&EPA
         United States
         Environmental Protection
         Agency
          Industrial Environmental Research  EPA-600/7-79-027
          Laboratory          January 1979
          Research Triangle Park NC 27711
Sampling Charged
Particles with Cascade
Impactors

Interagency
Energy/Environment
R&D Program  Report

-------
                  RESEARCH REPORTING SERIES


Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional  grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

    1. Environmental Health Effects Research

    2. Environmental Protection Technology

    3. Ecological Research

    4. Environmental Monitoring

    5. Socioeconomic Environmental Studies

    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 INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded  under the 17-agency  Federal  Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments  of, and development of, control technologies  for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
                        EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for  publication. Approval does not signify that the contents necessarily reflect
the  views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.

This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                    EPA-600/7-79-027

                                          January 1979
Sampling  Charged  Particles
   with  Cascade  Impactors
                       by
             W.E. Farthing, D.H. Hussey, W.B. Smith,
                   and R.R. Wilson, Jr.

                Southern Research Institute
                2000 Ninth Avenue, South
                Birmingham, Alabama 35205
                 Contract No. 68-02-2131
                  T.D. 10401 and 11301
                Program Element No. EHE624
              EPA Project Officer: D. Bruce Harris

           Industrial Environmental Research Laboratory
             Office of Energy, Minerals, and Industry
              Research Triangle Park, NC 27711
                     Prepared for

           U.S. ENVIRONMENTAL PROTECTION AGENCY
              Office of Research and Development
                  Washington, DC 20460

-------
                        TABLE OF CONTENTS
Figures	   iii
Tables   	   vii
Abstract	viii
Acknowledgment	    ix

Sections

   1.    Introduction  	     1

   2.    Procedures and Results  	     2
            Monodisperse Aerosol .  .  .	     2
               Experimental Procedure.  	     2
               Particle Charging 	     4
               Charge Measurement Methods. .  .  	     7
                  Method I	     9
                  Method II	     9
                  Method III	    11
                  Method IV	    12
               Evaluation of the Charged Particle
                 Generator .......  	    13
               Impactor Data	    18
                  MRI	    18
                  University of Washington 	    35
                  Andersen	_. .    35
               Effect of Charge on Efficiency Versus /ip  .    36
               Particle Deposition Patterns	    47
               Nonconducting Jet Plates	    53
            Polydisperse Aerosol Sampled with a Brink
              Impactor	    53
            Polydisperse Aerosol Sampled with Andersen
              Impactor	    61
               Experimental Apparatus	    61
               Experimental Procedure	    63
               System Equivalency	    63
               Charged and Neutralized Particle Sampling .    64
               Results	    68

   3.    Summary and Conclusions	:  .  . .    74
            Monodisperse Aerosol 	 .....    74
            Polydisperse Aerosol Sampled with a Brink
              Impactor	    75
            Polydisperse Aerosol Sampled with Andersen
              Impactor s	    75

         References	    77
                                 11

-------
                            FIGURES

 1.  Schematic of the VOAG charged particle generator
    and sampling arrangement used in sampling charged
    monodisperse aerosols	   3

 2.  Charged particle generator  orifice region with
    the parallel plates for  charge measurement	   5

 3.  Flow schematic and electronic block diagram of the
    Electrical Aerosol Analyzer.   From Sem3 	  10

 4.  Particle charge versus charging voltage for the
    VOAG charged particle generator as determined by
    Methods I-IV. (rp = 2.6  urn and rd = 23 ym)	14

 5.  Particle charge versus charging voltage for the
    VOAG charged particle generator 	  15

 6.  Control sampling run using MRI-Model 1502 Impactor
    with no charge	19

 7.  Sampling charged particles using MRI-Model 1502
    Impactor with no grounding wire—high particle
    charge	. . . .	20

 8.  Sampling charged particles using MRI-Model 1502
    Impactor with grounding wire—high particle
    charge. ....... 	 	  21

 9.  Sampling charge-neutralized particles using MRI- .
    Model 1502 Impactor with neutralizer at nozzle—
    np = 4 x 10*	22

10.  Sampling charge-neutralized particles using MRI-
    Model 1502 Impactor with nozzle losses corrected	23

11.  Sampling charged particles using MRI-Model 1502
    Impactor with grounding wire—moderate particle
    charge	24
12. Sampling charge-neutralized particles using MRI-Model
    1502 Impactor with neutralizer at nozzle—
    np
= 2.4 x 103	25
                                1X1

-------
13. Sampling charged particles using MRI -Model 1502
    Impactor with grounding wire — moderate particle
    charge ..........................   26

14. Sampling charge-neutralized particles using U. of
    W. Mark III Impactor with charge neutralizer at
    nozzle — np = 4 x 101*. ....... ....... ....   27

15. Sampling charged particles using U. of W. Mark III
    Impactor with no grounding wire — high particle
    charge ..... .....................   28

16. Sampling charged particles using U. of W. Mark III
    Impactor with grounding wire — high particle charge.  ...   29

17. Sampling charged particles using U. of W. Mark III
    Impactor with grounding wire — moderate particle
    charge ..........................   30

18. Sampling charge-neutralized particles using Andersen
    Mark III Stack Sampler with neutralizer at nozzle —
    np = 7 x 103 .......................   31

19. Sampling charged particles using Andersen Mark III
    Stack Sampler with grounding wire — moderate particle
    charge ...... ....................   32

20. Sampling charged particles using Andersen Mark III
    Stack Sampler with grounding wire — moderate particle
    charge ..... .....................   33

21. Reference curve (solid)  giving collection efficiency
    as a function of /ty  for the Andersen Stack Sampler
    and neutral particles ..................   38

22. Efficiency versus  /ij7 of  stages in Andersen Stack Sampler
    for neutral particles (solid curve)  and moderately
    charged  particles  (dashed curve) .............   41
23. Reference curve (solid)  giving collection efficiency as
    a function of /ij7 for the MRI-Model 1502 Impactor and
    neutral particles
24 •  Efficiency versus Sty of stages in MRI-Model 1502 Impactor
    for neutral (solid curve) ,  moderately charged (dashed
    curve), and highly charged  particles .  .  ........   44

25 •  Reference curve (solid) giving collection efficiency as
    a function of the /^ for the U. of W. Mark III Impactor
    and neutral particles .............  .....   45

26 •  Efficiency versus /^ of stages in the U.  of W. Mark III
    Impactor for neutral (solid curve) , moderately charged
    (dashed curve) , and highly charged  particles .......   46


                                iv

-------
27. Photographs of deposition patterns of ammonium
    fluorescein particles in MRI-Model 1502 Impactor	43

28, Photographs of deposition patterns of ammonium
    fluorescein particles in MRI-Model 1502 Impactor	49

29, Photographs of deposition patterns of ammonium
    fluorescein particles in Andersen Mark III Stack
    Sampler	50

30. Photographs of deposition patterns of ammonium
    fluorescein particles in Andersen Mark III Stack
    Sampler	51

31. Photographs of deposition patterns of ammonium
    fluorescein particles in Andersen Mark III Stack
    Sampler	52

32. Sampling charged particles using MRI-Model 1502
    Impactor with plexiglass jet plates, 2J, 3J,  and
    4J	54

33. Sampling system used to sample charged polydisperse
    aerosols with a Brink cascade impactor	55

34. Particle concentration at the outlet of a Brink cascade
    impactor with glass fiber substrates as collecting sur-
    faces for charged and neutralized conditions.	  57

35. Particle concentration at the outlet of a Brink cascade
    impactor with base metal plates as collection  surfaces
    for charged and neutralized conditions	58

36. Particle concentration at the outlet of a Brink cascade
    impactor operated without particle collection  surfaces
    (wall loss study) for charged and neutralized  conditions.  59

37. Particle losses in a Brink cascade impactor due to
    electrostatic forces	60

38. Polydisperse aerosol generating, charging, and
    sampling system 	  62

39. Particle size distribution—dM/dLogD versus
    particle diameter.  Both impactors sampling
    an uncharged aerosol simultaneously ... 	  66

40. Particle size distribution—dM/dLogD versus particle
    diameter.  Both impactors sampling a charged aerosol
    simultaneously	67

-------
41. Particle size distribution for moderate charging
    condition for a mass median diameter of 1.27
    micrometers	  70

42. Particle size distribution for moderate charging
    condition for a mass median diameter of 1.02
    micrometers	  71

43. Particle size distribution for moderate charging
    conditions for a mass median diameter of 0.692
    micrometers	  72

44. Particle size distribution for high charging
    conditions for a mass median diameter of 1.01
    micrometers	  73
                                VI

-------
                             TABLES

  I.  Average Values and Standard Deviations of Charging
        Parameters Observed by Reischl,  et al2.	8

 II.  Grouping of Surfaces for the Efficiency of Each
        Impactor Stage	40

III.  Summary of Polydisperse Aerosol Test Conditions.  ...  65

 IV.  Calculated Charge on Charged Aerosol Particles,
        Relative to that Expected in a Full Scale
        Precipitator	68

  V.  System Branch Equivalency Data 	  68
                                VI1

-------
                             ABSTRACT

     In performing particle size distribution measurements at
control devices operating on industrial process streams, investi-
gators are usually aware that in some cases charged particles
will be present in the gas stream.  In order to assess the in-
fluence of particle charge, three different experiments were
performed to determine whether or not cascade impactors sampling
charged aerosols can yield erroneous particle size distribution
measurements.  The commercially available cascade impactors utili-
zed in this study were the Andersen Mark III Stack Sampler, the
Meteorology Research, Inc. Model 1502 Cascade Impactor, and the
University of Washington Mark III Source Test Cascade Impactor.
In general, the measured distributions indicated more large par-
ticles and fewer small particles than actually existed.  The
deviations from the true size distribution was found to be a
function of the magnitude of charge.  This deviation was smaller
if glass fiber substrates were used as impactor collection sur-
faces instead of the metal collection plates alone.  For charge
levels representative of electrostatic precipitators operating
at normal charging conditions (an electric field strength of
400,000 V/m and a current density of 3xlO~4 A/m2), the differ-
ences between the true and measured polydisperse size distri-
butions with glass fiber substrates were small.
                              Vlll

-------
                         ACKNOWLEDGEMENT

     Mr. Don Johnson helped to take much of the data.  We are
also grateful for the suggestions and advice of the project
officer, Mr. D. Bruce Harris.
                                IX

-------
                           SECTION 1.
                          INTRODUCTION

     This report contains a summary of research conducted
on the collection of charged particles in cascade impactors.
The scope of this report does not include an examination of the
theories which would describe a charged particle's behavior as it
travelled through and was finally collected in a cascade impactor.
However, these experiments do indicate the relative deviation from
the true size distribution one might expect to find in cascade
impactor measurements of charged particles, especially those meas-
urements made at the outlet of an industrial electrostatic
precipitator.

     Three sets of experiments were performed:  one using mono-
disperse aerosols sampled by University of Washington Mark III,
Andersen Mark III, and the MRI Model 1502 cascade impactors;
a  second involving  sampling polydisperse aerosols with a
Brink cascade impactor;  and a third  involving  the sampling
of polydisperse aerosols with Andersen Mark III cascade impactors.
Section 2 contains a brief description of the experimental pro-
cedure and results, and  Section 3 contains a number of conclusions
that can be drawn from the data.

-------
                           SECTION 2.
                     PROCEDURES AND RESULTS

MONODISPERSE AEROSOL

Experimental Procedure

     Figure 1 shows the aerosol generating, charging, and sampling
system.  A monodisperse aerosol was generated with a Vibrating
Orifice Aerosol Generator (VOAG).  Ammonium fluorescein droplets
were charged, dried, and then sampled with a cascade impactor.
The collection efficiencies of the various impactor surfaces were
then found by washing each surface individually with an ammonium
hydroxide solution and measuring the optical absorbance of the
wash to find the total mass on each surface.  This procedure is
similar to that described by Gushing, et al.1

     Two particle sizes were used in the experiments, 5.2 and
2.1 ym.  The 5.2 ym particles were charged to two different
levels of charge, 40,000 and 7,000 electronic charges.*  The 2.1
ym particles were charged and sampled at two charge levels also:
2,600  and 600 electronic charges.  In each instance, the lower
charge values were representative of those to be expected at pre-
cipitator outlets.  The higher charge level was chosen in order
*    For large particles  (>2 ym in dia.) and common ESP operat-
ing conditions, the upper limit of charge is that of saturation
in field charging.  For a typical field strength  (4 x 103 volts/
cm), the number of charges per particle, n  , is given by
             np = 834  rp2                                      (1)
where rp is the particle radius in micrometers.  This means  that
for particles with a 5 ym diameter np is approximately 5600  and
for particles with a 2 ym diameter np is approximately 900.   For
smaller particles the  particle charge is somewhat larger  than
Equation  (1) predicts.

-------
                                                   TO FLOWMETER
                                                   AND PUMP
                      IMPACTOR
                      UNDER TEST
                                                GROUNDING CLIP
                                                  TO ELECTROMETER
 TO ELECTRICAL
 MOBILITY ANALYZER
    SHIELDED PLEXIGLAS
    COLUMN
                                COLLECTION
                                ELECTRODE
                 DRYING AND
                 LOFTING AIR
                                                        TO WATER MANOMETER
                                                        PRESSURE ABOVE AMBIENT
                                                          • CHARGING
                                                          VOLTAGE
SIGNAL
GENERATOR
                                                           ABSOLUTE SYRINGE
                                                           FILTER    PUMP
                                                                    „,	DRY
                                                                       AIR

                                                                     3630:013
         \Figure J. Schematic of the charged particle generator and sampling
                   arrangement.

-------
to investigate the potential for more serious effects if higher
than usual charge levels should be encountered.

     The operational procedure in the experiments was to sample
the charged aerosol with and without a charge neutralizer near
the impactor nozzle.  In each test the entire impactor was used,
and the impactors were operated at 0.5 cfm, the recommended flow
rate.  Tests were conducted with and without a grounding wire on
the impactor.

     The impactors investigated in this study were:

     1. Andersen Mark III Stack Sampler (Andersen)
        Andersen 2000, Inc.
        Atlanta, GA 30320
     2. MR! Model 1502 Inertial Cascade Impactor  (MRI)
        Meteorology Research, Inc.
        Altadena, CA 91001
     3. University of Washington Mark III Source Test
          Cascade Impactor
        (U. of W.)
        Pollution Control Systems, Inc.
        Renton, Washington 98055

     In this discussion, the method of charging particles and
measuring the charge is described first, then impactor data and
conclusions are presented.

Particle Charging

     The induction method of charging droplets produced with a
VOAG was introduced by Reischl, et al.2  The region of droplet
formation of a VOAG is depicted in Figure 2, with the entire
aerosol system shown in Figure 1.  Ammonium fluorescein solu-
tion is pumped through an orifice mounted on a piezoelectric cer-
amic.   Due to the applied  sine wave voltage of frequency f,  the

-------
                                            DROPLET
                                            STREAM
CHARGING PLATE
                                     W////W///////////////////A
W//////////////////////////A
                                                             3630-012
   Figure 2. Charged particle generator orifice region with the parallel
            plates for charge measurement.

-------
crystal oscillates, producing periodic perturbations of the
liquid jet formed at the orifice.  This causes the jet to be-
come detached and form droplets at regular intervals.  Droplet
formation occurs at a height L above the orifice.  The solvent,
ammonium hydroxide, evaporates and leaves spherical particles of
the solute, ammonium fluorescein.  The particle radius r  is then
given by

                     /3C'F   \  1/3
                     (JC  sy 1                           (2)
                     \ 4TTf

where F   is the liquid flowrate in ml per second and C1 is the
solute concentration.  In this study f was always set at 65 kHz
and F   at 0.201 ml per minute.

     If the liquid is electrically conducting, then charge can
be induced upon the droplet during its formation by applying an
electric field.  The expression for the induced charge per droplet,
n , given by Reischl, et al2 for a highly conducting liquid is
                               4TTe0r,  r
                       _     _ 	  a  La ,,             / -3 >
                     T> ~~  l-»       ^N    tj^i            \ *~ /
where    r^ = droplet radius =  .         ,
          H = height of top electrode plate above the orifice,
         V  = applied voltage,
          e = elementary unit of charge, 1.602 x 10~19 coulombs,
         e0 = dielectric constant of free space, 8.85 x 10~12
                coulomb /Newton m2
          a = constant determined from the geometry of the orifice
                region = 0.81 for the device shown,
        n   = charge on particle from spraying process, and
         "o
     4ireord = capacitance of a sphere.
     Equation (3)  is based on the model of the spherical droplet
immersed in the electric field with a conducting lead (liquid jet)
connected to it.  This expression has not been verified on an ab-
solute basis; however, the linear dependence of n  upon V  was
substantiated by Reischl, et al.2  Thus the measurement of charge

-------
was necessary to calibrate the instrument for this study although
r^j, L, H, a, and Vc were measured independently.  Comparisons of
the predictions of Equation (3) and the charge measurements are
given in this report.  The most difficult parameter to measure
is L.  A microscope  (125X) allowed visual observation of the jet
disturbance at the point of droplet formation and a rough measure-
ment of L.  However, this measurement cannot be made accurately
without a strobe and camera arrangement, which was unavailable.

      Reischl, et al,2 measured the charge on dry particles; no
measurements were made of the droplet charge.  During a single
run no variation in charge per particle was measured by them, in-
dicating a variation of less than ±2%.  From one run to another,
however, a variation in n   and the coefficient of V  was observed
(see Table I).            °

Charge Measurement Methods

      In this study four different methods were developed to
measure the charge per particle in order to provide independent
checks as well as to establish a convenient procedure.  Initially
two methods were devised to measure charge on the dry particles
of ammonium fluorescein:   (I) sampling with a modified, commer-
cial, electrical mobility analyzer3 (EAA) and  (II) absolute fil-
tering of dried particles on an electrometer electrode.   Diffi-
culties were encountered with the EAA due to wall losses and in-
ability to sample high concentrations of dry particles.  Method
II was time consuming.  Because of these difficulties  in charge
measurement of dry particles, two new methods were then developed
and applied which gave information about droplet charge:   (III)
deflection in a uniform field.with a parallel plate mobility
analyzer, and (IV) collection of undried aerosol droplets on an
electrometer electrode.  Previously, Zung and Snead1* observed that
the droplet charge was altered while drying.  In this  study a set
of measurements were performed using Methods II and IV to eval-
uate that possibility.  These methods are described below, and
results and evaluations are given in the next section.  Method
                                7

-------
   Table   I.  Average Values and Standard Deviations
             of Charging Parameters Observed
                    by Reischl, et al. 2
  Solute
Methylene
  Blue

Potassium
Biphthalate

Sodium
Chloride
Number
of runs

  12


  11
                                n
-1640
± 360

-5540
± 910

-2640
± 160
Coefficient of
Vc in Eg. (2)

    -1830
    ± 670

    -2900
    ± 770

    -1140
    ± 190

-------
IV turned out to be the most suitable one for this project, while
the others provided verification and had desirable possibilities
for any future work performed in this area.

Method I

      Method I employed a Thermo-Systems Model 3030 electrical
aerosol analyzer (EAA) (See Figure 3) .  This particular device
was designed to sample particles with diameters less than one
micron.  Several modifications were made to eliminate wall losses
so that the electrical mobility of larger particles could be
measured.  With this device np is given by

                 np = ZpGirnrp/eC                          (4)

where C is the slip correction factor,
     rp is the particle radius known from the VOAG operating
        conditions ,
      n is the viscosity of air, and
     Zp is the electrical mobility in in/volt-sec.  The mobility
        is determined by

                 zp = K/V = 3.98 x 10~Vv

where K is a constant depending upon the flow and geometry of the
        analyzer, and
      V is the applied voltage in the analyzer at which all par-
        ticles are collected; i.e., removed from the air flow so
        that the analyzer current goes to zero.

Method II
      The charge on dry particles versus charging voltage was
measured with a filter-electrode substituted for the impactor
shown in Figure 1 .  This method employed a filter membrane made
of silver and electrically connected to an electrometer.  The am-
monium fluorescein aerosol was sampled with the filter at the
same flow rate as an impactor.  An electrometer measured the total

-------
                                                                             CONTROL MODULE
                                                                              OATA R£AO COMMAND -
                                                                              CYCLE START COMMAND -
                                                                              CYCLC RlttT COMMAND -
                                                                           AEROSOL FLOWMCTEK READOUT
                                                                            CHAUJtH CUNRfNT KEAOOUT
                                                                            CHAKGCR VOLTAGE RC.AOOUT
                                                                      AUTOMATIC MICH VtXTAOt CCMTWX AMD READOUT
                                                                      ELECTROXETtR (ANiLTZER CURRENT I RCACOUT
                                                                      	TOTAL rLOWMtTW RIAOOUT
    rtmctt ag MRTICH
  t,.CLtCT«0>T«IIC
  td'ACROOYMHtIC OftAt
-• ilTCDHAL
-*   OAT*
—'ACQUISITION
—  tYSTCM
                                                                                     TO VACUUM PUH*
Figure 3.  Flow schematic and electronic block diagram of the
            Electrical Aerosol Analyzer.   From  Sem3.

-------
charge collected by the filter.  After sampling, the total mass
of ammonium fluorescein collected was determined by washing the
filter with an ammonium hydroxide solution and spectrophotometry
of the wash solution.  In this method n  is given by

                                    •
where   Q = charge collected,
      PAF = density of ammonium fluorescein,
       rp = particle radius, and
       M{- = total mass collected.

Method III

      The droplet stream passed between parallel plates as depicted
in Figure 2.  With a high voltage VHV applied between the plates,
the charged droplets could be deflected as indicated.  The equa-
tions of motion for the droplets are:
                                                          (7)

                                                          (8)
                n e V,
            x =
            x
                     HV
                  bd
t-t^ - £  1 - e "v° "B""1       (9)
                   -tn)/m\
where  Ux and Uy are the horizontal and vertical components of
                 the droplet velocity,

        t = time of flight, tg = elapsed time when droplet is at
            the bottom of the plates,
       y  = initial droplet velocity = F/TT (orifice radius)2,
        m = droplet mass = (density of liquid) x (F  /f),
        g = acceleration due to gravity,
                                11

-------
        b = Sirr^n, where n is the viscosity of air and  r^
            is the droplet diameter, and
        d = separation between the plates .
With VHV  =  0  (no deflection) and Uy  = 793 m/sec, it was found
that the droplets rise much higher  (y = 15 cm) than the equations
predict (y =  5 cm) using the viscosity of air  (182.7 x  10  6 poise)
at standard pressure and temperature.  The droplets rose higher
because of the flow of drying and dispersion air in the same di-
rection as the droplet stream.  Therefore, the droplets experi-
enced a lower viscous drag than they would in still air.  An ef-
fective viscosity of 57 x 10~6 poise was determined which gives
the correct height of the flight with VH  = 0.  Using the values
of Y  and YB  (the height of the top and bottom of the plates,
respectively) , t_ - tp. (where tT is the elapsed time when the
droplet is at the top of the plates) can be found from Equation
7.  Then, with VUT7 and d known, n  can be calculated from Equa-
                tl V              . ;P
tion 9.  Since the stream of charged droplets does not spread due
to the applied field VHV, the charge per droplet was uniform.

Method IV
      Another method employed to measure droplet charge collected
the droplets onto an aluminum foil electrometer electrode as they
passed through the charging plate.  The electrode was a hollow
cylinder into which the particles entered through the bottom (see
Figure  1 ) .  The top of the cylinder was covered with fine mesh
metal netting (pore size, 0.16 mm, and 36% open) to allow air to
flow through.  With this arrangement n  is given by

                 np=flT

where I = electrometer current,
      f = frequency of the oscillator driving the VOAG, and
      e = elementary unit of charge.
                                12

-------
The electrode was inserted and charge measured before and after
an impactor sampling run.

      Two variations of this method were attempted in search of a
measurement method which can be used during impactor runs.  An
electrode with the same cylindrical shape was formed entirely of
fine mesh metal netting.  In one variation, a high voltage wire
was placed along the cylinder axis.  This wire, having the same
polarity as the droplets, enhanced the collection efficiency of
the netting.  With the high voltage wire at zero volts a large
portion of the charged aerosol passed through the net electrode
and was sampled by the impactor.  Then, to intermittently measure
the droplet charge, the voltage on the wire was increased to col-
lect the aerosol on the electrode.  In the other variation of
this method, the cylinder, made of metal netting, was used without
the high voltage wire.  The dispersion and drying air flowrate
was lowered to give a high droplet collection efficiency for the
electrode during charge measurement and then was changed to values
which reduced the droplet collection of the electrode so that im-
pactor sampling could proceed.  Although both of these variations
showed promise, time did not permit sufficient refinement to
justify their use in the final analysis of impactor behavior.

Evaluation of the Charged Particle Generator

      The behavior of the charged particle generator was charac-
terized with the four methods of charge measurement described
above.  The results are depicted in Figures 4 and 5 where np is
plotted as a function of charging voltage Vc.  Figure 4 shows the
full range of Vc employed, while Figure 5 gives a better view for
small n .

      The open triangle in Figure 4 depicts the charge measured
on droplets using Method III.  The single point given was deter-
mined from the average of three V -values taken on different

                                 13

-------
    fr
    re
   in*
    o
    X
    Q.
                          I
                   O PARTICLES EAA (METHOD I)
                   • PARTICLES, FILTER (METHOD II)
                   A DROPLETS, PARALLEL PLATES (METHOD III)
                  	DROPLETS, FOIL ELECTRODE ( METHOD IV)
                  	THEORY (EQUATION 3)
T
T
T
                                   I
         I
                         -60      -90      -120
                        CHARGING VOLTAGE, volts
                -150
                -180
Figure 4. Particle charge versus charging voltage for the VOAG charged particle
         generator as determined by Methods l-IV(rp = 2.6 \im and rd = 23 i*m).
                                   14

-------
  50
3
0  30

c
o>

«  20
a
c
   10
  -100
    I         I        I        I        I
 O PARTICLES, EAA (METHOD I)
 • PARTICLES, FILTER (METHOD II) TEST A
— DROPLETS, FOIL ELECTRODE (METHOD IV)
 V DROPLETS, FOIL ELECTRODE (METHOD IV) TEST A
 D DROPLETS, MESH ELECTRODE (METHOD IV)
     IMPACTOR SAMPLING
 • DROPLETS, MESH ELECTRODE (METHOD IV)
     MAXIMUM
     DROPLET COLLECTION
                                                       8
                                                       a
                     I
                     I
                                      I
I
                    -10      -15      -20      -25
                       CHARGING VOLTAGE, volts
                                             -30
                                                               -35
        Figure 5. Particle charge versus charging voltage for the VOAG
                 charged particle generator.
                                 15

-------
occasions with V   =  2000V.  The procedure  in  this measurement
                HV
was  to  adjust the charging voltage V   to deflect the  stream  to
                                    c
the  top of one plate  as  shown  in Figure 2.  This value varied
from run to run by ±6%.  The most likely source of error  in  this
calculation is the estimation  of the effective viscosity  in
Equation  (6).

     The measurements using Method IV  (cylindrical electrode
made of foil) are depicted by  the solid line in Figures 4 and 5.
This line is a least  squares fit to Equation  (3) from 33  measure-
ments obtained in four different runs  over  a period of two weeks.
The  fluctuation of measured n  values  from  this line  were low as
judged  from the coefficient of determination (0.997  where
1.000 is a perfect fit).  The value of npo is  -3050 ±580 and the
coefficient of VG (slope of the line)  is 1280  ±20 volt-1.  These
results are comparable to those of Reischl, et al2 for sodium
chloride (see Table I).

      The measurements on dry particles, denoted by open circles,
were obtained with the modified EAA (Method I).  These results
are more easily viewed in Figure 5 where the lower quarter of
Figure  4 is reproduced.   These measurements with the EAA are con-
sistent with the results of the other methods.  Although the pre-
liminary results shown in Figure 4 and 5 were obtained by this
method, attempts to repeat this measurement were unsuccessful.
The problem was that no stable current of any significance
(>10 "llf amps)  was produced by charged particles passing through
the device.  Apparently, the wall losses were too great and  the
number of particles sampled too small.

      The data obtained by Method II for dry particles is depic-
ted by closed circles in Figure 4.  Although time consuming,
these measurements were important because Zung and Snead1*  observed
a change in the charge on the droplets while drying.  The agree-
ment between measurements using Methods II and IV eliminated
this possibility for our experiments and thus justified the use
of Method IV which determines particle charge from the droplet
charge measurement.
                               16

-------
      The dashed line in Figure 4 gives the predictions of Equa-
tion (3)  where the parameters L, H, and a were measured independ-
ently of n .   Since npo, the charge given to particles in the
spraying process, cannot be predicted, its value in Figure 4 was
estimated by extrapolating the Method IV foil electrode data to
yc = 0.  The charge values for Vc = 0 were too low to be measured
accurately.  The difficulty in  locating the end of the jet with-
out high speed photography caused additional uncertainty in the
measurements.
      The data depicted by squares in Figure 5 are presented to
illustrate data which may be obtained with the two described var-
iations of Method IV.  The data represented by open squares were
measured using the metal-mesh electrode with a high voltage wire
along the electrode axis.  The three values given for np at con-
stant Vc are for different values of this wire voltage:  0, 1000,
and 2000 volts.  The higher n^ values correspond to greater col-
lection efficiencies, none of which are 100%.  Total collection
was obtained by adjusting the dispersion and drying air flowrates
to maximize the electrometer current with the high voltage wire
at zero volts.  This single measurement is represented by the
solid square.  This data, concerning the variations of Method IV,
is given only to illustrate the possibility for its use in further
work if the ability to monitor particle charge is important.

      The solid line representing Method IV data in Figure 5
was used as the calibration curve for the charged particle genera-
tor.  Its agreement with Method II and the results of Reischl,
et al2 justified its use.  In addition to the calibration, fur-
ther measurements of particle charge with Method IV were per-
formed before and after most impactor sampling runs.
                                17

-------
Impactor Data

      Impactor sampling data are given in Figures 6 through 20 in
the form of histograms.  Each figure shows data from one sampling
test; the fraction of total particles collected is given for each
surface in the impactor starting with the nozzle and ending with
the housing of the absolute filter at the exit.  The surfaces are
distinguished according to nozzle, jet plate, substrate, and
housing if separable.  Each of these surfaces was washed separately
and the mass of fluorescein in the wash water determined with a
spectrophotometer.

      The parameters varied in these tests were particle size
(5.2 urn and 2.1 \im) , particle charge, electrical grounding, and
the use of a charge neutralizer.  The sampling times, varied be-
tween one to one-and-a-half hours.  However, the total number of
particles sampled was smaller for highly charged particles, a
result of greater wall losses in the drying chamber.

      All sampling tests were performed at ambient conditions at
a flow rate of 14 £/min.  The impactor nozzles were 1.27 cm
'(0.5 in.)  in diameter giving a gas velocity of 1.84 m/sec.  The
stream velocity where sampling occurred was approximately one
m/sec.  Glass fiber substrates were employed in each test.

MRI Impactor

      Figures 6 through 13 show MRI Model 1502 Cascade Impactor
data (Figures 6-11 for 5.2 ym diameter and 12-13 for 2.1 ym dia-
meter particles).

      Figure 6 shows data for a "reference" run with no particle
charge.  The data shown in Figure 7 is for particles with about
4 x 101* charges and electrically isolated impactor.  The increase
in the number of particles collected on the nozzle and metal jet
plates is quite significant.  Also there was a slight increase in
the number of particles reaching the stages below number three.
                                18

-------
   0.7
   0.6
   0.5
O   0.4
o
0  0.3
z
cc
u.
   0.2
   0.1
                                               MRI
                                               5.8 J£n AEROSOL
                                               vc = ov
                                          CHARGE NEUTRALIZER - ORIFICE    _
                                                J -JET
                                                S-SUBSTRATE
                                                H - HOUSING
                                                f • FINAL
                                                N - NOZZLE
111
J S  H
                   222
                   J  S  H
333
J S H
444
J S H
555
J S  H
666
J S H
7
J
77
S H
                                                                        FH
                              IMPACTOR SURFACE
                                                                    3630-014
         Figure 6.  Control sampling run using MRI-Model 1502 Impactor
                   with no charge.
                                        19

-------
oc
u.
   0.7
   0.6
   0.5
z
o
o
LU
-I
O  0.4
O
LL.
0  0.3

O

o
   0.2
   0.1
      n
                                             MRI
                                             5.8 /Mi AEROSOL
                                 np - 4 x 10

                                 J-JET
                                 S - SUBSTRATE
                                 H - HOUSING
                                 F - FINAL
                                 N - NOZZLE
-n
Jirf
           Ill  222   333   444   555  666   777  FSFH
           JSH  JSH   JSH   JSH   JSH  JSH   JSH
                             IMPACTOR SURFACE
                                                  3630-015
      Figure 7. Sampling charged particles using MRI-Model 1502 Impactor
              with no grounding wire — high particle charge.  (See Figure 6
              for comparison J

-------
   0.7
   0.6
   0.5
O
s
O  0.4
u.
O

O
oc
u.
0.3
   0.2
   0.1
                                          MRI
                                          5.2 /An AEROSOL
                                          V-  -33V
                                          np - 4
                                                 10
                                       J-SET
                                       S - SUBSTRATE
                                       H • HOUSING
                                       F • FINAL
                                       N - NOZZLE
                                       3£ x 1.0 ' PABTICLES
                    "kr
       N   111   222   333
           JSH  JSH   JSH
                           IMPACTOR SURFACE
                                                              FS SH
                                                                3830-0163
     Figure 8. Sampling charged particles using MR I-Model 1502 Impactor
              with grounding wire - high particle charge (See Figure 6 and 7
              for comparison.)
                                        21

-------
u.
O
CC
U.
   0.7
   0.6
   0.5
O
5
O  0.4
   0.3
   0.2
   0.1
                rk_r
      MR)
      5,2 fan AEROSOL
      Vc = - 33V

CHARGE NEUTRALIZES - NOZZLE

       J-JET
      S - SUBSTRATE
      H - HOUSING
      F - FINAL
      N - NOZZLE
      25 x 10s PARTICLES
          111   222   333444   565  666   777  FSSH
          JSH   JSH   JSHJSH   JSH  JSHJSH
                           IMPACTOR SURFACE
                                                               3630-0173
      Figure 9. Sampling charge-neutralized particles using M fit-Model 1502
              Impactor with neutralizer at nozzle - n _ = 4 x
                                   22

-------
   0.7
   0.6
   0.5
O
g 0.4-
   0.3
o
u
oc
u.
   0.2
   0.1-
      MRI
      5.2 pm AEROSOL
      Vc = -33V
CHARGE NEUTRALIZER - NOZZLE

      J-JET
      S- SUBSTRATE
      H - HOUSING
      F - FINAL
      18 x 10s PARTICLES
                               IMPACTOR SURFACE
                                  3630-018
     Figure 10. Sampling charge-neutralized particles using MRI-Model 1502
               Impactor vw'f/j nozzle losses corrected.  (Run data identical
               to Figure 9.)
                                       23

-------
   0.7
   0.6
   0.5
u
UJ
O  0.4
o
u.
O
   0.3
O

U
<
DC
   0.2
                                             MRl
                                             5.2 jUm AEROSOL
                                             np  -7x103
N - NOZZLE
J-JET
S - SUBSTRATE
H - HOUSING
f • FINAL
6.3 x 10s PARTICLES
                                IMPACTOR SURFACE
                     3630-0193
       Figure 11.  Sampling charged particles using MRl-Model 1502
                   Impactor with grounding wire-moderate particle charge.
                   (See Figures 6 and 8 for comparison.)
                                        24

-------
    .7
    .6
    .5
O
O
O

I
O
O
<
cc
    .3
   .2
  MRI
2.1 /An AEROSOL
VC = -4V
CHARGE NEUTRALIZER - NOZZLE
J -JET
S- SUBSTRATE
H - HOUSING
F • FINAL
N- NOZZLE
16 x 106 PARTICLES
                                                                 777
                                                                 J  S H
                                                                        FS  SH
                                IMPACTOR SURFACE
         Figure 12.  Sampling charge-neutralized particles using MRI-Model 1502
                    Impactor with neutralizer at nozzel - np = 2.4 x 10$.  (See
                    Figure 10 for comparison with 5.2 yjrt aerosol data.)
                                           25

-------
    .7
   .6
    .5
U
LLJ
_l
_l   .4
O
U
S- -3
    .2
                             MRI
                             2.1 /Jm AEROSOL
                             Vc = -4V
                             n  = 2.6 x 103
                             J-JET
                             S - SUBSTRATE
                             H - HOUSING
                             F-FINAL
                             N - NOZZLE
                             20 x  106 PARTICLES
                                          J
                                r-n
             J S H
222
J S H
333
J S H
444
J S H
555
J S H
666
JS H
777
J S H
                                                                      FS SH
                               IMPACTOR SURFACE
     Figure 13.  Sampling charged panicles using MRI-Model 1502
                Impactor with grounding wire - moderate particle
                charge.  (See Figure 11 for comparision.)
                                      26

-------
   0.7
   0.6
   0.5
O
8
   0.3
o
   0.2
   0.1
      rfLr
                U. of W.
                5.2 /Urn AEROSOL
                Vc = -33V

           CHARGE NEUTRALIZER - NOZZLE
                N- NOZZLE
                J -JET
                S- SUBSTRATE
                H - HOUSING

                F - FINAL
                15 x 10s PARTICLES
               Hi nrir-i
       N NH 1
           S
2 2
J S
3 3
J S
4 4
J S
5 5
J S
6 6
J S
               FS SH
                    IMPACTOR SURFACE
     Figure  14. Sampling charge-neutralized particles using U. of W. Mark HI
               Impactor with charge neutralizer at nozzle - n-= 4 x 1Cr.
                        27

-------
as
0.4
                     U. of W.
                     5.2 pm AEROSOL
                     VC--33V

                     n   • 4 x 104

                     N- NOZZLE
                     J-JET
                     S- SUBSTRATE
                     H - HOUSING
                     F - FINAL
                     7.2 x 10s PARTICLES
                             J  I .-THr-i
                                  r-n
    N NH 1
         S
2  2
J  S
3 3
J S
4 4
J S
S  5
J  S
6 6
J S
7
J
                                   FS SH
                  IMPACTOR SURFACE
   Figure  15. Sampling charged particles using U. of W. Mark Ml
             Impactor with no grounding wire - high particle charge.
            (See Figure  14 for comparison.)
                          28

-------
    0.7
    0.6
    0.5
O
1-
u
LU
8
o
O
u
<
tr
    0.3
    0.2
    0.1
         U. of W.
         5.2 fjm AEROSOL
         Vc  = -33V
         nD = 4 x 104
         N- NOZZLE
         J - JET
         S- SUBSTRATE
         H - HOUSING
         .F - FINAL
         13 x 10s PARTICLES
                                                   ii-n
        N NH I  22
             S  J  S
3344
J  S  J  S
5  5
J  S
6  6
J  S
77   SH FS
J S
                    IMPACTOR SURFACE
      Figure 16.  Sampling charged particles using U. of W. Mark III
                 Impactor with grounding wire - high particle charge.
                 (See Figures 14 and 15 for comparison.)
                        29

-------
    0.7.
    0.6
    0.5
Z
o
O   0.4
UL
O
DC
    0.3
    0.2
    0.1
                     U. of W.
                     5.2 fJm AEROSOL
                     VC = -8V
                     n  = 7 x 103
                     J-JET
                     S - SUBSTRATE
                     H - HOUSING
                     F - FINAL
                     N - NOZZLE
                     19 x 106 PARTICLES
           c
                                   n
        N H  1
             S
2 2
J S
3 3
J S
4 4
J S
5 5
J S
6 6
J S
7 7
J S
                                   FS FH
                       IMPACTOR SURFACE
       Figure 17. Sampling charged particles using U. of W. Mark III
                  Impactor with grounding wire - moderate particle
                  charge.  (See Figures 14 and  16 for comparison.)
                             30

-------
    0.7
    0.6
    0.5
                                          ANDERSEN
                                          5.2 /Jm AEROSOL
                                          Vc = -8V
                                          CHARGE NEUTRALIZER -   NOZZLE
                                          J - JET
                                          S—SUBSTRATE
                                          H - HOUSING
                                          F - FINAL
                                          N - NOZZLE
                                          42 x 106 PARTICLES
O

g
UJ
O
O
O
u.
O
0.4
    0.3
    0.2
    0.1
          N NH 11
               J  S
                  2  2
                  J  S
3344
J  S   J  S
5  5
J  S
6  6
J  S
7  7
J  S
8  8
 J  S
                                                                FS SH H
                           IMPACTOR SURFACE
           Figure  18.  Sampling charge-neutralized particles using Andersen
                      Mark III Stack Sampler with neutralizer at nozzle -
                      nn = 7 x 103.
                                         31

-------
     .7
     .6
     .5
Z
o
o
u
LL
o
o
     .4
.3
                                           ANDERSEN
                                           &2 /An AEROSOL
                                           VC=-8V
                                           n  =7x 103
                                           J-JET
                                           S - SUBSTRATE
                                           H - HOUSING
                                           F - FINAL
                                           N - NOZZLE
                                           16.2 x 106 PARTICLES
         rfl
         N NH  1  1
               J  S
                2  2
                J  S
3  3
J  S
4  4
J  S
S  S
J  S
6  6
J  S
7  7
J  S
8  8
J  S
                                                              FS  SH H
                             IMPACTOR SURFACE
       Figure 19.  Sampling charged particles using Andersen Mark III
                  Stack Sampler with grounding wire - moderate
                  particle charge.  (See Figure  18 for comparison.)
                                       32

-------
    .7
    .6
O
U
O
U
O
z
O
U
<
oc
    .4
.3
                                                ANDERSEN
                                                2.1  /Ltm AEROSOL
                                                VC = 4V
                                                n  =600
                                                 P
                                                J - JET
                                                S - SUBSTRATE
                                                H - HOUSING
                                                F - FINAL
                                                N - NOZZLE
                                                20 x 106  PARTICLES
          N NH 1  1
               J  S
                  2233
                  J  S  J  S
4455
J  S  J  S
6677
J  S J  S
8  8  FS  SH  H
J  S
                            IMPACTOR SURFACE
        Figure 20.  Sampling charged particles using Andersen Mark ///
                   Stack Sampler with grounding wire - moderate
                   particle charge.
                                     33

-------
The data represented in Figure 8 were taken with the impactor
electrically grounded, otherwise, the operating conditions were
identical to those listed in Figure 7.  The collection efficiencies
of the nozzle and stage one jet plate (U) are larger.  Only the
nozzle and 1J of the MRI were grounded from the outside due to an
anodized coating on each stage housing.  In these measurements a
grounded wire was wrapped around the threads between U and its
substrate housing.  The resistances between ground and IS and
between ground and 2J were measured to be of the order of 109 ohms
with the grounding wire and generally about 15% higher without
the grounding wire.

      Figure 9 presents a sampling run with identical operating
conditions to that of Figure 8 except that a charge neutralizer
was mounted upstream of the impactor nozzle.  The neutralizer is
a polonium 210 strip made by Nuclear Products Company.  For use
with an impactor it was bent to form a ring and placed at the
entrance to the nozzle.  In Figure 9 the high collection on the
nozzle was thought to be a result of turbulence induced by the
charge neutralizer.  Collection efficiencies were calculated ex-
cluding the nozzle losses.  These values are given in Figure 10.
Comparing this figure to Figure 6 shows that indeed the effect
of particle charge was eliminated by the ion source neutralizer.
After this run the ionizer mount was changed to disturb the flow less,

      Figures 7  and 8  showed an  effect  of  high  particle charge  upon
impactor  behavior  where the  charge  level was  higher  than encoun-
tered in most effluent streams.   In the run depicted in Figure 11
the charge per particle was approximately 7 x 103 elementary
charges,  approximately the same as that expected in an industrial
electrostatic precipitator.   Collection by the stage 3 substrate
(3S)  was nearly twice that with 4x10^ charges/particle (Figure 8)
but was still 15% less than with no charge (Figure 6).  These data
indicate that in sampling 5.2 pm diameter particles exiting a pre-
cipitator, approximately 25% of the particles collect on surfaces

                                34

-------
 due  to their charge and not size.  This conclusion is based oh
 the  assumption that all particles exiting a precipitator have
 the  level of charge predicted by Equation (1).

      Figures 12 and 13 show the effect of moderate charge on
 2.1 ym diameter particles.  3J and 4J again showed an increased
 collection when charged particles are sampled.   However, the
 effect for 2.1 ym particles charged to saturation was much smaller
 than with the larger particles at saturation.

 University of Washington Impactor

      Figures 14 through 16 show data obtained from the University
 of Washington Mark III Impactor in sampling 5.2 ym diameter par-
 ticles with 4 x 10"* charges per particle.   Figure 14 shows results
with the charge neutralizer and Figures 15 and 16 without the
 neutralizer.  The impactor was grounded electrically in the run
 depicted in Figure 16 and was not in Figure 15.  Particle charge
 affected the collection efficiency of the  various surfaces.  The
 behavior of the U. of W. impactor was similar to that of the MRI
 impactor except that the use of a grounding lead produced a
 greater change in the deposition of charged particles for the U.
of W. impactor.

      The results obtained with the moderate charge level, 7 x 103
 elementary charges, on 5.2 ym particles are shown in Figure 17.
 Comparison of these data with those in Figure 14 showed no signi-
 ficant effect on  surface collection efficiency due to the  in-
 creased charge level.

 Andersen Impactor

      Figures 18 and 19 show data obtained with the Andersen
 Mark III Stack Sampler when sampling 5.2 ym diameter particles
 with neutralized and moderate charge levels, respectively.
                                35

-------
Comparison of these two figures showed a significant difference
in the fraction of particles collected upstream of stages three
and four.  As with the MRI impactor about 25 percent of the
5.2 ym particles with moderate charge collected on wall and jet
surfaces because of their charge rather than size.

     Figure 20 shows sampling data for particles with 2.1 ym di-
ameters and moderate charge, 600 elementary charges.  The control
run for this data, that is, the sampling of neutral particles
with 2.1 ym diameters, had to be discarded because of an apparent
syringe pump malfunction.  However, the agreement of the data in
Figure 20 with impactor theory for neutral particles and a previ-
ous sampling study1 indicated  that  no  significant  effect  of  charge
was present in sampling the 2.1 ym particles.

     In the following section the data shown in Figures 6 through
20 are presented in terms of collection efficiency per stage
versus Stokes number for different levels of charge.

Effect of Charge on Efficiency versus /JfT

     Figures 6 through 20 depict impactor behavior under various
circumstances involving particle charge.  As discussed above, a
high charge affected the deposition of particles.  However, it
is not clear from those figures to what extent charge altered
the calculated size distribution in a collected sample.

     The size distribution inferred from impactor data is based
upon stage collection efficiency E. as a function of the square
root of inertial impaction parameter, /ijj\
                                36

-------
               t|> = D*Cp V /18nD.                      (11)

where D  = particle diameter  (can) ,
       C = Cunningham slip factor,
      DJ = jet diameter (cm),
      Pp = particle density (gm/cm3),
       n = gas viscosity  (poise), and
      VQ = jet velocity (cm/sec).

A representative example of efficiency versus /if is given  as  a
solid line in Figure  21.  The use of E^ versus /if gives  the de-
sired stage efficiency as a function of particle size  for  a
range of sampling conditions. In practice a stage is calibrated
by experimentally determining efficiency versus /if for a partic-
ular substrate material to obtain /ifTo"; that is the value  of  /if
at 50% collection efficiency.  Then, for particular sampling  con-
ditions the effective stage cut diameter D5o is obtained  from  /\jj 5 o.
It is assumed that all particles caught by an impactor stage  are
those particles having diameters equal to or greater than  the DSQ
of that stage, but less than the cut point of the preceding stage.
Therefore, the effect of particle charge upon the calculated  size
distribution depends on its effect upon /ifTV.  More sophisticated
deconvolution methods for impactor data which use  the entire effi-
ciency curves have been proposed5'6'7, but use of these techniques
is limited to low-noise sampling data.

     In order to completely characterize the effect of particle
charge, the function E^(/if,np)  [that is, efficiency expressed as
a function of the square root of the Stokes number and charge
 np  for stage i] is required.  For neutral aerosols,  E^(/if,0)
has been determined in an extensive calibration study1 for each
stage of the Andersen, MRI, and Univ. of Washington impactors at
                               37

-------
U)
00
              100


               90


               80
            U
            \L
            u.
            uj  so
            3  30


               20


               10


                0
T     \     i     i     rrrm     i     i     i     i      i     i     i     i     i     i     i
                         ^—.       *^ _
                 LJ    1     I    I     I      L    I     I     I     I     I     I    I     I
                 0    0.1    0.2    0.3   0.4    0.5   0.6   0.7   0.8   0.9   1.0   1.1   1.2    1.3   1.4   1.5    1.6   1.7    1.8   1.9
                                   Figure 21.  Reference curve (solid) giving collection efficiency as a function
                                              of \/W for the Andersen Stack Sampler and neutral particles.  The
                                              measurements ("O") of this study for neutral particles are plotted
                                              by the procedure given in the text.  The dashed lines depict
                                              the envelope of all stage efficiency curves superimposed.

-------
the sampling conditions used here, except that grease substrates
were used with the latter two.  To a first approximation,
E^(/ijJ,0) has the same form for all stages with the same substrate
material and impactor.  The reference curve which we denote by
E(/ij;,0) in Figure 21 was drawn by superimposing the stage cali-
bration curves of the Andersen from the previous study. 1  Since
the major differences in E. (/f,0) from one stage to another is
not in the shape of the curve, but /tyso , for the purpose of this
analysis the calibration curves were normalized so that /ijJTo" of
each coincided at 0.4.  The dashed lines in Figure 27 define the
envelope of those curves.

     The collection efficiency of each stage was calculated for
each run from the data given in Figures 6 through 20.  These
efficiencies were "corrected for wall losses" by combining the
particles collected on the various surfaces according to Table
II.  These groupings are based upon the procedure used in field
testing and the laboratory observations discussed above concern-
ing deposition on jet plates.  In the field, particulate mass on
the top of jet plates is combined with the preceding stage, and
that on the bottom is combined with the stage below the jet plate,
The results for sampling neutral particles with the Andersen are
given in Figure 21 as points.  For comparison of these data to
our reference curve, each measured efficiency is normalized and
plotted as described above.  The deviations of the data points
from the reference curve in Figure 2 1 is thus a measure of the
degree to which impactor calibration data for neutral particles
may vary.  In Figure 22 the same procedure was followed to relate
the measured efficiencies of the Andersen stages with charged
particles to its generalized efficiency curve for neutral parti-
cles.  The dashed curve is the resulting E(/iJ7,np) for moderate
np.  The effect of charge was to reduce the sharpness (slope) of
the curve at /ty values below the peak while not changing /\J7F7
significantly.  At /ij; values above the peak the effect of charge
was to reduce the efficiency in a manner similar to particle
bounce effects.

                                39

-------
Table  II .   Grouping of Surfaces for the
       Efficiency of Each Impactor Stage.

Stage
1




2


3


4


5


6


7




8



Andersen
surface
N
NH
1J
IS
2J
2S
3J

2S
4J

4S
5J

5S
6J

6S .
7J

7S
8J



8S
FS
SH
H
MRI
surface
N
1J
IS
1H

2J
2S
2H
3J
3S
3H
4J
4S
4H
5J
5S
5H
6J
6S
6H
7J
7S
7H
FS
SH




U. of W.
surface
N
NH
IS


2J
2S

3J
3S

4J
4S

5J
5S

6J
6S

7J
7S
FS
SH





                   40

-------
   100


    90


    80
    "
u.
u.
    50
    40
ui
_j

8   30


    20


    10


     0
I     I      I     I     I     I     I     I     1      I     I      I     I      I      I      I     I      I
            I     I     I     I     I      I     I      I     I     1     1      I      I      1     I     I
      0     0.1    0.2    0.3    0.4   0.5   0.6    0.7   0.8    0.9    t.O    1.1    1.2    1.3    1.4    1.5    1.6   1.7   1.8   1.9
                           Figure 22.   Efficiency versus V    of stages in Andersen Stack Sampler for
                                       neutral particles (solid curve) and moderate!/ charged particles
                                       (dashed curve).  The measurements of this study for moderately
                                       charged particles ("t>" - 5.2 urn and "9" - 2.1 \im diameter) are
                                       plotted by the procedure given in the text.

-------
     Figures 23 and 24, and Figures 25 and 26 present information
about the MRI and Univ. of Washington impactors analogous to that
in Figures 21 and 22 for the Andersen.  Stage efficiency curves
of the previous study1 for neutral particles were obtained using
grease rather than glass fiber substrates; therefore some adjust-
ments in the reference curves had to be made based upon the effi-
ciency data of the present study and comparisons of available
efficiencies of another impactor with these two substrates to
determine the differences.  The resulting reference curve of the
MRI impactor is shown in Figure 23.  For /if values up to 0.5,
following the same procedure as in Figure 21, the reference curve
of Figure 23 was obtained using MRI efficiency curves for grease
substrates.  The use of these was justified because, in this re-
gion, the efficiencies for neutral particles measured in this
study did not substantially deviate from this reference curve.

     In the region of /if values greater than 0.5, glass fiber sub-
strates produced substantial bounce, meaning that efficiency drop-
ped to much less than 100%.  This is indicated by the low measured
efficiencies shown in Figure 23.  Therefore in this region a ref-
erence curve was determined from experimental data for neutral
particles taken in the course of this study.

     Figure 24 shows the effect of high and moderate particle
charge upon the MRI stage efficiency versus /if.  For moderate
charge the effects were similar to those for the Andersen.  At
the higher charge level, however, the changes in E(/if,np) were
quite large at all /if values.

     The effect of particle charge upon Univ. of Washington stage
efficiency, shown in Figure 26, was small for moderate charge and
substantial for high charge.  In the "bounce" region, moderate
charge appears to increase the efficiency substantially, thus re-
ducing the effects of bounce.  However, data obtained for /if
values to the right of the peak in efficiency are inherently
                                42

-------
OJ
   100


    90


    80


*   70

£
UJ   gQ
o
II
iu   SO


|   *
u
_i

O   30


    20


    10


    0
                                                I     I     I     I     I     I     I     I     I     I     I      I      I      I
           0.1   0.2   0.3   0.4
                                               0.5  0.6   a?    0.8   0.9   1.0   1.1    1.2    1.3    1.4    1.5    1.6    1.7    1.8    1.9
                                    Figure 23.   Reference curve Isolid) giving collection efficiency as a function
                                                of V^ for the MRI-Model 1502 Impactor and neutral particles.
                                                The measurements ("&')  of this study for neutral particles are
                                                plotted by the procedure  given in the text.  The dashed lines
                                                depict   the envelope of all stage efficiency curves superimposed.

-------
  100
   00
   80
HI
o  60
   50
IL
IU
I
S  40

S  30

   20

   10

    9
          T     I     I     I     I     I     I
                                        I     l     I     I      i     I     l     I      I     I
I  ~t'A     I     I    I     I      I     I     I     I      I      I"  T  -1-
           0.1   0.2   0.3   0.4    0.5   0.6    0.7   0.8   0.9   1.0    1.1   1.2   1.3    1.4    1.5    1.6   1.7    1.8   1.9
                         Figure 24.   Efficiency versus \f& of stages in MR/-Model 1502 Impactor for
                                     neutral (solid curve), moderately charged (dashed curve), and highly
                                     charged (	) particles.   The measurements of this study for
                                     moderately charged ("&' - 5.2 urn and "A" - 2.1 \im diameters)
                                     and highly charged 5.2 urn diameter ("&" - grounded and "^fr -
                                     not grounded) particles are plotted by the procedure given in the
                                     text.

-------
UI
                 100


                  90


                  80


                  70
               r  eo
               O
ui  SO

O
P  40
O  w
ui
-i

g  30


   20


   10


    ff
           i
           0.1
                                  -^X    i    II
I     I     I     I     I      I      I      I     I     I
                                0.2    0.3   0.4   0.5   0.6   0.7   0.8   0.9   1.0    1.1    1.2   1.3    1.4   1.5   1.6    1.7    1.8    1.9
                                      Figure 25.  Reference curve (solid) giving collection efficiency as a function
                                                  of the \/3? for the U. of W.  Mark III Impactor and neutral particles.
                                                  The measurements ("O') of this study for neutral particles are
                                                  plotted by the procedure given in the text.  The dashed lines
                                                  depicted the envelope of all stage efficiency curves superimposed.

-------
  100


  90


  80


* 70
•a  so
u
   50


   40
ui

830


   20


   10
J,n-K  I                                                               I
           I      1-^1     I     I     I     I     I     I      I     I     I     I     II     III
     0    0.1    0.2    0.3   0.4   0.5   0.6   0.7   0.8   0.9    1.0   1.1    1.2   1.3    1.4   1.5   1.6   1.7    1.8   1.9
                     Figure 26.  Efficiency versus V^F of stages in the U. of W. Mark III Impactor
                                 for neutral (solid curve), moderately charged (dashed curve), and
                                 highly charged (—•) particles.  The measurements of this study
                                 for 5.2 psn diameter particles, moderately charged f "D'V and
                                 highly charged ("Q" - grounded and "0" not grounded) are
                                 plotted by the procedure given in the text.

-------
 subject to fluctuation because small numbers of particles are in-
 volved.  Therefore, exceptions to the overall trends in this re-
 gion are questionable.

     Figures 22, 24, and 25 show two obvious effects of particle
 charge.  At low values of /ijb the stage efficiency was increased
 due to the collection of particles on metal jet plates.  At high
 values of /ij;, past the peak, the efficiency was reduced as if
 repulsive forces of charge were significant.  Charge produced
 very small changes in D5 o.

 Particle Deposition Patterns

     The deposition patterns of fluorescein particles on impactor
 surfaces were observed by spraying them with a water mist and ex-
 posing them to a fluorescent light.

     Photographs in Figures 27 and 28 show representative deposi-
 tion patterns of particles on the top and bottom of jet plates
 and on substrates in the MRI impactor.  Most particles deposited
 on the jet plates were on the bottom or downstream side of the
 plate as shown in these photographs.  This pattern was visible
 with both charged and uncharged particles, but more particles
 were deposited on the bottom of the plates in the charged case.
 Apparently, particles passed in the vicinity of these surfaces
 whether charged or not.  As expected, the Univ. of Washington
 impactor, with a similar geometry, had deposition patterns sim-
 ilar to those on the MRI.

     Collection efficiencies of jet plates in the Andersen impac-
 tor were also slightly higher for charged particles than for neu-
 tral ones.   Also, substantial numbers of charged particles were
observed to be deposited on the upper surface around jets as well
 as on the bottom surface.  The photographs in Figures 29-31 il-
 lustrate the Andersen deposition data.
                                47

-------
FRONT OF SECOND JET PLATE
                                                 BACK OF SECOND JET PLATE
                            SECOND SUBSTRATE
                                                                  3630-024
    Figure 27.  Photographs of deposition patterns of ammonium fluorescein
                particles in MR I-Model 1502 Impactor.  Particle diameter was
                5.2 p.m.  Sampling test data are in Figure 8.
                                   48

-------
FRONT OF THIRD JET PLATE
                                               BACK OF THIRD JET PLATE
                         THIRD SUBSTRATE
                                                                 3630-023
Figure 28.  Photographs of deposition patterns of ammonium fluorescein
           particles in  MR I-Model 1502 Impactor.   Particle diameter was
           5.2 nm.  Samp/ing test data are in Figure 8.
                                  49

-------
 FRONT OF FIRST JET PLATE
                                                    BACK OF FIRST JET PLATE
                             FIRST SUBSTRATE
Figure 29.  Photographs of deposition patterns of ammonium fluorescein
           particles in Andersen Mark III Stack Sampler.  Particle diameter
           was 5.2 urn.  Samp/ing test data are in Figure 19.
                                   50

-------
FRONT OF SECOND JET PLATE
                                                  BACK OF SECOND JET PLATE
                           SECOND SUBSTRATE
Figure 30.   Photographs of deposition patterns of ammonium fluorescein
            particles in Andersen Mark III Stack Sampler.  Particle diameter
            was 5.2 fj.m.  Sampling test data are in Figure  19.
                                   51

-------
      FRONT OF THIRD JET PLATE
                                                        BACK OF THIRD JET PLATE
                                       ' ' :

                                  THIRD SUBSTRATE
Figure 31.   Photographs of deposition patterns of ammonium fluorescein
            particles in Andersen Mark III Stack Sampler.  Particle diameter
            was 5.2 /urn. Sampling test data are in Figure  19.
                                      52

-------
Nonconducting Jet Plates

     In an effort to study the effect that jet plates made from
an insulated material would have on deposition due to particulate
charge, plates 2J, 3J, and 4J were fabricated of plexiglass for
the MRI impactor in an attempt to eliminate the effect.  The data
for this experiment are shown in Figure 32.  Comparison with
Figures 6 and 11 revealed two significant differences.  First,
the stage cut points are shifted to larger sizes.  The origin of
this behavior is not known (measurements ruled out changes in jet
hole sizes as a source of discrepancy).  Secondly, the collection
efficiencies of 2J, 3J, and 4J were reduced, but the efficiencies
of the nozzle, 1J, IS, 1H, and especially 2S were increased.  The
reason the efficiency of 2S is so great might be attributed to
the presence of static charges on 2J.  It is unknown why the ef-
ficiencies of the nozzle, 1J, IS, and 1H were higher.  Obviously,
the use of plexiglass stage elements in impactors that sample
charged particles is undesirable.

POLYDISPERSE AEROSOL SAMPLED WITH A BRINK IMPACTOR

     Another set of experiments was done to investigate the pos-
sibility of charge effects at precipitator outlets.  Figure 33
shows a schematic of the setup of the impactor and particle
counter used in these experiments.  Two stages of a Brink cascade
impactor were employed with  a removable polonium  210  a-sourcre
attached to the nozzle of the impactor.  A flow rate of 2.83 £/min
(0.1 acfm) was maintained through the impactor using the pump in
the Climet Particle Analyzer, an optical particle counter, which
operates at 7.1 5,/min  (0.25 acfm).  Make-up air to the particle
counter was supplied by a controlled, filtered, air supply.  The
aerosol exiting the impactor was charge neutralized with a second
polonium 210 a-source to minimize sample and instrumental losses
due to electrostatic forces between the impactor and particle
counter.
                                53

-------
     0.7
     0.6
     0.5
u
UJ
O
O
LL
O
oc
     0.4
     0.3
     0.2
     0.1
              ffl
MRI
5.2 (Jm AEROSOL
Vc = - 8V
«p = 7x10*
N - NOZZLE
J-JET
S - SUBSTRATE
H - HOUSING
F - FINAL
38 x 10s PARTICLES
               111   222  333444555   666777  FSSH
               J  S  H  J   SHJSHJSHJSH   JSHJSH

                                IMPACTORSURFACE
           Figure 32.  Sampling charged particles using MRI-Model 1502
                      Impactor with plexiglass jet plates, 2J, 3J, and 4J.
                      (Other conditions are the  same as the test depicted
                      in Figure 11.)
                                          54

-------
IMPACTION PLATES
                                                         REMOVABLE POLONIUM
                                                              a-SOURCE
                                            3.49 urn
                                           2.03 nm D50
                                              POLONIUM a-SOURCE
                 VALVE
  ABSOLUTE
  FILTER
                             TO CLIMET
                             PARTICLE COUNTER
         Figure 33. Sampling system used to sample charged polydisperse
                  aerosols with a Brink cascade impactor.
                                   55

-------
     The aerosol was produced by a "hobby" paint sprayer producing
a polydisperse DOP aerosol which was sampled by the impactor at
the outlet of a model wet-wall precipitator.

     Four different configurations of the impactor were tested:
(1) with glass-fiber filter substrates,  (2) with bare metal
plates, (3) with no plates (in order to study the effect of wall
loss), and (4) extractive sampling with a 91 cm (3 ft), 6.35 mm
(114 in) I.D. copper probe.  Each condition was tested with and
without the charge neutralizer attached to the impactor nozzle.

     The data from this investigation are presented in Figures
34, 35, and 36.  Figure 34 indicates that there is no appreciable
effect of charge neutralization when glass-fiber filter substrates
are used.   Figure 35 indicates that there is some effect on the
number of particles exiting the impactor with and without charge
neutralization when bare metal plates are used.  Figure 36 indi-
cates that there may be substantial wall losses within the im-
pactor which can be attributed to the electrostatic effects.

     The results are plotted another way in Figure 37.  The data
are normalized to the concentration measured with the charge neu-
tralizer in place.  Curves are shown for the probe alone, for the
impactor with glass-fiber substrates (no probe), and for the im-
pactor operated with bare metal collection plates (no probe).
Losses in the probe and with bare substrates are seen to be rather
large.  For the tests made using glass-fiber substrates there was
no appreciable difference in the concentrations measured with and
without the charge neutralizer.

     The data from this study indicate that with a conducting
substrate, such as bare metal, particle charge can indeed cause
high increases in collection efficiency of an upper stage for
small particles.  However, with a nonconducting substrate little
change was produced by particle charge.
                              56

-------
s
at
•a
Si
3
'•5

c
re
re
£
o>
o
O
o
UI

o
    103
    102
    101
            O WITHOUT AEROSOL CHARGE NEUTRALIZATION PRIOR TO NOZZLE

            a WITH AEROSOL CHARGE NEUTRALIZATION PRIOR TO NOZZLE
      1.3
                1.1
0.85
0.65
0.51
0.39
0.25
                            PARTICLE DIAMETER, micrometers

           Figure 34. Particle concentration at the outlet of a Brink cascade
                    impactor with glass-fiber filter substrates as collecting
                    surfaces for charged and neutralized conditions.  Error
                    bars indicate one standard deviation.

                                         57

-------
    10*
•o
0)
s
I   103

I
0)
O
O
UJ
    101
            O WITHOUT AEROSOL CHARGE NEUTRALIZATION PRIOR TO NOZZLE

            O WITH AEROSOL CHARGE NEUTRALIZATION PRIOR TO NOZZLE
                                                              I
                  s
                 I
       1.3        1.1        0.85         0.65        0.51

                            PARTICLE DIAMETER, micrometers
0.39
0.25
            Figure 35. Particle concentration at the outlet of a Brink cascade
                     impactor with bare metal plates as collection surfaces
                     for charged and neutralized conditions.  Error bars
                     indicate one standard deviation.
                                        58

-------
    104
    103
S
0>

CO
•Q
8
3
'•S

c
to
0>
a>
o

O
111
o      £
o
o
p     0
cc   io2
    101
                  I          I            I          1           I            I

              O WITHOUT AEROSOL CHARGE NEUTRALIZATION PRIOR TO NOZZLE
              0 WITH AEROSOL CHARGE NEUTRALIZATION PRIOR TO NOZZLE
      1.3
1.1        0.85         0.65        0.51

            PARTICLE DIAMETER, micrometers
0.39
0.25
          Figure 36. Particle concentration at the outlet of a Brink cascade
                    impactor operated without particle collection surfaces
                    (wall loss study) for charged and neutralized conditions.
                    Error bars indicate  one standard deviation.
                                         59

-------
ui
N
    1*
O
b
O
o
o
g

o
cc
z
o
o
u
    0.5
                   I             I            1
               O   .   • EXIT OF TWO STAGE IMPACTOR,
                       GLASS FIBER SUBSTRATES
               A —.—. EXIT OF TWO STAGE IMPACTOR,
                       METAL SUBSTRATES
               D	- 1 METER, 6 mm I.D. METAL PROBE
                   \^
                  0.5           1.0           1.5


                      PARTICLE DIAMETER, ;um
                                                          2.0
     Figure 37. Particle losses in a Brink cascade impactor due to
                electrostatic forces.
                             60

-------
POLYDISPERSE AEROSOL SAMPLED WITH AN ANDERSEN IMPACTOR

     Another set of experiments was performed using two commer-
cial 'cascade impactors to sample simultaneously a polydisperse
ammonium fluorescein aerosol.  One impactor sampled a charged
aerosol while the other sampled a neutralized aerosol.

Experimental Apparatus^

    A schematic of the experimental apparatus is shown in Fig-
ure 38.  A polydisperse aerosol of ammonium fluorescein was
generated by an atomizer (RETEC X701N) and fed into a disper-
sion chamber.  A compressed air line supplied dry, filtered
air for dispersion, drying and atomization.  Two pairs of op-
posed air jets perpendicular to the direction of the aerosol
flow dispersed the air and initiated drying in the dispersion
chamber.  The aerosol flowed vertically through a 14 cm inner
diameter (ID) plexiglass cylinder which functioned as a mixing
and drying column, and two diffusional dryers.  The diffusional
dryers consisted of 90 mm ID, plexiglass cylinders with 30 mm
ID, coaxial, cylindrical screens shown in Figure 38 by dotted
lines.  The screen cylinders were surrounded by 6-16 mesh
silica gel desiccant. The particle charger as described by
Bush and Smith8 consisted of a disk-cylinder geometry.  The
disk electrode was 28.6 mm in diameter and 1.3 mm thick and was
surrounded by an electrically grounded, 7.37 cm ID cylinder.
Clear, plastic, 16-20 mm ID tubing connected the 25 mm ID, 90 mm
long, plexiglass chambers to the charger.  One chamber contained
four polonium 210 strips (Staticmaster Model 3C500) bent into
25 mm-diameter circles while the other was empty.  Each ion pre-
cipitator was 15 cm long and was constructed of a 9.5 mm outer
diameter (OD) stainless steel tube electrically isolated and co-
axial with a 17 mm ID stainless steel tube.  The outer tube was
electrically grounded and the inner tube was connected to a high
voltage power supply.  The ends of the inner tube were sealed
with a bullet-shaped plug to minimize particle loss due to im-
paction and turbulence.  Cascade impactors (Andersen Mark III
Stack Sampler) with glass fiber substrates were connected to the
                             61

-------
          DISPERSION
          CHAMBER
                 DILUTION ROTAMETER

i                     VALVE     DRYER     COMPRESSED AIRLINE
                   -®-
                                          <=p
                                     REGULATOR
                                          ABSOLUTE
                                    VALVE FILTER

                                ATOMIZER ROTAMETER
                     ATOMIZER
             MIXING-DRYING  COLUMN
                                            ELECTROMETERS
                                                   CASCADE IMPACTORS

                                                           METERING ORIFICES
             DIFFUSIONAL DRYERS
                          ELECTROSTATIC SHIELD
PO210 CHAMBERS

            ION PRECIPITATORS
                          INSULATED
                          FROM GROUND
HIGH VOLTAGE
POWER SUPPLY
                                        MERCURY MANOMETERS
                                                     WATER MANOMETERS

   Figure 38.  Polydisperse aerosol generating, charging, and sampling system.

                                 62

-------
ion precipitators with 9-im ID, polyethylene tubing 23 mm long.
Two calibrated metering orifices measured the flow exiting the
impactors.  The impactors were electrostatically shielded with
a small mesh brass .screen, and grounded through electrometers
(Keithley No. 612C and No. 600B) measured the charge on the
aerosol entering the impactors  (through the Faraday cup principle)

Experimental Procedure

     For each test, a polydisperse ammonium fluorescein aerosol
was generated, dried, and passed through the system at a flowrate
of 14.2 liters/min at the inlet of each impactor.  Electrometers
measured the charge collected on each impactor—a relatively low
measurement for the neutralized aerosol branch as compared to the
charged aerosol branch provided proof of sufficient neutraliza-
tion.  All tests were performed at ambient temperature with the
pressure at the inlet to the impactors about 0.1 atm above am-
bient.  After each test, the ammonium fluorescein particles col-
lected on each impactor stage were dissolved in an ammonium hy-
droxide solution and analyzed in a spectrometer to determine the
mass.  Although no mass was noticed on the jet plates for most
of the tests, each plate was rinsed and grouped with the sub-
strate below it for the mass determination.  The data from each
impactor test was reduced using the procedures described by
McCain, et al.*   This data is presented in the form of computer
plots of the differential mass concentration vs. particle diameter
in Figures 39-44.  Figures 39 and 40 show actual mass concentra-
tions in mg/m3 and Figures 41-44 list average, normalized mass
concentrations in arbitrary units.  Errors bars indicate 90 per-
cent confidence limits.

System Equivalency

     Initially, tests were performed to ascertain that the mass
concentrations and the particle size distributions entering both
impactors were identical.  In these tests both impactors sampled

                               63

-------
the same aerosol from alternate branches.  In the first four
tests the charger was not operating and the impactors sampled the
uncharged aerosol.  In the next three tests, the charger was
operated and the aerosol sampled was charged in an electric
field of 6.9 x 105 V/m   Impactors sampled alternately from each
branch to determine if there were any differences in the per-
formances of the impactors.

Charged and Neutralized Particle Sampling

     These tests were similar to the charged aerosol tests de-
scribed above, but with four polonium 210 sources in one branch
of the system.  It was experimentally determined that four
sources were sufficient to neutralize the charged particles exit-
ing the charger.

     Although the effectiveness of the charge neutralizers was
demonstrated, attempts to measure the charge level of the charged,
aerosol fraction accurately were unsuccessful.   The current level
was very low and stray currents in the charge collection/measure-
ment systems resulted in a low signal-to-noise ratio.  The ion
precipitators were used to remove ion charges produced by the
polonium 210 that might influence the particulate charge measure-
ments.  Electric fields of 13 and 2.7 kV/m were used and produced
identical results.  Particulate loss in the branch upstream of
an impactor was concentrated in the ion precipitators and amounted
to no more than 5 percent of the total mass of the particles enter-
ing the branch, being greater for the charged particles.

     The tests in which charged and neutralized aerosols were
sampled simultaneously can be divided into four conditions.  These
conditions are identified by the strength of the electric field
within which the particles were charged and by the mass median
diameter of the aerosol as given in Table III.
                               64

-------
      Table III. Summary of Polydisperse Aerosol Test Conditions
Condition
Tests
Mass Median
Diameter, ym
Electric Field
x 106 V/m
A
2-5

1.27

6.1
B
7-8

1.02

6.1
C
9-12

.692

6.3
D
15-18

1.01

8.3
An approximate calculation of the total charge accumulated by
an aerosol particle of radius r in an electric field E is given
by Pontius et al.10 to be.
     n  =
where
                      K
                      e
and
     t = particle residence time (sec),
     k = particle dielectric constant,
     e = electron charge (coul),
    EO = permittivity of free space (fd/m),
     y = ion mobility (m2/V*sec),
     T = temperature (°K)
     K = Boltzmann's constant  (j/°K),  and
     v = mean thermal ion speed  (m/sec).
     This calculation for five particle sizes indicates the rela-
tive difference between "moderate" and "high" charge levels, and
is given for each test condition in Table IV.  The data is pre-
sented as a ratio of the charge calculated for the conditions of
each test divided by the charge calculated for conditions repre-
sentative of those found at industrial electrostatic precipitators.
Conditions A, B, and C are considered moderate charge conditions
and condition D is considered a hiah charge condition.
                                65

-------
103,:
101,:
10'
             PARTICLE SIZE DISTRIBUTION
             BRANCH 1 -- ©      BRANCH 2 - X
                       *   x
10°: f     *
         H	1  I I Mill	1—I  I MMH
                                   -f—I  I Mill
icr1          ±cP           lo1
PARTICLE  DIAMETER  (MICROMETERS)
 Figure 39. Particle size distribution -— dM/dLog D versus particle diameter.
        Both impactors sampling an uncharged aerosol simultaneously.
                      66

-------
 ICAr
1CP--
10
                   PARTICLE SIZE DISTRIBUTION


                   BRANCH 1 -- ®  BRANCH 2- X
                      *
                             *
                             X
        H	1 -t I I 111
-I	1  I I  Mill    I	1  Mill
       i'1          10°           101           102



    PARTICLE DIAMETER  (MICROMETERS)


Figure 40. Particle size distribution -— d M/dLog D versus particle diameter.

        Both impactors sampling a charged aerosol simultaneously.
                      67

-------
    Table  IV. Calculated Charge on Charged Aerosol Particles
              Relative to that Expected in a Full Scale Precipitator

Diameter, ]jm       0.1       0.5       1.0       2.0       5.0
Condition
A
B
C
D
1.1
1.0
1.1
1.4
1.3
1.3
1.3
1.7
1.4
1.3
1.4
1.8
1 .4
1.4
1.5
1.9
1.5
1.5
1.5
2.0
Results

     The results of the system equivalency tests are summarized in
Table V and illustrations of typical results are shown in Figures
39 and 40.  The aerosol concentrations as measured by the cascade
impactors for the tests conducted at both conditions are given in
Table V.  The mass concentrations for the charged condition were
lower due to the loss of particles in the charger.  The figures
show that the particle size distributions in both branches were
identical.  Evidence that the two impactors performed identically
was demonstrated by the fact that the impactors sampled alternat-
ing branches for each consecutive test without causing an appre-
ciable difference in the size distribution plots.

             Table V. System Branch Equivalency Data
Date  Test   Condition              Aerosol Concentration
                                           (mg/m3)
                               Branch 1                  Branch 2
10/3/77 1
10/4 1
10/5 1
10/6 1
10/6 2
10/7 1
10/7 2
Uncharged
Uncharged
Uncharged
Uncharged
Charged
Charged
Charged
15.
13.
14.
14.
0.
0.
0.
2
8
4
2
91
97
96
16.0
13.1
13.6
13.7
0.88
0.79
0.97
                                68

-------
     Figures 41 through 44 show the results of the charged and
neutralized particulate. sampling.

     The ratio of the average calculated charge of the laboratory
aerosol to the typical charge on an aerosol exiting the outlet of
an industrial electrostatic precipitator is given for five parti-
cle sizes in the legend of each figure.  In each case, more mass
was caught on the upper stages of the impactor sampling the
charged aerosol than in the one sampling the neutralized aerosol.
This effect is most noticeable for particle sizes larger than
2.9 ym for conditions A,  B, and D, and sizes larger than 1.0 ym
for condition C.  The larger differences occur for condition D
where the particles have a higher charge than those in the other
conditions.  For all four conditions, approximately the same
amount of mass is collected on the stages with D50's close to
the mass median diameter of the aerosol for both neutral and
charged impactors.  For all conditions, relatively more mass was
collected on the lower stages of the impactor sampling the neu-
tralized aerosol than on the impactor sampling the charged aer-
osol.  For moderate particle charge, this mass difference is
very slight.  For the higher particle charge, the mass difference
becomes noticeable (Figure 44).
                                69

-------
         CALCULATED CHARGE ON CHARGED AEROSOL PARTICLES
         DIAMETER,/jm      0.1    0.5   1.0   2.0   5.0

         Qlab/Qesp          1-1    1-3   I-*   1-4   1.5
                         v
             I                 *TiP CHARGED

            i

•i fjl--     i                    A»* NEUTRALIZED
            i  i  i im|	1—i \ i mil	1—i  i  i mil
   1CT1          10°            101

   PARTICLE DIAMETER  (MICROMETERS)

   Figure 41.  Particle size distribution for moderate charging condition
            for a mass median diameter of 1.27 micrometers.
                      70

-------
           CALCULATED CHARGE ON CHARGED AEROSOL PARTICLES
           DIAMETER, pm    0.1    0.5   1.0   2.0   5.0

           Qlab/Qesp        1-°    I-3   I-3   !-4   1'5
c
a
IDS:
                i
               I
                                       CHARGED
                                       NEUTRALIZED
          —i—i  i i mil	1—i  i i mil	1—i  i i  mil
       icr1          10°           101

       PARTICLE  DIAMETER  (MICROMETERS)
        Figure 42.  Particle size distribution for moderate charging condition
                for a mass median diameter of 1.02 micrometers.
                          71

-------
            CALCULATED CHARGE ON CHARGED AEROSOL PARTICLES

            DIAMETER, /urn    0.1    0.5    1.0    2.0   5.0

            °WQeSp        1-1    1-3    1-4    1.5   1.5
a
                      «  i
                   i
                            1
3 '
a
•i IDS
4
i?
•
»
• <
                                  *»,.,*
                                        CHARGED
                                         NEUTRALIZED
          	1	1 I  I I lll|	1	1  I  I Hll|	-H	1  I I  IHI|
       lo"1           10°           lo1            la2
       PARTILO.E  DIAMETER  (MICROMETERS)

       Figure 43.  Particle size distribution for moderate charging conditions
                for a mass median diameter of 0.692 micrometer.
                           72

-------
    ±03-r
            CALCULATED CHARGE ON CHARGED AEROSOL PARTICLES
            DIAMETER, M™    0.1   0.5    1.0    2.0    5.0

            °WQesp       1-4   1-7    1-8    1.9    2.0
a
    10S:
    UOP-
                     I   V
            •4
           *$
                            I

                                       CHARGED
                                 b
                                   f
                                       NEUTRALIZED
      III! Mill	H—I  I I Mill	1—I  I  I llll|
10'
                      10°
101
       PARTICLE  DIAMETER  (MICROMETERS)
        Figure 44.  Particle size distribution for high charging conditions for
                a mass median diameter of 1.01 micrometers.
                          73

-------
                            SECTION 3.

                       SUMMARY AND CONCLUSIONS

     A review of the important parameters and conclusions for
each experiment are contained in the following lists:

MONODISPERSE AEROSOL

     1. Tests were done with the impactors operated at the recom-
mended flow rate and with particle charges equivalent to those
expected at precipitator outlets.  However, only two particle
sizes were used.

     2. Wall losses were measured along with the stage collection
efficiency.

     3. For charge levels approximately five times higher than
those expected at a precipitator outlet, very large effects were
noted.

     4. For samples taken using glass-fiber substrates at repre-
sentative particle charge levels, the effect of particle charge
on impactor performance was found to be minimal.

     5. Grounding the impactors generally increased stage collec-
tion efficiencies.

     6. The general effect of charge was to reduce the sharpness
(slope) of the curve at values of /if below the peak while not
charging /\J;5 0 (or the stage D50) significantly.  At large values
of /if, the effect of particle charge was to reduce the efficiency
in a manner similar to particle bounce.

                               74

-------
     7, The data from this experiment were not taken with particle
concentrations high enough to be representative of those to be
found at the outlet of industrial precipitators.

POLYDISPERSE AEROSOL SAMPLED WITH A BRINK IMPACTOR

     1. Tests were done with the impactors operated at the recom-
mended flow rate with a polydisperse aerosol and particle charges
equivalent to those expected at precipitator outlets.  Only two
stages of a Brink impactor were used.

     2. The particle concentration was less than would be expected
at a precipitator outlet.

     3. A test on wall losses showed that they play a significant
part in impactor performance when a charged aerosol is sampled.

     4. Tests using glass-fiber substrate did not show an appre-
ciable difference in the collection of charged and uncharged par-
ticles.  Tests using bare metal substrates, however, showed quite
a difference between the collection of charged particles and the
collection of uncharged particles.

POLYDISPERSE AEROSOL SAMPLED WITH ANDERSEN IMPACTORS

     1. Tests were performed with entire Andersen impactors oper-
ated at the recommended flow rate and with glass-fiber substrates
only.

     2. The particle concentration was similar to that expected
at the outlet of an industrial electrostatic precipitator.  The
aerosol contained polydisperse ammonium fluorescein particles.

     3. Wall losses were not considered separately, however, no
visible wall loss was observed for tests with particles having
moderate charge  (typical of ESP outlets).  The wall losses were
visibly evident in the high charge tests.
                                75

-------
     4. Moderate particle charge level? had little effect on
impactor performance for the middle and lower stages.  For stages
withi DSQ'S above 3 micrometers, up to about twice as much mass
was caught in the charged particle impactor than in the neutral
one.

     5. For charging conditons higher than those considered mod-
erate here, the effects of particle charge were more significant
both for the lower impactor stages and the upper impactor stages,
with the middle stages giving approximately true mass measure-
ments.
                                76

-------
                          REFERENCES

1. Gushing,  K.  M.,  G.  E.  Lacey,  J.  D.  McCain,  and W.  B.  Smith.
   Particle  Sizing  Techniques for Control Device Evaluation:
   Cascade Impactor Calibration.  EPA-600/2-76-280,
   U.S. Environmental  Protection Agency,  Research Triangle Park,
   N.C.,  1976.  79 pages.

2. Reischl,  G., W.  John,  and W.  Devor.  Uniform Electrical Charg-
   ing of Monodisperse Aerosols.  J. Aerosol Sci., 8:55-65, 1977.

3. Sem, G. J.   Submicron  Particle Size Measurement of Stack Emis-
   sions Using the  Electrical Mobility Technique.  In:  Proceed-
   ings of the Workshop on Sampling, Analysis,  and Monitoring of
   Stack Emissions, Electric Power Research Institute,  Dallas,
   Texas, 1975.  pp. 111-129.

4. Zung,  J.  T., and C. C. Snead.  Evaporation Kinetics of Liquid
   Droplets.  U.S.  Dept.  of Army, Edgewood Arsenal-DAAA-15-67-C-
   0151,  Field Evaluation Division, Edgewood Arsenal, Maryland
   21010, 1968.  116 pages.

5. Cooper, D.  W., and J.  W. Davis.   Cascade Impactors for Aero-
   sols:  Improved Data Analysis.  Amer. Ind. Hyg. Assoc., p.  79,
   1972.

6. Cooper, D.  W., and L.  A. Spielman.   Data Inversion Using Non-
   linear Programming with Physical Constraints: Aerosol Size
   Distribution Measurement by Impactors.  Atmos. Environ., 10,
   pp. 723-729, 1976.

7. Picknett, R. G.   A New Method of Determining Aerosol Size
   Distributions from Ministage Sampler Data.   Aerosol Sci.,
   1972.

8. Bush,  P.  V., and W. B. Smith.  An Electrostatic Precipitator
   Backup for Sampling Systems.   EPA-600/7-78-114, U.S.  Environ-
   mental Protection Agency, Research  Triangle Park,  N.C., 1978.
   35 pp.

9. Johnson,  J.  W.,  G.  I.  Clinard, L. G. Felix,  and J. D. McCain.
   A Computer-Based Cascade Impactor Data Reduction System.
   EPA-600/7-78-042, U.S. Environmental Protection Agency, Research
   Triangle Park, N.C., 1978.  601 pp.
                               77

-------
10. Pontius,  D.  H.,  L.  G.  Felix,  J.  R.  McDonald,  and W.  B.  Smith.
    Fine Particle Charging Development. .  EPA-600/2-77-173,  U.S.
    Environmental Protection Agency,  Research Triangle Park,  N.C.,
    1977.   240 pp.
                                78

-------
                                TECHNICAL REPORT DATA
                          (Please read Instructions on the reverse before completing)
 i. REPORT NO.
 EPA-600/7-79-027
                                                      3. RECIPIENT'S ACCESSION'NO.
4. TITLE AND SUBTITLE
Sampling Charged Particles with Cascade Impactors
                                6. REPORT DATE
                                January 1979
                                                      6. PERFORMING ORGANIZATION CODE
7. AUTHORIS)
W.E. Farthing, D.H.Hussey, W.B.Smith, and
   R.R.Wilson, Jr.
                                                      8. PERFORMING ORGANIZATION REPORT NO.
                                   SORI-EAS-79-024
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Southern Research Institute
2000 Ninth Avenue,  South
Birmingham, Alabama  35205
                                 10. PROGRAM ELEMENT NO.

                                 EHE624	
                                 11. CONTRACT/GRANT NO.
                                 68-02-2131
                        T.D. 10401 and
                              11301
 12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC 27711
                                 13. TYPE OF REPORT AND PERIOD COVERED
                                 Task Final; 4/77 - 9/78
                                 14. SPONSORING AGENCY CODE
                                  EPA/600/13
is. SUPPLEMENTARY NOTES
919/541-2557.
        project officer is D.  Bruce Harris, Mail Drop 62,
16. ABSTRACT
          The report discusses three sets of experiments which demonstrate that a
cascade impactor sampling a charged aerosol may yield a particle size distribution
measurement that deviates from the time distribution.  The distributions indicated
more large particles and fewer small particles than actually existed, due to the
particles' attraction to the grounded impactor plates  (stages).  Although higher
charge levels produced larger deviations from the true size distribution,  the magni-
tude of the deviation and the corresponding correction factor for any given charged
aerosol are unpredictable. Also, the error was  smaller if glass fiber substrates
were used as collection surfaces  instead of bare metal. For electrostatic precipita-
tors operating at normal charging conditions (an electric field of 400,000 V/m and a
current density of 0.0003  A/sq m/s), the size distribution (measured by the lower
stages of an Andersen cascade impactor with glass fiber substrates) was not signi-
ficantly different from  the true size distribution.
 7.
                             KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
                                          b.lDENTIFIERS/OPEN ENDED TERMS
                                             c. COSATI Field/Group
Pollution
Sampling
Measurement
Impactors
Charged Particles
Aerosols
Dust
Size Determination
Electrostatic Pre-
  cipitators
Glass Fibers
Substrates
Pollution Control
Stationary Sources
Cascade Impactors
Andersen Impactors
Time Distribution
13B
14B

131
20H
07D
11G
11B
11D
 8. DISTRIBUTION STATEMENT
 Unlimited
                                          19. SECURITY CLASS (ThisReport)
                                          Unclassified
                                                                   21. NO. OF PAGES
                                                   88
                                          20. SECURITY CLASS (This page)
                                          Unclassified
                                                                   22. PRICE
EPA Form 2220-1 (9-73)
                                         79

-------