     1200
      800
      600
                            NON- MAGNETIC
                                                 DENSITY (g/cm3)
                                                     2.1-2.5
                                                     2.5-2.9
                                                       >2.9
                                                         3
             <20
                         20-44          44-74
                         PARTICLE SIZE
   >74
     2800


   "^2400
                            MAGNETIC
     1600

   74
Figure 18.   Concentration of manganese in non-magnetic and magnetic coal
           fly ash fractions as  a function of particle size and density
           (Data of Loh, 1975.)
                                45

-------
Zinc—
     Zinc is concentrated in the small  particle size,  high density,  magnetic
fractions as shown in Figure 19.   Zinc  has been postulated to undergo vola-
tilization-condensation partitioning,  hence its concentration in small
particles.   Zinc also occurs with iron  as a sulfide mineral  (sphalerite)
present in coal and this could explain  its preference  for magnetic particles.

Chromium--
     Chromium is much more concentrated in the magnetic fractions and shows
trends similar to those of zinc (see Figure 19).

Cobalt-
     Cobalt is also more concentrated in the magnetic  particles, with concen-
trations increasing as the density increases.   The distribution of this
element also resembles that of zinc (see Figure 19).

Dysprosium, Europium, Hafnium, Lanthanum, Scandium, Samarium,
Tantalum, Thorium, and Uranium--
     These elements do not generally show any significant size/density
trends.  The data for Eu, which is fairly typical, is  shown in Figure 20.   In
general, there were some slight non-magnetic/magnetic  differences.  Thus,  the
elements La, Hf, U, Dy and Sm are more  concentrated in the nonmagnetic frac-
tions, the elements Sc and Th are more  concentrated in the magnetic fractions
and the elements Eu and Ta do not show  any significant variation in concen-
tration.  The reason why these elements are concentrated in certain fractions
is not known but is attributed to their presence as constituents of the
minerals in coal.  All of these elements have high boiling points and are  not
expected to volatilize to significant extents in the high temperature com-
bustion of coal.  Thus, after coal combustion, these elements would be
associated with the same mineral elements found in fly ash particles.

Gallium—
     Figure 21 shows that gallium is definitely more concentrated in the
small particle fractions.  Gallium and  its compounds can volatilize in the
high temperature combustion of coal and deposit on the surface of particles.

Antimony--
     Figure 22 shows that antimony is more concentrated in the smaller size
fractions and, to some extent, in the magnetic fractions.   Antimony has been
shown to exhibit volatilization-condensation partitioning, hence its high
concentration in small particles.  Antimony also appears to be more concen-
trated in the higher density fractions  as is arsenic but this trend is not
definitive.

     The concentration data for the different elements was also compared with
the specific surface area data given earlier in Table 6.  The comparison
indicates that even the volatilizable elements (Sb, Ga, As), which are
expected to be surface deposited, do not show simple linear relations with
the specific surface area as would be expected if these elements condensed on
the particles with the greatest specific surface area.  The largest particles
have high surface areas (by virtue of their porous structures), but this is


                                     46

-------
 §^1000




Z 800
  Ld
o
O
     400
     200
                            NON-MAGNETIC
                                               DENSITY (g/cm3)
                                                  2.1-2.5

                                                  2.5-2.9

                                                     >2.9
                               nm
                                           Jffll
                                       i M. . rnllll
            <20
                          20-44          44-74

                         PARTICLE  SIZE
                                                          >74
,O> 1200



^.lOOO


Z
O 800




<£ 600
UJ
O

o
o
     4OO
     2OO
                              MAGNETIC
             <20
                          20-44        44-74

                        PARTICLE SIZE
                                                        >74
Figure 19.   Concentration of zinc in non-magnetic and magnetic coal fly

           ash fractions as a function of particle size and density
           (Data of  Loh, 1975.)
                                47

-------
o>
^?^ X* X*
^6.0
3,
— • 5.0
24.0
^
Ct: 3.0
U] 2.0
O

o 1.0
0

NON-MAGNETIC
—

•
-

•
"r;



»*
•• *
'•\ :
*• •
—



v%
rr
^
[S







•*
,*
*







...






<20 20-44









sj
^









•* ,

—
DENSITY (g/cm3)
<2.l
2.1-2.5
2.5-2.9



~"


m >2.9
J
ITT
ni
•
—

~
rm —

IN*.'
Nf.:



44-74 >74
                         PARTICLE SIZE (/im)
                            MAGNETIC
O 4.0


o: 3-°


&-°
o
2 1.0
O
O
                                       rs
                                       S
            <20
                         20-44         44-74

                        PARTICLE  SIZE (/im)
>74
Figure 20.   Concentration of europium in non-magnetic and magnetic  coal fly

           ash fractions as a  function of particle size and density

           (Data of Loh, 1975.)
                                48

-------
^> izu
o>
0 80
j —
<
ct:
fe ^o
LlJ
O
O
o

NON-MAGNETIC
-


-


Hs
\
\
>
\
\
\
\
\
HMM
*
•
*.
."
•
•. .



DENSITY (g/cm3)
^.^ • ixl





<>C.. \
2.1-2.5
[1 2.5-2.9
>2.9
"^ __^ITTB
xil nrt
^f

in



rim

          <20
 20-44          44-74
PARTICLE SIZE (/xm)
 >74
 ,0>I20
  o>
 O 80
 fee
 a:

 S 40
 o
 o
                             MAGNETIC
            <20
 2O-44         44-74
PARTICLE SIZE (/im)
>74
Figure 21.   Concentration of gallium in non-magnetic and magnetic coal  fly
           ash fractions as a function of particle size and density.
           (Data of  Loh, 1975.)
                                49

-------
^60
? 50
••*»
Z 40
g
^ 30
o:
^ 20
UJ
g 10
8
NON-MAGNETIC

DENSITY (g/cm3)
<2.l
• FT_ 2.1-2.5
P™"*™


.

>
\
\
S
«^
*. .

*. .
.* •
:• •





	 2.5-2.9
fm

^F
1 KB
—



>2.9
TTfl

rE
•T
\


3
•"•
r—
Hi
           <20
20-44          44-74
PARTICLE SIZE (/xm)
   >74
^0) OU
^.50
g ^O
^30
1-
Z 20
LU
O
Z 10
O
O
MAGNETIC
-
-

_


-


—

















^mtm

*

'• •

*

—


i





-


TT)


r^
K
II KN
/*•
'.';.
' t " t
.'•'.'
*, ' *
";.'•








-

_

na
M
kN
N
             <20
 20-44         44-74
PARTICLE SIZE (/xm)
>74
Figure 22.   Concentration of antimony in non-magnetic and magnetic coal  fly
           ash fractions as a function of particle  size and density.
           (Data of Loh, 1975.)
                                 50

-------
only partially reflected in the elemental concentration data.   It is note-
worthy to point out that this fly ash sample was obtained from within the
plant, so that the particles are relatively large compared with particles
obtained from the stack.  Any dependence of concentration on surface area is
thus diminished.   Kaakinen et al.  (1975) also collected particles in bulk
from within the plant and obtained linear correlations between the elemental
concentrations and the specific surface area.  When the effects of particle
size, density and ferromagnetism are delineated, however, no simple correla-
tion is found.

     The dependence of the trace element concentrations on particle size,
density, ferromagnetism, matrix elements and specific surface area has
enabled the classification of three groups of trace elements:

     Group I—those not expected to volatilize during coal combustion and
showing either a preference (Sr, Co, Sc, Th, U) or no preference (Eu, Dy, Hf,
Ta, Sm, La) for certain particle matrices.  These elements are present in
trace amounts in fly ash and evidently remain associated with the same matrix
elements with which they are present in coal.

     Group II--those expected to be volatilized during coal combustion and
showing a preference for small particles and for certain particle matrices
(Na, Cs, Ga, Sb).

     Group Ill—those expected to be only partly volatilized during coal
combustion and showing a preference for certain particle matrices.   The
elements Zn and Cr exhibit a preference for small particles and have a
behavior similar to Group II.  The elements Mn, Ba and As showed no signifi-
cant preference for particle size and are postulated to be associated with
the same matrix elements in fly ash and in coal.  Arsenic has a distribution
similar to Mn and Ba and is included in this group even though As is expected
to almost completely volatilize during coal combustion.

     It is apparent from the foregoing remarks that these results provide
only a limited understanding of the distributional characteristics of the
constituents.  Consequently a second, more exhaustive, effort was undertaken
in order to better examine the relationship between particle size and
elemental concentration.

Dependence of elemental concentration on particle size

     Fly ash analyses were carried out for twenty-five elements as a function
of physical or equivalent aerodynamic particle diameter.  Samples included
both stack (airborne) and precipitator samples and reflected more narrow size
fractions than those reported earlier in part V.  The results are graphically
presented in Figures 23 to 32 and are also summarized in Appendix B.  Note
that the precipitator ash was physically sized while the stack ash was sized
aerodynamically.   These two quantities are related by Equation 1.  Note that
separation of density or magnetic fractions was not carried out as previ-
ously.  Fly ash particles larger than 74 urn exhibited no dependence of
element concentration on particle size, so the concentrations shown for this
fraction are averages over all larger fractions.  The figures show that for


                                     51

-------
      O>


     3 60
     §40
     u
     o

     o
     o
        20
                                THALLIUM


                 AIRBORNE FLY ASH    PRECIPITATOR FLY  ASH
             1.1 2.1 3.34.773II.3»I.3
                                   <5 8 10 15 203O>404474>74
Figure 23.   Concentration of thallium in coal fly ash fractions as a

           function of particle size.  (Data of Davison et al., 1974.)
    60
  o»
 040

 fee
 UJ
    20
 O
 o
                                ANTIMONY
               AIRBORNE FLY ASH
                                       PRECIPITATOR FLY ASH
                                           TTTL
         I.I 2.1 3.34.773II.3>II.3          <5 5 10 15 20 30>404474>74



                        PARTICLE  SIZE
Figure 24.
           Concentration  of antimony in coal  fly ash fractions as  a

           function of particle size.  (Data  of Davison et al., 1974.)
                                 52

-------
 ^60
 5 40
 o:
 LJ
 O
 o
    20
                               BERYLLIUM

              AIRBORNE FLY ASH      PRECIPITATOR FLY ASH
         I.I 2.1 3.3 4.7 7.31 l.3>l 1.3           <5 510 15 2030>404474>74

                        PARTICLE SIZE
Figure 25.  Concentration of beryllium in coal fly ash  fractions  as a
           function of particle size.  (Data of Davison et al.,  1974.)
 z
 g
 £80
 a:
 O 40
                                 COBALT

                AIRBORNE FLY ASH     PRECIPITATOR FLY ASH
          I.I 2.1 3.3 4.7 73II.3XI.3           <5 5 10 15 203O >404474>74

                        PARTICLE  SIZE
Figure 26.  Concentration of cobalt in coal fly ash fractions  as a
           function of particle size.   (Data of Davison et al., 1974.)
                                 53

-------

       1600
    Z 1200
    O
     I-  800
     UJ
     O

     8  400
                ARSENIC

AIRBORNE FLY ASH     PRECIPITATOR FLY ASH
               1.1 2J134.7 73II.3>II.3
                                       <5 5 K> I52030>404474>74
                            PARTICLE  SIZE

Figure 27.   Concentration of arsenic in coal  fly ash  fractions as a
            function of particle size.   (Data of Davison et a!., 1974.)
      o>
      UJ
      o
        1200
         800
      8  400
                                       LEAD

                _.,  AIRBORNE FLY ASH     PRECIPITATOR FLY ASH
                 2.1 3.3 47 7.3 II.3XI.3
                   <8 5 K> I82030>404474>74
                             PARTICLE  SIZE
Figure 28.   Concentration of lead in coal  fly ash  fractions as a function
            of  particle size.  (Data of Davison  et al., 1974.)
                                   54

-------
        1200
     £   800
     z
     UJ
     o
     O   400
                                      NICKEL

                     AIRBORNE FLY ASH    PRECIPITATOR FLY ASH
                1.1 2J 3.3 4.7 7.3 II3>II.3           <5 5 10 IS 20 30>40 44 74>74

                             PARTICLE SIZE
Figure 29.   Concentration  of nickel in coal fly ash fractions  as a function
            of particle  size.  (Data of Davison et al.,  1974.)
     z
     O
     tr
        2400
        IOOO
     o
     o   800
                                    CHROMIUM

                     AIRBORNE FLY ASH    PRECIPITATOR FLY ASH
                I.I 2J3.34.7T3IL3XI.3          <5 5 10 I62O3O>404474>74

                             PARTICLE  SIZE
Figure 30.   Concentration of chromium in coal  fly  ash fractions as a
            function of particle size.   (Data  of Davison et al., 1974.)
                                   55

-------
     Z
     g
     LU
     O


     O
     O
        600
        400
200
                                 MANGANESE


                    AIRBORNE FLY ASH    PRECIPITATOR FLY ASH
               I.I 2.I3.34.773II.3>II.3
                                      <8 8 10 15 20 30>404474>74
Figure 31.  Concentration  of manganese in  coal fly ash fractions as a

           function of particle size.   (Data of Davison et  a!., 1974.)
 ^» 600
 o>
 jt.


 O 400

 fee
 fr


 g 20°
 o
 z
 o
 o
                           VANADIUM


      —I  AIRBORNE  FLY ASH    PRECIPITATOR FLY ASH
             I.I 2.I134.773II.3>II.3
                               <5 5 K> 1520 30>4D4474>74
                           PARTICLE SIZE
Figure  32.  Concentration of vanadium in  coal fly ash fractions as a

           function of  particle size.  (Data of Davison et al., 1974.)
                                  56

-------
this particular ash the elements As, Be, Co, Cr, Ni,  Pb, Sb, and Tl  are
preferentially found in small particles.  The trends  for Mn and U are less
clear but suggest that similar trends exist for the  smallest particles.

     A large number of investigators have reported that many trace elements
show a preferential concentration in smaller particles.  Lee and von Lehmden
(1973) found an essentially similar dependence of element concentration on
fly ash particle size for Cd, Pb, Mn and Cr.  Toca (1973) found a similar
trend for Pd and Cd and noted that 70% of the Cd present in flue gases was
associated with particles <5 pro in diameter.  Sparks  (1973) reported that the
elements Pb, Ba, Sr, Rb, As, and Zn in fly ash particles collected on a
0.4 urn filter following the last stage of a Brinks impactor were enriched on
a weight for weight basis by at least an order of magnitude over those
deposited on the last impactor stage.

     Recently several reports in which concentration  trends for a large
number of elements were studied have appeared in the  literature.   The most
comprehensive of these, in addition to the data just  presented (Davison et
al., 1974), are those of Block and Dams (1976), Klein et al. (1975), Cambell
et al. (1978), Coles et al.  (1979), and Smith et al.  (1979).  The results are
generally consistent in classification of elements to Group I (no trends),
Group II (increasing concentration with decreasing size) and Group III
(intermediate behavior or concentration decreases with decreasing size).   A
comparison of the classification of different elements is given in Table 8.

     The comparison of the classifications of the various elements given in
Table 8 reveals several interesting facts.  First, while most of the matrix
elements consistently fall within Group I and many of the more volatile
elements into Group II, many elements either exhibit  intermediate behavior or
else fall into different categories depending on the  particular study.  The
discrepancies between results are probably a function of the many different
coals, combustion conditions and collection and analysis systems utilized by
the different workers.

     In spite of the problems inherent in classifying the Group III  elements,
the result of this and the other studies on size dependence do support four
significant points.

      1.  A coal fired power plant produces enrichment of certain elements in
          the smallest emitted particles.

      2.  The highest concentrations of these trace  elements are found in
          particles which may deposit in the pulmonary region of the
          respiratory system.

      3.  Existing particle control devices are least effective for  removing
          the particles bearing some of the most toxic elements.

      4.  Estimates of toxic element emissions based  on analyses of  undif-
          ferentiated fly ash collected from particle precipitators  will  be
          much lower than actual emissions.


                                     57

-------
 TABLE 8.   SUMMARY OF ELEMENTAL CONCENTRATION DEPENDENCE ON PARTICLE SIZE*
A. Elements
As
Br
Cd
Cu
showing
clear inverse concentration-particle size dependence
Cl Mo Sb
Ga Nb§

B. Elements showing
size dependence


C. Elements
Al
Ba
Bit
Ca
AgIT
AulF
Be|
Cs
showing
Ce
Co
Dy
Eu
IH Pb
InlF S
less well establi
Cr
Mn
Na
Ni
Se
Tl
Zn
shed inverse concentration particle
Ut|i
U
W
Zn§
no concentration-particle size dependence
Fe Mg
Hf Nd
K Rb
La Sc
Si Ta Yb
Sm Tb
Sn<|> Th
Sr Ti
* Based on the data of Loh (1975); Davison et al.  (1974); Coles et al.
  (1979); Klein et al. (1975); Cambell et al. (1978); and Block and Dams
  (1976).  Elements listed reflect the classifications of a majority of
  the authors cited.

t Data of Klein et al. (1975) and Block and Dams (1976) only.

IT Data of Block and Dams (1976) only.

§ Data of Cambell et al. (1978) only.

<)> Data of Davison et al. (1974) only.

4 Data of Davison et al. (1974) and Coles et al. (1979) only.

t|i Data of Coles et al. (1979) and Klein et al.  (1975) only.
                                    58

-------
     In fact, only a small fraction of the total fly ash mass has particle
diameters less than 10 pm and by no means all of this is emitted to the
atmosphere.   However, the fraction emitted undoubtedly presents a greater
potential health hazard per unit weight than that retained (see also the
discussion of element solubilities below).

Explanations of Preferential Elemental Concentration on Small Particles

     Several theories has been suggested to explain the preferential associa-
tion of different metals with small particles.  These are generally based on
one of three general explanations; volatilization and subsequent condensation
or adsorption of the particular element, geochemical fractionation, or the
"bursting" of larger particles.

     In support of the volatilization and adsorption/condensation hypothesis
it is noteworthy that all the elements (except Nb) listed in Table 8A have
boiling points comparable to or below the temperature of the coal combustion
zone (1300-1600°C) (see Table 9).  This is also true of Ba, Sr, and Rb as
determined by Sparks (1973).  This implies that metal compounds are reduced
to the element before volatilization.  However, while reduction in the com-
bustion zone is certainly feasible, such reduction is not necessary to the
basic hypothesis.  It is suggested that these elements have access to the
vapor phase as sul fides, halides, oxyhalides or even as carbonyls, whose
highly transient formation during coal combustion has been postulated
(Schroeder,  1971).  Mercury, of course, undoubtedly volatilizes as the
element and is predicted to exhibit a dependence of concentration on particle
size if the proposed mechanism is valid.

     A simple model can be constructed by considering a single particle in

which an element, X, is uniformly deposited on the particle surface at a

concentration C  (ug/cm).  In addition, X is assumed to be uniformly distri-

buted throughout the particle with a concentration C  (ug/g).  The total

concentration of X, Cx ((jg/g). is then given by:
                                  o
                                  °      pV

 where X  = element, C  = bulk concentration of X, C  = surface concentration

 of X, GX = total concentration of X, V  = particle volume, A  = particle

 surface area, and  p  = particle density.  By summing over all fly ash

 particles and assuming spherical particles, the average concentration of X,

 Cy, is given by
                         _   _  _       6CS     1
                         L   —  L   +    -   •
                                         p      D

where the bars denote average values and D the particle diameter.


                                     59

-------
  TABLE 9.   BOILING POINTS OF POSSIBLE INORGANIC SPECIES EVOLVED
                      DURING COAL COMBUSTION


   Species  boiling or                        Species boiling or
   subliming < 1550°C                        subliming > 1550°C


As, As205,  As203, As2S3                     Al,  A1203

Ba                                          Be,  B&O

Bi                                          Bi203

Ca                                          C

Cd, CdO, CdS                                CaO

Cr(CO)6, CrCl3, CrS(1550)                   Co,  CoO, CoS

K                                           Cr,  C,  Cr203

Mg                                          Cu,  CuO

Ni(CO)4                                     Fe,  Fe203, Fe304, FeO

Pb                                          MgO, MgS

Rb                                          Mn,  MnO, Mn02

S                                           Ni,  NiO

Se, Se02, Se03                              Si,  Si02

Sb, Sb2S3,  Sb203(1550)                      Sn,  Sri02

SnS                                         Ti,  Ti02, TiO

Sr                                          U,  U02

Zn, ZnS

Tl, T120, T1203
                               60

-------
     In order to substitute the values in Figures 23-32 in Equation (4) it is
necessary either to make the assumption that p is not a function of particle
diameter, D, or to determine p for each size fraction.   Appropriate values of

D for each size fraction can be obtained by assuming

                                (ECD)n  +  (ECD)p
                          D  =	                      (5)
                                        2

where (ECD)  and (ECD)- are the upper and lower 50%-effective cutoff

diameters for each stage of the Andersen cascade impactor stack sampler.

Equation (3) thus assumes a symmetrical distribution of the mass of X over

the diameter range (ECD)  to (ECD)0.  Incorporating these assumptions
                        U        ^jL>         	
allows construction of a plot of (L versus D l, as depicted in Figure 33,

from which it can be seen that the results are in at least qualitative

agreement with the proposed model.

     The thickness, £, of the deposition layer can be estimated from the
expression

                                   S- = Cs/p'                          (6)

where p' is the density of the deposition layer, which was assumed to

approximate 3 g/cm3.   Values of C , C , and £ are presented in Table 10.

These values, with the notable exception of those for sulfur, are considered
reasonable for a thin surface-deposited layer.  The 600A layer thickness
obtained for sulfur is considered too great to be accounted for by a simple
adsorption process.  Indeed, the high concentrations of S obtained for the
0.65-1.1 urn size fraction (Appendix C) can be accounted for only if sulfur is
present as the element.  This suggestion is at variance with the findings of
Hulett (1973), who has shown, using electron spectroscopy, that S predom-
inates as sulfate.  We, therefore, consider that the sulfur values listed in
Appendix C are all proportionately high due to lack of a fly ash standard
having sulfur deposited on the surface of appropriately sized particles as
required for X-ray fluorescence analysis.

     If observed dependences of element concentration on particle size are,
in fact, due to volatilization followed by adsorption or condensation as
suggested, one would expect the same phenomenon to be exhibited by particles
derived from all high-temperature solid-combustion operation.  Data for
combustion sources other than coal-burning power plants are not widely avail-
able to substantiate this suggestion.  If correct, however, it suggests that
many sources may preferentially emit small particles enriched in toxic ele-
ments or their compounds.  The mass median diameters of such elements in the
emitted particle distribution will thus be reduced as a direct result of
preferential surface deposition.   The extent of reduction can be determined
by combining Equation (4) with the mass distribution function appropriate for


                                     61

-------
      TABLE  10.   ELEMENTAL  SURFACE CONCENTRATIONS AND  LAYER THICKNESSES
                CALCULATED* USING THE SURFACE DEPOSITION MODEL
Element
Pb
Tl
Sb
Cd
Se
As
Zn
Ni
Cr
S
Sample
pairs
6
6
6
6
6
6
6
6
6
4
Cs
(ug/cm2)
0.04
0.003
0.003
0.002
0.004
0.009
0.6
0.1
0.3
19
Co
(Mg/g)
1000
40
20
10
0.7
600
6000
100
-300
5 x 10
Linear
correlation
coefficient
0.73
0.80
0.93
0.99
0.92
0.97
0.60
0.98
0.94
--
Estimated
surface
thickness (A)
1
0.1
0.1
0.7
0.1
0.3
20
4
9
600
     Calculations were based on the  airborne  particle  data  in  Appendix  C
     assuming p = 3 g/cm3.   The slope  of the  line  in Figure 33 can  be
     obtained from C  by multiplying by 104.
a given particle source.   In the case of a log-normal  distribution this  gives
the following:
     dM
       X   _
    d(lnD)
                         CQ exp  -
     (In D/DJ2

     2(ln ag)2
6C
-7- exp
(In ag)2
                                                                         (7)
                       exp
(In D/D  + In2  ag)

     21n2a_
where a  = geometric standard deviation, MY = mass of X,  D  = mass median
       g                                  ^9
diameter of original distribution, C  = bulk concentration of X,  Cg = sur-
face concentration of X, p  = particle density, and D  =  particle diameter.
                                     62

-------
              1500
           o»
           v>
           9  looo -
           L_
           O
           UJ
           O

           O
           O
              500
                             0.25        0.50

                      (PARTICLE DIAMETER)'' (microns)"
Figure 33.   Dependence of As,  Ni,  and Cd concentrations in coal  fly ash on
            particle size.   (Reprinted by permission of D.  F.  S.  Natusch and
            the American Association for the Advancement of Science from
            Natusch et al.,  1974.   Copyright American Association for the
            Advancement of Science 1974.)
                                   63

-------
     Equation (7) does not provide a simple analytical  expression for the

mass median diameter D (X) except when C  = 0,  i.e.,  X  is present only in the

deposited layer.   In this case,  it can readily  be shown (Butcher and Charl-

son, 1972) that


                          InD (X) = InD  - In2a                        (8)
                             y         y       y

Equations (7) and (8) demonstrate that the mass median  diameter of a surface-
deposited species, X, is considerably less than that  of the median diameter
based on total particulate mass.

     Several authors (Block and Dams, 1975; Coles et  al. , 1979; Klein et al.,
1975) have suggested that the preferential concentration of certain elements
in small particles is a result of their lithophilic or  chalcophilic proper-
ties.  Lithophilic elements are those generally found associated with
alumino-silicate materials, while chalcophiles  are associated with sulfide-
containing minerals.  This classification of elements generally relies on the
scheme proposed by Mason (1966),  although slight differences exist with other
reports (Krauskopf, 1972; Nicholls, 1968).

     Most reports indicate that Al, Ba, Be, Ca, Cs, Hf, K, Mg, Mn, Na, Rb,
Ta, Th, Ti, U, V, and the rare earths (Ce, Dy,  Eu, La,  Nd, Sm, Tb, and Yb)
are classified as lithophiles.  Similarly, As,  Cd, Ga,  Mo, Pb, Sb, Se, and Zn
are chalcophiles.  Elements that may be either  lithophilic or chalcophilic in
their behavior include Cr, Co, Cu, Fe, Ni, and  W.  In terms of small particle
behavior, the lithophiles belong mainly to Group I, showing little prefer-
ential concentration on small particles.  The Group II  elements, which are
much more easily volatized during combustion, are generally chalcophiles.
Group III elements include, but are not limited to, those elements that may
belong to both classes.

     In general, this geochemical classification scheme is consistent with
the volatilization-condensation theory suggested.  The  ultimate behavior of
the different elements is still  determined by the coal  and combustion facil-
ity, but prior analysis of the coal based on its geochemistry may permit
better predictions concerning which elements, if any, will be preferentially
concentrated in small particles.

     Recently, a third hypothesis concerning trace metal-small particle
relationships was proposed (Smith et al., 1979).  In  this study it was found
that concentrations of trace elements that increased  with decreasing particle
size in the range of 1 to 10 urn became independent of size below 1 urn.  As as
result, the authors suggested that gas expansion caused a "busting" of larger
particles, with the resulting production of a large number of small parti-
cles.  These small particles then coagulate into the  0.1-1.0 ym size
particles.  Condensation would occur concurrently with  the busting and coagu-
lation process.  Data to support this hypothesis are  largely circumstantial.
                                     64

-------
Elemental Enrichment in Coal Fly Ash

     If the concentration of toxic trace metals in small particles is a
widespread phenomenon, as has been suggested, then one would expect volatile
elements present in urban aerosols to have significantly lower mass median
diameters than nonvolatile elements.  In fact, data obtained by the National
Air Surveillance Network (NASN) have shown that the more easily volatilized
elements Zn, Ni, Pb, Cd, and Ba have statistically lower mass median dia-
meters than common nonvolatilizable (particle matrix) elements.  In addition,
a number of toxic elements, including Pb, Se, Sb, Cd, Ni,  V, Sn, and Zn, in
urban aerosols have been reported to have equivalent mass  median diameters of
the order of one micrometer or less, which is considerably less than those
reported for common matrix elements such as Fe, Al, and Si, whose mass median
diameters lie in the range of 2.5-7.0 micrometers (Rahn et al., 1971; Lee et
al., 1972; Colovos et al. , 1974).   Ondov et al. (1979) also reported this to
be true for particles emitted to the atmosphere from electrostatic precipi-
tators and especially from scrubber systems or coal-fired power plants.
Also, although not substantiated statistically, Se and Sb have been shown to
have small mass median diameters in ambient aerosols (Gladney et al., 1973;
Ondov et al., 1979).  Lead, of course, is not expected to be typical, since
it is derived mainly from a single source (the automobile) known to produce
small particles.

     The predominance of certain elements in small particles is also signifi-
cant in determining the degree of enrichment of these elements in an urban
aerosol, since the smallest particles have the longest atmospheric residence
time.  Indeed, Gordon and Zoller (1973) have shown enrichment factors of
greater than ten times over natural crustal abundance for Tl, Cr, Ni, Cu, Zn,
As, Cd, Sn, Pb, Se, S, Cl, and Br in the Boston aerosol and have established
substantial correlations with enrichment patterns in coal  fly ash, municipal
incinerator fly ash, and residual  fuel oil ash.  In the present context, it
is noteworthy that the majority of these elements could be volatilized during
combustion.

     A number of authors (Kaakinen et al., 1975; Gladney et al., 1976; Ondov
et al., 1977) state that enrichment of toxic trace elements in coal fly ash
may be quite significant.   Some of the enrichment factors  compared with the
original coal samples are given in Table 11; these tend to exceed a factor of
three enrichment.

Volatile Species in Coal Fly Ash

     Although we have considered only trace elements present in particulate
matter, the importance of vapor species, especially of selenium, arsenic, and
mercury, should not be overlooked.   Thermodynamic data (Hodgeman, 1963)
indicate that, at 25°C, as much as 80 ug/m3 of Se as Se02  and 70 ug/m3 of As
as As203 can exist as vapors.  These levels are much greater than normally
observed for Se and As in urban aerosols (^ 10 ng/m3).  It is possible,
therefore, that additional amounts of these elements may be emitted as
vapors.  Consistent with this suggestion, Pi 11 ay and Thomas (1971) have
reported that at least 50% of the Se present in urban air passes through a
filter designed to collect all particles greater than 0.1 urn in diameter.


                                     65

-------
          TABLE 11.   ENRICHMENT  FACTORS  FOR  IN-STACK TOTAL  SUSPENDED
                              PARTICULATE  MATERIAL
Element
As
Cr
Ga
Pb
Sb
Se
Zn
Gladney*
6.8
1.1
1.2
4.0
4.7
6.2
1.5
Kaakinent
3.1
NR§
NR
5.8
3.2
5.7
3.2
Ragaini and OndovlF
7.3
2.0
3.1
NR
6.7
2.5
6.4
      * Gladney et al.  (1976).

      t Kaakinen et al.  (1975).

      IT Ragaini and Ondov (1977).

      § NR = Not reported.
Klein et al.  (1975) and Andren et al.  (1975) report that 20-32% of the
selenium that passes through a coal-fired power plant does so as a non-
parti culate-associated vapor and that much of the selenium is present in the
zero oxidation state.   Kl^in et al.  (1975) also suggest that, based on mass
balance information, a minimum of 60% and probably about 90% of the mercury
entering the coal stream is emitted as the vapor.  Little information is
presently available for arsenic, but its behavior would be expected to
parallel that of Se and Hg.

INORGANIC CHEMICAL SPECIATION AND LEACHING CHARACTERISTICS

Chemical Compounds in Coal Fly Ash

     Studies of fly ashes derived from the oxidative combustion of coal have
defined, in part, the chemical forms of sulfur and carbon.  In some cases the
salts in which trace metals occur can be inferred.

     Electron Spectroscopy for Chemical Analysis (ESCA) shows that sulfur is
primarily present in the +6 oxidation state (Linton et al., 1977).  Particu-
lates derived from coal conversion processes that involve reducing conditions
                                     66

-------
also contain sulfur in the -2 oxidation state (Keyser et al.,  1978).   Neither
result is unexpected.   Time Resolved Solvent Leaching (TRSL)  studies of coal
fly ash, in which analyses of soluble anions are performed by  means of ion
chromatography, indicate that sulfate is the only sulfur-containing anion
leached by water.   It is probably, therefore, that the sulfur  species present
in the surface layer of coal fly ash is, at least predominantly and probably
exclusively, in the form of sulfate.   However, this must be tempered by the
fact that most elemental sulfides are highly insoluble.

     Some evidence is available regarding the cations associated with sulfate
species in coal fly ash.  Thus, X-ray powder diffraction analyses of some fly
ashes indicate the presence of either anhydrite (CaS04)  or gypsum (CaS04 •
2H20).  Fisher et al.  (1976) also suggested crystalline  anhydrite or gypsum
was present on the surfaces of some ashes, based on ion  microprobe analysis.
The two CaS04 species are present most commonly in fly ashes  derived from
western U.S. coals, which contain especially high levels of calcium.   The
second form results, apparently, from exposure of the highly hygroscopic
anhydrite to moisture.  In a sense, therefore, the occurrence  of gypsum is
probably artifactual.

     Quite strong indications have also been obtained for the  existence of
several trace metal sulfates in coal  fly ash.  Thus, both Fourier transform
infrared Spectroscopy (FTIR) and Time Resolved Solvent Leaching (TRSL) pro-
vide evidence for the presence of Cd, Co, Cr, Mo, and Ni sulfates in coal fly
ash.  The alkali metals Ba, Cu, and Ca are also present, at least partly, as
sulfates.  Even stronger leaching evidence is available  for the existence of
Al and Fe as sulfates in the surface layer of fly ash (Natusch and Tompkins,
1978).

     In the case of carbon, a number of alkali and alkaline earth carbonates
have been quantified at trace levels.  Part of the evidence comes from carbo-
nate decomposition as fly ash is heated with a slowly increasing temperature
ramp.  Evolved C02 was monitored by plasma emission spectrometry using an
atomic carbon wavelength.  The temperature of decomposition and associated-
C02 evolution was characteristic of the metal to which it is  bound.  In an
additional experiment using the same emission detector,  C02 was evolved by
acidification of the fly ash with aqueous HC1.  The total amounts of C02
released from each fly ash by thermal and aqueous methods were comparable.

     The thermal evolution data alone does not allow unambiguous identifica-
tion and quantitation of the carbonate compounds.  However, combined with the
acid evolution procedure, an accurate determination can  be made.   Thus, the
analyses of the leachate obtained from treating a fly ash with the amount of
acid just necessary to dissolve all the carbonate coupled with the leaching
time enables identification of all of the metals that were present as carbo-
nates.  The leachates were analyzed by atomic spectrometry for Fe, Li, Na, K,
Rb, Mg, Ca, Sr^and Ba.  These results, when combined with the thermal C02
evolution datawdefined concentrations or concentration  limits on carbonate
species (Table 12).  Note that the existence of carbonates seems to be a
general phenomenon, regardless of the source of the ash.
                                      67

-------
 TABLE  12.   UPPERLIMIT  CONCENTRATIONS  (ppm)* OF  CARBONATE  COMPOUNDS  IN  COAL
        FLY ASH.   NUMBER  IN  PARENTHESES  REPRESENTS  PERCENT DEVIATION
Compound
Li2C03
Na2C03
K2C03
Rb2C03
FeC03
MgC03
CuC03
SrC03
BaC03
Fly ash from indicated coal
Illinois #1
^220 (15)t
759-1090H (15)
570H (12)
^0. 32 (15)
2630§ (4)
191011 (14)
22701F (14)
1060-152011 (15)
8161T (12)
Illinois #2
^45.3 (14)
456 (19)
474 (19)
--§
438 (12)
^1710 (14)
1770-3870 (14)
797-1330 (8)
177 (19)
New York
^202 (26)
S438 (26)
g!680 (26)
--§
819 (14)
596§ (14)
487 (27)
^150 (26)
^676 (26)
Montana
^85.7 (1)
n29 (8)
^332 (8)
-§
517 (6)
1670-1770§ (1)
2000-2100U (1)
--§
3300 (11)
 * ppm of ash.

 t Percent deviation.

 II For the following sets of elements:   Mg and Fe in "New York";  Mg and Cu
   in "Montana"; Mg, Cu,  and Fe in "Illinois #1"; and Li, Na,  and Sr in
   "Illinois #1.  Contributions from the individual  metals could  not be
   resolved.  The numbers listed are maximum values  assuming no contribution
   from the others in the set.

 § No meaningful data.
     In several cases the mass balance indicates more metal  is Teachable than
can be accounted for by carbonates alone.   Table 13 presents the percent of
each metal that is bound in fly ash as the carbonate, as defined by the
leaching procedure.   Note that the procedure is not exhaustive; therefore,
these percentages represent upper limits.

     At present, there is no evidence to indicate whether the carbonate
compounds are associated with the surfaces of coal fly ash.   Nevertheless,
since the flue gas in which the ash is entrained contains high concentrations
of C02 and H20, a surface-preferred reaction is quite likely.
                                     68

-------
           TABLE 13.   CARBONATE SPECIES CONTRIBUTION TO LEACHABLE
                        ELEMENTAL CONCENTRATIONS. (%)
Fly ash from indicated coal
Element
Li
Na
K
Rb
Fe
Mg
Ca
Sr
Ba
Illinois #1
0-100
0.6-0.8H (16)t
7.811 (14)
0-100
4.911 (30)
3.011 (18)
0.811 (16)
9.3-13.311 (19)
5.611 (12)
Illinois #2
0-100
100 (19)
100 (19)
"§
7.2 (19)
0-100
1.7-3.8 (15)
"§
100
New York
0-100
0-100
0-100
"§
40. 1§
91.4 (45)
4.0 (28)
0-100
0-100
Montana
0-100
0-100
0-100
"§
4.3 (17)
1.6-1.711
0.711 (3)
23.2-38.7
58.3 (14)






(3)

(9)

 * ppm of ash.

 t Percent deviation.

 II For the following sets of elements:   Mg and Fe in "New York";  Mg and Ca
   in "Montana"; Mg, Ca, and Fe in "Illinois #1"; and Li, Na, and Sr in
   "Illinois #1; contributions from the individual metals could not be
   resolved.  The numbers listed are maximum values, assuming no
   contribution from the others in the set.

 § No meaningful data.
     Present indications are that a large fraction of the elements present
in fly ash are bound as sulfates and carbonates and that, while definitive
evidence is lacking, these elements are present in the so-called surface
layer.  It must be recognized that the actual compounds may be mixed salts.
For example, the existence of alkali iron tri-sulfates has been suggested by
numerous authors (Covey et al., 1945; Nelson and Cain, 1960; Weintraub et
al.,  1961), and various combinations of Ca, Mg, and Fe occur in naturally
occurring carbonate minerals.   No evidence has been found for the presence
of free H2S04 in fly ash particles.
                                     69

-------
AQUEOUS LEACHING OF COAL FLY ASH

Introduction

     The foregoing sections have concentrated on a description of research
directed towards establishing the physical  and chemical  characteristics of
coal fly ash.   In order to obtain some assessment of the probably environ-
mental impact of this material,  however,  it is necessary to obtain informa-
tion about the characteristics and behavior of coal  fly  ash between its point
of discharge into the environment and its adverse interaction with an
environmental  or biological system.

     In this last regard, it is  suggested that few,  if any, pollutant species
can have an adverse effect upon  any biological organism  unless that species
is capable of entering into solution either in water or  into one or more of
the several types of body fluids that may be encountered.   If one accepts
this hypothesis, it follows that determination of the solubility of various
pollutant species is of fundamental  importance to an assessment of their
capability for exerting a toxic  effect.

     It has been reported (Shannon and Fine, 1974; Theis and Wirth, 1977;
James et at., 1977; Dreesen et al.,  1977) that the bulk solubility of coal
fly ash in water is very low and rarely exceeds 2-3 percent by weight.   The
low bulk solubility is clearly a result of both the glassy and crystalline
matrix materials identified earlier, and one would expect elements that are
either chemically or physically trapped within this matrix to exhibit low
solubility.  On the other hand,  at least some of the material present in the
surface layer is readily soluble in water (Linton et al., 1977).  Indeed, it
is now quite well established (Linton et al., 1977;  Natusch, 1978a; Fisher et
al., 1978) that most of the soluble fraction of fly ash  is derived from this
surface layer and is thus quite rich in trace elements.

     A number of authors have concentrated on laboratory leaching (James et
al., 1977; Shannon and Fine, 1974; Fisher et al, 1979; Theis and Wirth, 1977;
Brimblecombe and Spedding, 1975; Cox et al., 1978; Eggett and Thorpe, 1978;
Phung et al., 1979), but few studies exist of natural systems (Talbot et al.,
1978).  Unfortunately, there is considerable confusion involved in inter-
preting and understanding results obtained from different studies of fly ash
solubility.  Specifically, quite different results are obtained by workers
using apparently similar laboratory leaching techniques and few are readily
transferable to field studies.   However, the work of Dressen et al. (1977)
is an exception.  In this study, laboratory leaching of fly ash was related
to effluent water measurements made at the same power plant.  Their results
showed that, among the elements found to be most extractable in the labora-
tory by either water (B, F, Mo,  and Se) or acid (As, B,  Cd, F, Mo, and Se)
leaching, many were also elevated in the effluent water (As, B, F, Mo, and
Se).  In general, however, factors that control leachate composition under
both laboratory and field conditions have not been well  studied, nor has an
attempt been made to standardize or compare different leaching techniques.
                                     70

-------
     Some insight into leaching behavior can be obtained by recognizing that

a soluble inorganic species,  M-A. ,  associated with fly ash particles can

dissociate into its component cations,  M- ,  and anions, A., in aqueous solu-

tion and that both dissolution and  deposition (e.g.,  precipitation)  processes

can occur in competition with each  other.   Furthermore, cations and  anions

present in solution can interact so as  to set up multiple equilibria, which

may involve ion pairing, complexation,  precipitation, or acid-base behavior

(e.g., M.A- or M.A.)-   The result can be expressed simplistically by the

following equations:
                               n         n
                           P + I M.  +   I A.
                              i=l 1     1=1 ]
                                                                      (9)
                               m         m
                               I A.      I M.
                               =  J      =  J
                         k_2 t 4- k2     k_3 t 4- k3

                         m   n           m   n
                         I   I M.A.  +   11 M-A.
                        1=1  =l 1 J     1=1  =l J n
Here P represents the parent (host) particles and k , k  ,  are the rate

constants for forward and reverse reactions, respectively.   It is apparent
from Equation (9) that, when leaching studies are conducted under batch
conditions such that the amount of fly ash and solvent are  maintained
constant, an equilibrium will be established between parti cul ate and solution
species.  Consequently, only a fraction of the potentially  soluble material
will enter solution, and some of this may be precipitated by other ionic
reactions.  On the other hand, if conditions are such that  soluble material
is continuously removed by providing fresh solvent or by providing a large
solution sink in the form of complexing ligands or acids, then all poten-
tially soluble material will ultimately be dissolved.

                                     71

-------
     It is apparent from the equilibrium outlined in Equation 9 that any
factor capable of altering the position of the general equilibria indicated
will directly affect the composition of the leachate solution from a fly
ash-solution contact.   Furthermore, it is reasonable to suppose that many of
these factors will also control the rate at which soluble species are leached
from the fly ash mass.  It has been the primary objective of the present
work, therefore, to identify those parameters that are capable of influencing
the position of equilibrium and thus of affecting leachate composition.   Such
factors have been shown to be the ratio of fly ash mass to solution volume,
the temperature of the solution, the pH of the solution, the presence of
complexing or precipitating agents capable of promoting or suppressing solu-
tion of individual species, the particle size of the fly ash, the origin of
the fly ash, and, perhaps most important, the method by which the leaching
solution and the fly ash are brought into, and maintained in, contact.   A
large body of data has been gathered with respect to establishing the
influence of these factors on leachate composition.   Specifically, emphasis
has been placed on the following objectives:

 1.  The need to establish standard methods whereby leaching can be conducted
     in different laboratories and the same results reproducibly obtained.

 2.  Determination of the characteristics of leaching and the composition of
     leachates under conditions closely representing those likely to be
     encountered in real environmental situations.

 3.  Establishment of a predictive model capable of determining the composi-
     tion of a given leachate under known field conditions.

     In order to address these objectives, several types of leaching proce-
dures were developed and the influence of the parameters listed above
determined for each.  The experimental methods, as well as the actual data,
are presented in Appendices A and D, respectively.  The fly ash used in these
studies was obtained from the Corrette Plant in Billings, Montana, which uses
a sub-bituminous coal  from Colestrip, Montana.

Continuous Single-Pass Leaching

     This method involves passing the leaching solution from a reservoir
through a bed of coal  fly ash and analyzing aliquots of the collected
leachate as a function of time (volume).  Emphasis is placed on maintaining
the fly ash bed in the configuration of a flat disk to reduce the possibility
of equilibrium being obtained between the fly ash and the solution actually
present in the bed.  For the same reason, a comparatively fast flow rate is
employed.  Details of the parameters actually employed in these studies are
given in the several tables describing the results (see Appendix D).

     It will be noted that in continuous single-pass leaching, fresh solution
is continuously being brought into contact with the fly ash so that the
equilibria indicated in Equation (9) are established at most only momen-
tarily, with the result that there is a net transfer of material from the
left- to the right-hand side of Equation (9).  Consequently, all readily
soluble material will  be removed from the fly ash so that a so-called

                                     72

-------
"exhaustive" leaching process is achieved.   Such a leaching process, when
carried to its logical conclusion of extremely long time periods, is, of
course, capable of completely dissolving all of the fly ash material away.

     Table 14 gives the pH of the leachate emerging from the column as a
function of leaching time (volume) through the column.   Three different pH's
were employed.  The results establish quite clearly that the material leached
from the coal fly ash contributes a net alkalinity to the leachate but that
the ability to leach alkaline material into solution falls off quite rapidly
with time (volume).

     The anion concentrations in the column leachate increased with decreas-
ing pH, as expected.  Furthermore, changing the fly ash solution contact time
showed that, for some species (e.g., sulfate), contact time can have a sig-
nificant effect upon leaching solution composition.  The data given here are,
however, not sufficiently definitive to establish whether its effect results
from partial establishment of equilibrium in the solution contained in the
bed or simply from different rates of solubility.  In the case of sulfate,
one inclines marginally to the latter view.

     Analysis of the changes in cationic concentrations indicate that

 1.  A number of elements, including cadmium, phosphorus, molybdenum, manga-
     nese, sodium, aluminum, calcium, barium, potassium, chromium, strontium,
     lead, and lithium, appear either to be present at higher concentrations
     near the external surface of the fly ash particles or are present in
     readily soluble form in fly ash.  If high solubility is, in fact, the
     case, it is expected that most of the material is near the surface, as
     has been discussed.

 2.  In contrast, several elements (notably magnesium) seem either to achieve
     equilibrium rather slowly with the leaching solution or to be exposed as
     leaching proceeds.  It is noteworthy that this behavior is observed
     under high, rather than low, pH conditions.

 3.  As a general rule, larger amounts of most of the trace elements listed
     are leached by more acidic leaching solutions.  Since many of the com-
     pounds are likely to be oxides, this result is not unexpected.

     Column leaching was also applied to different fly ash size fractions
with distilled water.  The data, however, proved to be inconclusive with
respect to the dependence of leachate concentration on particle size.  Thus,
consideration of only two smaller size ranges (20-30 and 30-45 urn) indicates
that the amount of material that can be leached increases with decreasing
particle size, as would be expected if the hypothesis of volatilization
followed by condensation of several trace metal species onto particle sur-
faces is valid.   On the other hand, the data presented for the large (150-180
um) sample would seem to refute this, and repeated experiments indicate that
the results are generally valid.

     Consideration of the morphologies of coal fly ash particles described in
an earlier section and of the specific surface areas listed in Table 6


                                     73

-------
















z
C£
LU
H-
CO
LU
3:
	 l
00
X
-
X LL.
O.

O «t
LU O
rv c_j
3
co en
LU O
t— t
Z £
»-H 3
t—
CO 1— t
LU CO
z i
X CO




rH
LU
_ 1
s
1—



























































/•™\

e
1 '

O)
E

^~
O

1
*
C

E


CD
E

r—

































O
CO
oo
1
o
oo
00
CM
O
00

ID
1
O
rH
CM
O
O

in
t
o
o
00
rH
O
rH

^>
1
O


i— t
O
LO
rH
CM
1
O
CM
r*-

o
CM
UD
rH
1
O

in

o
00
o
rH
1
O
CO
CO

o

in

o
3

o
LO
CO
1
0
CM
rH


rH
1
O
CO

m
i
in





o








0) C
C O

-C 4->
O 3
« r-
CU O
^ U>

o
oo

cri





o
co

o



in
ro

o
rH



in
o

o
rH




O
in

o
rH




0
r-

o
rH




in
O

r-i
rH



0
LO

r-i



O


r-J
rH


O
rH

CM
rH

m
O
CM



O
ro

CM
rH

O
CM
X


01


•^
+J
u>
.r—
•a

o
rH

CO





O
rH

CO



in
rH

CO




O
CM

CO





O
CO

CO





O
ro

ro





o


CO




o


o
rH


O
CM

r-J
rH


in
00

I—I
rH

in
CD
CM



0
CM

CM
rH









j.
CO

X
O.

o
CD

r-i





O
O

r-i



O
O

rH




O
O

r-i





O
O

rH





O
O

rH





O
O

r-i




0
O

r-i



O
0

r-i



O
O

r-J


O
O
rH



m
rH

r-i










^B
rH

X
O.


























































rt
o
z
X


in
rH
^

O
CM • "
X eg
X
^ ,_
O
0 0
o o
rH O
in
C
••- c
n '"~
O ro
Z O
X Z
X
u
c u
c o c
•r- O O
E U
\ •*-
r- 0 H-
e o
co 'e i—
u r~
00
CD * in
+J o t
a
i.

3 01 rH
0
f— X I
LL. 0. Q.

<•!-*•
74

-------
indicates that the large particles may have significantly different charac-
teristics than the small, predominantly spherical  particles.   It is conceiv-
able, therefore,  that these characteristics may overshadow the size trends
that would otherwise be expected.

Soxhlet Extraction

     The use of conventional soxhlet extraction offers a useful  method for
the preparation of leachates under standard conditions.   Since the process of
soxhlet extraction washes the solid fly ash sample with fresh aliquots of
solvent, it is possible, at least in principle, to quantitatively remove all
soluble material  and, as indicated for the case of continuous single-pass
leaching in the previous section, to dissolve the fly ash completely if
sufficient time is allowed to elapse.   Since each aliquot is  of comparatively
small volume (on the order of 10 ml),  the equilibrium (Equation 9) between
solid and dissolved species may be established in many cases  during the
period a given aliquot is in contact with a fly ash.   On the  other hand, the
overall soxhlet extraction process is an equilibrium situation in that the
material in the solvent leaving the soxhlet will be in equilibrium with the
batch solvent material below.

     Two types of analytical measurement are of potential interest with
respect to soxhlet extraction of coal  fly ash.  The first involves determina-
tion of the concentrations of individual species present in each aliquot
drained from the region of the extraction thimble.  Such measurements should
be roughly, though not rigorously, comparable to the measurements made in
continuous single-pass leaching described in the previous section.  The
second type of measurements involves determination of the concentrations of
individual species present in the bottom flask of the soxhlet extractor,
which accumulates all leached material.  In the second case,  however, it
should be noted that, as the concentration of many ion-cation pairs (e.g.,
sulfate and barium) in solution increases, precipitation may  soon result so
that both soluble and colloidal material may be present in the bottom flask.

     Soxhlet extraction has the advantage of being a simple,  well-
characterized procedure for the extraction or leaching of solid materials.
It is, however, limited in use to constant boiling mixtures,  so that, for
most purposes, only pure liquids can be used for leaching.  This may be a
severely limiting factor in situations where leaching with acidic solutions
or with solutions containing specific complexing agents may be desired.  In
addition, the temperature of the leaching liquid is generally only a few
degrees below its boiling point, which hardly represents realistic environ-
mental conditions.  This problem can be partly overcome by performing soxhlet
extraction under reduced pressure so as to lower the boiling  point of the
leaching liquid.   In the case of water, it has not proven practical to
operate below about 35°C because of difficulties encountered  with "bumping"
within the boiling flask.  For this reason, we have developed a useful
modification of the soxhlet extraction procedure whereby any  constant boiling
liquid can be presented to the material in the extraction thimble at a
desired temperature close to ambient.   The apparatus used for this purpose,
the operation of which is fairly self-explanatory, is illustrated in Figure
34.

                                     75

-------
Figure 34.   Apparatus for ambient temperature soxhiet leaching.   A,  bottom
            flask; B, heating tape;  C,  cooling condenser;  D,  extraction
            thimble; E, interval  sampling region;  F,  condenser to isolate
            soxhiet.
                                   76

-------
     Several samples of the bulk Corrette fly ash were subjected to soxhlet
extraction at room temperature with the apparatus depicted in Figure 34.   In
all cases, 10 gm of coal fly ash were leached at 25°C, and selected aliquots
of the leachate were sampled and analyzed for both cationic and anionic
species as a function of time throughout the extraction.   In addition,
samples of the solution in the bottom flask of the extractor were also
analyzed as a function of time and corrected for the dilution necessitated by
removal of these aliquots from the bottom flask.

     The results are noteworthy primarily because they illustrate great
differences between the concentrations of material found in individual
aliquots of leachate and those aliquots collected from the soxhlet bottom
flask.  Thus, concentrations measured in individual aliquots collected
directly from the solution that has been in immediate contact with the
soxhlet thimble give results generally similar to those obtained for
continuous single-pass leaching, as would be expected.  On the other hand,
analyses of material collected from the bottom flask rather dramatically
illustrate the operation of solubility equilibrium for several of the species
present, e.g., iron and fluoride.

     In general, however, the soxhlet extraction procedure provides no
particular advantage over the previously described method of continuous
single-pass leaching other than that of simplicity of unattended operation.
It is expected that the equilibria attained in the bottom flask of a soxhlet
extraction apparatus are likely to resemble those likely to operate in a real
environmental system.

Batch Leaching with Sonic Agitation

     This procedure involves sonic agitation of a suspension of the solid fly
ash in solution.  In contrast with the procedures described previously, this
method maintains the ratio of fly ash mass to solution as a constant and thus
achieves equilibrium between the solid and the leachate.   The information
obtained, therefore, is likely to be pertinent to situations (e.g., ponding
of fly ash) in which a constant volume of fly ash and water are brought into
contact.  On the other hand, if a very small amount of fly ash is sonically
agitated with a very large volume of water, it is possible to determine
approximately the total amount of leachable material.

     Not unexpectedly, sonic leaching results in the greatest variability in
concentrations of almost every element.  The multiple equilibriua previously
discussed clearly have considerable effects on the solution concentrations.
Changes in the amount of leaching solution did not result in a linear or even
uniform response in terms of increased concentrations.  This method lacked
reproducibility as well, indicating that it is is a generally less desirable
laboratory method for studying fly ash solubility.

SURFACE CHARACTERIZATION

     The surface predominance of potentially toxic elements has been demon-
strated recently in coal fly ash particles emitted to the atmosphere (Linton
et al., 1976).  Preliminary studies also indicate that elemental surface

                                     77

-------
predominance may be a general  phenomenon occurring in airborne particles
derived from high-temperature  combustion operations (Liriton et al.,  1976;
Linton et al.,  1977).

     These findings are of significance in part because they provide direct
confirmation of previous predictions (Davison et al., 1974; Kaakinen et al.,
1975) that surface predominance should occur.   The phenomenon is attributed
to the condensation of species previously volatilized in the high-temperature
combustion zone of a particulate emission source (Davison et al.,  1974;
Kaakinen et al., 1975).

     The existence of elemental surface predominance is also important in
demonstrating that bulk analysis techniques, which give only average concen-
tration values, provide little insight into the actual chemical  nature of the
particles.  For example, results obtained solely on the basis of bulk analy-
sis markedly underestimate the potential environmental impact of airborne
particles derived from high-temperature combustion operations in that (1)
surface regions of the particles have enhanced concentrations of elements,
including some that are potentially toxic, and (2) the smallest particles
will have much higher bulk (ug/g) concentrations of such elements because of
larger surface area to volume  ratios (Davison et al., 1974).

     On the basis of the discussion in Section V, it is apparent that there
is a need for analytical techniques capable of surface and in-depth elemental
characterization of individual particles.  Several approaches to obtain this
analytical information comprise the subject matter of this section.   Princi-
pal methods employed include the surface analytical techniques of secondary
ion mass spectrometry (SIMS) and Auger electron spectrometry (AES) as prac-
tices for microanalysis, i.e., ion microprobe mass spectrometry and scanning
Auger microscopy.

     This section also illustrates the benefits of a multitechnique approach,
not only by the application of several complementary surface microanalytical
techniques but also by the use of solvent leaching in conjunction with both
surface and bulk (spark source mass spectrometry) analyses.  Specifically,
the combination of leaching, bulk, and surface techniques demonstrates that
the observed elemental surface predominance is not the consequence of arti-
facts, especially those that may result from the use of ion sputtering to
obtain depth profiles.  It also permits semi-quantitation of elemental
concentrations in the surface region.

     In the case of coal fly ash, the information obtained by this approach
provides substantial insight into the physicochemical characteristics of fly
ash surfaces and permits assessment of its potential environmental impact.

     Ion microprobe depth profiles for 15 elements in fly ash are shown in
Figures 35-40.   These elements include all of the major and minor elements in
fly ash (greater than 1% by weight), as well as six trace elements with
concentrations below 1000 ppm, as determined by bulk analysis.  Other
elements of possible surface predominance and toxicity (Davison et al., 1974;
Natusch et al., 1974) could not be characterized, including Co, As, Ni, Zn,
                                     78

-------
                   160
  Approximate Depth (A)
320     480     640
800
960
                            80      120     160
                                  Time (sees)

                «=Unextrocted   * = DMSO Extracted   '*H20 Extracted
                                                              - 4000    2
                                                              -{2000
                                                                        £
                                                                        x D.
                                                                        o
                                                              -200000 ••=
                                   - IOOOOO
                                                                        o _^

                                                                        I *
                                                                          E
                                                                        Q.
                                                                        a.
Figure 35.   Ion  microprobe depth profiles  of the Group A elements  (Ti,  Al,
             Si)  in  unleached and leached  fly ash.   (Reprinted by permission
             of R.  Linton and The American  Chemical Society from Linton  et
             al.,  1977.)
                                      79

-------
           30
         in

         4)
         & 20
         o
         w
         o:
            10
                     160
             0
40
          Approximate Depth(A)
        320     480     640
                                                      fitOO
                                 960
                                                                   IOOOOO
                                              50000  |
80
                                                       2OO
                    120      160
                  Time (sees)
*=Unextraded  *=DMSO Extracted  »=H20 Extracted
                                          240
                                                                          x
Figure 36.  Ion microprobe depth  profiles of the Group B elements (Fe, S)  in
            unleached and leached fly ash.   (Reprinted by permission of  R.
            Linton and the American Chemical Society from Linton et al.,
            1977.)
                                    80

-------
                                Approximate Depth(A)
0 160 320 480 640 . 800 960



x 15
in
£

5 10
8

o
IT



>,40
in
1 30
c
»-*
> 20
o
a 10

0


>>40
c
|30
4>
.i 20
£ 10

• l i 1 < 1 i 1 i 1 i 1
39.x •*•
•»

-



-
V
^
t\
^E^f^-^ 	 x 	 	 x 	 * 	 *
1 , 1 1 I 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1
2w -
-
\
\
t\
, y^*- — ~^__ 	 „ x „

1 * 	 1 * » + 	 ; 	 i 	 4
	 I 	 | 	 | 	 1 	 1 	 I 	 1 	 1 	 1 	 1 I 	 1 	 !
	 1 	 1 — i | 1 | ' I ' 1 ' 1
7, .+
Ll
_

r
\
-^_A ....

* ? " ? J ?.,... i




300000



200000


100000


75000

50000


25000


0


800


400

O
°0 4O 80 120 160 200 240


c
o
0
"c
v "•;
<
Si
x g
o °-
&


                                    Time (sees)
                   *=Unextracted  »=DMSO  Extracted   «=H20 Extracted
Figure 37.   Ion microprobe depth profiles of the  Group B elements  (K,  Na,
             Li) in unleached and leached fly ash.   (Reprinted by permission
             of R.  Linton and the American Chemical  Society from Linton et
             al., 1977.)
                                     81

-------
                                Approximote Depth (A)
c

240

£.

c
"c
~ 160
>
^
%
•"
80



0

36
> 160 320 480 640 800 960
	 l • i ' 1 ' I
Pb



_


-

•
\
\
-\
\
\I^_A 	 ^
<^____^ "^ 	 * 	 -« 	 	 	 x K "

1 | 1 | 1 | 1 | 1 | 1 |
205Tr
-

>» I
1
c I
«> 1
~ 24
V

.-
o
o>
o:
12



°C




. V n ,.

*-i
E
Q.
Q.
40000 -
o
o
30000 ^
o>
o
c
o
20000 °
9)
O
E
10000 g
k-
o.
o.
0 <
$
E
0.
Q.

4000 o

o
'c
a)
0
C
5
2000 o>
"5
X
8
ex
o.
n
) 40 80 120 160 200 240
                                     Time(secs)

                    =Unextrocted   *=DMSO Extracted   » = H20 Extracted
Figure 38.  Ion microprobe depth profiles of the Group C elements (Pb, Tl)
            in unleached and leached fly ash.   (Reprinted by permission of
            R. Linton  and the American Chemical Society from Linton et al.,
            1977.)
                                     82

-------
          O)
          >
            15
         I "0
         5  5
         o>  ^
         o:
V)

£ 60

V
.2 40
o

                                                              o
                                                              §
                                                              CL
                                                              a.
                                                               c
                                                               o
                                                                        s a
                                                                        a.
                                                                        a
                                                                  0
>«80
in
| 60
.1 40
0
& 20
r>
	 1 	 | 	 1 	 1 	 r 	 1 	 1 	 1 i | i |
-
vv
^^r~:r~x — x— - x
1,1,1,1,1. ....i...

1200
800
400
O
                                                                        o
                                                                        is
                                                                        a>
                                                                        8-
                                                                        0 i.
             '0       40      80
                                    Time (sees)
                  *=Unextracted  *=DMSO Extracted  »=H20 Extracted
Figure 39.  Ion microprobe depth profiles  of the Group C elements (Cr, Mn,
            V) in unleached and leached fly ash.   (Reprinted by permission
            of R. Linton and the American  Chemical  Society from Linton et
            al., 1977.)
                                    83

-------
      v>

      V
      >
      "a
        20
      >>

      c
      B
                  160
                  —1—
  Approximate Depth (A)
320     480     640
         T
80O
960
                                                                60000
                                                                       >0
                                             a.
                                             a.

                                             o
                                                                40000 o
                                      20000
                  MW      OW      lt.W

                                 Time(secs)

               t = Unextracted  * = DMSO Extracted  • =H20 Extracted
                                                                       u
                                                                       §
                                                                       X
                                                                       8


GO


40



20


Q
i 1 i I i 1 i i i I i i
2W
^ _ n 	 ,-
-X 	 X 	 K 	 «
*\
-V
x>^
^^O- — *— ^
*-^ 	 L_ : :
.
-
-
-
i , i , i , i . i i i

$
8OOO ^

60OO ^
o
c
4OOO ?J
s
x
2000 e
a
n
"0 40 80 \20 \6O 20O 240 ~
Figure 40.  Ion microprobe depth profiles of the  Group  D  elements (Ca, Mg)
            in unleached and leached fly ash.   (Reprinted by permission of
            R. Linton and the American Chemical Society from Linton et al.,
            1977.)
                                     84

-------
Se, Cd, Hg, and Sb.   The last elements were present at low bulk concentra-
tions ranging from approximately 10 to 1000 ppm by weight.  Secondary ion
mass spectral interferences from species such as molecular ions,, hydrocar-
bons, and isotopes of major elements were common in the fly ash mass spectra
and generally required the use of high spectral resolution (Bakale, 1975;
Williams and Evans,  1975), with an accompanying loss in sensitivity.  The
above eight elements were not of sufficient intensity to be characterized.
Elemental identifications were made only when multi-isotopic species
exhibited correct isotope ratios or when mono-isotopic species could be mass
resolved from interferences and identified by mass difference measurements.
Most of the elements listed above also have intrinsically poor detection
limits, since they have relatively low positive secondary ion yields (Town-
send et a!., 1976).

     Figures 35-40 indicate that the elements Fe, Na, Li, K, S, Pb, Tl, Mn,
Cr, and V all show a surface predominance.  However, it was necessary to
obtain additional experimental confirmation that elemental surface predomi-
nance was not the consequence of analysis artifacts.  This was accomplished
by the use of the following:  (1) complementary surface analytical techniques
such as Auger Electron Spectrometry (to be discussed later) and (2) solvent
leaching of the fly ash particles.

     Semi quantitative values for bulk elemental Teachabilities were obtained
by ratioing the total mass of an element in a solution leachate to its total
mass in the original fly ash sample as determined using spark source mass
spectrometry (SSMS).  Analytical precision for the solution or ash sample
were about ±20%.  The percent leached values (Table 15) were accurate to
within about a factor of two of the true value.  This can be attributed to
the small variations in the sensitivity of the same element in different
matrices (Ahearn, 1967).  (That is, the sensitivity of an element in the
leachate solution doped onto graphite differs somewhat from its sensitivity
in the fly ash-graphite mixtures.)

     High elemental  Teachability should, in part, reflect high surface
accessibility.   Results indicated that the extent of Teachability (Table 15)
did correlate with the extent of surface predominance (depth profiles of
unextracted particles in Figs. 35-40).  Specifically, the elements that the
ion microprobe indicated were not surface predominant (e.g., Si, Mg, Ca, Ti)
generally had the lowest Teachabilities.

     Of the instrumental methods used, the ion microprobe was the only one
that detected trace-level elements, including Pb, Tl, Cr, Mn, and V.  Solvent
leaching was used to substantiate the surface predominance of these elements
by comparison of their depth profiles before and after solvent leaching
(Figs. 38 and 39).  Normalized secondary ion intensities for a given element
were reproducible to about a factor of two for different particles in either
the leached or unleached samples.  Following leaching, secondary ion
intensities of Pb, Tl, Cr, and Mn near the surface were reduced by more than
a factor of four (Figs. 38 and 39).  Thus, the depletions in the surface
region concentrations of these elements following leaching were indeed
significant.
                                     85

-------
   TABLE 15.   ELEMENTAL CONCENTRATIONS IN COAL
         FLY  ASH LEACHATES (DATA OF LINTON
                  ET AL.,  1977).
Element
T1
Al*
Si
Fe
S
K
Na*
Li
Pb
Tl
Cr
Mn
V
Ca*
Mg
Concentration
(ppm by wt)
in unleached
particles
4,700
>6,600
120,000
92,000
7,100
39,000
>5,300
200
620
30
380
310
380
>12,000
12,000
% Leached
DMSO
1.5

0.1
0.7
27
0.2
>5.2
17
3.4
25
2.4
7.1
1.5
>1.3
0.6
(wt%)
H20
0.5

0.3
5.8
36
6.0

14
0.4
11
4.5
35
3.9
>5.8
2.7
* Could not be quantitated because of inter-
  ferences, or analytical lines were too
  intense to fall within the working range
  for the internal standards.
                       86

-------
     The factor of two irreproducibility in the depth profiles was largely
the result of variations in the chemical composition of individual particles.
Because of this imprecision, solvent leaching cannot be considered to have a
significant effect on the depth profiles of the other elements (Figs. 35-37).
As was anticipated, depth profiles of elements (Al,  Si, Ti,  Ca, Mg) with
negligible surface concentrations and only low to moderate Teachability
generally showed no significant change following leaching (Figs.  35 and 40).
However, the highly Teachable and surface-predominant elements (Na, K, Li, S,
Fe) showed no significant reduction in peak secondary ion intensity after
leaching (Figs. 36 and 37).   There are several possible explanations for
this.   First, Na, K, Li, and S all diffuse to the surface during ion bombard-
ment,  accounting for at least some, if not most, of the surface predominance
(Coburn, 1976; McCaughan and Kushner, 1974).   Second, these  elements (except
for Li and S) are also present at high concentrations in less soluble forms
as part of the bulk matrix composition of fly ash.  Such high "interior"
concentrations may obscure the effect of the surface leaching of Fe, K, and
Na on the shape of depth profiles within the limits of reproducibility for
different particles.  These elements also may be present to  some extent in
soluble species near the particle surface.  Thus, dissolution of such ele-
ments may be limited in part by solubility product considerations.

     The results obtained by the ion microprobe and leaching studies were
supplemented by the results of analyses using Auger Electron Spectrometry
(AES).  Depth resolutions are roughly comparable to those of the ion micro-
probe, and microfocusing of the primary electron beam permitted analysis of
individual particles.

     Because of basic limitations discussed in Appendix A, Auger electron
spectrometry was used only to obtain qualitative elemental depth profiles of
Al, Ca, Fe, K, Na, S, and Si contained in fly ash particles.  For unleached
particles, Auger depth profiles showed agreement with the ion microprobe
results in that only K, Na, and S exhibited major surface enhancement.
Examples of Auger elemental depth profiles are shown in Figure 41.  The
qualitative agreement between the Auger and ion microprobe data obtained
using different surface sputtering conditions offers additional evidence that
both techniques yield depth profiles that reflect the actual elemental compo-
sition of the surface region.

     It is evident that the combination of solvent leaching with bulk multi-
elemental analysis and with surface micro-analysis provides  the information
necessary to construct a composite picture of the physicochemical character-
istics of the surface regions of fly ash particles.   Based on the Teachabil-
ity and the extent of surface predominance, the elements studied can be cate-
gorized into four general groups (Table 16).

     The Group A elements (Si and Al) are major constituents of the glassy
fly ash particle matrix (Natusch et al., 1975).  The refractory nature and
low solubility expected for Al and Si coalesced as oxides in the fly ash
matrix account for the lack of surface predominance and low Teachabilities
observed for these elements.  Similar analytical results for Ti and its close
association with complex silicates in the original coal (Ruch et aT., 1974)


                                     87

-------
                      30
     Approximate Depth (&)
60    90    120    150    180
210


^40
"tt>
o: 20
1
^40
4>
o: 20



^40
120



c40

c 20


2(40


^'40
120
n
i i i i i i i
K(LMM:
-\
V
A
1 ^^l 	 (— x 	 1 	 1 	 | 	 — *

\lllll!
NQ(KLL:



I I ii 1 	 1 	 ¥
1
252 eV)
—

i

I
990 eV)
•~


i

VI I 1 1 1 1
S(LMM:


i i i i I i i
I
152 eV)

-
i

I i i I i I i
Fe(LMM

i
598 +
-
^x-^-x 	 " 651 -i- 703 eV )
- x
i i I I i I I

i i i i I i i
Al (KLL
i i i i i I 1
—
i

i
M396eV)
i

1 i i i I i i
I
x: x— ___ ____^ Si (KLL: 1619 eV)
- x-x 	 x
-
1 I 1 1 I I 1

-
i
                                  345
                                     Time (min)
Figure 41.  Auger Electron  Spectrometry (AES) elemental depth profiles  for
            unleached  fly ash.   (Reprinted by permission of R. Linton and
            the American Chemical  Society from Linton et al., 1977.)
                                    88

-------
     TABLE 16.   SUMMARY OF ANALYTICAL RESULTS-SURFACE CHARACTERIZATION
                              OF COAL FLY ASH
                      Significant
                        surface
                     concentration
            Effect of leaching
            on concentrations
             in the immediate
Group
A
B
C
Elements
Al
Fe
K,
Pb
Cr
, Ti,
, Na,
S
, Tl,
, v
Si
Li
Mn
observed?
No
Yes
Yes
surface region
Insignificant
Insignificant
Depletions observed
following solvent
Leachability
Low
Moderate
Moderate
Moderate

to
to
to

high
high
high
         Mg, Ca
No
leaching

Generally not
significant
Moderate
indicate that Ti is also incorporated as a minor component in the fused fly
ash matrix.

     The Group D elements (Ca, Mg) exhibit behavior similar to that of the
Group A elements.  The alkaline earths are often present largely as carbon-
ates in coal and subsequently decompose to form oxides during combustion
(Reid, 1971).  The refractory nature of the alkaline earth oxides apparently
explains their lack of surface predominance in fly ash.   The oxides are
potentially quite soluble, as are alkaline earth sulfates, which form by
reaction with sulfur oxides in the flue gas (Ruch et al., 1974; Miguel, 1976;
Fisher et al., 1976).

     The surface predominance observed for the Group B and C elements (Table
16) apparently results from the volatilization of these species in the high-
temperature combustion zone and their subsequent deposition on the surfaces
of refractory particles as the temperature falls inside the stack (Linton et
al., 1976; Davison et al., 1974).   Most of these elements fall within either
Group II (concentration dependent on particle size) or Group III (intermedi-
ate behavior, may or may not show a concentration dependence) as outlined in
the section on trace elements.  It should be noted, however, that the higher
Teachability of the Group B elements coupled with the relatively minor change
in observed surface concentrations suggests that much of the leached material
is not derived from the surface.   The surface migration of Na, Li, K and S
under ion bombardment may be responsible for the observed surface concentra-
tions of these elements.
                                    89

-------
     As has been discussed previously,  certain trace metals,  for example Hg
and Se, are so volatile that a large percentage of their total  emission from
coal-burning power plants is in the vapor phase (Kaakinen et  al.,  1975; Klein
et.,  1975).  Enrichments of elemental  concentrations observed in fly ash
relative to the residual slag or bottom ash are in excellent  agreement with
the results of the depth profiles.   Specifically,  elements which were not
surface predominant (Groups A and D) have shown negligible concentration
enhancements in fly ash relative to the slag (Klein et al., 1975).   However,
elements which exhibited significant surface predominance in  this  study
(Groups B and C) also have been observed to show minor to major enrichments
in fly ash relative to the slag (Klein et al., 1975).   In summary,  the re-
sults of the ion microprobe and AES depth profiles, the enrichment of ele-
ments in fly ash relative to the slag (Klein et al., 1975), and the inverse
dependence of bulk elemental concentrations on particle size  (Davison et al.,
1974) strongly suggest that a volatilization-condensation process  occurs for
the Group B and C elements.

HEALTH EFFECTS

     Possibly the greatest impact of high trace metal  concentrations in
smaller particles is in the area of health,.  This is the case since most of
the small particle fractions' mass is concentrated in the size range 0.5 to
10.0 micrometers (urn) which is inhaled and deposited in the human respiratory
system (Natusch and Wallace, 1974; Miller et al.  1979).

     A number of workers (Morrow, 1964; Hatch and Gross, 1964; U.S. Dept. of
H.E.W., 1969) have shown that inhaled airborne particles are  deposited in
different regions of the body depending on their aerodynamic  size.   From a
toxicological standpoint, the smallest particles (less than 1 pm) which
deposit in the pulmonary region of the respiratory tract are  of greatest
concern.  This is because the efficiency of extraction of toxic species from
particles deposited in the pulmonary region is high (60-80%)  (Langham, 1960;
Hatch and Gross, 1964; Patterson and Sal via, 1968; Dautreband, 1968; Schroe-
der, 1971), whereas the extraction efficiency from the larger particles which
deposit in the nasopharyngeal and tracheo-bronchial regions and are removed
to the pharynx by cilia! action and swallowed is low (5-15%).  Consequently,
toxic species which predominate in submicrometer-sized particles will have
their entry to the bloodstream enhanced over those which predominate in
larger particles.

     Potentially toxic elements including Pb, Tl, Cr, and Mn  have been shown
to have surface concentrations much higher than hitherto supposed on the
basis of conventional bulk analytical data.  For example, Pb  and Tl were
present at concentrations of only 620 and 30 ppm on the basis of bulk analy-
sis (Table 16), but reached respective concentrations of 4% and 4500 ppm near
the particle surfaces (Figure 35).  In addition, the surface  predominant
region for such elements was highly Teachable.  The region of surface predom-
inance is  thus very likely to be accessible to the environment either by
washout processes in the atmosphere or ground waters, or by solubilization  in
lung and digestive fluids of higher organisms.
                                     90

-------
PARTICULATE ASSOCIATED ORGANIC COMPOUNDS

Introduction

     Participate associated organic material emitted from fossil fueled power
plants is known to contain both aliphatic and aromatic compounds (Committee
on Biological Effects, 1972).   To date, essentially all studies have been
directed towards the latter class of compounds with special emphasis being
given to polynuclear aromatic (PNA) species which include many well estab-
lished carcinogens (Committee on Biological Effects, 1972; Searle, 1976).
Even within this group, primary emphasis has been placed on hydrocarbons and
little attention has been paid to heterocyclic compounds containing oxygen,
nitrogen, or sulfur.  Similarly, derivatives containing substituents such as
carboxylic, nitro, sulfonic acid, or phenolic groups (if, indeed, they occur)
have received little attention.  At this time, only eighteen polycyclic
organic compounds have been uniquely identified as being associated with fly
ash emitted by fossil fuel power plants.  These compounds are listed in Table
17 (Committee on Biological Effects, 1972; Hangebrauck et al., 1967).  It
should be noted that many more compounds have been tentatively identified but
have not yet received full confirmation.
             TABLE 17.  POLYCYCLIC AROMATIC COMPOUNDS IDENTIFIED*
                            IN EMITTED COAL FLY ASH
            Fluorenelf                          1-methylpyrene

            PhenanthrenelF                      Benzophenanthrenelf

            AnthracenelF                        Benzo[a]anthracene

            9, 10-dimethylanthraceneH          Perylene

            Fluoranthenel!                      Benzo|.a]pyrene1I

            TriphenyleneU                      Benzo[e]pyrene1F

            PyreneU                            Benzo[ghi]perylene

            Chrysenef                          Anthanthrene

            Benzofluorene                      Coronene
         By gas chromatography or literature data (Committee on
         Biological Effects, 1972; Hangebrauck et al,, 1967).

         Identity confirmed by GC/MS using reconstructed ion
         chromatograms.

                                     91

-------
     A number of studies of particulate polycyclic organic matter (POM)
emitted by fossil fuel power plants have concluded that total  emissions are
negligibly small compared with those from other sources (Committee on Bio-
logical Effects, 1972).   A summary of reported emission factors for several
coal combustion operations is presented in Table 18.   These figures translate
to a total annual emission of 500 tons of benzo[a]pyrene from all heat and
power generating plants in the United States.   The much higher emission
factors associated with hand stoked furnaces are attributed to inefficient
combustion.  The latter observation is consistent with previous work (Diehl,
et al., 1967) in which it was found that combustion conditions were of far
more importance in determining POM emissions than was the source of fuel.

     The detailed mechanism(s) of POM formation are not well understood;
however, it is widely believed that POM is formed via a free radical mecha-
nism (Badger, 1962).   Formation occurs as the result of combustion of any
carbonaceous material, is promoted by reducing conditions, and similar
relative amounts of individual compounds are produced irrespective of the
nature of the fuel (Committee on Biological Effects,  1972).   Although the
free radical mechanism suggests a vapor phase origin for POM there is a large
body of data which attests to the fact that POM present in the atmosphere is
almost invariably found in particul ate form (Committee on Biologial Effects,
1972).   It is apparent, therefore, that vapor-to-particle conversion takes
place between the points of formation of POM in a combustion source and its
determination in the atmosphere.  It is the mechanism of this vapor-particle
transformation and the subsequent oxidative transformation of particulate POM
which constitute the subject of this section.
             TABLE 18.  POLYCYCLIC AROMATIC HYDROCARBON EMISSION
                      FACTORS FOR COAL-FIRED FURNACES*


                              POM emission factors (Ib/ton coal x 104)
Species
Benzo[a]pyrene
Pyrene
Benzo[e]pyrene
Perylene
Fluoranthene
Pulverized
firing
0.2-0.52
0.8-1.6
0-2.1
0-0.6
—
Chain grate
stoker
0.3
3.5
1.1
—
6.0
Hand
fired
3520
5260
880
526
8800
       * Committee on Biological Effects (1972) Data.
                                     92

-------
VAPOR-PARTICLE TRANSFORMATION

Introduction

     A number of workers have reported the presence of vapor phase POM in
combustion sources where elevated temperatures are encountered (U.S.
Environmental Protection Agency, 1976a, 1976b, 1976c; Bertsch et al . ,  1974;
Pupp et al., 1974; Hangebrauck, 1967).  In other words, the saturation vapor
pressure or dew point of POM must be attained for these processes to take
place (Castellan, 1971).  On the other hand, adsorption of POM vapor onto the
surface of particulate material present in stack or exhaust gases can  cer-
tainly take place and could account for the almost exclusive occurrence of
particulate POM at ambient atmospheric temperatures.   In order to investigate
the possible operation of an adsorption mechanism we have conducted a  study
which involves theoretical modelling, field measurement, and laboratory
simulation experiments.   These are described briefly, together with the
results, in the following sections.

Theoretical Model

     The theoretical model is designed to consider the active adsorption of
POM (primarily polynuclear aromatic hydrocarbons [PAH]) from the vapor phase
onto the surface of particulate material (coal fly ash) as the two move down
a nonuniform temperature gradient (the stack system).  The treatment is,
however, quite general and applies to any vapor species and any particulate
adsorbant.

     The basic assumption is that a given POM, P, present in the vapor phase,
can adsorb onto the surface, S, of particulate material to provide an
adsorbed entity, P'S, and that the reverse process can also take place.
Thus,
                                      ki
                             P + S    .      P-S                       (10)
                                      k_i

where kx and k_i are the rate constants for adsorption and desorption,

respectively.  If one assumes, a priori , that kx and k_! in Equation 10
represent first-order processes, then the rate of adsorption of P can  be
written


                                             k.itp-s]                 (ID
In Equation 11, 6 is the fraction of the total available adsorption sites
which are occupied and A is the surface area of the particulate material.

     The rate constant for desorption, k_i, is given by


                         k_i  =  ^ exp [-E/RT]                      (12)


                                     93

-------
where Ed is the activation energy for desorption, T is the absolute tempera-
ture, and k, h and R are Boltzmann's constant,  Planck's constant, and the
universal gas constant, respectively.  The rate constant for adsorption, klf
is
                      k,  =  c
        RT
          d
exp [-Ea/RT]               (13)
where c is the so-called "sticking coefficient" (the probability that the
orientation of a molecule with the spherical particle surface will result in
adsorption), M  is the molecular weight of the adsorbate species, and E  is
              O                                                        d
the activation energy for adsorption.   A number of assumptions are inherent
in equations (12) and (13).   These are presented briefly (Natusch and Tom-
kins, 1978) and discussed in detail in a forthcoming publication (Miguel,
Schure, and Natusch, 1979).
     In order to evaluate the rate and extent of adsorption as a function of
temperature it is convenient to compute the mole fraction, X, of total POM
adsorbed and the time (ti) taken to achieve one half of the equilibrium
adsorption at a given temperature.  The equations used to evaluate the
adsorption process are beyond the scope of this report; however, Figure 42
gives the temperature-dependent plots of X and T, assuming that the frac-
tional surface coverage  is dependent on temperature and the values of E  and
                                                                       d
E , (energy of desorption).  The plots assume (1) that the adsorption process
is zero order in its dependence on the fractional surface area, (2) that the
available literature values for p and W , the density and mass per unit
volume of the adsorbing  particles, respectively, are reasonable, and (3) that
d  and d  , the mass and  surface median diameters of the particle size distri-
bution, respectively, are also reasonable.
     In the case of a log NORMAL distribution of particle sizes, dm and d
are related by
                                         m
In ds  =  In dm - In2a                       (14)
where a is the geometric standard deviation of the distribution (Butcher
and Charlson, 1972).
     Consideration of the data presented in Figure 42 shows that, over a wide
range of conditions, POM present at combustion source temperatures (>150°C)
is predicted to occur mainly in the vapor phase, whereas at the ambient
temperatures following emission (<40°C), essentially quantitative adsorption
is predicted.  Variations from this general behavior are, however, different
for individual compounds and different particle size distributions and mass
loadings.
                                     94

-------
                    250 300 350 400 450 500 550 600
                             T,K
                                250 300 350 400 450 500 550 600
                                        T,K
          300 350 400 450 500 550 600
                IK
                                            250 300 350 400 450 500  550 600
                                                     IK
Figure 42.  Dependence  of mole fraction (X) of  PAH  adsorbed on temperature.
            (Reprinted  by permission of D.F.S.  Natusch  and Raven Press from
            Natusch  and Tonkins, 1978.)
Reference values:   c  =  1;  mg = 3.360 (10~22) g/molecule;  w  =
                                           2 = 10"4  cm;  E   =
            106 Mg/m3 =  10"6  g/cm3; p = 3 g/cm3; dm3
            -10 kcal/mole;  Ed = -30 kcal/mole; 6 = \
            (a) Variation  of  mole fraction of PAH adsorbed  as  a function of
                temperature and C = 0.001, 0.01, 0.1, 1.0.
            (b) Variation  of  mole fraction of PAH adsorbed  as  a function of
                temperature and dm3/ds2 = 0.1, 1.0, 10.0 urn.
            (c) Variation  of  mole fraction of PAH adsorbed  as  a function of
                temperature and w  = 102, 104, 106, 108 (jg/m.
            (d) Variation  of  mole fraction of PAH adsorbed  as  a function of
                temperature and Ea-Ed = -15, -20, -25 kcal/mole;  Eg = -10
                kcal/mole.
                                      95

-------
     The temperature dependencies shown in Figure 42 are generated with the
assumption that adsorption equilibrium is achieved at all temperatures.  This
requires that the rate of attainment of equilibrium is fast compared with the
rate of change of temperature experienced by a given vapor-particle combina-
tion.   The rate of attainment of equilibrium is indicated in terms of the
half-time for reaction in Figure 43.  These data show that reaction times
depend primarily on the activation energies of adsorption, E ,  and desorp-
                                                            a
tion,  E ., and that achievement of sub-second reaction times at both in-stack

and ambient temperatures requires activation energy values to be at the lower
end of the range investigated.  Indeed, reduction of the activation energy
for adsorption by only a few kcals results in achievement, of equilibrium
within a few seconds even at ambient temperatures.

     While the foregoing brief discussion is far from rigorous or conclusive,
it does validate the possibility of vapor-to-particle conversion of POM
occurring via an adsorption mechanism.  Further, it points out the very
strong temperature dependence of such a process.

Field Measurements

     In order to investigate the occurrence of vapor-to-particle conversion
in a combustion source, measurements were made in the stack system and in the
emitted plume of a small coal fired power plant.  This plant was specially
chosen because it was of obsolete design utilizing a chain grate stoker known
to produce high POM emissions (Hangebrauck, et al., 1967); it possessed no
particle control equipment, thereby facilitating collection of large amounts
of particulate material; and it had a relatively short stack such that sample
collection from the emitted plume was convenient.

     Fly ash samples were collected during the same time periods both inside
the stack (temp. -v-2900C) and from the emitted plume (temp. -v5°C) using both
cascade impaction and total collection on glass fiber filters.   Collected
material was extracted with benzene and analyzed for POM using gas chromato-
graphy (GC), gas chromatography/mass spectrometry (GC/MS), and fluorimetry.
All necessary analytical precautions were taken and it was established by
chemical and microscopic means that material collected from the plume was
derived exclusively from the power plant.

     The results of these analyses are presented in Table 19, which lists the
individual compounds uniquely identified and provides a quantitative measure
of the specific concentrations (ug/g) of several compounds associated with
fly ash at the two sampling points.  Preliminary identification was estab-
lished using GC retention time data and in many cases was confirmed using
either a single-ion mass chromatogram or a mass spectrum from a GC peak.
Only crude vapor traps were employed during sample collection so no quanti-
tative measure of vapor phase POM was obtained.  Fluorescence measurements of
condensation trap residues did, however, indicate that a considerable quan-
tity of gaseous POM was present from in-stack sampling but none was in resi-
dues from plume sampling.
                                     96

-------
                                  log 11,
J33 150 «CO •»» !OC S53 500
                                     °ao XO JSO «03 4M WJ iiO WC
                                            T.K
               109 ' 'A =
                                  log
Figure 43.  Dependence of half-time of  adsorption  on temperature.
            (Reprinted by permission of D.F.S.  Natusch and Raven Press
            from Natusch and Tomkins, 1978.)
            Reference values:  c = 1; mg =  3.360  (10~22)  g/molec;  w  = 106
            ug/m3 = 10"6 g/cm3; p = 3 g/cm3;  d  3/d 2 = 10"4 cm;  E, = -10
                                              Hi   S              a
            kcal/mole; E, = -30 kcal/mole;  6  =  %.

            (a) Dependence of half-time for adsorption on temperature and
                C = 0.001, 0.01, 0.1, 1.0.
            (b) Dependence of half-time for adsorption on temperature and
                dm3/ds2 = 0>1' 1<0' 10'° Mm-
            (c) Dependence of half-time for adsorption on temperature and
                w  = 102, 104, 106, 108 ug/m3.
            (d) Dependence of half-time for adsorption on temperature and
                Ed = -25, -30, -35 kcal/mole.
                                      97

-------
                TABLE  19.   MEASUREMENT  OF  POM  EMITTED  FROM  A
                        COAL-FIRED  POWER PLANT STACK*


                                     Inside  stack         Outside  stack
         Compound                       (ng/g)             (Mg/g)

      Fluorene                            <1                   0.5

      Phenanthrene                        <1                   12

      Fluoranthene                        <1                   17

      Pyrene                              <1                   12

      Benzofluorene                       <1                   2

      1-methylpyrene                      <1                   0.6

      Benzophenanthrene                   <1                   3

      Benzo[a]pyrene                      <1                   5

      Fluorescence/g total           3.6 x  10-3 units       370 units


      * Data of Korfmacher et al.  (1977).
     These results establish quite firmly that,  in this power plant,  consid-
erably more POM is associated with fly ash collected from the plume at a
temperature of 5°C than with that collected from the same stream at a temper-
ature of 190°C.  This behavior is in exact accord with that predicted in
Figure 42.  Furthermore, since the two collection points were only ^100 ft.
apart, quite rapid vapor-to-particle conversion  is indicated.   Unfortunately,
while the full range of aerodynamic equivalent particle sizes accessible to
Anderson-Stack and Hi Vol samplers was collected, this represented only a
small range of specific surface area due to the  considerable particle irregu-
latiry encountered.  Nevertheless, correspondence between specific concentra-
tion of POM and specific surface area of fly ash fractions was noted.  This
further suggests the operation of a surface adsorption mechanism.

Laboratory Experiments

     In an attempt to obtain direct measurements of the rate(s) and extent(s)
of POM adsorption and to evaluate the quantities c, Ea, E^ (equations 12 and

13), a series of laboratory simulation experiments were set up.  Fresh coal
fly ash that had previously been shown to contain no detectable POM was
fluidized in an expanded bed through which a stream of air containing pyrene
                                     98

-------
vapor was passed (Natusch and Miguel, 1975; Miguel  et al., 1979).   The objec-
tive was to expose all particles to the same constant concentration of pyrene
for different times and to determine the specific concentrations of pyrene as
a function of time at different temperatures.

     The experiments showed that the uptake of pyrene was so rapid that a
uniform vapor phase concentration could not be achieved.   The amount of
pyrene required to saturate the fly ash was, however, shown to increase
significantly with decreasing temperature.   Furthermore,  attempts  to remove
adsorbed pyrene by heating in a stream of clean air were  unsuccessful.
(Miguel, 1976).  While these experiments were essentially qualitative they do
establish the facts that coal fly ash will  strongly adsorb pyrene  (and
probably other POM) and that the saturation capacity is a strong inverse
function of temperature.

     Overall, therefore, the results of these studies point strongly towards
the idea that POM, formed initially as vapor, is adsorbed onto co-entrained
particulate material as the temperature falls, consistent with the curves in
Figure 42.  There is some doubt about the rate at which this process takes
place, but the evidence is in favor of rapid (on the order of seconds) ad-
sorption even at ambient temperatures, under most conditions encountered in
or near combustion sources.

Ramifications of POM Transformation

     The gas-to-particle transformation studies discussed above, while far
from definitive, do point out several important considerations which should
be borne in mind in assessing the environmental/health impact and methodology
for the measurement and control of POM.   These may be categorized as follows:

 1.  From the standpoint of the environmental and potential human health
     impact of POM, it is important to note that adsorption of vapor phase
     POM onto particulate matter should result in the predominance of POM on
     small particles, which provide the largest available surface area per
     unit mass.  Thus, POM should be preferentially concentrated in particles
     whose aerodynamic size falls in the range which can  remain airborne for
     several days and which is capable of being deposited in the pulmonary
     region of the human respiratory system when inhaled.  (Davison et al.,
     1974; Natusch and Wallace, 1974).  This prediction is in accord with the
     results of measurements of the atmospheric aerosol size distribution of
     POM (Natusch and Wallace, 1974).

 2.  If POM is capable of converting rapidly from vapor to particulate form,
     then the relative amounts of each would be expected  to vary with the
     position (i.e. temperature and particle surface density) in a combustion
     system.  Consequently, separate measurements of vapor and particulate
     POM concentrations will apply only to a specific point in a specific
     plant generated under specific conditions and cannot be extended to
     other power plants in general.  Furthermore, the predictions  of Figures
     42 and 43 suggest that considerable vapor-to-particle conversion may
     actually occur within the sampling device—especially when it is main-
     tained at a temperature that is different from that  of the stream being

                                     99

-------
     sampled.   As a result,  measurements  of the  particle  size  distribution of
     POM and of the particulate  to vapor  phase POM  concentration  ratio will
     not represent true quantities but will  reflect sampling artifacts.

 3.   Of more practical  importance, however,  is the  very clear  need to  collect
     both vapor and particulate  POM in order to  establish total POM emission
     factors.   Current  methodology for sampling  emission  sources  does  employ
     vapor collection devices;  however, some doubt  has  been expressed  con-
     cerning their collection efficiencies,  (Jones  et al., 1976)  and it seems
     likely that present POM emission estimates  may be  low where  measurements
     are made at elevated temperatures.  Certainly  emission estimates  based
     on analyses of particulate  material  alone,  when it is collected from
     within an emission source,  are likely to be grossly  in error (Natusch,
     1978b).

 4.   The vapor-to-particle transformation behavior  of POM has  two important
     consequences in terms of emission control.   First, the fraction of the
     total POM which is in the vapor phase during passage through particle
     control  devices will  not be collected.   Secondly,  the effect of adsorp-
     tion is to move the aerodynamic equivalent  mass median diameter of the
     adsorbed species to values  which are significantly smaller than those of
     the substrate particle mass as described by equation (7)  (Section V).
     Just as this enhances the ability of POM to penetrate the human respira-
     tory system, it will  also increase the difficulty  of controlling  POM
     emissions since small particles in the range of interest  (0.1 to  5.0 urn
     aerodynamic diameter) are collected  with reduced efficiency  by most
     particle control devices (White, 1963).

 5.   One postive point  which can be made  is that collection of POM vapor by
     scrubbers may be quite efficient insofar as the low  temperatures  encoun-
     tered in liquid systems may promote  adsorption and at least  partial
     collection of vapor species.

 6.   Finally, it is important to recognize that  the processes  descirbed
     herein, although presented in terms  of a coal-fired  power plant,  should
     occur quite generally.   They must, therefore,  be taken  into  account when
     considering any combustion source.

Oxidative Transformation

     Information about  the fate of particulate  PAH  compounds  released to the
atmosphere is presently fragmentary and unclear.  It is generally assumed,
however, that photochemical oxidation processes  play an important role
(Committee on Biological Effects, 1972).   There  is  ample  evidence that most
PAH's will undergo photo-oxidation in solution,  as  the  pure  solid, and when
adsorbed onto certain solid substrates such as  alumina  (Committee on Biologi-
cal  Effects, 1972) and  it has been inferred that similar  processes take place
when the compounds are  adsorbed on airborne particulate material  (Committee
on Biological Effects,  1972; Falk et al., 1956;  Mukai et  al.,  1968).  Indeed,
it has been suggested that the half-lives of such PAHs  in the  presence of
sunlight may be "only hours or days" (Committee  on  Biological  Effects, 1972).
Contrary to this expectation, the present studies show  that the  rate and


                                     100

-------
extent of photodecomposition of PAHs may be decreased substantially by
adsorption onto coal fly ash particles.  Certain PAHs are, however, found to
undergo rapid oxidation in the dark when adsorbed onto fly ash.

     Studies were conducted using a model system in which individual PAHs
were adsorbed onto the surface of fly ash collected from the electrostatic
precipitators of several coal fired power plants.  Known gas phase concen-
trations of individual PAHs were generated by a simple diffusion cell
(Miguel, 1976; Miguel et al., 1979) and passed through an expanded bed of fly
ash.  Regulation of the bed temperature and the time of exposure provided a
means of controlling the amount of a PAH adsorbed.   All fly ash samples were
size fractionated by sieving; the 44-74 urn physical size range was utilized.
Exhaustive pre-extraction with several solvents (cyclohexane, benzene,
methanol) showed that none of the fly ashes employed contained more than
50 mg/g of total PAH.

     Sample irradiation was achieved using outdoor sunlight and several
artificial light sources (150W Xenon arc lamp, 275W commercial sun!amp,
5-OT3/C1 General Electric "quartz line" lamps; all  unfiltered).   The fly ash
was tumbled in a quartz container to provide equal  exposure of all surfaces
and radiant fluxes were estimated to be equal to or greater than that of
midday summer sunlight (45°N lat.) over the wavelength range 300 to 800 nm.
Portions of each fly ash sample were exhaustively extracted directly follow-
ing exposure to gaseous PAH and then after irradiation or dark storage.
Analyses for the original PAH and for degradation products were performed by
high performance liquid chromatography (ultraviolet [UV] absorbance detec-
tion) and by UV absorption and fluorescence spectrometry.  The extent of
decomposition was determined both from the disappearance of the original PAH
and from the appearance of its reaction products.  (Reproducibility of
multiple analyses was established to be ±8% relative standard deviation
RSD).  Experiments were conducted to determine the influence of PAH type and
concentration, irradiation time, intensity, and wavelength distribution, dark
stability, and the nature of the adsorbing substrate on the rate and extent
of PAH decomposition.

     Rather unexpectedly none of the PAHs irradiated following adsorption
onto fly ash exhibited significant photodegradation (Table 20 and Figure 44).
The data indicates (Figure 44) that the % change was not significantly dif-
ferent from zero at the 95% confidence level.  Furthermore, no significant
dependence on irradiation intensity nor amount of PAH adsorbed was observed.
Under similar illumination conditions, however, extensive photodecomposition
was observed for all PAHs irradiated in solution (Figure 44) and for both
anthracene and benzo[a]pyrene adsorbed onto alumina from methanolic solution.
Additional evidence for the apparent influence of adsorbent substrate on
photodecomposition was obtained by adsorbing anthracene and benzo[a]pyrene
vapors onto both fly ash and alumina, which were then coated on thin layer
chromatographic plates and exposed to identical irradiation.  In a typical
experiment of this type benzo[a]pyrene present on the alumina surface under-
went 50 per cent decomposition upon exposure for 80 minutes to an unfiltered
150W Xenon arc lamp.  By contrast, only 15 per cent decomposition was
observed for the benzo[a]pyrene adsorbed on fly ash and irradiated under
identical conditions.


                                    101

-------
                                     CU






o

<















o












•>






0 0





o
o
0 o

o


r8





.




Q
•oo



CO
fo*
.8i

UJ


	
i 	
t—

g
i
^"
0 <
* Q
g
(E

•




^O
CM









-o
DO O O O O O
r cu  o
* ,,,,..
Q.
£
to
10 >>
JD
-o
CU T3
-4-> CU
to +J
•r- C
IT3 5-
S- Ou
1- CU
•r- a;
1 ^^^
c
0
c .
c
ra o
o -13
•o 'o
CU I/)
S- /— N
e < r*
O CD
O • «rH
-C
CU I/) -
C (C S-
ai cu
s- >»-c
>>i— U
Q.M- 03
i — i E
(O O «(-
1 	 !-(-> S-
o c o
N 0 ^i
^
ai ID e
o at o
f^ ^
"O S- *4—
CU 0
-l-> U) S-
ro -a cu
•r- (0 .C
•a o
mi, g
1- O •»-
•i- £-
- 0
 o o
C. •!-
O •)-> C
U U 0
C -r-
•(-> 3 10
C M- (fl
cu ••-
U (0 E
S- t-
cu 10 cu
o_ n3 Q.


v^
39NVHO  39VlN3Dd3d
                                     0)
                                     O)
          102

-------
          TABLE 20.   DECOMPOSITION OF IRRADIATED POLYCYCLIC AROMATIC
                     HYDROCARBONS ABSORBED ON COAL FLY ASH
                              (KORFMACHER, 1978)
Compound
Pyrene
Phenanthrene
Fluoranthene
Anthracene
Benzo[a]pyrene*
Benzo[a]pyrene1I
Irradiation
time
(hr)
5-126
2-113
2.5-18
3-24
2.5-36.5
21-100
Cone. PAH
adsorbed
(M9/9)
67-291
129-186
49,115
300
44-110
15-44
No. of
Sampled
19
5
2
3
5
4
Average
change
(%)
-6.3 (±14.7)
+10 (±12.3)
-7.2 (±2.0)
-11 (±7.0)
-8 (±8)
-14 (±7)
 * Quartzline lamp.

 II Xenon arc lamp.
     While the extent of photodegradation of PAHs adsorbed on fly ash is
apparently small, a number of the compounds studied (Table 21) were found to
undergo quite extensive oxidation in the absence of light.  In particular,
different fly ashes were found to give different oxidation rates.  In all
cases, however, the major oxidation product was the same ketone or quinone as
obtained from the photo-oxidation process.   As illustrated in Figure 45 for
fluorene, which oxidizes to 9-fluorenone, considerable conversion occurs
during the actual PAH adsorption step but subsequent irradiation has no
effect upon the rate or extent of further reaction.   The extent of initial
reaction, which occurs at the elevated adsorption temperature, varies from 7
to 90 percent decomposition depending on the type of fly ash used as the
adsorbent substrate and was shown not to be due simply to passage of a PAH
through the adsorption apparatus.  It is apparent, therefore, that the nature
of the adsorbent is an important factor in determining the rate of this dark
reaction.  This observation is borne out by the findings that nonphotochemi-
cal conversion does not occur for solid fluorene, for fluorene dissolved in
methanol or cyclohexane, nor for fluorene adsorbed onto alumina, silica gel,
Linde 13X molecular sieve, glass and Ambersorb XE-340 adsorbent resin.
However, fluorene adsorbed onto activated carbon and graphite did decompose
in a manner similar to that of fluorene adsorbed on coal fly ash.

     Finally, it is noteworthy that all of the compounds that undergo spon-
taneous oxidation as a result of adsorption onto coal fly ash (Table 21)
                                    103

-------
                                                            CD
                                                            E
                                                            •*-> 00
                                                            <_> r*.
                                                            c. en
                                                            ro s-
                                                               0)
                                                            (0 U
                                                               (0
                                                               s-
                                                               o
(0 <
O
U

c
o
                                                               o
                                                               J-
                                                               OJ
                                                               u
                                                               (0
                                                            •O E
                                                            O) 4-
                                                            JD J-
                                                            t. o
                                                            O i^
                                                            to
                                                            -o  .
                                                            ro 
                                                             (/) Q.
                                                             S-
                                                             >
                                                             > ^3
                                                             c
                                                             o ~a
                                                             o o>
                                                               4->
                                                            -P C
                                                             c •>-
                                                             O) S-
                                                             U D.
                                                             S- (U
                                                             a) o;
                                                            a. •>— '
                                                            in
                                                             cu
                                                             s-
            104

-------
      TABLE 21.  DECOMPOSITION OF NON-IRRADIATED POLYCYCLIC AROMATIC
         HYDROCARBONS ADSORBED ON COAL FLY ASH (KORFMACHER, 1978)
       Compound
Change
Compound
Change
Acenaphtene                     None

7,12-Dimethylbenz[a]anthracene  None

Carbazole                       None

Acridine                        None

Phenazine                       None

Fluoranthene                    None

Benzo[a]pyrene                  None

Phenanthrene                    None
          Fluorene                  Decomposed

          Benzo[b]fluorene          Decomposed

          Benzo[a]fluorene          Decomposed

          9,10-Dimethylanthracene   Decomposed

          9,10-Dihydroanthracene    Decomposed

          4-Aza fluorene            Decomposed

          Anthracene                Decomposed
contain a benzylic carbon atom, indicating that this structure is especially
susceptible to non-photochemical oxidative attack.  Possible exceptions to
this rule are afforded by anthracene and benzo[a]pyrene, which exhibit very
slow non-photochemical oxidation when adsorbed onto fly ash.

     It is apparent from the foregoing results that PAH adsorbed onto the
surface of coal fly ash exhibits quite different chemical behavior from that
of PAH adsorbed onto other solid substrates or present either as the pure
solid or in solution.  The net result is stabilization of PAH against photo-
chemical oxidation, on the one hand, but promotion of spontaneous non-photo-
chemical oxidation for a limited class of compounds, on the other.  The
reasons for this behavior are not clear.  It is conceivable that the limited
photochemical decompostion is the result of either efficient competive absor-
bence of radiation by the fly ash matrix (Wehry, 1971) or the PAH being
largely present in pores, where light cannot penetrate.   A more convincing
suggestion, however, is that energetic adsorption, such as apparently occurs
on fly ash, activated carbon, and graphite surfaces, effectively stabilizes
PAHs against photo-oxidation, perhaps by molecular orbital rearrangment,
which either increases the electronic excitation energy or decreases the
lifetime of the excited state (Wehry, 1971).   Any rapid photo-oxidation that
did occur could then be due to a fraction of the PAH bonded not directly to
the active surface but, rather, to another PAH molecule, or to a weakly
adsorbing mineral impurity.  The promotion of non-photochemical oxidation
could, in turn, be due either to extensive molecular orbital rearrangement or
simply to the ready availability of active oxygen on strongly adsorbing
                                    105

-------
surfaces (Lao et al.,  1973).   Not unexpectedly,  the points of highest Bron-
sted acidity (e.g.,  position  9 in fluorene)  are  most susceptible to oxidative
attack.

     These findings  are of considerable importance from an environmental
standpoint because they show  that:

 1.   The adsorption  association of PAHs with coal  fly ask stabilizes these
     compounds against photochemical  degradation on the one hand, thereby
     preserving their potentially carcinogenic properties.  On the other
     hand, however,  a limited number of PAHs are quite rapidly converted to
     their corresponding ketones or quinones, whose toxic and carcinogenic
     potentials are  largely unknown (Jones and Freudenthal, 1978; Searle,
     1976).   In this regard it is noteworthy that fluorenone is normally
     encountered at  much higher levels in the atmosphere than is fluorene
     (Lao et al., 1973).

 2.   Based on the evidence presented here, one would not expect to find
     diurnal variations in the concentration of  PAHs present in plumes emit-
     ted from coal-fired power plants.  One  would, however, expect to find
     significant differences  between PAH transformation rates in plumes
     derived from different power plants because of the different surface
     characteristics of the fly ashes emitted.

 3.   If, in fact, all  PAHs containing a benzylic carbon atom are susceptible
     to non-photochemical oxidation,  as suggested here, then the probable
     oxidative behavior of a large number of PAHs can be predicted (Jones and
     Freudenthal, 1978; Korfmacher, 1978).  For  example, one would expect
     several strongly carcinogenic compounds such as dimethyl benz[a]anthra-
     cene and 3-methyl chloranthene (Committee on Biological Effects, 1972)
     to be rapidly oxidized and thus detoxified.  Possible evidence for this
     prediction is afforded by the fact that neither of these compounds has
     been found in atmospheric particulate samples (Committee on Biological
     Effects, 1972;  Lao et al., 1973).

     In summary, then, the widespread belief that particulate association of
PAHs will promote their photochemical conversion is not substantiated by
experiment.  Nevertheless, substantial non-photochemical conversion of PAHs
adsorbed on coal fly ask can occur and may result in significant detoxication
of several particulate PAHs.

Chemical Transformations

     In addition to photochemical or  non-photochemical oxidation, it is also
possible for POM adsorbed onto particulate surfaces to react with other gases
present in the atmosphere.  Of special concern are those gases that often are
present as the result of fossil fuel  combustion or conversion processes,
namely, NO, N02, S02, and S03.  A report  indicating that  reactions between
adsorbed POM and other  atmospheric gases  can produce compounds with increased
or previously unobserved mutagenic activity  (as indicated by the Ames test)
has appeared in the literature (Pitts et  al., 1978).   In  these studies,  it
was found that both benzo[a]pyrene (BaP)  and perylene  dispersed on glass
                                     106

-------
fiber filters reacted to form nitro derivatives when exposed to an atmosphere
containing 1 ppm N02 and that the products were mutagenic (Pitts et al.,
1978), even though one of the materials (perylene) did not previously exhibit
such behavior.

     We have studied the reactions of two PAHs adsorbed cito different sub-
strates with a number of gases known to exist in the plumes of fossil fuel
combustion and conversion sources in the laboratory and have found that both
the gas and the particulate surface significantly affect the reactions
observed.

     Pyrene and benzo[a]pyrene (50-150 ug/g) were adsorbed onto the surfaces
of several different particulate substrates (alumina, coal fly ash, and
silica gel) using the same techniques that were previously described (Miguel,
1976; Miguel and Natusch, 1975; Miguel et al., 1979).  Samples were then
exposed to the individual gases NO, N02, S02, or S03 at room temperature for
a period of 6 (NO) or 12 (N02, S02, S03) hours.  Concentrations of approxi-
mately 100 ppm were used in all cases except that of S03, where precise
quantification proved impossible because of experimental difficulties.
Exposure was conducted both in the presence and in the absence of ultraviolet
irradiation provided by a vitalite, quartzline lamp and natural daylight in
order to observe whether the reactions might be photochemically induced.
Following exposure the samples were Soxhlet extracted for 12 hours with
benzene and the extract subjected to high-performance liquid chromatography
using a Bondapak CIS reverse phase column and a water-methanol gradient
elution program for the mobile phase.  The chromatographic peaks were identi-
fied by high-resolution mass spectrometry following collection of fractions
containing each separated compound.  Electron Spectroscopy for Chemical
Analysis (ESCA) was employed to examine particle surfaces both before and
after exposure to the reactant gases.

     In general, only two of the gases studied (N02 and S03) reacted with the
two compounds used in this study.  Neither NO nor S02 reacted significantly
under the experimental conditions employed.  None of the PAH-gas reactions
was influenced by exposure to uv irradiation.  Chromatograms of the extract
resulting from the exposure of pyrene and benzo[a]pyrene (BaP) adsorbed on
fly ash to N02 and S03 are shown in Figures 46 and 47, respectively, together
with control chromatograms of extracted but unexposed material.  Before gas
exposure, only one major peak, the parent compound peak, A, is evident in
both cases.  Following exposure to N02, however, a second peak, labeled B in
both figures, is observable.  Mass spectrometric analyses indicate a mono-
nitro derivative in both cases.  The smaller peaks, labeled C-F in Figure 47,
appear to be either a benzene impurity (peak C) or small amounts of oxidized
BaP.  Insufficient material was available to permit mass spectrometric
identification of these smaller peaks.

     The results of the PAH-S03 reactions were not as clear.  Although
several peaks (C-E in Figure 46 for pyrene and G-I in Figure 47 for benzo-
[a]pyrene) were not present following S03 exposure, it was unclear whether
these peaks were actually products of the reaction or were decomposition/
oxidation products.  It does appear that these substances are not simple
substitution products analogous to the previously observed mono-nitro


                                    107

-------
                              PYRENE
 FL
                                              B
                              PYRENE-NO,
 UV
 FL
                              PYRENE-SO,
  FL
Figure 46.  Chormatograms of extracts  from pyrene adsorbed on coal  fly ash
           before and after exposure  to N02 and S03.   Both uv and
           fluorescence traces are  shown.  For peak identification see
           text.  (Reprinted by permission of D. F. S.  Natusch.)
                                  108

-------
                                                                             U
O_

,0.

CD
              U
  CM
O


Q_

-9.
QD
cf
to
 I
5^
3.
co
                                                                                                    -o
                                                                                                     c
                                                                                                     to  s-
                                                                                                        o
                                                                                                     O)  U_
                                                                                                     S-
                                                                                                     o
                                                                                                    4-    •
                                                                                                     cu  c
                                                                          O  -P
                                                                      J=. -C  rO
                                                                       in  in  Z
                                                                       ro
                                                                          cu   •
                                                                       >> 1-  i/)
                                                                      i—  rO
                                                                      4-
                                                                          in  u_
                                                                      i—  0>
                                                                       ro  U   •
                                                                       O  ro  Q
                                                                       u  i-
                                                                          -P  4-
                                                                             O
                                                                          CU
                                                                          U  C
                                                                      "O  C  O
                                                                       O)  CU  •!-
                                                                          u  in
                                                                          in  in
                                                                          cu  •«-
                                                                                                     c
                                                                                                     o
                                                                                                      £-
                                                                                                      o
                                                                                                      in
                                                                                                     -a
                                                                                                      ro
       1-
    3  O)
       Q-
 O) 4-
 C     >>
 Ol T3 J3
 ^  C
 >, rO -O
 Q.    O>
i—i > -P
 (0  3  C
1_J   •!—
 O -C  1-
 N -P  O-
 C  O  O)
 O> C3Q Q£
-Q    *—'
                                                                       O   co  •
                                                                       S- O -P
                                                                       4- CO  X
                                                                              OJ
                                                                       in "O -P
                                                                       •P  C
                                                                       (J  (0  CU
                                                                       (O     <1J
                                                                       S-   CM in
                                                                       -P O
                                                                       X Z  C
                                                                              O
                                                                           O •!-
                                                                          +J -P
                                                                              ro
                                                                           O)  U
                                                                           S- T-
                                                                                                      01
                                                                                                      in
                                                                                                      E 3 4-
                                                                                                      ro in -r-
                                                                                                      s- o +->
                                                                                                      C75 a. c
                                                                                                      o x  01
                                                                                                      -p 01 -a
                                                                                                      (IJ    •!-

                                                                                                      o cu -^
                                                                                                      S- -p  ro
                                                                                                      .C 4-  Ol
                                                                                                      O rO  O.
                                                                                                      CU
                                                                                                      s-
                                                                                                      OJ

                                                                                                      U-
                                               109

-------
  CM
O
z<
LUI—

cT
ZLLJ
                                                                            CO I
                                                                            C •!-
                                                                            .c c
                                                                            to 
   S-
TD O
 0) u_
 10
 o
 O. • ^
 X C  .
 
   QJ tO
 fC S- Z
 U «

i — in CO
•^  v-
 rO S- 1/5
   O to
 >
     -Q

     -0
      O)
                                                                            tO -4-J C
                                                                            -P  O •!-
                                                                            O QQ S-
                                                                            (TJ    O.
                                                                            S-    0)
                                                                            -i->    cc:
                                                                            X   • ^-'
                                                                            0)  en
                                                                            O  O3-P
                                                                               re x
                                                                            U)    O)
                                                                            S. v- 0)
                                                                            en 5   o u
                                                                           oo
                                                                            Ol
                                                                            s-
                                                                            3
                                                                            O)
                             110

-------
       B
      500      400     300     200       100      0
                  BINDING  ENERGY  (EV)
Figure 49.  Low-resolution ESCA spectra of a (A) coal fly ash surface
          following pyrene adsorption and (BJ the same coal fly ash
          surface following adsorption of pyrene and exposure to N02,
          (Reprinted by permission of D. F.  S. Natusch.)
                              Ill

-------
                                              CD
                                              Of

                                              8
                                              CD
                                                           c
                                                           o <->

                                                           •P -C
                                                           Q. U
                                                           s-  in
                                                           O  13
                                                           in -P
                                                           TD  to
                                                           m z

                                                           en  .
                                                           C CO


                                                           o u!
                                                            O Q
                                                            Ol  O
                                                            o
                                                            to  c
                                                           M-  O
                                                            S- •!-
                                                            3  in
                                                            U)  I/)
                                                            U)  t-
                                                            (0  d»
                                                               Q.
<— -a
 to  01
 o -P
 u  c
                                                               a.
                                                           4-  O)
                                                            o a:
                                                           -P   >
                                                           T:  o.
                                                            OJ
                                                            s-
112

-------
compounds.   If reactions with S03 are occurring, it appears that the PAH
itself must be decomposing.

     The observed reactions appear to be highly surface dependent.   Three
different surfaces were investigated in this regard, viz,  alumina,  fly ash,
and silica gel.   Only the reactions of adsorbed pyrene and N02 were examined
in these experiments.  The results indicate that pyrene reacts similarly when
adsorbed on alumina and fly ash but that the reactions on the silica gel
surface are significantly different.  Figure 48 shows the chromatogram of
pyrene-silica gel extract obtained immediately after exposure to N02 and
following aging for 3 weeks.   Initially, three strong peaks are evident, none
of which corresponds to pyrene.   Peaks A and C have been identified as
different dinitro pyrene products, and peak B appears to be the mono-nitro
product previously observed in the fly ash and alumina systems.  After 3
weeks, however,  the mono-nitro pyrene has disappeared, but the two other
products remain, and it appears that at least one of the dinitro products has
increased by a comparable amount, suggesting that conversion, rather than
decomposition, of the mono-nitro pyrene has occurred.  Since the acidic
surface of silica gel and existence of HN03 on the surface would be expected
to promote nitration, it is probably that nitration is at least partly
responsible for the reactions, as suggested by Pitts et al. (1978).

     Further information concerning PAH-surface interactions were obtained
from surface analytical studies using ESCA.  A low-resolution scan of a fly
ash surface after exposure to pyrene is shown in Figure 49A, and a scan
following exposure of the same material to N02 is shown in Figure 49B.  As
can be seen in Figure 49, the principal difference between the two scans is
the appearance of a nitrogen peak following N02 exposure.   A high-resolution
scan of the nitrogen region at liquid nitrogen temperature is presented in
Figure 50.   Three distinct regions are evident in this spectrum.  The largest
peak, at ^408 ev, corresponds to an inorganic nitrate (probably nitric acid),
the peak at ~404 ev corresponds to that expected for an organic nitrate, and
the peak at ~399 is probably that of a reduced nitrogen species, such as a
nitrile, ami no,  or pyridino compound.  Observation of the behavior of these
three regions at both room temperature and 80°C shows a gradual depletion of
the high binding energy nitrogen species and a slight increase in the reduced
nitrogen species with increasing temperature.

     These results suggest that nitric acid exists on the surface of the fly
ash and are consistent with the idea that nitric acid is either a nitrating
agent or a catalyst in these systems as in those studied by Pitts et al.
(1978).  No reduced nitrogen species were extracted or identified,  but the
formation of such species has been reported on soot by Chang and Novakov
(1975).  No useful information about PAH reaction products was obtained from
examination of the high-resolution carbon and sulfur peaks.

     The environmental implications of these results can be summarized as
follows:

 1.  Particulate-associated PAH can undergo a variety of reactions when
     exposed to normal plume gases.  Reactions appear to proceed most readily
     with N02 and S03, with NO and S02 being generally unreactive.


                                     113

-------
2.   The extent of the reactions appears to be highly surface dependent.   The
    nitration of PAH appears to be facilitated by the presence of inorganic
    nitrate on the surface.

3.   The formation of new PAH species,  at least in some instances, makes  some
    PAHs carcinogenic where previously they displayed little such behavior.
    This increases the potential problem due to particulate-associated PAH.
                                    114

-------
                                   SUMMARY

     As this report has demonstrated, coal fly ash is a complex mixture of
particles whose properties may vary considerably from source to source.
However, there exist many similarities between these samples, and the report
has attempted to stress them.  Most important, it is the surface and the size
of the fly ash that will be the primary factors in calculating the environ-
mental impact.  Thus, it is the smallest, most respirable particles that have
the highest concentrations of inorganic and organic materials and that are
most likely to present a health hazard.  The accessibility of potentially
toxic materials on or in coal fly ash to the environment and to man is not
yet well understood but is an area that requires much future work.
                                    115

-------
                                  REFERENCES

Ahearn, A. J.  Spark Source Mass Spectrometric Analysis of Solids.   In:
     Trace Characterization - Chemical and Physical, W. W., Meinke and B. F.
     Scribner, eds.  Natl. Bur.  Stand. (U.S.), Monogr. 100.  1967.   pp.
     347-376.
American Society for Testing Materials.
     pp.
X-ray Power Data File.  1970.  368
Ames, B. N., J. McCann, and E. Yamasaki.  Methods for Detecting Carcinogens
     and Mutagens with the Salmonella/mammalian-microsome Mutagenicity Test.
     Mutat. Res., 31(6):347-363, 1975.

Anderson, C.  A., H. J. Roden, and C. F. Robinson.  Negative Ion Bombardment
     of Insulators to Alleviate Surface Charge-up.  J. Appl. Phys., 40(8):
     3419-3420, 1969.

Andren, A. W., D. H. Klein, and Y. Talmi.  Selenium in Coal-Fired  Steam  Plant
     Emissions.  Environ. Sci. Technol., 9(9):856-858, 1975.

Armstrong, J.  T., and P. R. Buseck. Interaction between Natural and Anthro-
     pogenic Airborne Microparticles - An Electron Microprobe/SEM  [Scanning
     Electron Microscope] Study.  Abstracts of the First Chemical  Congress of
     the North American Continent, Mexico City, Mexico, Nov. 30 -  Dec. 5.
     Envir. Paper 48, 1975a.

Armstrong, J.  T., and P. R. Buseck.  Quantitative Chemical Analysis of
     Individual Micro Particles Using the Microprobe:  Theoretical.  Anal.
     Chem., 47:2178-2192, 1975b.

Badger, G. M.   Mode of Formation of Carcinogens in Human Environment. In:
     Analysis of Carcinogenic Air Pollutants, E. Sawicki and K. Cassel,  Jr.,
     eds.  Nat. Cancer Inst. Monogr. 9.  U.S. Dept. of Health, Education and
     Welfare.   1962.  pp. 1-16.
Bakale, D. K., B. N. Colby, and C. A. Evans, Jr.  High
     Microprobe Mass Spectrometry of Complex Matrices.
     1532-1537, 1975.
              Mass Resolution Ion
               Anal. Chem., 47:
Bertsch, W., R. C. Chang, and A. Zlatkis.  The Determination  of Organic
     Volatiles in Air Pollution Studies:  Characterization of Profiles.   J.
     Chromatogr. Sci., 12(4):175-182, 1974.
                                     116

-------
Bickelhaupt, R. E.  Electrical Volume Conduction in Fly Ash.  J. Air  Poll.
     Control Assoc., 24:251-255, 1974.

Bickelhaupt, R. E.  Surface Resistivity and the Chemical Composition  of  Fly
     Ash.  J. Air Poll. Control Assoc., 25(2):150-152, 1975a.

Bickelhaupt, R. E.  Volume Resistivity - Fly Ash Composition Relationship.
     Environ. Sci. Technol., 9(4):336-342, 1975b.

Block, C., and R. Dams.  Study of Fly Ash Emissions During Combustion of
     Coal.   Environ. Sci. Technol., 10:1011-1017, 1976.

Bolton, N.  E. , R. I. Van Hook, W. Fulkerson, W. S. Lyon, A. W. Andren, J. A.
     Carter, and J.  F. Emery.  Trace Element Measurements at the Coal-fired
     Allen Steam Plant.  Progress Report June 1971 to January 1973, Oak  Ridge
     National Laboratory, Oak Ridge, Tennessee, 1973.

Brecher, G., M. Schneiderman, and G. Z. Williams.  Evaluation of Electronic
     Red Blood Cell  Counter.  Am. J. Clinical Pathology, 26:1439-1449, 1956.
     Chem.  Abstr.  52:4943i.

Brimblecombe, P., and D. J.  Spedding.  The Dissolution of Iron from Ferric
     Oxide and Pulverized Fuel Ash.  Atmos. Environ., 9:835-838, 1975.

Brindley, G. W., and M. J. Nakahira.  The Kinetics of Dehydroxylation of
     Daolinite and Halloysite.  J. Am. Ceram. Soc. , 40:346-350, 1957.

Brown, J., N. J.  Ray, and M. Ball.  The Disposal of Pulverized Fuel Ash  in
     Water Supply Catchment Area.  Water Research, 10:1115-1121, 1976.

Bull, H. I., M. H. Hall, and W. E. Garner.  Reaction Between Carbon and
     Oxygen at Low Pressures and Room Temperature.  J. Chem. Soc., 837-847,
     1931.

Butcher, S. S., and R. J. Charlson.  An Introduction to Air Chemistry.
     Academic Press, New York, 1972.  241 pp.

Cabrera, J. G., and M. N. Gray.  Specific Surface Pozzolanic Activity, and
     Composition of Pulverised-fuel Ash.  Fuel, 52(3):213-219, 1973.

Cambell, J. A., J. C. Laul, K. K. Nielson, and R. D. Smith.  Separation  and
     Chemical Characterization of Finely Sized Fly-ash Particles.  Anal.
     Chem., 50:1032-1040, 1978.

Cambell, J. A., R. D. Smith, L. E. Davis, and K. L. Smith.  Characterization
     of Micron-size Fly Ash Particles by X-ray Photoelectron Spectroscopy
     (ESCA).  Sci. Tot. Environ., 12(l):75-85, 1979.

Castellan, G. W.    Physical Chemistry, Second Edition.  Addison-Wesley
     Publishing Co., Reading, Mass., 1971.  866 pp.
                                     117

-------
Chang, C. C.  Analytical Auger Electron Spectroscopy.  In:  Characterization
     of Solid Surfaces, P. F. Kane and G. B. Larrabee, eds.  Plenum  Press,
     New York, New York, 1974.  pp. 509-575.

Chrisp, C. E., G. L. Fisher, and J. E. Lammert.  Mutagenicity of Filtrates
     from Respirable Coal Fly Ash.  Science, 199(4324):73-75, 1978.

Coburn, J. W.  Sputtering in the Surface Analysis of Solids:  A Discussion  of
     Some Problems.   J. Vac. Sci. Techno!., 13(5):1037-1044, 1976.

Coles, D. G., R. C.  Ragaini, J. M. Ondov, G. L. Fisher, D. Silberman,  and B.
     A. Prentice.  Chemical Studies of Stack Fly Ash from a Coal-fired Power
     Plant.   Environ.  Sci. Technol., 13:455-459, 1979.

Colovos, G., G. S. Wilson, and J. L. Moyers.   Particle Size Determination of
     Trace Elements in Urban Air.  Int. J. Environ. Anal. Chem.,
     3(3):239-252, 1974.

Committee on Biologic Effects of Atmospheric Pollutants:  Particulate  Poly-
     cyclic Organic Matter.  National Academy  of Sciences, Washington, D.C.,
     1972.  361 pp.

Corey, R. C.,  B. J. Cross, and W. T. Reid.  External Corrosion of Furnace
     Wall Tubes - II.   Significance of Sulphate Deposits and Sulphur Trioxide
     in Corrosion Mechanism.  Trans. Am. Soc.  Mech. Eng., 67: 289-302, 1945.

Cox, J. A., G. L. Lundquist, A. Przyjazny, and C. D. Schmulbach.   Leaching
      of Boron from Coal Ash.  Environ. Sci. Technol., 12:722-723, 1978.

Craig, N. L., A. B.  Marker, and T. Novakov.  Determination of the  Chemical
     States of Sulfur in Ambient Pollution Aerosols by X-ray Photoelectron
     Spectroscopy.  Atmos. Environ., 8(1):15-21, 1974.

Dalmon, J., and E.  Raask.  Resistivity of Particulate Coal Minerals.   J.
     Inst.  Fuel., 45(374):201-205, 1972.

Dalmon, J., and D.  Tidy.  Comparison of Chemical Additives as Aids to  the
     Electrostatic  Precipitation of Fly Ash.   Atmos. Environ.,  6:721-734,
     1972.

Dautreband,  L.  Microaerosols.  Academic Press, New York, 1962.  360 pp.

Davison,  R.  L., D.  F. S. Natusch, J. R. Wallace, and C. A. Evans,  Jr.   Trace
     Metals  in Fly  Ash:  Dependence of Concentration on  Particle Size.
     Environ. Sci.  Technol., 8:1107-1113, 1974.

Diehl, E. K., F. du Breuil,  and  R. A. Glenn.   Polynuclear Hydrocarbon  Emis-
     sion from Coal-fired  Installations.  Trans. Am. Soc. Mech.  Eng. J.  Eng.
     for  Power, 89:276-282,  1967.

Ditl,  P., and R. W. Coughlin.  Improving  Efficiency  of  Electrostatic Pre-
     cipitation  by  Physicochemical Modification of the  Electrical  Resistivity
     of  Fly  Ash.  J. Am.  Inst. Chem.  Eng.,  22(4):730-746, 1976.

                                     118

-------
Dreesen, D.  R.,  E.  S. Gladney, J. W. Owens, B. L. Perkins, C. L. Wienke, and
     L.  E.  Wangen.   Comparison of Levels of Trace Elements Extracted from Fly
     Ash and Levels Found in Effluent Waters from a Coal-Fired Power Plant.
     Environ.  Sci.  Techno!. , 11:1017-1019, 1977.

Eggett,  J.  M. ,  and T. M. Thorpe.  Mobilization of Chromium from Fly Ash
     Particulates by Aqueous Systems Modeling Natural Waters.  J. Environ.
     Sci.  Health, A13(4):295-313, 1978.

Evans, C.  A.,  Jr.  Secondary Ion Mass Analysis:  A Technique for Three-
     dimensional Characterization.  Anal. Chem., 44(13):67A-80A, 1972.

Evans, C.  A.,  Jr.  Surface and Thin Film Compositional Analysis:  Description
     and Comparison of Techniques.  Anal. Chem., 47:818A-829A, 1975.

Falk, H. L., I.  Markul, and P. Kotin.  Aromatic Hydrocarbons. IV.  Their Fate
     Following Emission into the Atmosphere and Experimental Exposure to
     Washed Air and Synthetic Fog.  Am. Med. Assoc. Arch. Ind. Health,
     13:13-17,  1956.

Farrar,  H., IV.   Relating the Mass  Spectrum to the Solid Sample Composition.
     In:  Trace Analysis by Mass Spectrometry, A. J. Ahearn, ed.  Academic
     Press, New York, 1972.   pp. 239-95.

Fisher,  G.  L.   Personal Communication.  Radiobiology Laboratory, University
     of California, Davis, California, 1978.

Fisher,  G.  L.,  D. P. Y. Chang, and  M. Brummer.  Fly Ash Collected from
     Electrostatic Precipitators: Microcrystalline Structures and the Mystery
     of the Spheres.  Science, 192:553-555, 1976.

Fisher,  G.  L.,  B. A. Prentice, D. Silberman, J. M. Ondov, A. H. Biermann, R.
     C.  Ragaini, and A. R. McFarland.  Physical and Morphological Studies of
     Size Classified Coal Fly Ash.  Environ. Sci. Techno!.,  12:447-451, 1978.

Fisher,  G.  L.,  D. Silberman, R.  E.  Heft, and J. M. Ondov.  Filtration Studies
     with Neutron-activated Coal Fly Ash.  Environ. Sci. Technol.,
     13:689-693, 1979.

Flesch,  J.  P.,  C. H. Norris, and A. E. Nugent, Jr.  Calibrating Particulate
     Air Samplers with Monodisperse Aerosols:  Application to the Anderson
     Cascade Impactor.  Am.  Ind. Hyg. Assoc. J., 28:507-516, 1967.

Gladney, E. S., G. E. Gordon, and W. H. Zoller.  Abnormally  Enriched Trace
     Elements in the Atmosphere.  Seventh Annual Conference  on Trace Sub-
     stances in Environmental Health, University of Missouri, Columbia,
     Missouri,  June, 1973.

Gladney, E. S., J. A. Small, G.  E.  Gordon, and W. H. Zoller.  Composition and
     Size Distribution of a In-stack Particulate Material at a Coal-fired
     Power Plant.  Atmos. Environ., 10:1071-1077, 1976.
                                      119

-------
Gladney, E. S., W. H. Zoller, A. G. Jones, and G. E. Gordon.  Composition and
     Size Distribution of Atmospheric Particulate Matter in Boston Area.
     Environ.  Sci. Techno!., 8(6):551-557, 1974.

Gluskoter, H.  J.  Clay Minerals in Illinois Coals.  J. Sed. Pet., 37:205-214,
     1967.

Gooch, J. P.,  and S. Oglesby.  A Theoretically Based Technique for Estimating
     the Size of Electrostatic Precipitators.  Southern Research Institute
     Report to the Environmental Protection Agency, October, 1974.  pp.
     84-88.

Gordon, G. E.   Study of the Emissions from Major Air Pollution Sources and
     their Atmospheric Interactions, First Year Progress Report, Nov. 1, 1972
     to Oct.  31, 1973.  University of Maryland, College Park, Maryland, 1973.

Gordon, G. E.   Atmospheric Impact of Major Sources and Consumers of Energy.
     NSF-RANN Progress Report October 1974 to August 1975, University of
     Maryland, College Park, Maryland, 1975.

Gordon, G. E., and W. H. Zoller.  Normalizations and Interpretation of At-
     mospheric Trace-element Concentration Patterns.  Oak Ridge Trace
     Elements Conference, Oak Ridge, Tennessee, 1973.

Hangebrauck, R. P. , D. J. Von Lehmden, and J. E. Meeker.  Sources of Poly-
     nuclear Hydrocarbons in the Atmosphere.  Public Health Service Pub-
     lication 999-AP-33, U. S. Dept. of Health, Education and Welfare, 1967.
     44 pp.

Hatch, T. F.,  and P. Gross.  Pulmonary Deposition and Retention of Inhaled
     Aerosols.  Academic Press, New York, 1964.  192 pp.

Heindrycks, R., and R. Dams.  Continental, Marine, and Anthropogenic Con-
     tribution to the Inorganic Composition of the Aerosol of an Industrial
     Zone.  J. Radioanal. Chem., 19(2):339-349, 1974.

Hodgeman, C.  D. (ed.).  Handbook of Chemistry and Physics, 44th Edition.  The
     Chemical  Rubber Publishing Co., Cleveland, Ohio, 1963.  pp. 24-29.

Hopke, P. K.   Chemistry Progress Report.  Laboratory for Nuclear Science,
     Massachusetts Institute of Technology,  Cambridge, Massachusetts, 1969.

Hulett,  L. D.   Personal Communication, 1973.

Hulett,  L. D.   Progress Report.  Oak Ridge National  Laboratories, Oak Ridge,
     Tennessee, 1973.

Hulett,  L. D., and A. J. Weinberger.  Some Etching Studies of the Micro-
     structure and Composition of  Large Aluminosilicate Particles in Fly Ash
     from Coal Burning Power Plants.  Environ. Sci.  Technol., Submitted,
     1979.
                                      120

-------
James, W. D., M. Janghorbani, and T. Baxter.  Leachability of Neutron  Irradi-
     ated Fly Ash.  Anal. Chem., 49:1994-1997, 1977.

Jones, H. R.  Fine Dust and  Particulates Removal.   Pollution Control Review,
     No.  11.  Noyes Data Corp., Park Ridge, New Jersey, 1972.

Jones, P. W., and R. I. Freudenthal (eds.).  Carcinogenesis--A Comprehensive
     Survey, Vol. 3.  Raven  Press, New York, 1978.  512 pp.

Jones, P. W., R. D. Giammar, P. E. Strup, and T. B. Stanford.  Efficient
     Collection of Polycyclic Organic Compounds from Combustion Effluents.
     Environ. Sci. Technol., 10(8):806-810, 1976.

Judikis,  H. S., and S. Siegel.  Particle-catalyzed  Oxidation of Atmospheric
     Pollutants.  Atmos. Environ., 7(6):619-631, 1973.

Kaakinen, J. W., R. M. Jorden,  J. H. Lawasani, and  R.  E. West.  Trace  Element
     Behavior in Coal Fired  Power Plants.  Environ. Sci. Technol., 9:862-869,
     1975.

Karolinska Institute.  Proceedings of General Air Pollution and Human  Health
     with Special Reference  to  Long-Term Effects Meeting, Stockholm, March
     1977.  Environ. Health  Perspectives, 22:1-12,  1978.

Keyser, T. R., D. F. S. Natusch, C. A. Evans, Jr.,  and R. W. Linton.   Charac-
     terizing the Surfaces of Environmental Particles.  Environ. Sci.
     Technol., 12(7):768-773, 1978.

Klein, D. H., A. W. Andren,  J.  A. Carter, J. F. Emergy, C. Feldman, W.
     Fulkerson, W. S. Lyon,  J.  C. Ogle, Y. Talmi, R. I. Van Hook, and  N.
     Bolton.  Pathways of Thirty-seven Trace Elements  Through Coal-fired
     Power Plant.  Environ.  Sci. Techno!., 9:973-979,  1975.

Korfmacher, W. A.  The Decomposition of Polycyclic  Organic Matter Adsorbed on
     Coal Fly Ash.  Ph.D. Thesis, University of Illinois, Urbana, Illinois,
     1978. 272 pp.

Kotrappa, P., and M. E. Light.  Design and Performance of the Lovelace
     Aerosol Particle Separator.  Rev. Sci. Instrum.,  43(8):1106-1112,  1972.

Krauskopf, K. B.  Thermodynamics of Geochemistry, Chapter 3, In:  Handbook of
     Geochemistry, Vol. I, K. H. Wedepehl, ed.  Springer-Verlag, New York,
     1972.  pp. 37-76.

Kropp, L. I., I. N. Shmigol, G. S. Chekanov, S. Oglesby, Jr., and R. E.
     Bickelhaupt.  J. Air Poll. Control Assoc., 29(6):665-669, 1979.

Lamb, R.  E.  The Chemical Characterization of Airborne Particles.  Ph.D.
     Thesis, University of Illinois, Urbana, Illinois, 1975.  168 pp.
                                     121

-------
Langham, W.  W.   Radioisotope Absorption and Methods of I: limi nation:  Relative
     Significance of Portals of Entry.   In:  Symposium on Radioisotopes in
     the Biosphere, R.  S.  Caldecott and L.  A.  Synder, eds.  University of
     Minnesota Press, Minneapolis, Minnesota,  1960.  pp,  489-513.

Lao, R.  C.,  R.  S. Thomas,  H. Oja, and L. Debois.  Application of a Gas
     Chromatograph - Mass  Spectrometer-Data Processor Combination to the
     Analysis of the Polycyclic Aromatic Hydrocarbon Content of Airborne
     Pollutants.  Anal.  Chem.,  45(6):908-915,  1973.

Lee, R.  E. ,  Jr., H. L.  Crist, A.  E. Riley,  and K. E. MacLeod.  Concentration
     and Size of Trace Metal Emissions from a Power Plant, a Steel Plant, and
     a Cotton Gin.  Environ. Sci. Technol., 9:643-647, 1975.

Lee, R.  E.,  Jr., S. S.  Goransen,  R. E.  Enrione, and G. B. Morgan.  National
     Air Surveillance Cascade Impactor Network.  II.  Size Distribution
     Measurements of Trace Metal  Components.   Environ. Sci. Technol.,
     6:1025-1030, 1972.

Lee, R.  E.,  Jr., and D.  J. Von Lehmden.  Trace Metal Pollution in the
     Environment.  J. Air Poll. Control Assoc., 23(10):853-857, 1973.

Linton,  R.  W., A. Loh,  D.  F. S. Natusch, C. A. Evans, Jr., and P. Williams.
     Surface Predominance of Trace Elements in Airborne Particles.  Science,
     191:852-854, 1976.

Linton,  R.  W., P. Williams, C.  A. Evans, Jr.,  and D. F. S. Natusch.  Unpub-
     lished Results, 1976.

Linton,  R.  W., P. Williams, C.  A. Evans, Jr.,  and D. F. S. Natusch.  Determi-
     nation of the Surface Predominance of Toxic Elements in Airborne Parti-
     cles by Ion Microprobe Mass Spectrometry and Auger Electron Spectro-
     metry.   Anal. Chem.,  49:1514-1520, 1977.

Littlejohn,  R. F., and J.  D. Watt.  The Distribution of Mineral Matter in
     Pulverized Fuel.  In:  The Mechanism of Corrosion by Fuel Impurities, H.
     R.  Johnson and D.  J.  Littler, eds.  Butterworths, London, 1963.  pp.
     102-112.

Lofroth, G.   Mutagenicity Assay of Combustion  Emissions.  Chemosphere,
     7(10):791-798, 1978.

Loh, A.   Some Environmental Applications of Analytical Microscopy.  Ph.D.
     Thesis, University of  Illinois, Urbana, Illinois, 1975.  257 pp.

Levering, J. F.  Application of SIMS Microanalysis Techniques to Trace
     Element and  Isotopic Studies  in Geochemistry and Cosmochemistry.  In:
     Secondary  Ion Mass Spectrometry, K. F. J. Heinrich and D. E. Newburg,
     eds.  Natl. Bur. Stand. (U.S.), Spec.  Publ. 427,  L975.  pp. 135-178.

Maney, J. P., J.  L. Fasching, and  P. K. Hopke.  A Versitile and Comprehensive
     Analysis Code for Automated Reduction of  Gamma  Ray Spectral Data.
     Computers  and Chemistry, 1:257-264, 1977.

                                     122

-------
Mason, B.   Principles of Geochemistry, Third Edition.  Wiley, New York,  1966.
     pp.  55-59.  329 pp.

Matthews,  B.  J., and R. F. Kemp.  Development of Laser Instrumentation for
     Particle Measurement.  TRW Report No. 14103-6003-RO-OO, 1971.

Matusiewicz,  H., and D. F. S. Natusch.  To be submitted for publication.

McCaughan, D. V., and R. A. Kushner.  Impurity-Movement Problems in Analysis
     Methods Using Particle Bombardment.  In:  Characterization of Solid
     Surfaces, P. R. Kane and G. B. Larrabee, eds.   Plenum Press, New York,
     1974.  pp. 627-640.

McFarland, A. R., R. W. Bertch, G.  L. Fisher, and B. A. Prentice.  A Frac-
     tionator for Size-classification of Aerosolized Solid Particulate
     Matter.   Environ. Sci. Techno!., 11:781-784, 1977.

McHugh, J. A.  Factors that Influence an Elemental Depth Concentration Pro-
     file.  In:  Secondary Ion Mass Spectrometry, K. F. J. Henrich and D. E.
     Newburg, eds.  Natl.  Bur. Stand. (U.S.), Spec.  Publ. 427, 1975.  pp.
     179-189.

Miguel, A. H.  Studies of Gas Solid Reactions of Environmental Significance:
     I.  Pyrene Adsorption onto Fly Ash; II.  Thiol Oxidation and Adsorption
     by Activated Charcoal.  Ph.D. Thesis, University of Illinois, Urbana,
     Illinois, 1976. 136 pp.

Miguel, A. H., W. A. Korfmacher, E. L. Wehry, G. Mamantov, and D. F. S.
     Natusch.  Apparatus for Vapor-phase Adsorption  of Polycyclic Organic
     Matter onto Particulate Surfaces.  Environ. Sci. Technol., 13(10):
     1229-1232, 1979.

Miguel, A. H., M. R. Schure, and D. F. S. Natusch.  Adsorption of Vapor  Phase
     Polycyclic Organic Matter onto Fly Ash.  Submitted to Anal. Chem.,  1979.

Miller, F. J., D. E. Gardner, J. A. Graham, R. E. Lee, Jr., W. E. Wilson, and
     J. D. Bachman.  Size Considerations for Establishing a Standard for
     Inhalable Particles.   J. Air Poll. Control Assoc. , 29(6):610-615, 1979.

Morabito,  J.  M.  A Comparison of Auger Electron Spectoscopy (AES) and
     Secondary Ion Mass Spectrometry (SIMS).  In:  Secondary Ion Mass
     Spectrometry, K. F. J. Henrich and D. E. Newberg, eds.  National Bur.
     Stand. (U.S.), Spec.  Publ. 427, 1975.  pp. 191-224.

Morey, G.  W.   The Properties of Glass, Second Edition.  Reinhold, New York,
     1954.  591 pp.

Morrow, P. E.  Evaluation of Inhalation Hazards Based on the Respirable  Dust
     Concept and the Philosophy and Application of Selective Sampling.   Amer.
     Ind.  Hyg.  Assoc. J.,  25:213-236, 1964.
                                     123

-------
Mukai, M., J. F. Thomas, and B. D. Tebbens.  Fate of Airborne Benzo[a]pyrene.
     Environ. Sci. Technol., 2(l):33-39, 1968.

Muller, R. 0.  Spectrochemical Analysis by X-ray Fluorescence.  Plenum  Press,
     New York, 1972.  pp. 72-86.  326 pp.

Natusch, D. F. S.  Characterization of Fly Ash from Coal Combustion.  Work-
     shop Proceedings on Primary Sulfate Emissions from Combustion  Sources,
     US EPA 600/9-78-020b, Vol. 2, 1978a.  pp. 149-163.

Natusch, D. F. S.  Potentially Carcinogenic Species Emitted from  Fossil
     Fueled Power Plants.  Environ. Health Perspectives, 22:79-90,  1978b.

Natusch, D. F. S., C. F. Bauer, H. Matusiewicz, C. A. Evans, J. Baker,  A.
     Loh, R. W.  Linton, and P. K. Hopke.  Characterization of Trace Elements
     in Fly Ash.  Proceedings of the International Conference on  Heavy  Metals
     in the Environment, Toronto, Vol. II, Part 2, 1975.  pp. 535-575.

Natusch, D, F. S., and H. Matusiewicz.  Unpublished results, 1976.

Natusch, D. F. S., and A. H. Miguel.  A Diffusion Cell  for the  Preparation of
     Dilute Vapor Concentrations.  Anal. Chem., 47:1705-1707, 1975.

Natusch, D. F. S., and M. R. Schure.  Manuscript in preparation,  1979.

Natusch, D. F. S., and B. A. Tomkins.  Theorectical Considerations  of the
     Adsorption  of PAH Vapor onto Fly Ash in a Coal Fired Power Plant.   In:
     Carcinogenesis -  A Comprehensive Survey,  Vol. 3, Polynuclear Aromatic
     Hydrocarbons, P. W. Jones and R. I. Freudenthal, eds.  Raven Press, New
     York, 1978.  pp. 145-153.

Natusch, D. F. S., and J. R. Wallace.  Urban Aerosol Toxicity:  The Influence
     of Particle  Size.  Science, 186:695-699,  1974.

Natusch, D. F. S., J. R. Wallace, and C. A. Evans, Jr.  Toxic Trace Elements:
     Preferential Concentration in Respirable  Particles.  Science,  183:
     202-204, 1974.

Nelson, W., and  C. Cain, Jr.  Corrosion of Superheaters and Reheaters of
     Pulverised-Coal-Fired Boilers.  Trans. Am. Soc. Mech. Eng. J.  Eng.  for
     Power, 82:194-204, 1960.

Nicholls, G.  D.   The Geochemistry of Coal Bearing Strata.  In:  Coal and Coal
     Bearing  Strata, D. Merchison and P. Westoll, eds.  American  Elsevier,
     New York,   1968.  pp. 269-307.

Novakov, T. ,  S.  G. Chang, and A. B. Harker.  Sulfates as Pollution  Particu-
     lates:   Catalytic Formation on Carbon (Soot) Particles.  Science 186:
     259-261, 1974.

O'Gorman, J.  V.,  and P.  L. Walker, Jr.  Mineral Matter  Characteristics  of
     some American Coals.  Fuel, 50:135-151, 1971.
                                     124

-------
Olson, K. W., and R. K. Skogerboe.  Identification of Soil  Lead Compounds
     from Automotive Sources.  Environ. Sci. Technol. , 9:227-230,  1975.

Ondov, J. M., R. C.  Ragaini, and A. H. Biermann.  Comparison of Particulate
     Emissions from a Wet Scrubber and Electrostatic  Precipitator  at  a  Coal-
     fired Power Plant.  Presented at American Chemical Society Meeting, Div.
     of Environmental Chemistry, San Francisco, August 29 - September 3,
     1976.

Ondov, J. M., R. C.  Ragaini, and A. H. Biermann.  Elemental Emissions from a
     Coal-fired Power Plant.  Comparison of a Venturi Wet Scrubber System
     with a Cold-side Electrostatic Precipitator.  Environ. Sci. Technol.,
     13:598-606, 1979.

Ondov, J. M. , R. C.  Ragaini, R. E. Heft, G. L. Fisher, D. Silberman,  and B.
     A. Prentice.   Interlaboratory Comparison of Neutron Activation and
     Atomic Absorption Analyses of Size-classified Stack Fly Ash.   Natl. Bur.
     of Stand. (U.S.) Special Publication 464.  Proceedings of 9th IMR  Sym-
     posium, Gaithersburg, Maryland, September 20-24, 1976, 1977.

Ondov, J. M., W. H.  Zoller,  I. Olmez, N. K. Aras, G.  E. Gordon, L.  A. Ranci-
     telli, K. H. Able, R. H. Filby, K. R. Shah, and  R. C.  Ragaini.   Ele-
     mental Concentrations in the National Bureau of  Standards Environmental
     Coal and Fly Ash Standard Reference Materials.   Anal.  Chem.,  47(7):
     1102-1109, 1975.

Owens, E. B., and N. A. Giardino.  Quantitative Mass  Spectrometry  of  Solids.
     Anal. Chem., 35:1172-1178, 1963.

Patterson, C. C., and J. D.  Salvia.  Lead in the Modern Environment.  How
     Much is Natural.  Sci.  and Citizen, 10:66-79, 1968.

Peuschel, R. F., C.  C. Van Valin, and F. P. Parungo.  Effect of Air Pollu-
     tants on Cloud Nucleation.  Geophys., Res. Lett., l(l):51-54,  1974.

Phung, H. T., L. J.  Lund, A. L. Page, and G. R. Bradford.   Trace Elements in
     Fly Ash and their Release in Water and Treated  Soils.  J. Environ.
     Qual., 8(2):171-175, 1979.

Pillay, K. K. S., and C. C.  Thomas, Jr.  Determination of the Trace Element
     Levels in Atmospheric Pollutants by Neutron Activation Analysis.   J.
     Radioanal. Chem., 7(1):107-118, 1971.

Pitts, J. N., Jr.,  K. A. Van Cauwenberghe, D. Grosjean, J.  P. Schmid, D. R.
     Fitz, W. L. Belser, Jr., G. B. Knudson, and P. M. Hynds.  Atmospheric
     Reactions of Polycyclic Aromatic Hydrocarbons:   Facile Formation of
     Mutagenic Nitro Derivatives.  Science, 202:515-519, 1978.

Price, N. B., and G. R. Angell.  Determination of Minor Elements in Rocks by
     Thin Film X-ray Fluorescence Techniques.  Anal.  Chem., 40(3):  660-663,
     1968.
                                     125

-------
Pringle, W.  J.  S. ,  and E. Bradburn.  Mineral Matter in Coal.  (II)  Compo-
     sition of the  Carbonate Minerals.  Fuel, 37:166-180, 1958.

Pupp, C.,  R.  C.  Lao, J.  J.  Murray, and R.  F. Pottie.  Equilibrium Vapour
     Concentrations of some Polycyclic Aromatic Hydrocarbons, As.CL and SeCL
     and the Collection Efficiencies of these Air Pollutants.  Almos.
     Environ.,  8(9):915-925, 1974.

Raabe, 0.  G.   Aerosol Aerodynamic Size Conventions for Inertia! Sampler Cali-
     bration.   J.  Air. Poll. Control Assoc., 26(9):856-860, 1976.

Raask, E.   Cenospheres in Pulverized-fuel  Ash.   J. Inst.  Fuel, 41:339-344,
     1968.

Rahn, K. A.,  R.  Dams, J.  A. Robbins, and J.  W.  Winchester.  Diurnal Varia-
     tions of Aerosol Trace Element Concentrations as Determined by Non-
     destructive Neutron Activation Analyses.  Atmos. Environ., 5:413-422,
     1971.

Ranz, W. E.,  and J. B. Wong.  Jet Impactors for Determining the Particle Size
     Distributions  of Aerosols.  Ind. Hyg. Occup. Med., 5:464-477, 1952.

Rao, C.  P., and H.  J. Gluskoter.  Occurrence and Distribution of Minerals  in
     Illinois Coals.  Illinois State Geological Survey Circular No. 476,
     1973.   56 pp.

Reid, W. T.   External Corrosion and Deposits:  Boilers and Gas Turbines.
     Elsevier, New York, 1971.  199 pp.

Ruch, R. R. ,  H.  J.  Gluskoter, and N. F. Shimp.   Occurrence and Distribution
     of Potentially Volatile Trace Elements in Coal.  Illinois State Geo-
     logical  Survey Report No. 72, 1974.

Sarofim, A.  F., J.  B. Howard, and A. S. Padia.   Physical Transformation of
     the Mineral Matter  in Pulverized Coal  Under Simulated Combustion Condi-
     tions.   Combust. Sci.  Technol., 16:187-204, 1977.

Schmidt, F.  J., and J. L. Royer.  Submicrogram Determination of Arsenic,
     Selenium, Antimony, and Bismuth by Atomic Absorption Utilizing Sodium
     Borohydride Reduction.  Anal. Letters, 6(1):17-23,  1973.

Schroeder, H. A.  Metals in the Environment.  Environment, 13(8):18-32, 1971.

Searle, C. E. (ed.).  Chemical Carcinogens.  American Chemical Society,
     Monog. 173, 1976.   788 pp.

Shannon, D. G., and  L. 0. Fine.   Cation Solubilities of  Lignite  Fly Ashes.
     Environ. Sci.  Technol., 8(12):1026-1028, 1974.

Simons, D. S., J.  E.  Baker, and C. A. Evans, Jr.  Evaluation of  the  Local
     Thermal Equilibrium Model  for Quantitative  Secondary  Ion  Mass Spectro-
     metric Analysis.  Anal. Chem., 48(9):1341-1348, 1976.

                                      126

-------
Simons, H.  S.,  and J.  W. Jeffery.  An X-ray Study of Pulverised Fuel Ash.  J.
     Appl.  Chem.,  10:328-336, 1960.

Smith, R.  D.,  J.  A.  Cambell, and K. K. Nielson.  Concentration Dependence
     upon Particle Size of Volatilized Elements in Fly Ash.  Environ. Sci.
     Technol.,  13(5):553-558, 1979.

Southern Research Institute.  Survey of Information on Fine Particle Control.
     Electric Power Research Institute, Report No. 259, Palo Alto, Califor-
     nia, 1975.   247 pp.

Sparks, C.  J.   Personal communication, 1973.

Stahley, S.   Personal  communication, 1976.

Steyermark,  A.   Quantitative Organic Microanalysis.  Academic Press, New
     York,  1961.   p. 222.  665 pp.

Strauss, W.   Industrial Gas Cleaning, Second Edition.  Pergamon Press,  Elms-
     ford,  New York, 1975.  471 pp.

Talbot, R.  W.,  M.  A. Anderson, and A. W. Andren.  Qualitative Model of
     Heterogeneous Equilibria in a Fly Ash Pond.  Environ. Sci. Technol.,
     12:1056-1062, 1978.

Theis, T. L., and J. L. Wirth.  Sorptive Behavior of Trace Metals on Fly Ash
     in Aqueous Systems.  Environ. Sci. Technol., 11(12):1096-1100, 1977.

Theriault,  G.,  T.  Barry, and M. Thomas.  Technique of Preparing Powder
     Samples for AES (Auger Electron Spectroscopy) and/or ESCA (Electron
     Spectroscopy for Chemical Analysis) Analysis.  Anal. Chem., 47(8)1492-
     1493,  1975.

Toca,  F. M.   Lead and Cadmium Distribution in the Particulate Effluent  from a
     Coal Fired Boiler.  Ph.D. Dissertation, University of Iowa, July,  1972.
     See also Dissertation Abstracts International, 33:31566 (1973).  174 pp.

Townsend, P.  D.,  J.  C.  Kelly, and N. E. W. Hartley.  Ion Implantation,  Sput-
     tering and Their Applications.  Academic Press, New York, 1976.  pp.
     125, 233.   344 pp.

Urone, P.,  and W.  H. Schroeder.  Sulfur Dioxide in the Atmosphere:  A Wealth
     of Monitoring Data, but Few Reaction Rate Studies.  Environ. Sci.
     Technol. ,  3:436-445, 1969.

U. S.  Department of Health, Education, and Welfare.  Air Quality Criteria for
     Particulate Matter.  National Air Pollution Control Administration Publ.
     No. AP-49, Washington, D.C., 1969.

U. S.  Environmental  Protection Agency.  Emissions from Residential and  Small
     Commercial Stoker-Coal-Fired Boilers under Smokeless Operation.  Publ.
     EPA-600/7-76-029.  Office of Research and Development, Washington, D.C.,
     1976a.   78 pp.

                                     127

-------
U. S. Environmental Protection Agency.  Environmental Assessment Sampling and
     Analysis:  Phased Approach and Techniques for Level 1.  Publ. EPA-600/
     2-77-115.  Office of Research and Development, Washington, D.C., 1977.
     34 pp.

U. S. Environmental Protection Agency.  Field Testing:  Trace Element and
     Organic Emissions from Industrial Boilers.  Publ. EPA-600/2-76-086b.
     Office of Research and Development, Washington, D.C., 1976c.  146 pp.

U. S. Environmental Protection Agency.  Preliminary Emissions Assessment of
     Conventional Stationary Combustion Systems; Volume 1 - Executive Sum-
     mary.  Publ. EPA-600/2-76-046a.   Office of Research and Development,
     Washington, D.C., 1976b.   3 Volumes.

U. S. Government Printing Office.  The Clean Air Act as Amended August 1977,
     Serial No. 95-11, Washington, D.C., 1977.

Von Lehmden, D. J., R. H. Jungers, and R.  E. Lee, Jr.  Determination of Trace
     Elements in Coal, Fly Ash, Fuel  Oil,  and Gasoline.  Preliminary Compari-
     son of Selected Analytical Techniques.   Anal. Chem., 46(2):239-245,
     1974.

Wallace, J. R.  The Physical and Chemical  Characterization of Airborne Parti-
     culate Matter.  Ph.D. Thesis, University of Illinois, Urbana, Illinois,
     1974.  162 pp.

Watt, J. D., and D. J. Thome.  The Composition and Pozzolanic Properties of
     Pulverised Fuel Ashes.  J. Appl.  Chem., 15:585-604, 1965.

Wehry, E. L.  Analytical Problems in the Study of Photochemical Reactions.
     In: Analytical Photochemistry and Photochemical Analysis, J. M. Fitz-
     gerald, ed.  Marcel Dekker, New York, 1971.  pp. 167-212.

Weintraub, W., S. Goldberg, and A. A.  Orning.  A Study of Sulfur Reactions in
     Furnace Deposits.  Trans. Am. Soc. Mech. Eng. J. Eng. for Power, 83:
     444-450, 1961.

White, H. J.  Industrial Electrostatic Precipitation.  Addison-Wesley,
     Reading, Massachusetts, 1963.  376 pp.

Williams, P., and C. A. Evans, Jr.  High Resolution Secondary Ion Mass
     Spectrometry.   In:  Secondary Ion Mass Spectrometry,, K. F. J. Heinrich
     and D. E. Newbury, eds.  Natl. Bur. Stand. (U.S.), Spec. Publ. 427,
     1975.  pp. 63-68.

Williams, P., and C. A. Evans, Jr.  A Simple Electronic Aperature for
     Rastered-beam Depth Profiles.  Int. J.  Mass Spectrom. Ion Phys.,
     22(3/4): 327-331, 1976.

Yeh, H. C. , R. F. Phalen, and 0. G. Raabe.  Factors Influencing the Depo-
     sition of Inhaled Particles.  Environ.  Health Perspectives, 15:147-156,
     1976.


                                     128

-------
APPENDICES
      129

-------
                            APPENDIX A

EXPERIMENTAL PROCEDURES AND INSTRUMENTAL TECHNIQUES

Sample Sources and Fractionation Procedures

     A number of different fly ashes were studied.   The sources included two
coal-fired power plants burning midwestern bituminous coals,  the Iowa/
Illinois Gas and Electric Plant in Davenport,  Iowa, and the Vermilion Plant
near Danville, Illinois.   The former utilized  a chain grate stoker,  while the
other used a pulverized coal feed system.

     Samples were collected in the stack systems at temperatures in  the range
of 200-300°C.  Ash from the chain grate stoked plant was studied most exten-
sively because it had generally higher trace element bulk concentrations,
facilitating detection by the various analytical techniques.   Large  particles
(45-180 urn physical diameter) were analyzed to facilitate both particle
handling and the characterization of individual particles using surface
microanalytical techniques.  The trace element studies of the aerodynamically
sized fly ash studied in the trace metal section came from the pulverized
feed system.  Samples for that study represented a) fly ash retained in the
cyclone precipitation system and b) airborne fly ash collected in the ducting
approximately 10 feet from the base of the stack.   The western fly ash was
collected from the electrostatic precipitators of the Corrette plant in
Billings, Montana, which burns a Montana sub-bituminous coal.

     Size separations were made in three different ways.  Aerodynamic labo-
ratory separations on the bulk ash were made using a Rollar particle size
analyzer (American Instrument Co.).  Physical  separations were made  by
sieving with a sonic sifter (ATM Corp., Milwaukee, Wisconsin).  Airborne fly
ash samples were collected and size differentiated HI situ using a stainless
steel Andersen stack sampler which is designed to operate at the stack
temperature  (Andersen Samplers, Inc., Atlanta, Georgia).

     Particle size calibrations were based on data supplied by the sampler
manufacturers.  These data are established in terms of equivalence to the
aerodynamic  diameter of spherical particles of unit density (Ranz and Wong,
1952; Flesch et al., 1967).  The size distributions of particles collected on
the third and fourth plates (4.6-7.1 urn and 3.0-4.6 urn) of the Andersen stack
sampler were also examined using a Coulter counter (Coulter Electronics,
Inc., Hialeah, Florida) in the timed analysis mode with a 100 urn aperture
(Brecher et  al., 1956).  Milligram portions of the fly ash were dispersed in
a 50% mixture of methanol  in water and ultrasonically agitated for five
minutes before adding the  suspension to the counter.  (Since the major con-
stituents of the fly ash are aluminum, silicon and iron oxides, little, if


                                     130

-------
any, of the fly ash particles dissolves in this solution; hence, one would
not expect any systematic error due to solvation effects.)  Volume median
diameters of 6.2 urn and 4.3 urn were obtained assuming a particle density of
2.5 g/cm3 in order to convert physical diameters to aerodynamic diameters.
These results indicate the relative validity of the aerodynamic diameters
quoted.  Physical diameters were confirmed by the use of a scanning electron
microscope.

Morphology Studies

     A JSM U-3 SEM (Jeol, Inc., Medford, Massachusetts) equipped with an
ORTEC 7000 Series Si(Li) X-ray detector and an ORTEC Model 6200 Multi-channel
Analyzer (Ortec, Inc., Oak Ridge, Tennessee) were used to study the morphol-
ogy and elemental distribution of the matrix elements, respectively.  The
particles were deposited onto Cambridge type SEMS stubs (Structure Probe,
Inc., Westchester, Pennsylvania) using one of the following supporting media:

 1.  Scotch double-sided adhesive tape (Minnesota Mining and Manufacturing
     Co., St. Paul, Minnesota).

 2.  Dag 154 (Acheson Colloids Co., Port Huron, Michigan), a suspension of
     colloidal graphite in alcohol.

 3.  Dotite Paint Type D-550 (Fujikura Kasei Co., Ltd., Japan), a suspension
     of silver powder in alcohol.

     The particles are then made conducting by depositing on their surface a
layer of carbon about 1,000ft thick under vacuum using a Denton DV-503 vacuum
evaporator (Denton Vacuum, Inc., Cherry Hill, New Jersey).  All X-ray analy-
sis were carried out with the specimen stage tilted at a 30° angle towards
the X-ray detector.  This enabled X-rays to emerge from the sample at a high
take-off angle and thus decreased topographical effects.

Elemental Analyses

     The chemical analyses of the different samples were obtained through a
variety of techniques.  These techniques fall into two classes, viz. , those
that analyzed the fly ash directly as the solid (Instrumental Neutron Acti-
vation Analysis, X-ray Fluorescence, Spark Source Mass Spectrometry and DC
Arc emission spectroscopy), and those that analyzed the sample in solution
following wet digestion (Atomic Absorption).  The former methods retain
sample integrity but involve calibration uncertainties; the latter allow easy
calibration but are susceptable to possible formation of analytically intrac-
table compounds during digestion.

     The matrix elements in the separated fly ash samples (Fe, Ti, Al, Si,
Ca, K, S and Mg) were analyzed with wavelength dispersive X-ray fluorescence
spectrometry.  A vacuum path single crystal Phillips Norelco X-ray spectrom-
eter (Phillips Electronic Instruments, Mount Vernon, New York) was used with
a 1500 watt chromium X-ray tube as the source.  A thallium acid phthalate
(TLAP) wavelength analyzing crystal was used to disperse Si, Al and Mg
X-rays, an ethylene diamine d-tartrate (EDDT) crystal for K, Ca and S X-rays,


                                     131

-------
and a lithium fluoride (LiF) crystal  for Ge and Ti  X-rays.   Thin samples were
used to approximate an infinitely thin film so that matrix effects can be
neglected.   To ensure the thinness of the samples,  all  samples of nominal
particle diameter >4 urn were ground further so as to minimize surface samp-
ling and inhomogeneity effects.   The  powders were suspended in propanol and
dispersed ultrasonically before deposition by filtration onto 0.4 urn mini-
pore membrane filters (Price and Angell, 1968).   National  Bureau of Standards
Certified Fly Ash and Rock Samples were used as standards.   K  radiation was

monitored for all elements and a vacuum radiation path  maintained for all
elements except Fe and Ti.  A lithium fluoride crystal  was employed for
detecting Fe and Ti, EDDT was used for Al, Si, Ca,  K and S and ADP was used
for Mg.  For this method, precisions  of ±5% were achieved,.

     Trace elements in some of the separated fly ash fractions were deter-
mined by instrumental neutron activation analysis.   The samples were irradi-
ated for one hour in the Illinois Reactor Facility which had a neutron flux
of 2 x 1012 neutrons-cm-2-sec-1 in the irradiation region..   The resultant
y-emissions were measured on a multichannel analyzer equipped with a Ge(Li)
detector.  National Bureau of Standards Certified Fly Ash and Rock Samples
were run simultaneously for calibration.  The resulting spectra were recorded
on magnetic tape and analyzed with the "GAMANL" Program (Hopke, 1969).

     An A.E.I, model MS-7 spark source mass spectrometer was used for the
qualitative determination of all elements of atomic number greater than Li
and for quantitative determination of Bi, Pb, TI, Sb, Sn, As, Zn, Cu, Ni, Fe,
V, Ca, K, and Si.  One part of fly ash was mixed by weight with two parts of
spectroscopic graphite for five minutes in a Wig-L-Bug and the mixture
pressed into an electrode.  Electrodes were manually positioned and sparked
with a 25-usec spark duration and a repetition rate of 300 sec-1 at 10-6 torr
source pressure.  Mass spectra were recorded photographically.

     Internal standardization of the mass spectra was achieved by referencing
line intensities both to the Pb in the sample and to 60 ug/gm of solution-
doped Au.  The Pb was determined independently by atomic absorption spectros-
copy.  The 197Au  ion was at least two orders of magnitude more intense than
i8iyai6Q  from source contamination.   Element concentrations were calculated
from the expression (Farrar, 1972)
                          'X
                            = C,
                                 Jx
(A-l)
where  I = peak intensity of ion beam, D = isotope abundance, M = mass of
analyte, C = concentration of analyte, X = analyte, S = internal standard, k
= sensitivity factor for a given element relative to the standard.

     This expression assumes that the line width on the photographic plate is
                 \f.                                              "~is
proportional to M  and emulsion sensitivity is proportional to M   (Owens and
Giardino, 1963).  Values of k were determined by doping increasing amounts of
Pb, TI, Sb, Sn, As, and Ni into the graphite before forming a series of


                                     132

-------
electrodes with fly ash.   For these elements values of k ranged from 1.0 to
1.8.   For the remaining elements k was set equal to unity, an assumption
usually valid within a factor of three (Farrar, 1972).  Precisions of ±20%
were achieved.

     Carbon present as SiC, FeC, and free C was determined as C02 after
combustion with 02 on a V205 catalyst (Steyermark, 1961).

     Sample digestion was achieved by heating 0.5 g of fly ash, 3.5 ml of 3:1
concentrated HC1/HN03 (aqua regia), 0.5 ml water, and 2.5 ml of an aqueous
solution containing 48% HF for two hours at 110°C in a 25 ml teflon-lined
Parr pressure bomb (Parr Instrument Co., Moline, Illinois).  After cooling,
2.5 g of boric acid was added to neutralize the HF.  The small amount of
black, solid residue remaining was removed by centrifugation and was shown by
spark source mass spectrometry to contain mainly Ca, F and Al in addition to
carbon.  At least 95% extraction of the elements of interest was achieved.

     Atomic absorption analyses were performed by direct aspiration of dilu-
tions of the original digest for Pb, Tl, Cd, As, Ni and Be.  Air-acetylene
flames were employed for all elements except Be, for which nitrous oxide-
acetylene was used.  A Jarrel Ash 8-10 dual beam, double monochromator
instrument was employed.   Background corrections were achieved by monitoring
a non-absorbing wavelength within 40A of the analytical wavelength.  Se was
determined by its atomic absorption after conversion to volatile H2Se accord-
ing to the method of Schmidt and Royer (1973).  Standard addition calibra-
tions were performed in all cases, and a precision of ±10% was achieved.

     The elements Pb, Be, Cr, Mn, Co and Ni were determined by DC arc emis-
sion spectroscopy with a Baird-Atomic 3 meter spectrograph.  Samples coarser
than 325 mesh (Tyler series) were ground to pass through a 325-mesh sieve.
One part by weight of fly ash was mixed with four parts of spectroscopic
graphite for one minute in a Wig-L-Bug mechnical shaker (Spex Industries).
Spex mix A-7 pure graphite standards doped with 49 elements were used for
comparative standards.  Approximately 50 mg of graphite diluted sample were
burned to completion in a cup electrode operating with a 4 mm gap and 10 amp
current.  Element concentrations were obtained with a precision of ±30%.

Method Evaluation

     All the techniques have their strengths and limitations in terms of the
number of elements that can be analyzed, interferences, amount of sample
required, sample handling procedures, etc.  A detailed discussion of these
factors is beyond the scope of this report.  However, several comments are
especially relevant to particulate analyses.  Instrumental Neutron activation
analysis (INAA) is probably the best overall technique.  It is a non-
destructive multielement technique with good precision and accuracy.  How-
ever, a reactor, considerable analysis time, and skilled personnel are
required.  In addition, since not all elements can be analyzed by INAA, a
supplemental technique is necessary.

     Of the other techniques utilized, spark source mass spectrometry
undoubtedly affords the greatest advantage for multielemental determination


                                     133

-------
of solid fly ash, although the analyses are extremely time consuming.   In
addition, interferences prevent the analysis of Te,  Cd,  Se, Co,  Mn,  Cr, and S
and permit only a semi quantitative estimate for Be.   D.C.,  arc emission spec-
troscopy exhibits no significant advantages over spark source mass spectros-
copy other than detection of Co, Mn, Cr and Be and wider availability.  The
X-ray fluorescence method employed has the advantages of high speed and
precision but is somewhat limited by calibration difficulties and matrix
effects.  The shallow penetration depths of soft X-rays  from the lighter
elements, such as sulfur and magnesium, necessitate high matrix identity
between samples and standards and a very small particle  size (Muller,  1972).

     Atomic absorption spectrometry is considered the most accurate technique
employed in this work.  However, large amounts of sample are required if more
than a few elements are determined.  Also, great care must be taken to
achieve good background corrections because of the large number of elements
present.  Atomic absorption spectrometry displayed no evidence of loss of
sample integrity as a result of the fly ash digestion. One possible exception
was Tl, for which analyses were consistently 3-5 times lower than those
obtained by spark source mass spectrometry and anodic stripping voltammetry.

X-ray Powder Diffraction

     Diffraction patterns were obtained with a Phillips  Norelco X-ray Dif-
fraction system (Phillips Electronic Instruments, Mt. Vernon, New York).  The
samples were ground to a particle size of less than 44 |jm and a thin film of
particles was placed in a monochromatric beam of copper K  (8.05 keV) X-rays.

The intensity of the diffracted X-rays was measured between 15°-60° 26 angles
with a proportional counter.

Surface Techniques

     The variation of elemental concentrations as a function of depth for
both leached and unleached fly ash particles was determined by ion microprobe
mass spectrometry and Auger electron spectrometry combined with ion etching.
The procedure for particle mounting involved placing large particles (45-180
pm) in a folded strip of indium foil and hand pressing to imbed them into the
foil.  This mounting technique has the advantages that indium is electrically
conducting; presents few spectral  interferences; is soft and malleable,
allowing imbedding of particles without physical alteration; has a low vapor
pressure for ultra-high vacuum compatibility; and is inexpensive (Theriault,
et al., 1975).

     An AEI Model IM-20 ion microprobe was used to obtain elemental depth
profiles.  Basic features of the instrumentation have been described pre-
viously (Bakale, et al., 1975).  A 25-keV negative oxygen primary beam of
40 nA and 20-um diameter was rastered  rapidly over an area of 100 urn by 100
|jm.  The rastering procedure was important primarily because it enabled
fairly uniform current densities to be maintained over the entire area of the
particle being analyzed.  Mass spectrometer resolving powers used ranged from
250 to 1800 (10% valley definition) as required to resolve molecular  ion
interferences.  Depth profiles were acquired by two methods:  (1) contin-
uously monitoring the positive secondary  ion  intensity of one element  vs.


                                     134

-------
time, using electrical  detection, and (2) using photographic detection of the
positive secondary ions as dispersed by the double-focusing Mattuch-Herzog
analyzer.   Results reported herein were obtained only by electrical detec-
tion, since the use of photographic detection was limited by a lack of
sensitivity.

     Elemental depth profiles were also obtained with a Physical  Electronics
Industries Model 545 Scanning Auger Microprobe.  The indium foil  containing
the imbedded particles was mounted on the standard carrousel at 30° grazing
incidence to a 5-keV, 50-(jm primary electron beam.   Elemental depth profiles
were obtained by recording entire Auger spectra after incremental periods of
sputtering with 2-keV Ar .  Sputtering rate calibrations were achieved in an
manner analagous to that used for the ion microprobe depth profiles.

     On the basis of its inherently high sensitivity, good depth  resolution,
and ability to perform microanalysis of samples (including those  which are
insulators), ion microprobe mass spectrometry was a priori one of the most
useful surface analytical techniques for the in-depth characterization of
airborne particles.  Specific analytical concerns or objectives involved
using the ion microprobe to obtain elemental depth profiles of individual fly
ash particles were the following:  (1) to establish which elements of envi-
ronmental  interest could be characterized, (2) to analyze particles before
and after solvent leaching to reduce concerns about possible sputtering
artifacts, and (3) to use solvent leaching and bulk analysis to substantiate
and to quantitate the ion microprobe results, i.e., to estimate surface
constraints on the analysis such as limits to depth resolution or possible
effects due to charging.

     The major potential constraints on the depth resolution of the ion
microprobe elemental depth profiles of fly ash (Figures 32-37) are variations
in the sputtering rate over the area of the particle being analyzed (Townsend
et al., 1976), crater edge effects (Evans, 1972; Williams and Evans,  1976),
escape depths of secondary ions (McHugh, 1975), and cascade mixing of sub-
surface layers resulting from penetration of the primary ion beam into the
sample (Morabito, 1975; Coburn, 1976).

     In order to convert sputtering time to depth, sputtering rates for Si02
were measured under primary ion beam conditions identical to those used for
the analysis of fly ash.  The time required to sputter a 3000-ft Si02  layer
on Si was determined by monitoring the Si intensity that corresponded to a
sputtering rate of 4 A (+0.5 A) per second.  Since fly ash is predominately
a silicate material, the rates were considered comparable.  The major limita-
tion in applying this calibrated rate to fly ash was that the irregular
surface of a fly ash particle may cause substantial variations in sputtering
rate over the region sampled.  That is, microtopographical features approach-
ing grazing incidence with respect to the primary beam (e.g., the edges of a
particle) will sputter faster (up to about a factor of 3) than those  at
normal incidence (Townsend et al., 1976).  Therefore, the calibrated sputter-
ing rate was considered a minimum rate for fly ash, with secondary ions
actually being produced from a range of depths approximately 1-3  times
greater than the calibrated depth.   Thus, the remaining factors mentioned
                                     135

-------
above affecting depth resolution are considered insignificant relative to the
effects of sputtering rate variation and would in any case tend to diminish
and broaden, rather than enhance, surface peaks.

     One of the most important analytical requirements was to establish that
the observed elemental surface predominance was not the consequence of sput-
tering or analysis artifacts.   Conventional methods to assure that variations
in secondary ion intensities reflect relative differences in element concen-
trations are to (1) perform the profile under conditions that ensure a high
oxygen concentration in the sample surface, (2) reference the profiled
element intensity to a uniformly distributed element, and (3) apply quantita-
tion procedures to the data (Morabito, 1975).   All  of these methods have been
utilized and will be discussed as follows.

     First, a high oxygen concentration in the sample surface was ensured by
using an oxygen primary beam.   Specifically, the purpose of using a negative
oxygen primary beam was twofold:   (1) oxygen promotes the production of
stable, high-intensity yields of positive secondary ions (Lovering, 1975) and
(2) a negative primary beam has been shown to maintain charge neutrality on
insulating surfaces (Andersen et al., 1969).  The absence of large variations
in the secondary ion intensities of matrix elements (Al, Ti in Figure 32)
beyond the intial rise in the first 100 A did indicate that secondary ion
yields are relatively constant over the greater part of the depth profile.

     All secondary ion intensities (Figures 32-37)  were referenced to Si in
order to permit a direct comparison of depth profiles for different elements
and to compare the same element in the leached vs.  the unlcached sample.  Use
of Si as the internal standard allowed compensation for variations in the
amount of sample within the rastered area and for variable ion extraction
efficiency.  Silicon was considered the most appropriate element, since (1)
it was homogenously distributed in all particles at depths greater than
several hundred angstroms, (2) the Si profiles showed minor variations for
different particles in the same sample, and (3) Si  was negligibly leached by
either DMSO or H20.  The use of Si as an internal standard was also necessary
in the quantitation procedures to be discussed.

     With regard to methods for quantitation of surface region concentra-
tions, it is well known that secondary ion yields are highly variable from
element to element and exhibit strong matrix effects.  For fly ash, the
surface elemental composition apparently is very different from the interior,
and standards suitable for surface analysis are not available.  For the
estimates of elemental concentrations shown in Figures 32-37, the essentially
constant secondary ion intensities observed for all elements at depths
greater than about 500 to 1000 ft were taken to approximate bulk concentration
levels.  As summarized below, this assumption was verified with calculated
relative elemental sensitivity factors, the bulk concentration of Si in fly
ash as an internal standard, and the results of bulk analysis with SSMS.

     Our quantitation procedure was based upon the empirical observation that
a simplified form of the Saha-Eggert equation can be used to reduce secondary
ion yields to elemental concentrations (Simons et al., 1976):


                                     136

-------
                                     _  .                              (A-2)
                         N9     Ne


where N- and N
-------
     An alternative approach to the quantitation of surface region concentra-
tions was to ratio the mass of a given element leached from the particles to
the total mass removed from the particles during leaching.   This procedure
was deemed unsatisfactory for the following reasons:   (1) it integrates the
elemental concentration over the entire surface region that is leached, and
(2) it is not known how uniformly,  or to precisely what depth, the particles
are leached.

     Auger Electron Spectrometry--The application of Auger electron spec-
trometry (AES) for in-depth characterization was of interest,  since Auger
depth resolutions are roughly comparable to those obtainable with the ion
microprobe, and microfocusing the primary electron beam also permits analysis
of individual particles.   Unfortunately, the application of AES was severely
limited by electron beam charging effects (Chang, 1974) and the inability to
detect elements at concentrations below about 0.1-1% atomic in the analytical
volume.

     The charging problem was the result of the insulating nature of fly ash.
Charging effects are normally minimized by using grazing or glancing inci-
dence to increase secondary electron emission coefficients.  However, for
irregular surfaces such as those of fly ash particles, grazing incidence
cannot be readily maintained over the entire area being analyzed.  The resul-
tant charging caused occasional loss of particles during analysis, instabil-
ities and/or shifts in the positions of Auger peaks (especially those having
energies below a few hundred eV), and shifts in the position of the electron
beam.

     The solvent-leaching experiments were of minor value in substantiating
the Auger results for two reasons:   (1) the highly Teachable,  surface-
predominant trace elements could not be detected and (2) the leached parti-
cles were more susceptible to charging.  No pronounced changes in the Auger
depth profiles were observed following leaching except for Na and K.  The
stronger surface enhancement shown by Na and K following leaching may have
resulted from charge-induced migration of alkalies to the particle surface
(McCaughan and Kushner, 1974).  Leached particles were generally observed to
be more susceptible to charging because of the apparent removal of a more
conducting surface layer by solvent leaching.

     Despite the limitations discussed, scanning Auger microscopy does have
several advantages over the ion microprobe for particle characterization:
(1) The initial surface spectrum is obtained without sputtering and thus is
not subject to the effects of sputtering artifacts.  (2) Multielement depth
profiles of the same particle are more readily obtained.  (The IM20 ion
microprobe does not have rapid magnetic peak switching or multiple electrical
detection capabilities.)  (3) Spectral interferences may be less severe for a
few elements.  For example, sulfur was routinely determined by Auger, but the
ion microprobe mass spectrometric analysis required high mass resolution.

     Other Techniques (Scanning Electron Microscopy, Electron Spectroscopy)
for Chemical Analysis—The major disadvantage of scanning electron microscope
or electron microprobe methods for in-depth characterization is the lack of
depth .resolution; i.e., x-ray escape depths are on the order of a micron


                                      138

-------
relative to the 10-50 ft analytical escape depths for the ion microprobe or
Auger techniques.

     Qualitative differences between surface and interior composition can be
discerned, however, by comparing X-ray spectra before and after the removal
of material in the surface region by ion sputtering (Linton et al., 1976;
Loh, 1975; Hulett, 1973).   X-ray intensities can also be measured from the
sputtered and unsputtered sides of the sami* particle (Linton et al. ,  1976;
Loh, 1975).  Experimental  results using Ar  sputtering of fly ash were pre-
viously reported using scanning electron microscopy-energy dispersive x-ray
spectrometry (SEM/EDS) (Linton et al., 1976; Loh, 1975).  Results indicated
enhanced surface region concentrations of Pb, Zn, K, Na, P, and S,  with no
detectable surface predominance of Ca, Si, Al, and Fe.   Thus, the SEM/EDS
technique is significant for providing a qualitative corroboration of the
results obtained with the true surface microanalytical  techniques for the
mutually detectable elements.

     Other possible approaches to the surface characterization of particles
using the electron microprobe include the following: (1) varying the primary
electron beam energy to produce a variation in the analysis depth (Armstrong
and Buseck, 1975b) or (2) ratioing X-ray intensities of elements to matrix
elements as a function of particle size (Linton et al. , 1976; Peuschel et
al., 1974).  Both methods suffer from the poor depth resolutions obtainable
with electron probe techniques.  Using the first technique above also
requires a very stable electron beam current and extensive X-ray quantitation
methods, including corrections for particle shape (Armstrong and Buseck,
1975a; Armstrong and Buseck, 1975b).  Elemental detection limits also
increase at lower electron beam energies, thus limiting the extent to which
beam energies may be reduced.   The second approach assumes that the smaller
the particle, the greater the relative surface area being sampled by the
electron beam (Peuschel et al., 1974).  However, this approach oversimplifies
the mechanisms of X-ray production and really requires detailed corrections
for the effects of particle shape on the emission of X-rays.

     Electron spectroscopy for chemical analysis (ESCA) is another technique
used for surface characterization of environmental particles (Craig et al. ,
1974; Gordon, 1975; Cambell et al., 1979).  ESCA not only has depth resolu-
tion and detection limits generally comparable to those of AES, but also has
advantages over AES for surface characterization in that (1) chemical infor-
mation is more available, since ESCA chemical shifts are usually more readily
related to oxidation states (Evans, 1975), and (2) charging of insulating
particles is minimized, since X-rays, rather than electrons, serve as the
source of ionization (Linton et al., 1976).  However, ESCA is not favored
over AES for in-depth characterization, because (1) chemical information is
potentially lost, since destructive ion sputtering may alter the forms of the
elements present (Evans, 1975), and (2) individual particles cannot be depth
profiled, since ESCA microanalysis is not yet feasible Linton et al., 1976).
Thus, heterogeneity and variations in sputtering rate within the field of
particles being analyzed are major difficulties in obtaining meaningful ESCA
depth profiles.
                                     139

-------
Leachate Analysis

     Leaching studies were undertaken in two experiments.   In the comparison
tests of the different leaching methods, four techniques  were used.   These
are described in the text in Section V.

     The techniques used in the surface  study leaching experiments were
performed with Fisher reagent grade dimethyl sulfoxide (DMSO) and triply
distilled water.  DMSO was chosen after  inorganic species were found in DMOS
extracts during studies investigating the extraction of trace organics, while
H20 was employed to simulate leaching in the natural environment.   Leaching
with DMSO was carried out for 48 hrs in  a Soxhlet extractor at a temperature
of 40°C and a pressure of -vO.l mm Hg.  A DMSO blank was prepared by leaching
an empty Soxhlet thimble in identical  experimental  conditions.   Leaching with
H20 was performed by placing fly ash in  a glass jar containing triply
distilled water for a 12-hr period.   The jar was also placed in a DISONte-
grator Model 320 and sonicated for 1 hr  in a water bath in order to maximize
interaction of the solvent with fly ash  surfaces.   The leachate solution was
then filtered twice through Whatman No.  41 filter paper,  as was an H20 solu-
tion serving as the blank.  The aqueous  leaching procedure was used primarily
to indicate relative elemental Teachabilities under the conditions employed,
and thus it may not solubilize the entire avilable fraction of most elements
(Natusch and Matusiewicz, 1976).
                                     140

-------
                   APPENDIX B

MATRIX ELEMENT DISTRIBUTION FOR A SIZE, MAGNETIC
        AND DENSITY FRACTIONATED FLY ASH
                       141

-------
 TABLE B-l.   CONCENTRATION OF SILICON (WT %) IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (|jm]
<20
20-44
44-74
>74
<20
20-44
44-74
>74
Density
<2.1 2.1-2.5
26.43 23.97
22.93 23.17
22.70 23.70
23.77 24.14
__
__
22.82 21.93
22.28
(g/cm3)
2.5-2.9
18.90
19.73
19.02
22.44
9.99
21.01
17.78
21.69

>2.9
	 *
18.37
14.65
15.73
10.51
6.49
12.25
11.56
No meaningful data.
                                   142

-------
 TABLE B-2.   CONCENTRATION OF ALUMINUM (WT %) IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (pm"
<20
20-44
44-74
>74
<20
20-44
44-74
>74
Density
<2.1 2.1-2.5
13.37 15.15
18.03 13.12
15.47 13.04
14.35 14.86
__
--
13.65 11.54
12.20
(g/cm3)
2.5-2.9
10.38
10.66
10.88
11.04
6.04
10.11
9.95
12.26

>2.9
__*
11.06
8.78
8.14
5.98
3.94
7.48
6.19
No meaningful  data.
                                  143

-------
   TABLE B-3.   CONCENTRATION OF IRON (WT %) IN FRACTIONATED FLY ASH
Particle size (MHI]
Non-magnetic- <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74
Density
1
<2.1 2.1-2.5
5.34 3.32
5.56 8.01
8.08 8.77
8.12 6.68
__
--
11.44 14.46
13.67
(g/cm3)
2.5-2.9
18.40
9.74
11.43
12.00
41.75
14.80
20.51
13.46

>2.9
__*
14.38
18.25
19.72
41.36
41.31
35.22
38.61
No meaningful data.
                                  144

-------
 TABLE B-4.   CONCENTRATION OF POTASSIUM (WT %) IN FRACTIONATED FLY ASH
Particle size (pm)
Non-magnetic <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74

<2.1
2.69
2.63
1.89
1.79
__*
--
1.78
1.37
Density
2.1-2.5
2.34
2.28
1.63
1.48
--
1.92
1.60
1.62
(g/cm3)
2.5-2.9
2.22
1.33
1.05
1.06
0.76
1.48
1.27
1.49

>2.9
1.73
1.09
0.45
0.13
0.70
0.73
0.85
0.83
No meaningful data.
                                 145

-------
    TABLE B-5.   CONCENTRATION  OF  SODIUM  (pg/g)  IN  FRACTIONATED  FLY ASH
Particle size (pm)
Non-magnetic <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74
1
<2.1
3621
2494
1984
2249
__*
--
1869
1761
Density
2.1-2.5
2903
3011
2447
2940
--
2554
1952
—
(g/cm3)
2.5-2.9
2500
1709
1443
1574
1059
1860
1618
—

>2.9
3492
1473
738
1275
881
883
--
--
* No meaningful data.
                                    146

-------
  TABLE B-6.   CONCENTRATION OF CALCIUM (WT %) IN FRACTIONATED FLY ASH
Particle size ([im]
Non-magnetic <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74
V
)
10.0
6.0
3.2
3.3
__*
--
11.0
6.0
Density
2.1-2.5
5.0
5.3
4.0
3.4
--
10.0
11.0
7.0
(g/cm3)
2.5-2.9
6.0
4.6
4.1
4.4
6.0
9.0
12.0
14.0

>2.9
3.0
7.0
8.3
5.2
5.0
6.0
10.0
7.0
No meaningful data.
                                  147

-------
  TABLE B-7.   CONCENTRATION OF SULFUR (WT %)  IN FRACTIONATED FLY ASH
Particle size (pm]
Non-magnetic <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74
)
<2.1
0.24
0.11
0.21
0.31
— *
—
0.10
--
Density
2.1-2.5
0.40
0.48
0.37
0.12
--
—
0.21
0.43
(g/cm3)
2.5-2.9
0.22
0.82
1.26
0.48
0.16
0.45
0.34
0.20

>2.9
__*
0.43
1.02
0.71
0.19
0.09
0.28
0.14
No meaningful data.
                                  148

-------
 TABLE B-8.   CONCENTRATION OF TITANIUM (WT %) IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (pm]
<20
20-44
44-74
>74
<20
20-44
44-74
>74
Density
<2.1 2.1-2.5
0.80 0.50
0.97 1.08
1.19 3.92
0.94 0.65
__*
__
0.94 1.10
0.79
(g/cm3)
2.5-2.9
2.39
1.10
0.88
0.77
0.70
1.12
0.94
0.69

>2.9
__*
1.34
1.13
0.73
0.72
0.55
0.68
0.55
No meaningful data.
                                  149

-------
                  APPENDIX C

TRACE ELEMENT DISTRIBUTIONS FOR A SIZE DENSITY
  AND MAGNETICALLY FRACTIONATED COAL FLY ASH
                      150

-------
   TABLE C-l.   CONCENTRATION OF MAGNESIUM (WT %) IN FRACTIONATED FLY ASH
Particle size (jjm>
Non-magnetic <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74
Density
<2.1 2.1-2.5
0.77 0.97
0.80 0.83
0.87 1.22
1.10 0.90
__
--
0.93 0.92
0.87
(g/cm3)
2.5-2.9
0.71
0.79
1.21
1.19
0.48
0.86
0.84
0.84

>2.9
__*
1.21
0.99
0.80
0.49
0.28
0.62
0.36
* No meaningful data.
                                    151

-------
 TABLE C-2.   CONCENTRATION OF MANGANESE (pg/g) IN FRACTIONATED FLY ASH
Particle size (|jm)
Non-magnetic <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74

<2.1
608
313
280
288
--
--
401
477
Density
2.1-2.5
643
729
546
615
--
904
493
—
(g/cm3)
2.5-2.9
648
1083
1101
916
722
899
749
--

>2.9
504
1570
2399
1596
611
543
--
—
No meaningful data.
                                  152

-------
  TABLE C-3.  CONCENTRATION OF ARSENIC (jjg/g)  IN  FRACTIONATED  FLY ASH
Particle size (pm)
Non-magnetic <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74

<2.1
46
12
19
30
--
__
—
41
Density
2.1-2.5
90
77
42
49
—
79
59
54
(g/cra3)
2.5-2.9
86
133
129
93
70
111
97
80

>2.9
74
164
126
156
63
40
48
62
No meaningful data.
                                  153

-------
  TABLE C-4.   CONCENTRATION OF BARIUM (pg/g)  IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (pm)
<20
20-44
44-74
>74
<20
20-44
44-74
>74

<2.1
216
309
409
448
--
—
535
574
Density
2.1-2.5
209
390
364
492
--
373
366
161
(g/cm3)
2.5-2.9
287
714
559
500
300
525
609
249

>2.9
522
3143
7432
1070
240
269
168
91
No meaningful data.
                                  154

-------
 TABLE C-5.   CONCENTRATION OF STRONTIUM (pg/g) IN FRACTIONATED FLY ASH

Non-magnetic


Magnetic


Particle size (nm)
<20
20-44
44-74
>74
<20
20-44
44-74
>74

<2.1
<75
120
210
248
__
198
251
Density
2.1-2.1,
192
261
139
237
129
164
180
(g/cm3)
2.5-2.9
180
377
329
148
118
182
101
44

>2.9
214
363
274
222
91
85
75
136
No meaningful data.
                                  155

-------
    TABLE C-6.   CONCENTRATION OF CESIUM (jjg/g) IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (pm}
<20
20-44
44-74
>74
<20
20-44
44-74
>74
Density
<2.1 2.1-2.5
__*
0.31
0.33 0.27
0.29 0.20
—
5.70
5.60 4.40
__
(g/cm3)
2.5-2.9 >2.9
—
0.14
0.23 0.07
0.19 0.14
2.00
3.50 1.50
3.50
__
* No meaningful data.
                                    156

-------
 TABLE C-7.   CONCENTRATION OF RUBIDIUM (pg/g) IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size ([im]
<20
20-44
44-74
>74
<20
20-44
44-74
>74
Density
<2.1 2.1-2.5
__*
92
87 102
83 <146
—
<137
<100 <55
__
(g/cm3)
2.5-2.9
--
--
69
78
<55
<76
<155
--

>2.9
--
40
32
18
<49
<44
--
—
No meaningful data.
                                 157

-------
     TABLE C-8.   CONCENTRATION OF  ZINC  (M9/g)  IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (ym]
<20
20-44
44-74
>74
<20
20-44
44-74
>74
Density
1
<2.1 2.1-2.5
__*
48
42 39
37 40
--
460
110 350
—
(g/cm3)
2.5-2.9
—
--
64
93
1100
480
480
—

>2.9
--
69
203
327
1100
500
--
--
* No meaningful  data.
                                    158

-------
   TABLE C-9.   CONCENTRATION OF CHROMIUM (ng/g) IN FRACTIONATED FLY ASH

Non-magnetic


Magnetic



Particle size (urn
<20
20-44
44-74
<20
20-44
44-74
>74
Density (g/cm3)
<2.1 2.1-2.5 2.5-2.9 >2.9
__*
10.7 — 10.0
6.0 6.1 6.2 8.9
312 300
214 200 214
211 189 216
__
* No meaningful data.
                                    159

-------
  TABLE C-10.   CONCENTRATION OF COBALT (ng/g)  IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (pm]
<20
20-44
44-74
>74
<20
20-44
44-74
>74
Density
1
<2.1 2.1-2.5
__*
10.2
7.0 8.7
7.8 8.6
__
41
48 45
__
(g/cm3)
2.5-2.9
--
--
11.0
15.7
119
52
57
—

>2.9
--
13.5
12.3
14.8
113
84
--
--
No meaningful data.
                                  160

-------
 TABLE Oil.   CONCENTRATION OF EUROPIUM (pg/g) IN FRACTIONATED FLY ASH
Particle size (pm)
Non-magnetic <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74

<2.1
1.9
1.4
1.0
1.0
—
1.4
1.3
Density
2.1-2.5
2.3
1.7
1.0
1.2
2.2
1.5
1.0
(g/cm3)
2.5-2.9
2.1
1.4
1.4
0.9
1.4
2.0
1.6
0.8

>2.9
1.7
1.9
3.3
2.0
1,2
1.5
0.7
0.6
No meaningful data.
                                  161

-------
   TABLE C-12.   CONCENTRATION OF SCANDIUM (pg/g)  IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (pm)
<20
20-44
44-74
>74
<20
20-44
44-74
>74
Density
1
<2.1 2.1-2.5
__*
11.0
7.8 8.5
6.8 5.8
—
49
47 41
__
(g/cm3)
2.5-2.9
--
--
9.3
10.1
40
54
47
--

>2.9
—
10.4
10.7
9.5
38
38
--
--
* No meaningful data.
                                    162

-------
 TABLE C-13.   CONCENTRATION OF THORIUM (pg/g) IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (urn)
<20
20-44
44-74
>74
<20
20-44
44-74
>74
Density
<2.1 2.1-2.5
__*
5
16 16
14 9
__
23
22 18
__
(g/cm3)
2.5-2.9 >2.9
__
5
13 12
17 12
17 15
23 15
__
—
No meaningful  data.
                                  163

-------
 TABLE C-14.   CONCENTRATION OF URANIUM (|jg/g) IN FRACTIONATED FLY ASH
Particle size (pm)
Non-magnetic <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74

<2.1
15
13
9
6
__*
—
6
12
Density
2.1-2.5
15
16
9
13
—
5
6
9
(g/cn.3)
2.5-2.9
15
17
16
15
7
7
7
10

>2.9
12
28
15
17
6
6
8
9
No meaningful data.
                                  164

-------
TABLE C-15.   CONCENTRATION OF DSYPROSIUM (ng/g) IN FRACTIONATED FLY ASH
Particle size (pm)
Non-magnetic <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74

<2.1
9
7
9
8
__*
--
14
5
Density
2.1-2.5
10
37
9
9
--
15
13
7
(g/cm3)
2.5-2.9
9
17
12
12
12
17
13
7

>2.9
7
14
11
10
10
10
7
6
No meaningful data.
                                  165

-------
 TABLE C-16.   CONCENTRATION OF HAFNIUM (pg/g)  IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size ((jm.
<20
20-44
44-74
>74
<20
20-44
44-74
>74
Density
)
<2.1 2.1-2.5
	 *
5.0
4.5 4.1
4.6 2.0
__*
8.5
6.2 6.4
__
(g/cm3)
2.5-2.9
—
6.0
3.8
4.9
4.0
6.7
12.9
--

>2.9
--
9.9
11.1
3.2
3.9
4.2
--
--
No meaningful data.
                                  166

-------
 TABLE C-17.   CONCENTRATION OF TANTALUM (jjg/g) IN FRACTIONATED  FLY  ASH

Non-magnetic


Magnetic


Particle size (urn)
<20
20-44
44-74
>74
<20
20-44
44-74
>74
Density
<2.1 2.1-2.5
__*
0.7
0.5 0.6
0.7 0.4
__*
1.0
1.4 0.9
__
(g/cm3)
2.5-2.9 >2.9
0.8
0.7 0.7
0.8 0.6
0.5
0.8 0.7
1.0
__
No meaningful data.
                                  167

-------
 TABLE C-18.   CONCENTRATION OF SAMARIUM (pg/g)  IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (pm)
<20
20-44
44-74
>74
<20
20-44
44-74
>74

<2.1
9.1
8.5
6.4
6.4
__*
__
8.6
6.4
Density
2.1-2.5
12.1
10.4
6.5
7.0
—
9.8
7.9
--
(g/cm3)
2.5-2.9
13.0
9.8
8.0
8.3
7.5
10.0
—
--

>2.9
8.7
13.4
9.1
8.9
6.3
6.7
--
--
No meaningful data.
                                  168

-------
TABLE C-19.   CONCENTRATION OF LANTHANUM (pg/g) IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (pm)
<20
20-44
44-74
>74
<20
20-44
44-74
>74

<2.1
51
51
41
39
__*
--
43
39
Density
2.1-2.5
53
53
36
35
--
41
39
25
(g/cn.3)
2.5-2.9
51
42
33
34
25
40
36
27

>2.9
41
55
34
34
22
24
21
20
No meaningful data.
                                  169

-------
 TABLE C-20.   CONCENTRATION OF  GALLIUM (Mg/g)  IN FRACTIONATED FLY ASH

Non-magnetic



Magnetic



Particle size (pm)
<20
20-44
44-74
>74
<20
20-44
44-74
>74

<2.1
90
30
8
11
__*
—
14
23
Density
2.1-2.5
93
22
13
17
...
29
23
24
(g/cm3)
2.5-2.9
70
22
21
29
58
34
31
48

>2.9
55
24
20
28
50
30
46
11
No meaningful data.
                                  170

-------
   TABLE C-21.   CONCENTRATION OF ANTIMONY (pg/g) IN FRACTIONATED FLY ASH
Particle size (urn)
Non-magnetic <20
20-44
44-74
>74
Magnetic <20
20-44
44-74
>74

<2.1
24
7
4
5
__*
--
11
13
Density
2.1-2.5
30
11
7
8
--
29
22
--
(g/cm3)
2.5-2.9
28
20
17
15
38
38
33
--

>2.9
19
23
14
16
33
21
--
--
* No meaningful data.
                                    171

-------
TABLE C-22.   DEPENDENCE  OF  ELEMENTAL CONCENTRATIONS ON PARHCLE SIZE
                      FOR SIZE-FRACTIONATED COAL FLY ASH
Particle
diameter

Pb
Ti
Sb Cd
Se
As
Ni
Cr
(|jg/g)


A.
Zn S
(wt %)
fraction
Precipitator fly ash
Sieved fractions
>74
44-74
140
160
Aerodynamical ly
>40
30-40
20-30
15-20
10-15
5-10
<5
90
300
430
520
430
820
980
7
9
sized
5
5
9
12
15
20
45
1.5 <10
7 <10
fractions
8 <10
9 <10
6 <10
19 <10
12 <10
25 <10
31 <10
<12
<20

<15
<15
<15
<30
<30
<50
<50
180
500

120
160
200
300
400
800
370
B. Airborne
>11.3
7.3-11.3
4.7-7.3
3.3-4.7
2.1-3.3
1.1-2.1
n cc_i 1
1100
1200
1500
1550
1500
1600
29
40
62
67
65
76
17 13
27 15
34 18
34 22
37 26
53 35
13
11
16
16
19
59
680
800
1000
900
1200
1700
100
140

300
130
160
200
210
230
260
fly ash
460
400
440
540
900
1600
100
90

70
140
150
170
170
160
130

740
290
460
470
1500
3300
500
411 1.3

730 <0.01
570 0.01
480
720
770 4.4
1100 7.8
1400

8100 8.3
9000
6600 7.9
3800
15000 25.0
13000
d« 8
66.30
22.89

2.50
3.54
3.25
0.80
0.31
0.33
0.08







                                172

-------
TABLE C-23.   DEPENDENCE OF ELEMENTAL CONCENTRATIONS ON PARTICLE SIZE
                 FOR SIZE-FRACTIONATED COAL FLY ASH
Particle
diameter Fe Mn V Si
(Mm) (wt %) (|jg/g) (pg/g) (wt %)
Sieved fractions
>74
44-74 18
Aerodynamically sized
>40 50
30-40 18
20-30
15-20
10-15 6.6
5-10 8.6
<5

>11.3 13
7.3-11.3
4.7-7.3 12
3.3-4.7
2.05-3.3 17
1.06-2.06
0.65-1.06 15
A.
700
600
Mg
(wt %)
C
(wt %)
Be
(pg/g)
Al
(wt %)
Precipitator fly ash
150
260
--
18
--
0.39
--
11
12
12
--
9.4
fractions
150
630
270
210
160
210
180

150
210
230
200
240
470
—
250
190
340
320
320
330
320
B. Ai
150
240
420
230
310
480
—
3.0
14
--
--
19
26
--
rborne fly
34
--
27
--
35
—
23
0.02
0.31
--
--
0.16
0.39
--
ash
0.89
--
0.95
—
1.4
—
0.19
0.12
0.21
0.63
2.5
6.6
5.5
--

0.66
0.70
0.62
0.57
0.81
0.61
—
7.5
18
21
22
22
24
24

34
40
32
55
43
60
—
1.3
6.9
--
--
9.8
13
--

19.7
--
16.2
--
21.0
--
9.8
                                 173

-------
TABLE C-24.   DEPENDENCE OF ELEMENTAL CONCENTRATIONS ON PARTICLE SIZE
                 FOR SIZE-FRACTIONATED COAL FLY ASH
Particle
diameter
(fjm)

Sieved fractions
>74
44-74
Aerodynamically
>40
30-40
20-30
15-20
10-15
5-10
<5

>11.3
7.3-11.3
4.7-7.3
3.3-4.7
2.06-3.3
1.06-2.06
0.65-1.06
Bi
Sn
Cu
Co
(pg/g)
A.

<2
<2
Precipitator

<2
<2

120
260
fly ash

28
27
Ti Ca K
(wt %)


__
0.61 5.4 1.2
sized fractions
<2
<2
<2
<2
<2
<2
<2

<1.7
<3.5
<4.0
<4.8
<4.5
<4.4
--
<2
<2
<2
<2
<2
<2
<2
B.
7
11
18
19
16
18
--
220
120
160
220
220
390
490
75
76
55
50
55
46
54
0.01 2.5 2.54
0.64 6.3 6.26
--
4.5 4.46
0.66 4.0 4.04
1.09
__
Airborne fly ash
270
390
380
—
330
300
--
60
85
90
95
90
130
__
1.12 4.9 4.9
__
0.92 4.2 4.2
—
1.59 5.0 5.0
—
1.08 2.6 2.6
                                 174

-------
                      APPENDIX D



RESULTS OF DIFFERENT COAL FLY ASH LEACHING EXPERIMENTS
                          175

-------
    I   <\J rn
i    —  O
U-  C_>  O
s?
                                    176

-------
CJ O

2: LJ

O
                                                                                                                         W)  -i—   -Q
                                                                                                                             e   ™
                                                                                                                         e   5
                                                                                                                         3       u
                                                                                                                        •,-   
-------
i/J  z
              178

-------
                                                                                 rH   ifi  r-t

ce >-
t/l U_
CD

CO

o


rH
CO
0

en
CO
o





o
CM





rH
Cn







in
(Ti







CO

o

en
CO
o



^-
o

^.
m
o




m
00
o




i —
CO
o







m
r-t






-a


o o csj
V
rH P-"
o *t o
o o m
V

rH O
o **• co
O O rH
V

rH CO
o CM r-*
O O CSJ
V



rH CO
O O *JD
O O ro
V



rH f"-
O O CSJ
CD O m
V V






0 O
o o ^j-
V V rH





Q* CT
CO 3EI CL.

CM O
V
rH

rH O
V

rH
CM O
«fr 0
V

rH
O
CO O
rH V



rH
O
O O
CM



CM
CM
CM O







rH
CO *f
rH





•i— O
«-O E

O

in
CO
o


^
CD
O

rH
rH
O




CO
o
0




ro
O
o







CO
o





-H-

z

o
V
rH
O
o
V

rH
O
O
V

rH
O
O
V



rH
CD
CD
V



rH
O
O
V





rH
CD
0
V






z.

CM


Uft
CM



o
LT)


*»•
iT>





CO
VO






t-M








rv






 O
V rH

rH m
CD CO CO ^I
O CO O CSJ CM O
v r- CM

rH CM
O U"> Lf>
O rH O {D ^D CD
V rH r* CM
r~t


rH CM
o •«• r+*
o in o r^ r*^ o
V CM U3 CO
CM


rH ro
O P-
CD CM O ro «* O
V if) ^O rH in





rH
CD m o
CD 00 CD in iT> C*l
V rH O P-
CO
rH



4— -t— 4—

O 

O O CD
V
Cn cn rH
O ro O
o o o


CO 
T3
0)
U
!_
x
QJ

C
at
E
, 	
0)
4_
O

O)

m

at

Ul
QJ
t-
CL
X
01

c
o
•^
4-1
c
cu
u
c
o
CJ
*

















a>
r—
a.
m


c
0
u
QJ
l-
0
o


_c
en

r—
>^

i~
e
3

U


CJ
+-
                                                         179

-------
        TABLE  0-5.   ANION  CONCENTRATIONS* RESULTING FROM COLUMN LEACHING AT 3 ML/MIN OF
               DIFFERENT SIZE  FRACTIONS OF CORETTE FLY ASH WITH A pH 7 SOLUTION
Time (min)
150-180 pin
Anions
F"
Cl"
sof

F"
Cl"
c°r
sof

F~
Cl"
cot
s°f
0
3.2
5.6
930

1.8
1.9
300
560

1.8
0.85
280
540
15
0.95
0.50
44.

0.55
0.15
220
21

0.31
0.15
200
28.
60
0.60
0.36
18.

0.17
0.10
175
4.8

0.10
0.09
155
10.
120
0.
180
25 0.11
0.20 0.07
18.

0.
0.
<100
2.

0.
0.
<100
6.
14.
45-74 Mm
15 0.11
09 0.05
<100
8 1.9
20-30 Mm
05 <0.01
02 <0.01
<100
5 5.1
360
0.09
0.05
13.

0.06
<0.01
<100
1.0

<0.01
<0.01
<100
4.9
540
0.07
<0.01
12.

0.05
<0.01
<100
0.80

<0.01
<0.01
<100
3.8
720
0.05
<0.01
9.8

<0.01
<0.01
<100
0.60

<0.01
<0.01
<100
1.9
1440
<0.01
<0.01
7.9

<0.01
<0.01
<100
0.50

<0.01
<0.01
<100
1.1
* Concentrations expressed as  ^9  of  anion  extracted per gram of fly ash.
                                          180

-------
     TABLE 0-6.   ELEMENTAL CONCENTRATIONS* RESULTING FROM COLUMN LEACHING  AT  3
                  ML/MIN OF THE 20-30 \M SIZE FRACTION OF CORRETTE
                            FLY ASH WITH A pH 7 SOLUTION
Time (min)
Element
B
Cd
Be
Mg
P
Si
Mot
Mnt
Nit
Na
Cu
Alt
Ca
Bat
K
Crt
Srt
Pbt
0
30.
<0.03
<0.01
0.57
1.9
8.0
1.3
<0.01
<0.01
65.
<0.05
37.
1400
14
140
0.40
42.
1.3
15
20.
0.44
<0.01
0.61
<0.10
6.8
0.43
<0.01
<0.01
2.0
<0.01
29.
480
12
51
0.20
25.
0.35
60
6.

2
0.60
<0.
0.
3.
6.
0.
01
77
77
7
27
<0.01
<0.
1.
<0.
19.
220
6.
6.
0.
10.
0.
01
9
01


2
4
15

25
120
1.
0.
<0.
0.
<0.
3.
0.

2
74
01
70
10
5
08
<0.01
<0.
<0.
<0.
13.
120
2.
3.
<0.
4.
<0.
01
05
01


3
4
01
0
01
180
0.81
0.79
<0.01
0.62
2.0
2.0
0.01
<0.01
<0.01
<0.05
<0.01
10.
77.
1.2
1.6
<0.01
2.0
<0.01
360
0.23
0.91
<0.01
0.51
1.0
1.1
<0.01
<0.01
<0.01
<0.05
<0.01
7.2
30.
1.1
0.78
<0.01
1.1
<0.01
540
0.09
0.99
<0.01
0.47
<0.10
1.0
<0.01
<0.01
<0.01
<0.05
0.01
6.8
12.
0.83
0.28
<0.01
0.48
<0.01
720
0.02
0.97
<0.01
0.21
<0.10
1.0
<0.01
<0.01
<0.01
<0.05
<0.01
5.8
10.
0.64
0.09
<0.01
0.27
<0.01
1440
<0.02
0.32
<0.01
0.31
<0.10
1.0
<0.01
<0.01
<0.01
<0.05
<0.01
4.1
8.1
0.21
0.13
<0.01
0.18
<0.01
* Concentrations expressed as (jg of element extracted per gram of fly ash.

t Calcium stray light corrections applied.
                                      181

-------
      TABLE D-7.   ELEMENTAL  CONCENTRATIONS*  RESULTING  FROM  COLUMN  LEACHING AT  3
                   ML/MIN  OF THE  45-74 \tm  SIZE  FRACTION OF  CORRETTE
                            FLY ASH  WITH A pH 7 SOLUTION
Time (min)
Element
B
Cd
Be
Mg
P
Si
Mot
Mnf
Nit
Na
Cu
Alt
Ca
Bat
K
Crt
Srt
Pbt
0
19.
<0.
<0.
<0.
1.
12.
1.
<0.
<0.
23.
<0.
9.
2400
9.
280
0.
51.
1.


03
01
07
1

1
01
01

01
8

7

26

4
15
18.
<0.
<0.
<0.
<0.
7.
0.
0.
<0.
7.
<0.
4.
1100
18.
120
0.
24.
0.


03
01
07
10
0
59
37
01
1
01
7



20

49
60
4.1
0.13
<0.01
<0.07
0.21
2.3
0.28
1.4
<0.01
1.6
<0.01
4.2
350
6.5
18.
0.14
6.9
0.31
120
1.1
0.27
<0.01
<0.07
5.3
2.2
0.24
<0.01
<0.01
1.5
<0.01
4.1
190
2.8
7.3
0.12
3.1
0.21
180
0.68
0.32
<0.01
<0.07
3.0
2.0
0.11
<0.01
<0.01
1.0
<0.01
4.0
110
1.2
4.8
0.07
?.2
0.02
360
0.28
0.48
<0.01
<0.07
<0.10
1.9
0.08
<0.01
<0.01
0.92
<0.01
4.0
39.
0.88
0.57
0.03
2.0
<0.01
540
0.03
0.52
<0.01
<0.07
2.0
1.2
<0.01
<0.01
<0.01
0.63
<0.01
2.2
21.
0.24
0.28
<0.01
1.5
<0.01
720
<0.02
0.50
<0.01
<0.07
<0.10
1.3
<0.01
<0.01
<0.01
0.21
<0.01
2.0
17.
0.25
0.29
<0.01
1.2
<0.01
1440
<0.02
0.49
<0.01
<0.07
<0.10
1.2
<0.01
<0.01
<0.01
0.20
<0.01
1.3
13.
0.21
0.30
<0.01
1.0
<0.01
* Concentrations expressed as (jg of element extracted per gram of fly ash.

t Calcium stray light corrections applied.
                                        182

-------
      TABLE 0-8.   METAL CONCENTRATIONS*  RESULTING  FROM  COLUMN  LEACHING AT 3 ML/MIN
         OF THE 45-74 pm SIZE FRACTION OF  CORRETTE FLY  ASH WITH A pH 7 SOLUTION
Time (min)
Element
B
Cd
Be
Mg
P
Si
Mot
Mnf
Nit
Na
Cu
Alt
Ca
Bat
K
Crt
Srt
Pbt
0
23.
<0.03
<0.01
<0.07
<0.10
18.
1.2
0.29
<0.01
11.
<0.01
0.86
4200
14.
190
0.17
53.
1.4
15
16.
<0.03
<0.01
<0.07
<0.10
31.
0.7
0.42
<0.01
7.7
<0.01
3.0
2700
32.
50.
0.08
22.
0.67
60
4.9
<0.03
<0.01
<0.07
<0.10
40.
0.59
1.2
<0.01
5.4
<0.01
7.0
1800
8.5
44.
0.06
8.3
0.35
120
1.9
<0.03
<0.01
<0.07
<0.10
41.
0.13
0.22
<0.01
3.0
<0.01
14.
550
2.2
29.
0.03
4.0
<0.01
180
0.5
<0.03
<0.01
<0.07
<0.10
46.
<0.01
<0.01
<0.01
1.0
<0.01
15.
280
1.0
4.9
<0.01
2.9
<0.01
360
0.2
<0.03
<0.01
<0.07
<0.10
38.
<0.01
<0.01
<0.01
0.51
<0.01
16.
120
0.78
3.1
<0.01
2.0
<0.01
540
0.08
<0.03
<0.01
<0.07
<0.10
37.
<0.01
<0.01
<0.01
<0.05
<0.01
17.
78.
0.56
2.1
<0.01
1.4
<0.01
720
<0.02
<0.03
<0.01
<0.07
<0.10
40.
<0.01
<0.01
<0.01
<0.05
<0.01
17.
45.
0.11
1.8
<0.01
1.3
<0.01
1440
<0.02
<0.03
<0.01
<0.07
<0.10
47.
<0.01
<0.01
<0.01
<0.05
<0.01
17.
29.
0.08
1.1
<0.01
1.1
<0.01
* Concentrations expressed as  |jg of element extracted  per  gram  of  fly  ash.

t Calcium stray light corrections applied.
                                        183

-------
 TABLE D-9.   ANION CONCENTRATIONS* RESULTING FROM MODIFIED SOXHLET LEACHING
               OF BULK CORRETTE FLY ASH WITH A pH 7 SOLUTION
Anion
Time (min)
1st cycle
15
60
120
180
360
540
720
900
1080
1260
1440
1800
2160
2880

S.
1.
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
F
E.t
3
0
37
29
18
17
13
11
14
10
09
07
13
07
06

B.F. 11
15
12
6.8
7.6
8.4
9.6
11.
11.
12.
13.
13.
14.
16.
16.
17.

S.
1.
1.
0.
0.
0.
0.
0.
0.
0.
<0.
<0.
<0.
<0.
<0.
<0.
Cl
E.
8
5
73
55
22
30
15
10
04
01
01
01
01
01
01

B.F.
33
26
12
12
16
15
16
16
19
21
21
23
24
22
23
NO
S.E.
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
0.01
0.03
0.06
0.07
0.07
3
B.F.
<0.01
<0.01
<0.01
0.40
2.0
4.8
11.
12.
14.
15.
16
16.
18.
18.
17.
sof
S.E
103
70
6.
2.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
•


6
4
88
51
42
29
23
22
20
18
35
20
20
B.F.
760
600
240
.310
300
340
340
330
350
350
340
340
380
370
350
* Concentration expressed as ug of anion extracted per gram of fly ash.

t Soxhlet extractor (interval).

If Bottom flask (cumulative).
                                    184

-------

QQ
\CHING Of
IXTRACTOf
LU
—1 Z
1 — 1
H-
1 1 1 /— V
31 
CO LU
h-
Q Z
LU 1— 1
1 — 1 ^^
U_
Q O
O l-H
«=c
0 r-
Q£ Z
U_ LU
O
CD Z
Z O
i— i O
I—
CO Z
LU O
Q£ I— i
1—
•x :D
CO —J
O CO
t— i
r— !*"•*
h- Q-
z
LU -H
-
LU —1
— 1 U-
LU
1 1 1
0 LU
Q O
C_>
LU
_J
QQ
 •=*•
0 O
CM 00
CM rH
0 i-H
0 O
rH rH
0 O
O O
V V
O 00
rH LO
O O
i — uo
0 0
0 O
uo oo
CM CM
o o
CM O
r^» cn
i-H i-H
0 0
0 0
V V
rH rH
0 0
0 0
V V
CD O
0 0
m o
cn cn
o o
CD O
CM l-H
CD O
CO CD
i-H i— 1
cn
00 O
0 i-H
i-H rH
0 0
CD CD
V V
o o
rH CM
CM CD
CM OO
0 0
rH 0
CM CM
O O
00 CD
O CM
i-H rH
rH
0
O
V
rH
O
0
V
O
o
CD
cn
0
rH
O
0
V
co
rH
in
rH
l-H
O
O
V
cn
CM
o
LO
0
cn
1— 1
o
rH
rH
CD
O
V
i-H
CD
CD
V
rH
i-H
0
O
rH
0
O
V
CD
i-H
cn
rH
rH
0
O
V
00
oo
oo
00
o
cn
rH
O
O
00
rH
rH i-H
0 0
CD O
V V
rH rH
O O
0 0
V V
CM r~-
o o
0 0
^- oo
co co
0 0
rH rH
O O
0 0
V V
LO LO
rH i-H
*J- CM
CM CM
rH rH
0 O
O O
V V
00 CO
OO LO
00 00
rH rH
O O
CO 00
rH 00
CM CM
185

-------




/—\
-TJ
QJ
3
C
+J
C
0
c_>
o
rH
a
LLJ
1
CO

C
*r—
E
s_^
0)
t—
r-i 01
CM r- 1
00 rH
IT) rH
0 0
o o
rH IX)
O 0
V
00 0
«*• CO
oo r-N
CM rH
in ix>
rH rH
0 0
ID CM
ro ^f
0 0
^f 05
rH rH
O O
rH 00
0 0
0 0
V
CM rH
IX) un
O O
CO
U
>> ui
U rH
-P
in
rH
CM O
CM «3"
i— 1 CM
O 0
0 O
rH (Ti
CM UT>
O CO
CM r-t
rH
CM CM
rH 0
rH CM
CM «*
0 0
0 rH
ID UT>
O 0
r^. ix>
rH CO
O O
CM 00
rH CM
0 0
r^- o
*d- ro
O 0
0 O

<*• ro
O 0
CO ID
CO CM
0 O
f~ rH
CO CO
O O
OO CD
CM CM
O O
rH ID
CO CNJ
O 0
O 0
CO ID
rH CO
CD «3-
•=f CM
00 rH
CM O
0 0
V
Csj CM
rH rH
oo <-
**• «3-
O O
CO CM
ro co
o o
0 r-t
m <*•
O 0
^ ID
ID CM
O 0
O ID
«*• co
O O
rH CM
m ^j-
O O
O ID
CM CM
0 0
O O
*i- CM
m r-.
p~- m
rH rH
CM «*
0 0
0 0
CO CO
rH rH
rH rH
CM rH
O 0
ID CT>
CO **•
O O
CM l£>
m un
0 0
O rH
CO CO
0 0
o r--
«*• co
o o
ID ^J-
•* m
0 O
rH in
CO CO
o o
o o
O 00
01 o
i— i
oo m
rH rH
CM CO
rH CM
O O
m ID
rH rH
00 CO
rH O
O 0
V
ID rH
«3- m
0 O
0 OT
UD in
o o
CM CO
co co
O 0
en o
CO •*
o o
00 CD
in m
0 0
in rH
CM CO
0 O
O 0
ID «3-
CM *1-
rH rH
r~- co
rH CM
in rH
CO O
O O
V
r^ r>-
rH rH
CO CO
O O
O O
V V
vx> m
^l- co
o o
us r^
ID m
0 0
vo -3-
CM 00
0 0
r- m
co co
o o
O CM
r~- ID
O 0
co r«-
co «*•
CO O
o o
O ID
00 rH
i— t CM
LO
rH
Csj
Csl
0
LO
rH
CO
CO
CO
V
cn
CM
CD
LD
Ln
CD
CM
't
CD
«d-
CM
CD
in
in
CD
rH
CO
O
0
!§

Q.
t extracte<
c.
O)
0)
'oJ
<*-
Q
O5
3.
in
03
•^
OJ
in
in

.C
O5 C
•r- O
+j
>> Q.
ro i-
S- 0
-P to
m .a
m
3 U
U E
r— O
ro -t->
0 
-------



_J
mx-/1
^L_
OO
0 —1
u_
z s:
i— i O
0 H-

^ LU
X >•
O I — i
00 H-
0 —1
LU ID
1 — I ^"
LL. ID
i— i O
o
s: z
o
S-. ^""^
O 1—
a: -
	 1
•— t LU
rH r—
1 H-
O LU
o:
LU a;
— I O
CQ C_J
<^
I —


S=
s:






£_





ro
QQ



s=
s-
OO


-p
c
O> ro
E 0
a>
r^
LU

cn





cu






NX





no
Z


:*r
	 i

^— ^
< —
• r—
E


a>
E
£
LD
o

cn
CM
0

^.
CO
o



CO
CM




.
ID
rH



O
O





CM
rH



rH
0
O
V



CM
ID



ID
O
.
O
V

^:
CM


01

U
>^
U

-p
10
rH
CD f^*
rH rH

r-. r-
in cn
o o

^
cn m
0 rH



CM <3-
^ r~^




. ,
co in
co r-~



o o
0 0
O *D
rH rH



CO ID
CM Ln



rH rH
0 0
O O
V V


. .
cn ID
ID CO





, .
co m
rH in

in cn
CM CM





in o
rH 
-------

















.
f 	 s
T3
Ol
^
C
-P
C
0
U
^"^
1— 1
rH
CD.

LU
	 1
CO

E
CD
i i i
T3
0





C
M






3
O






•r—
Z






o>
LJ_








,, — s
f~
•r—
E


CD
E
•r~
i—
rH rH
O O
O O
V V

o o
CO CM
rH
00 LO
rH CO




CTI
ID O
CM






O O
rH rH





00 ID
CO CO





rH rH
O CD
O O
V V




rH ID
0 O
O O
V





ID rH
CM CM










CD

U
>> LO
U rH

-P
rH
rH
CO O
O O
v

o o
CM **•
CM CO
CTI 00
LO ID





CM rH
^- LO






rH CTI
rH CM





r-. LD
CO CO





rH CO
0 rH
O O
V




^j-
CO rH
0 rH






CO CO
rH CM













0 0
ID CM
rH


CTI r**
UD r--


O 0
CT) «3-
CM rH
rH CO
CTl CTI





•3- 0
r^*1* f~~i
rH





LD LO
^f ID





CO CTI
CO CM





cn
r^ LD
O rH






^ LD
rH CM






ID 00
CM CM













O 0
00 ID
rH CO


rH
O
CM O
rH V
rH
0 0
LD V

CD «*•
ID LO





0 0
co r~~
rH rH





<=t CD
ID CT)





•3- o
^ LO





rH rH
O CD
0 0
v v




rH
<3" CD
CO CD
V





r-- CD
CM CM













0 O
•* CM
LD r —


o o
r-. oo

rH rH
CD 0
V V

«3- CO
ID r^*





0 0
00 O
rH CM





00
CTl 0
rH




rH CTI
«tf- co





rH CD
O CM
O O
V





«3- CTl
rH rH






O CM
CO CO













0 0
0 00
CT) O
rH

CM
OO 00
CO
rH rH
0 CD
V V

CO «3-
00 ID





0 0
rH CO
CM rH






O ID
rH CM




cn oo
CO CM





O rH
ID CD
O O
V





CD
CM CT)
rH





«* 0
CO «3"













O O

CM •*
rH rH


tD ID
CO LD
rH rH
0 0
V V

ID 00
00 CTI





O 0
^1- LD
rH rH






00 rH
rH CM




CO O
LD «=|-






0 tD
**• LO







LD LD
rH rH





0 CO
*d- LO













O O
O 
ro

3-*
<4_

tf_
o

E
ro
j-

i-

u
Ol
i-
i-
0
u
+J
.C
cn c
•r- O
'~ '-P
>> Q-
ro i-
i- 0
-P in
in o
ro
E
3 U
•1 	 •(—
U E
i— O
ro -P
CJ eC
6= CO3
188

-------



I
ID
CO

LL.
0

CJ3 ^
•z. co
1-1 
LL! P
t-l 
y y
-~^
s: c-5
CD ^-^
Qi
Ll_ Z
O
C_3 i— i
Z 1—
i— i  z
CO LU
LU O
rv ~y
O
* CJ
CO
Z
o •
P 0
-
Z —1
LU U_
LU LU
-J t—
LU h-
LU
fV
- Q;
oo o
rH C_3
1
LU
CO
«^
r—






;jr



4*—
O
s:




4—
S-
t_)




4*~
ro
CO



4*~~
4-J £—
C CO

^>"







CO








'^






ro
Z



c
• r-
E
^^/

0)
£
£
rH
O
CD
V

rH
0
.
O
V



rH

O



rH
.
rH



CD

co






0
0
o
rH


CD
^^
.
0




rH
CD
•
O
V




oo

f-^.






.
o
rH


0)
r—
^
u

+J
I/)
rH
rH
0
CD
V

rH
O
.
CD
V


cn
00
.
CD



i— I
.
00




.
rH
rH





0
O
j1^.
rH


r--
0
.
0
V



rH
0
*
O
V




oo

CD






,
OO
^J-





ID
t-H



rH
O
O
V

rH
CD
.
O
V



cn
.
CD



LO

•=*




.
^~
oo





CD
O
p^.
oo



oo
.
rH




rH
0
.
0
V





,
^"
rH






O
rH
rH




O
CD



t— 1 OO
O LO
CD O
V

CD
cn CD

O CD




00 CD

00 00



cn i —
. .
LO r^




, ,
^j- oo
oo ^i-





0 0
0 0
oo oo
oo oo


CD r^~
CD O
. .
0 0
V



rH rH
0 0
. .
O O
V V





, ,
r-t OO
oo cn






^~ ^~
P-* p***





0 0
OO 00
rH rH


rH
O OO
O i— 1
V


oo O
. .
rH OO




CD 00
. .
«3" CD




, .
*d- CD
i— I cn



. .
0 00
r**** 0*1





0 0
0 0
cn co
LO 1 —



oo o
. .
rH rH




rH rH
0 0
• .
0 0
V V





,
CD O
cn cn
rH




,
O OO
CD cn
rH




0 O
CD -3-
OO LO



CO
O



00
.
00




CD
.
r^




.
LO
00




o
OJ
rH




0
O
ro
cn



CO

^~




rH
0
*
O
V






CD
O
oo





o
rH
rH




0
OO
r^.



CD
CD



00
.
00




CD
,
00




.
CD
OO




CD
^~
rH





1
1




O

P--




rH
O
.
0
V






CD
CO
rH





O
00
rH




0 0
0 00
cn o
rH

rH
CD
CD
V


CD
.
OO





*
rH
t-H



*
^>
ro




o
r*^
rH





1
1





.
OJ
rH



rH
0
.
0
V






CD
co
rH





O
ro
rH




0
co
OJ
rH

CD rH
00 0
0 0
V


OJ CD
• .
r^ r^





. .
cn oo
rH OJ



* .
oo o
ro oo




CD O
cn oo
rH OO




O O
0 0
0 0
ID OO
t-H rH


OO
. .
LO rH
rH



rH rH
0 0
. .
0 0
V V






CD CD
•51- •*
«3- -3-


LO ID
0 0

0 O
V V





0 0
«tf- o
*J- 00
rH rH

rH t-H
0 0
0 0
V V



. .
CD OO
rH rH




. .
CD rH
OO OO


CD

o cn
0
1
1 1


.
CD CD
*d- cn
oo




o o
0 0
0 0
CD 1^
OO rH



. .
oo co
rH rH



rH rH
0 0

CD 0
V V






CD CD
LO cn
^~ oo





CD CSJ
LO rH





0 0
CD CO
rH CO
00 00

189

-------

















.
^*^s
T3
CU
3
C
,f—
•P
C.
0
(J
v~*/
CM
rH
Q

LLJ
i
QQ

r—













































-P
C
CU
E
CU
r—
LLJ


















































OO


t
^^

















•a
0











3
CJ









_-r-
Z









.^^
C
• ^
E


CU
E
• r-
| —
CD
rH
CD
\/
cn
,_!




CM
0
O
\f









rH
.
O









rH
O
*
0
•v







0

o










cu

(J
>>
(J

4_>
in
rH
O
rH CM
o cn
\s
ID
r^ 06
CM



CM CM
O O
O O
•V \f









00 00

0 CD









rH rH
O O
. •
O 0
v •v







ID ID
•vf 00
O O













un o
rH ID




O
rH
0
\S

^
<^-



CM
O
O
V









rH

r-i









rH
CD
.
0
V








m
CM













o
CM
rH




ID
CM


ID
ID



CM
O
0
V









CD
.
CM









rH
O
*
O
•v








un
CM













0
00
rH





cri CM CM o
ID *D CO i — 1
rH

O O CD O
oo r~« CM i —
rH rH CM CM


CM CM CM CM
0000
O O O O
V V \S V









rH  V \/ V \/

0 0 0 0 CD
«tf- CM ID O ID
OO OO OO ^" ^J"


CM CM CM CM CM
0 O 0 O CD
O O CD O CD
V V V \S V









^>
• • • • •
CO CO OO 1 — CM
CM CM CM CM








rH 0 rH O
O CO O CO
• • » • •
0 0 0 0 CD
\S V rH








•*"
ID CM ro ID r~
rH rH rH rH












O O O O CD
ID ^" *"~i ID 00
CM «3- CO rH 00
rH rH rH CM eg










-C
in
re
^
M-
M-
O
E
re
i-
O)

i-
QJ
O-

"O
cu
_l_)
u
re
-P
X
cu
+J
c
cu
E
cu

a>

4-
o

2
re

•o
0)
in
in
i-
OL
X
01
U5 C
c o
O •!-
•r- in
-P in
re T-
i- E
-P CU
c
cu cu
U E
c re
o •—
o u_

•X -h-
































.
-a
cu
• r—
i —
Q.
a.
re

in
o
u
0)
t-
j_
o
u
4_>
r~
cn
•>-
T~~
^^
re
S-
4_>
in

E
3
• p—
U
^—
re
c_>

t=


















































.
c
0
-P
Q.
S-
o
in
.a
re

u
*r~
E
O
-P
*3^


-------











, — ^
r—
LU
^

O
IT)


LU
— i Z
<_> O
c_> r—
LU ^>
ce. —i

co a
O LJ-
=> 0
z
P— 1 Z
r- 0
Z
21 =>
O Li-
Li- -

Li-

on LU
Z h-
1 — t LU
< ce
a: o
r- O

LU ^
C_) —I
Z =>
O CQ


—1 O
<£
h- CJ3
LU 1— 1
E: 3:
LU O
UJ UJ
	 1


CO
rH
1
O

LU
	 1
CQ
<^
r-






















































^^

's
' — '
c
0

4_>



O













































O
O
CD
0
i — 1


0
0
o
I—t





CD
O
CD







0
O
CM





O
0
rH






in









CD
in








in
CM










o
rH









in






C
01
E
01
LU
in in
0 0
rH O CD
ro V V


in in
0 O
o o d





in in
CO 0 0

r~i d d







m in
rH O 0
co o d
V V




in LO
CM 0 O
r- o o





in
cn CD

in ro o
rH V





in
CM O

CO CM O
rH V







cn ro
iri cn r-!
rH










d CD co
rH r~-







ID in

^ CM cn
in







-r- m
— 1 Z K:

o o
O CD
cn CD
rH 0
rH


0 0
0 O
in o
CM
CM




CD CD
CO O
rH O
ro




o
CO
0 O
O
o
in
CM

o
CD
O O
O
O
rH
rH




CO

rH 0
O
O




CD
CO

O O
0

m





CD
O CD
V O
m
co








CM O
o
CM





r—
o

d o
V CD
^f-
CM





cn re
El cj

CD O
O 0
cn CO
CM ro


o o
CD ID
co in






0 0
CO rH








0 O

•3- LD





O O
CO 1 —







O 0
rH F^—
m ro







0 0
CO CM
CM ro







o o
CM in
ro in









o o
rH rH
CM in








O 0
in CM
rH ro






t_ |TJ
l/l CQ

ro r-J
rH rH



ro O
CM Csi





CO
r- ro

rH O







CO
r- **•
d CM





in in
rH O
O O





ID rH
rH CD

d d






rH CO
rH 0

o d







in ro
rH CM
o d








O CD
CM rH

o d







in cn
0 0

o d








01 C.
u_ r-g
o in
CM CD



in
o o
rH O





in *3-
rH O

d d







CM ro
•3- 0
d o





CM ro
in o
o o
V




CD ro
rH O

d o
V





<3- ro
CM CD

O O
V






tjD ro
0 O
o o
V







cn ro
O O

O O
V






*rf- ro
O O

d o
V







3 -a
o o
CD
r-^




CD
^






rH

<^J-








in
CM







o






CM

o






CNJ
CM

O







ro
rH
O








i — 1
O

O
V






rH
O

O
V







£-
C_3
O
[^




ro
^





O
cn

d








i— t
o





CD
O
O





CM
O

d






CM
O

0







CM
O
O








ro
CD

d







rH
O

d








c
31
in o
r~~ ro



rH
o o
rH CD
V




r^ o
rO rH

d o







in
CM O
ID O





in rH
rH O
d d





rH rH
O 0

O O
V





CM iH
O O

O O
V






rH CD
CM O
0 O








CM rH
O CD

d o
V






CM rH
0 O

d o
V







J3 "-
a. z
CM
CO




in
06






in

CO








m
rH
rH





m
rH
i-H
ifi
-^

4—

00 4-
O
O
r-t 1 E
ro
01
J_
QJ
CL

CO T3
QJ
•— 1 -*-J
i-H U
to


X
OJ
^J
c
C\J OJ
c\i aj
rH r—
OJ

•4-
O

OT
3.

ID l/l
fO
CSJ
i-H "D
OJ
i/l
l/>
0)
Q.
X
OJ

CO
CSJ C
i-H O

4->
t-
C
QJ
U

^ O
CL <->
191

-------
TABLE D-14.   ELEMENTAL CONCENTRATIONS* RESULTING FROM CONTINUOUS RECYCLE
               (SOXHLET) LEACHING OF BULK CORRETTE FLY ASH
Elemen
Li
Na
K
Mg
Ca
Sr
Ba
Fe
Zn
Cu
Cd
Cr
Mn
pH

5
0.19
<0.05
--
1.4
190
1.6
3.1
0.34
0.24
0.02
<0.03
0.01
0.06
9.2
Time (min)
15
0.78
5.3
--
0.60
500
14.
12.
0.09
0.21
0.02
<0.03
0.02
0.02
10.0
60
1.9
17.
1.6
0.70
1100
47.
41.
0.11
0.40
0.02
<0.03
0.01
0.02
10.8
180
1.9
12.
--
0.80
2200
74.
91.
0.07
0.54
0.02
<0.03
0.08
0.01
11.0
360
1.1
6.6
0.6
1.4
2800
69.
170
0.08
0.59
0.05
<0.03
0.06
<0.01
11.1
720
0.56
<0.05
--
1.5
3900
86.
450
0.07
0.89
0.06
<0.03
0.06
<0.01
11.0
1440
0.50
6.9
3.1
1.5
4700
73.
720
0.09
1.0
0.07
<0.03
0.06
<0.01
11.8
 Concentrations expressed as |jg of element extracted per gram of fly ash.
                                   192

-------
    TABLE D-15.   ELEMENTAL CONCENTRATIONS* RESULTING FROM CONTINUOUS AND
      INTERMITTENT RECYCLE (SOXHLET) LEACHING OF BULK CORRETTE FLY ASH
                     FOR 24-HOUR AND 7-DAY TIME PERIODS
Quantity
Element
Li
Na
K
Mg
Ca
Sr
Ba
Fe
Zn
Cu
Cd
Cr
Mn
PH
Specific conductivity (pmhos/cm)
Alkalinity (mg/liter CaC03)
Hardness (mg/liter CaC03)
SofCpg/ml)
F~ (pg/ml)
Continuous
24 hr 7 days
6.2
16
--
1.1
5600 12000
280 550
270 720
0.05
1.2
0.09
<0.03
0.26
<0.01
11.9 12.0
5900 3200
1000 690
1100 780
<0.01 <0.01
0.33 0.24
Intermi
24 hr
7.0
48
--
7.9
15000
370
1500
0.85
3.9
0.28
0.04
0.28
0.10
--





ttent
7 days




12000
500
520






--
6600
1400
1500
--
--
* Concentrations expressed as pg of element extracted per gram of fly ash.
                                    193

-------
 TABLE  D-16.   ELEMENTAL  CONCENTRATIONS*  RESULTING FROM 24-HOUR CONTINUOUS
                RECYCLE  (SOXHLET)  LEACHING OF BULK CORRETTE
                   FLY ASH AS A  FUNCTION OF  PARTICLE SIZE
Element
Li
Na
Mg
Ca
Sr
Ba
Fe
Zn
Cu
Cd
Cr
Mn

>180
1.3
1.9
5.8
46000
240
1400
0.77
1.5
0.77
0.05
0.74
0.09

150-180
2.2
2.2
1.9
46000
290
1900
0.99
8.3
0.98
0.05
0.63
0.01
Particle
74-150
5.3
23.
2.7
48000
370
3200
0.70
1.6
0.64
0.05
0.87
0.02
diameter (pm)
45-74
6.8
59.
10.
13000
380
1500
1.1
5.3
0.16
0.04
0.36
0.02
30-45
11.
67.
14.
11000
420
680
0.89
0.85
0.15
0.03
0.60
0.04
20-30
13.
82.
12.
6600
430
600
2.5
1.4
0.16
0.04
0.73
0.02
* Concentrations expressed as pg of element extracted per gram of fly ash.
                                    194

-------
 TABLE D-17.   ELEMENTAL CONCENTRATIONS* RESULTING FROM 24-HOUR
              CONTINUOUS RECYCLE (SOXHLET) LEACHING
                  AS A FUNCTION OF TEMPERATURE
Element
Li
Na
K
Mg
Ca
Sr
Ba
Fe
Zn
Cu
Cd
Cr
Mn

25°C
7.0
48.
5.3
7.9
15000
370
1500
0.85
3.9
0.28
0.04
0.28
0.10
Temperature
65°C
12.
140
91.
8.8
25000
610
2000
1.4
2.1
0.43
0.04
1.5
0.08
* Concentrations expressed as (jg of element leached per gram of
  fly ash.
                               195

-------
•=c

>•
o
CJ
o
o
o
o
O H-.
ni o
                                                       ro


                                                       CM
z o
O 1-1
                                                                o


                                                                E


                                                                !_

                                                                O)


                                                                S-

                                                                01

                                                                D.
o
o
                                                O


                                                C
                                                ra
                                                                o

                                                                en
00
1-t

O
                                        IQ


                                        CM
                              00


                              00
                c
                o
                                             CM "3-
                                               O
                                                u
                                                c
                                                o
                                                       re
                                                       O.
                                  196

-------
QC O
^ i-< .,
Qh- I' cn
                                                 S
                                 K  £
                                                        197

-------
     TABLE D-20.   ELEMENTAL CONCENTRATIONS* RESULTING FROM A 3-HOUR
             SONIC EXTRACTION OF BULK CORRETTE FLY ASH WITH
                      WATER AND WITH EDTA SOLUTION
Element
Alt
B
Bat
Ca
Cd
Crt
Cu
K
Mg
Mnt
Mot
Na
Nit
Pbt
Si
Srt
Lilf
Fe§
Zn§
H20
260
160
320
20,000
0.31
0.14
0.22
10.
<0.07
0.11
5.0
18.
0.12
4.9
87.
220
3.5
0.85
0.10
0.1 M EDTA
3,500
340
1,400
53,000
2.0
2.6
11.
130
2,900
210
15.

0.10
18.
5,600
310
7.0
440
1.9
* Concentrations expressed as |jg of element extracted per gram
  of fly ash.
t Calcium stray light correction applied.
1T Flame emission.
§ Atomic absorption.

                                    198

-------
            TABLE D-21.   ELEMENTAL AND  ANIONIC  CONCENTRATIONS*  RESULTING
              FROM 3-HOUR SONIC EXTRACTIONS OF  DIFFERENT  COAL FLY  ASHES
Anion or
Element
PH
F"
cr
C03~
N03
sof
B
Cd
Be
Mg
Si
Mot
Mnt
Nit
Na
Cu
Alt
Ca
Bat
K
Crt
Srt
Pbt
LiTT
Fe§
Zn§
Corrette
12.5
47.
18.
7600
15.
1300
160
0.31
<0.01
<0.07
87.
5.0
0.11
0.12
18.
0.22
260
20000
320
10.
0.14
220
4.9
3.5
0.85
0.10
North
Dakota
11.9
94.
18.
15000
10.
26000
760
0.13
<0.01
<0.07
75.
19.
0.13
0.2
6700
0.34
65.
14000
19.
2000
0.47
310
7.8
5.0
0.40
0.80
Niagara
Mohawk
9.0
8.0
12.
<150
8.0
4400
44.
0.48
<0.01
54.
29.
12.
0.06
0.18
350
0.36
5.2
1600
0.82
400
0.35
9.9
1.2
36.
0.35
0.45
Will
County
11.5
12.
190
26000
6.2
41000
110
0.05
<0.01
<0.07
54.
21.
0.20
0.19
21000
0.48
130
9100
47.
460
0.46
350
7.9
2.0
2.0
0.42
Corrette
(EDTA)
4.8





340
2.1
<0.01
2900
5600
15.
210
0.10

11.
3500
53000
1450
130
2.6
310
18.
7.0
440
1.9
* Concentrations expressed as pg of element or anion extracted per gram of  fly  ash.

t Calcium stray light correction applied.

II Flame Emission.

§ Atomic absorption.
                                        199

-------
                       APPENDIX E

                 FLY ASH REPORT GLOSSARY


AES       Auger electron spectrometry

ESCA      Electron spectroscopy for chemical analysis

FTIR      Fourier transform infrared

GC        Gas chromatorgaphy

GC/MS     Gas chromatography/mass spectrometry

INAA      Instrumental Neutron Activation Analysis

MMED      Mass Median Equivalent Diameter

NBS       National Bureau of Standards

PAH       polynuclear aromatic hydrocarbon

POM       polycyclic organic matter

RSD       Relative Standard Deviation

SEM       Scanning electron microscopy

SEM/EDS   Scanning electron microscopy with energy dispersive
          X-ray spectrometry

SIMS      Secondary Ion Mass Spectrometry

SRM       Standard reference material

SSMS      Spark source mass spectrometry

ug        microgram (10 6 g)

urn        micrometer  (10 6 meter)

uv        ultraviolet

VMD       volume mean diameter

XRF       X-Ray Fluorescence

                                 200

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
  EPA 600/3-80-093
                                                           3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
                                                           5. REPORT DATE
 Environmental
 IV  -  Chemical
 Ash
Effects of Western  Coal  Combustion:  Part
and Physical Characteristics of Coal Fly
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
                                                           8. PERFORMING ORGANIZATION REPORT NO.
 David F.  S.  Natusch and David  R.  Taylor
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Department of Chemistry
 Colorado State University
 Fort Collins, Colorado  80523
                                            10. PROGRAM ELEMENT NO.
                                               1NE625, EHE625
                                            11. CONTRACT/GRANT NO.


                                               R803950
12. SPONSORING AGENCY NAME AND ADDRESS
U.S.  Environmental  Protection Agency
Environmental  Research Laboratory-Duluth
6201  Congdon Boulevard
Duluth,  Minnesota  55804
                                                           13. TYPE OF REPORT AND PERIOD COVERED
                                              Final
                                            14. SPONSORING AGENCY CODE
                                              EPA-600/03
15. SUPPLEMENTARY NOTES
16. ABSTRACT
      Coal  fly ashes from different  sources were extensively  studied.   Fly ash consists
of as many as 12 different particle types, made up mainly of Si,  Al,  Ca, K, Na, and Fe.
Many  potentially volatile trace  elements (As, Se, Ga, In, Pb,  Cd) are always preferen-
tially associated with more  respirable  particles.  Many other  minor  and trace elements
may also show an inverse particle size  dependence, depending on  the  coal fly ash.
Elements showing no dependence include  Al, Ba, Ca, Co, Fe, K,  Mg, Rb, Sc, Si, Sr, Ta,
Ti, and the rare earths.  Many of the more volatile and potentially  hazardous trace
elements are surface associated.  Most  potentially hazardous inorganic species in fly
ash can be leached into aqueous  solution but are unlikely to give rise to solution
concentrations of great concern.  The physical and chemical  characteristics and be-
havior of polycyclic organic matter (POM)  associated with coal fly ash was also
studied.  Theoretical models show that  adsorption of POM on  coal  fly  ash will occur
in seconds in a typical emission stream and that the process is  highly temperature
dependent.  Surface-adsorbed POM is highly reactive to some  nitrogen  and sulfur
oxides at levels expected to be  found in power plant plumes.   It  was  found that some
POM adsorbed on fly ash are  stabilized  against photochemical  decomposition, while
other compounds decompose readily upon  adsorption.  Although fly  ash  emitted to the
atmosphere contains much more POM than  precipitator ash, there is an  insufficient
amount to make a discernible impact on  any terrestrial aquatic system.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS
                                                          c. COSATI Field/Group
Fly  ash
Particulates
Trace  metals
                               Energy development
                               Coal combustion
                               Air emissions
                    10/B
18. DISTRIBUTION STATEMENT


RELEASE  TO PUBLIC
                               19. SECURITY CLASS (This Report)
                                 UNCLASSIFIED
                                                          21. NO. OF PAGES
                   215
                                              20. SECURITY CLASS (Thispage)
                                                UNCLASSIFIED
                                                                         22. PRICE
EPA Form 2220—1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE
                                            201
                                                      US GOVERNMENT PRINTING OFFICE 1981 -757-064/0 197

-------