t h e particle size being c o n s t a n t .
I n m o s t
T h e Charging o f Small Particles f o r
Electrostatic P r e c i p i t a t i o n
G. W . HEWITT
MEMBER AIEE
C H A R G I N G C O N D I T I O N S a n d par­
ticle size a r e i m p o r t a n t factors in
electrostatic gas cleaning. Particle m o ­
bility, t h e r a t e a t which t h e particle
moves in a n electric field of u n i t strength,
is a function of t h e charge a n d size of t h e
particle. T h e m i n i m u m mobility and t h e
particle size a t which it occurs, a r e func­
tions of t h e charging conditions.
Charging generally is accomplished b y
exposing t h e particles t o a stream of ions
of one sign in t h e gas. T h r e e i m p o r t a n t
conditions which affect t h e charging of a
particle a r e electric field strength, ion
concentration, a n d time of exposure t o t h e
ions. T h e acquired charge depends also
upon physical properties of t h e particle,
particularly its size.
Electrostatic precipitation now is used
extensively in air-conditioning applica­
tions for t h e purpose of greatly reducing
the blackening of walls, furnishings, a n d
other exposed surfaces, b y t h e a t m o ­
sphere. 1
Fractional-micron particles are
responsible for a large p a r t of this blacken­
ing. A s t u d y of city atmosphere m a d e a t
this laboratory using a n electron micro­
scope, indicates t h a t a b o u t half t h e black­
ening is a t t r i b u t a b l e t o particles between
0.1 a n d 0.6 micron in radius. This is a n
i m p o r t a n t size range in t h e cleaning of t h e
atmosphere.
Existing e q u a t i o n s do n o t permit ac­
c u r a t e calculation of t h e charging of these
small particles, a n d further t h a n this,
there h a v e been very few experimental
d a t a which h a v e been published, particu­
larly for practical charging conditions.
T h e purpose of this paper is t o present
experimental information on t h e charging
of fractional-micron particles in air, as a
function of particle size, field strength, ion
concentration, a n d charging time.
Paper 57-90, recommended by the AIEE Elec­
tronics Committee and approved by the AIEE
Technical Operations Department for presentation
at the AIEE Winter General Meeting, New York,
Ν. Y., January 21-25, 1957. Manuscript sub­
mitted August 28, 1956; made available for
printing February 6, 1957.
G. W. H E W I T T is with the Westinghouse Electric
Corporation, Pittsburgh, Pa.
The author acknowledges the assistance of Mrs.
G. Lachmanek and M. C. Werst in carrying out
this project, and the advice of Dr. Α. V. Phelps,
all of the Westinghouse laboratories; also, the
advice of Professor G. W. Penney, Carnegie
Institute of Technology.
300 Hewitt—Small
Experimental Methods
T h e isolation of particle size effects h a s
been one of t h e difficulties in particle
charging studies in t h e p a s t . T h e effects
of particle size in t h e fractional-micron
range were difficult t o d e t e r m i n e because
it was necessary t o m a k e m e a s u r e m e n t s on
individual particles. T h e s t u d y of clouds
of particles is simpler a n d less tedious, a n d
it a u t o m a t i c a l l y provides a n average
result for a large n u m b e r of particles. F o r
studies on clouds of particles, i t is r e ­
quired t h a t t h e size b e nearly uniform a n d
controllable over t h e desired range.
AEROSOL G E N E R A T I O N A N D
MEASUREMENT
Cloud-type m e a s u r e m e n t s were m a d e
possible b y t h e use of a n aerosol generator
a n d light-scattering i n s t r u m e n t s developed
a t Columbia University during W o r l d W a r
II by La Mer and Sinclair. " 2 4
T h e s e in­
s t r u m e n t s a r e described in Appendix I .
T h e generator produces particles of con­
trollable a n d very uniform size. T h e
particles a r e t i n y liquid spheres. T h e
liquid used in these t e s t s w a s diocytl
p h t h a l a t e , which h a s a v e r y low v a p o r
pressure a t room t e m p e r a t u r e , a n d pro­
vides a v e r y stable aerosol. T h e gas in
which t h e particles were suspended w a s
a m i x t u r e of nitrogen, oxygen, a n d w a t e r
vapor in t h e proportion in which t h e y
occur in air a t 2 0 % relative h u m i d i t y .
T h e degree of particle size uniformity
produced b y t h i s t y p e of generator is
given b y L a M e r . Generally, m o r e t h a n
7 0 % of t h e particles a r e within ± 1 0 %
of t h e average radius, a n d a negligible
percentage differs from t h e average b y
more t h a n ± 2 0 % .
T h e size a n d concentration of t h e aerosol
particles were m e a s u r e d b y light-scatter­
ing m e t h o d s developed a t C o l u m b i a
University, a n d described in Appendix I.
T h e i n s t r u m e n t used t o d e t e r m i n e par­
ticle size, code-named t h e * O w l , " provides
simple a n d r a p i d size m e a s u r e m e n t s of
uniform aerosols.
I n t h e aerosol concentration meter,
light scattered b y t h e particles is focused
upon a photomultiplier t u b e . T h e re­
sponse of t h i s i n s t r u m e n t is linear with
particle concentration over a wide range,
Particles Charging for Electrostatic
of t h e tests only t h e relative concentra­
tions, a t c o n s t a n t particle size, were re­
quired. I n some cases, t h e absolute
concentration in t e r m s of liquid weight per
u n i t volume of air was required. F o r
this purpose, t h e i n s t r u m e n t w a s cali­
b r a t e d experimentally w i t h respect t o
particle size.
PARTICLE CHARGER
T h e charging device w a s designed so as
to provide i n d e p e n d e n t control of t h e
three i m p o r t a n t charging variables, field
s t r e n g t h E, ion concentration N, a n d
charging t i m e /. T h i s device is shown in
Fig. 1. T h e charging region C is b o u n d e d
on opposite sides b y electrodes A a n d 6*
between which a voltage is applied t o o b ­
tain t h e desired field s t r e n g t h . Electrode
5 is a fine screen of large percentage open
area. Ions of one sign (positive in t h e
case of t h e polarity shown), produced b y
corona a t coaxial wire W, are caused t o
flow t h r o u g h t h e screen a n d across t h e
charging region. T h e aerosol flowing
axially between A a n d 5 is subjected t o
these ions a n d t o t h e voltage gradient in
this region, for a t i m e / which is deter­
mined b y t h e aerosol velocity a n d t h e axial
length D of t h e charging region. T h e
cylindrical s h u t t e r H, slidable inside
cylinder B, is used t o a d j u s t D. T h e ion
concentration in region C is controlled
b y adjusting t h e voltage between t h e
coaxial wire a n d t h e cylinder B.
As soon as t h e aerosol particles pick u p
some of t h e ions, t h e y experience a n
electrostatic force m o v i n g t h e m in t h e
direction of t h e field. F o r m a n y charg­
ing conditions, this would cause t h e loss of
a large percentage of t h e particles b y
precipitation on electrode A if t h e volt­
age between S a n d A were continuous.
T o reduce loss b y precipitation greatly, a n
a l t e r n a t i n g voltage is used between .S a n d
A so a s t o limit t h e t r a n s v e r s e motion of
y
t h e particles t o small a m p l i t u d e s . D u r ­
ing t h e reverse half-cycle, t h e direction of
t h e field in C p r e v e n t s ions from passing
t h r o u g h t h e screen, so n o charging occurs.
A s q u a r e w a v e of voltage is desirable in
order t h a t Ε a n d Ν remain c o n s t a n t dur­
ing each charging period.
Variations in Ε a n d Ν w i t h radial posi­
tion in region C were less t h a n 1 0 % of
t h e average values in m o s t of t h e tests.
F u r t h e r information on t h i s a n d o t h e r
details of t h e charger a r e given in A p p e n ­
dix I I I .
MOBILITY ANALYZER
T h e mobilities of t h e charged particles
were determined b y measuring their
velocities in a n electric field. Fig. 2
Precipitation JULY 1957
 A C,
S Q U A R E WAVE
6 0 CYCLE
DC
INSULATION
Β light-scattering c h a m b e r of t h e concen­
Fig, 1 . Particle charger. Charging r e g i o n
is s h o w n at C b e t w e e n e l e c t r o d e A a n d screen
S. Ions are furnished b y corona at w i r e W
in c y l i n d e r B. L e n g t h D o f charging r e g i o n
is adjustable b y p o s i t i o n i n g of shutter Η
shows the analyzer a n d the a r r a n g e m e n t
of a p p a r a t u s . Clean air flows axially in
a laminar stream between two coaxial
cylinders which are held a t different poten­
tials. T h e charged aerosol enters t h e
analyzer through a narrow annular pas­
sage A a t t h e circumference of t h e outer
cylinder. Downstream from this, a
narrow circumferential sampling groove
is provided in t h e wall of t h e inner
cylinder a t B. This sampling groove
connects to an exit t u b e of small diameter
inside t h e inner cylinder, leading o u t of
the analyzer. T h e inner cylinder is
movable axially for a d j u s t m e n t of t h e
distance L. T h e mobility of t h e par­
ticles entering t h e sampling groove a t B,
is a function of the analyzer dimensions,
.lean AEROSOL
AIR GENERATOR
CHARGER
-o VARIABLE
+ DIRECT
Ô VOLTAGE
AEROSOL
CONCENTRATION
MEASUREMENT
AEROSOL
DISCHARGER
Fig, 2 . Coaxial cylinder m o b i l i t y analyzer.
C h a r g e d aerosol enters a t A . S a m p l i n g
g r o o v e is at Β in inner c y l i n d e r , w h i c h is
adjustable a x i a l l y
JULY 1 9 5 7 Hewitt—Small
t h e distance L, t h e air-flow rate, a n d t h e
voltage V between cyclinders, as given b y
equation 11. I t is i n d e p e n d e n t of t h e
radial distribution of air velocity between
t h e cylinders. T h i s calculation a n d other
information on t h e analyzer are given in
Appendix IV.
U p o n leaving t h e analyzer, t h e aerosol
sample is passed t h r o u g h a c h a m b e r in
which it is exposed to a l p h a radiation.
T h e purpose of this is t o bring t h e n e t
charge on t h e aerosol to near zero in order
to minimize t h e loss of particles b y space-
charge precipitation, particularly in t h e
t r a t i o n meter, t h r o u g h which t h e aerosol
then flows.
0 0.1 0.2 0.3 0.4 01
F o r a n y given aerosol entering a t A, t h e
MOBILITY ANALYZER SETTING,
mobility t e s t is m a d e b y holding t h e air
CM STATVOLT
flow a n d distance L c o n s t a n t a t suitable SEC
PER
CM
values a n d recording t h e concentration of
F i g . 3 . T y p i c a l a n a l y z e r response curve.
t h e sample, while v a r y i n g voltage V in
D a t a o b t a i n e d o n particles o f 0 . 1 3 - m i c r o n
steps. F r o m these d a t a , a c u r v e showing
radius c h a r g e d a t E = 1 2 , N t = 1 0 7
relative particle concentration is plotted
against mobility; Fig. 3 is a typical curve.
T h e average mobility is t a k e n as t h e aver­ same size, for various charging conditions.
age of t h e mobilities a t half t h e peak For example, since Figs. 3 a n d 4 a r e for
ordinate, as indicated b y t h e broken lines. the same particle size, t h e average par­
T h e a c t u a l spread in particle mobility, ticle charge for t h e test shown in Fig. 3
caused b y statistical variations in charg­ is 2 1 .
ing a n d a small variation in particle size,
is smaller t h a n t h e spread of t h e corre­ CHARGE-MASS MEASUREMENTS
sponding analyzer curve, because some Another m e t h o d of determining par­
broadening is caused b y conditions in the ticle charge, independent of mobility
analyzer. T h i s is discussed in Appendix measurements, is t h a t of measuring t h e
IV. charge per u n i t m a s s of liquid in t h e
T h e relation between mobility a n d droplets. F r o m this t h e particle charge
particle charge can b e determined from can be calculated readily for a uniform
mobility m e a s u r e m e n t s on particles of t h e
same size carrying very few charges.
P A R T I C L E C H A R G E η , E L E M E N T A R Y UNITS
One w a y of charging particles very weakly 1 2 3 4
is to subject t h e aerosol to ionizing radia­
tion. T h i s provides b o t h positive a n d
negative gas ions, so some particles become
weakly charged negatively a n d some
weakly charged positively. A preferable
m e t h o d is to use ions of only one sign in
order t o minimize agglomeration of t h e
particles. T h i s is done easily with t h e
charger shown in Fig. 1 b y using very
small values of Ε, N, a n d /.
T o measure t h e mobilities of these par­
ticles, low air velocity a n d large distance
L in t h e analyzer a r e required. A
typical t e s t result is shown in Fig. 4.
N o t e t h a t t h e mobilities corresponding to
t h e three peaks, are related in a l-to-2-to-
3 proportion. Since electric charge exists
0 0.01 0 . 0 2 0 . 0 3 0 . 0 4 0 . 0 5 0 . 0 6 0 . 0 7
in discrete quantities, it is evident t h a t MOBILITY ANALYZER SETTING,
t h e three peaks correspond to particles
CM ^ STAT VOLT
PER
0
carrying one, two, a n d three elementary SEC CM
charges. T h i s provides t h e linear rela­ F i g . 4 . A n a l y z e r response curve for v e r y
tion between charge a n d mobility, which w e a k l y c h a r g e d particles, 0 . 1 3 - m i c r o n radius.
can b e used to calculate charge from Equally spaced peaks indicate t h e particle
mobility m e a s u r e m e n t s on particles of t h e charge
Particles Charging for Electrostatic Precipitation 301
 Fig. 7 . E x p e r i m e n t a l results
R A D I LJS = 0 . 6 / 1 0.42 for N t = 1 0 , s h o w i n g effects
7

1.0 of charging gradient

_ι 0.30
_) ο
> Έ 0.8
CD ο
Ο
Έ < ·<-

ST
0.6
r ^Ό . Ι 4

LE
ce
Lϋ 0.4
ο \
F i $ . 5 . D e v i c e for measuring t h e current CL
ι- 0.19 0.07
carried b y a n aerosol stream. A e r o s o l enters ο : «cr Ο
«2: 0.2

ΡΑ

SE
U ;
through p o l y s t y r e n e t u b i n g Β b e t w e e n metal
t u b e A a n d c y l i n d e r C w h i c h is s u p p o r t e d
from t h e terminal of electrometer E. Particles 0 5 10 15 20 25 30 35 40
are r e t a i n e d b y h i g h - e f f i c i e n c y filter D . CHARGING G R A D I E N T E, S T A T V O L T S P E R CM.
S h i e l d i n g is p r o v i d e d b y c y l i n d e r F a n d
screen G

aerosol of known particle size (equation
12). Or, if t h e charge is determined b y
mobility tests, this provides another
method of measuring particle size.
T h e a p p a r a t u s which was used for
measuring t h e electric current carried by
the aerosol stream is shown in Fig. 5.
T h e charged particles in a measured
flow of t h e aerosol are collected b y t h e
high-efficiency filter. T h e mass of liquid
per u n i t volume of air is measured b y t h e
concentration meter which has been cali­
brated b y precipitating t h e particles on
a l u m i n u m foil and weighing. F r o m these
measurements, the charge per unit mass
is determined.
Test Results
T h e particle sizes obtained from charge-
mass measurements, equation 12, agree
closely with those measured b y the Owl.
A third m e t h o d of size determination is
the use of the Stokes-Millikan relation,
t h e other two m e t h o d s for particles larger
t h a n 0.28-micron radius. F o r smaller par­
ticles, a discrepancy appeared, reaching a
m a x i m u m of a b o u t 1 2 % a t 0.15-micron
radius. T h e Stokes-Millikan relation in­
dicates a larger size t h a n t h e other two
m e t h o d s . T h i s raises t h e question of t h e
applicability of t h e Millikan correction
factor for dioctyl p h t h a l a t e particles.
T h e sizes given in this paper are those
measured b y t h e other two m e t h o d s .
T h e experimental results presented here
are for positive ion charging, t h a t used in
precipitators for air-conditioning applica­
tions.
T h e experimental points shown in Figs.
6 to 14, a n d 16, r e p r e s e n t t h e average
particle mobility or charge. M o s t of these
points were obtained from mobility a n a ­
lyzer curves similar t o t h a t shown in Fig.
3. Some were obtained from charge-mass
d a t a , as indicated in t h e captions.
Fig. 6 shows t h a t for each charging field,
there is a particle size a t which t h e
mobility reaches a m i n i m u m . T h e mini­
m u m mobility increases r a t h e r rapidly
with field strength. F o r practical charg­
ing conditions, t h e particles m o s t difficult
to precipitate are those between 0.1- a n d
0.2-micron radius.
I n Fig. 7, t h e d a t a are plotted to show
directly t h e variation with field strength.
F o r a given particle size, t h e intercept a t
E = 0 indicates t h e mobility which could
be obtained b y t h e r m a l charging alone.
T h e additional mobility produced b)^ a
charging field is approximately pro­
portional to t h e s t r e n g t h of t h e field. I t
is evident t h a t t h e field is an i m p o r t a n t
factor even for particles as small as 0.07
micron.
T w o mechanisms b y which particles
become charged b y gas ions are field
charging a n d t h e r m a l c h a r g i n g . The 5 - 8

equation 14, along with t h e single-charge electric field drives ions onto t h e particle.
mobility. This was found to agree with T h e second mechanism is t h a t of diffusion

ι—
_i
Ο 2 300

PARTICLE RADIUS, MICRONS

Fig. 8 . Particle charge versus size for N t =

1 0 a n d l o w gradient
7

) 1 I Ι UILIJ : I ! 1 I I I II PARTICLE RADIUS, MICRONS

03 .05 0.1 0.2 0 . 3 0.5 1.0
A — E x p e r i m e n t a l results
PARTICLE RADIUS, MICRONS Β—Theoretical curve for field charging alone,
Fig. 9 . Particle charge versus size for N t =
b y e q u a t i o n 19
1 0 a n d moderate gradient
7

Fig. 6 . E x p e r i m e n t a l m o b i l i t y curves for G — T h e o r e t i c a l curve f o r thermal charging

N t = 1 0 . N o t e m i n i m u m mobilities
7
alone, by equation 2 3 For e x p l a n a t i o n o f c u r v e s , s e e F i g . 8

302 Hewitt—Small Particles Charging for Electrostatic Precipitation JULY 1 9 5 7

 CO
or 70 T— 3 5 0
0.14 M I C R O N RADIUS
/ E = 3 0

RY
3 0 0 --4 • -
Έ |e=3o[ 1
UJ
< / j 0 . 4 6MiCRON RADIUS

EN
_1 5 0 2 5 0

Έ
CO 4 0 LU
Η _! 2 0 0
UJ
' ; E - 12 -
3 30 C
_
[Ι )
150 \

(S>
20 or
<
X
100 .... . ..{

Ο ο
ΙΟ LU
or I 5 0
< Ο
Ο­ ι—
or
N t x I O " ! ^ ) (SEC) ( I 0 f 7 < 0 I 2 3 4 5 6

7 /I0NS\
Nt χ 10 S)(SEC)(.0)- T

CC
0.1 0.2 0.3 0 4 0.5 0.6 0 7 F i g . 1 2 . Effect o f N t o n charging o f 0 . 1 4 -
PARTICLE R A D I U S , MICRONS micron particles. D a t a o b t a i n e d b y t h e
F i g . 1 4 . Effect o f N t o n charging o f 0 . 4 6 -
m o b i l i t y m e t h o d for E = 1 2 , a n d b y t h e
Fig. 1 0 . Particle charge versus size for micron particles. D a t a o b t a i n e d b y m o b i l i t y
charge* mass m e t h o d for t h e other t w o curves
N t = 1 0 a n d high gradient
7 method

For e x p l a n a t i o n o f c u r v e s , see Fig. 8

of i o n s onto the particle because o f their
t h e r m a l energies. Theoretical equa­
tions, considering t h e two m e c h a n i s m s
separately, are given in Appendix V.
Theoretical curves, from equations 19 a n d
Figs. 8 , 9, a n d 10 show particle charge as
itself on t h e particle in such a way as to
a function of size for low, m e d i u m , a n d
allow t h e particle to a c q u i r e a greater
high charging g r a d i e n t s . A t low gradient,
charge. T h e u p p e r limit of this effect is
t h e e x p e r i m e n t a l results agree reasonably
equivalent t o infinite dielectric c o n s t a n t
well w i t h theory. A t t h e higher gradients,
in equation 16, which is a 4 0 % increase in
a n d for t h e smaller particles, t h e experi­
charge. Since t h e resistivity of dioctyl
m e n t a l charge is considerably greater t h a n
p h t h a l a t e is only 3 ( 1 0 ) ohm-centimeters 10

2 3 , are shown in some o f t h e figures for t h e theoretical charge, either field or

comparison with t h e experimental curves.
Addition of t h e held a n d t h e r m a l charges
is not allowable because each equation
was developed on t h e basis of only one
mechanism acting. R a t h e r , the charging
rates, equations 18 a n d 22, should be
added. T h e charge calculated from t h i s
(cm), t h e c o n d u c t i v i t y effect m a y account
t h e r m a l . F o r t h e larger particles, t h e
for a n appreciable p a r t of t h e discrepancy
discrepancy decreases.
for t h e larger particles in Figs. 9 a n d 10.
T h e field charging equation 19 does not
T h e variables Ν a n d t a p p e a r in the
t a k e into a c c o u n t t h e effect of t h e con­
charging equations only as t h e product
d u c t i v i t y of t h e particle. C o n d u c t i v i t y
NL A test was m a d e in which t was
t e n d s to p e r m i t t h e charge t o distribute
varied from 0.014 to 0.13 second, with Nt
held c o n s t a n t a t 0.25 (10) . T h e r e was 7

is only slightly greater t h a n t h e field

no significant change in mobility for this
or t h e r m a l charge, whichever is larger.
variation in / a n d N. I n a n o t h e r test, t h e
mobility remained c o n s t a n t for variation
of t from 0.07 to 0.13, a t Nt = 2 (10) . 7

T h e charging t i m e used in m o s t of t h e
tests was 0.07 second. F o r values of Nt
greater t h a n 2 (10) , it was necessary to7

increase

T a b l e I. Effects o f N t

Experimental
Particle Charging Charge Ratios*
Radius, Gradient,
Microns Ε

0.07 2. .0.81 1.23

12. .0.75 1.18
30. .0.72 1.15
0.14 1. .0.80 1.22
12. . .0.81. .1.19
30. . .0.80. .1.19
0.28.. 12. . .0.83. .1.18
N.'x.o-^KSECK.O)" 7 6
36. . .0.81. .1.25
0 1 2 3 4 5 0.46.. 2. . .0.80. .1.22
12.
N t χ 10 â
)H('°f Fig. 1 3 . Effect o f N t o n charging o f 0 . 2 8 - 30.
. .0.83.
. .0.84
.1.16

micron particles 0.66.. ....12. ..0.82. .1.17

Fig. 11. Effect o f N t o n charging o f 0 . 0 9 -
* A is the ratio of the particle charge at Nt — 0.5
micron particles. D a t a o b t a i n e d b y t h e 0 = charge-mass d a t a (10)7 to that at Nt = 107 d β is the ratio of the
a n

charge-mass m e t h o d Δ = mobility data particle charge at Nt = 2 (10)7 to that at Nt = 10?.

JULY 1 9 5 7 Hewitt—Small Particles Charging for Electrostatic Precipitation 303


2 3
ίΟ" ,7
(-^)(sec)00)
T h e variation of charge with t h e prod­
u c t Nt is shown in Figs. 1 1 t o 1 4 . N o t e in
Fig. 13 t h e close agreement of d a t a ob­
tained b y t h e two independent methods,
mobility m e a s u r e m e n t s a n d charge-mass
measurements.
These curves show a greater slope a t
A f t = 10 t h a n theory indicates. I n T a b l e
7
I, A a n d Β are, respectively, t h e ratios
of t h e charge a t A / = 0 . 5 ( 1 0 ) a n d
7 7
Nt=2
(10) to t h a t a t i W = 1 0 .
7
Theoretically,
7
for field charging . 4 = 0 . 8 9 a n d 5 = 1 . 0 6 ,
a n d for t h e r m a l charging A varies from
0 . 8 4 t o 0 . 9 0 a n d Β from 1 . 1 3 t o 1 . 1 6 for
this particle size range. T h e experimental
curves do n o t flatten off as quickly as
theory predicts, particularly field charg­
ing.
T h e effect of Nt upon t h e statistical
O.C3 0.05 O.I 0.2 0.3 0.5 0 7
PARTICLE RADIUS, MICRONS
F i g . 1 6 . M o b i l i t y curves for N t = 1 0
A — E x p e r i m e n t a l data f o r Ε = 1 2
Β—Theoretical curve f o r field chargins a l o n e ,
E = 1 2 , b y equation 1 9
C — T h e o r e t i c a l curve f o r thermal charging
alone, by equation 2 3
D — T h e o r e t i c a l c u r v e c o n s i d e r i n g b o t h field
a n d thermal charging b y a d d i t i o n o f their
rates g i v e n b y e q u a t i o n s 1 8 a n d 2 2
B r o k e n lines s h o w m o b i l i t y o f particles
carrying o n e a n d t w o charges, according t o
equation 1 4
304 Hewitt—Small
Fig. 1 5 . W i d t h -of a n a l y z e r r e ­
sponse curve as a f f e c t e d b y N t .
Data obtained o n 0.14-micron
particles a t Ε = 1 2
4
variation of particle charge is indicated in
Fig. 1 5 . T h e a c t u a l spread in mobility
is less t h a n t h a t of t h e analyzer response
curve b y a nearly c o n s t a n t a m o u n t which
is a t t r i b u t a b l e t o conditions in t h e a n ­
alyzer. A rough e s t i m a t e of t h e a m o u n t
to b e s u b t r a c t e d from t h e ordinates
of Fig. 1 5 to obtain spread in mobility,
is 0 . 1 . T h e statistical variation is small
for large values of Nt. A s Nt decreases
below 1 0 , t h e statistical spread increases
7
rapidly.
Fig. 1 6 provides a comparison of ex­
perimental a n d theoretical mobilities a t a
m o d e r a t e charging gradient. C u r v e D
takes into a c c o u n t b o t h charging m e c h a -
isms operating simultaneously b y addi­
tion of t h e charging r a t e s , equations 1 8
a n d 2 2 . F o r t h e smaller particles, which
includes t h e region of m i n i m u m mobility,
the experimental mobilities a r e consider­
ably greater t h a n t h e theoretical, even if
t h e m a x i m u m correction for particle con­
d u c t i v i t y were applied.
Conclusions
1 . The most difficult particles to pre­
cipitate are those in t h e size range from
0 . 1 - to 0.2-micron radius.
2 . Field strength is an important charging
factor even for particles as small as 0 . 0 7 -
micron radius.
3. For thermal charging alone, t h e experi­
mental results agree reasonably well with
theory.
1.0
4. For t h e larger particles and high field
strengths, the discrepancy between experi­
7
ment and present charging equations is not
large, particularly when the effects of
particle conductivity are considered.
5. For small particles the effect of the
field is considerably greater than present
charging equations indicate.
6 . At Nt = 1 0 the particle charge changes
7
more rapidly with Nt t h a n present charging
equations predict.
7 . The statistical variation o f particle
charge rises rapidly as Nt is decreased below
10 .
7
Particles Charging for Electrostatic
A p p e n d i x I. A e r o s o l G e n e r a t i o n
and M e a s u r e m e n t
T h e La Mer-Sinclair t y p e of generator
produces particles of uniform size by
condensation of a suitable vapor upon
condensation nuclei under carefully con­
trolled conditions. T h e size which the
particles attain depends upon t h e ratio of
t h e mass of vapor to t h e number of nuclei.
T h e size is controlled by regulating the
rate of vapor production and the rate of
nuclei production.
Particle size of a uniform aerosol can
be determined from its light-scattering
properties. The angular positions of spec­
tral bands of light scattered from the par­
ticles, are a function of particle size. For
particles smaller than 0.2-micron radius,
this method is not applicable b u t there
is a polarization method which can be used
t o extend t h e range t o 0.1-micron radius.
Also, by exposing the aerosol t o t h e vapor
of a suitable solvent under proper condi­
tions, t h e particles can be " g r o w n " by a
known factor. 9
Then t h e size o f the
grown particles can be measured b y the
Owl and t h e original size calculated from
the known growth factor. Particles smaller
t h a n 0.1-micron radius can be measured
by this method.
T h e aerosol concentration meter is a
forward-angle type of Tyndallometer which
employs dark-field illumination in order to
exclude direct light from t h e collecting lens
system. There is provision for inserting
an opal glass arbitrary standard into the
light-scattering position when desired. The
amount of light scattered depends upon
particle size as well as upon particle con­
centration. This instrument was cali­
brated in terms of mass concentration,
with respect t o particle size, by electro­
statically precipitating t h e particles from
measured volumes of aerosols onto alu­
minum foil and weighing. T h e device for
this is similar t o t h a t described by Barnes
and Penney, except t h a t an aluminum foil
liner was used inside t h e cylinder. 10
A p p e n d i x II. Nomenclature
a = radius of particle, cm
e = elementary charge = 4 . 8 ( 1 0 ) ~
1 0
stat-
coulomb
k = dielectric constant of particle
?« = ion mobility, cm per second per stat-
volt per cm
η = number of elementary charges on
particle
q = air flow in analyzer, cubic cm per second
r, 0, ζ = cylindrical co-ordinates in analyzer
(Fig. 1 8 )
/ = time of exposure of particle t o charging
conditions, seconds
v~ velocity of particle with respect to the
gas, cm per second
O , v — components of particle velocity in
r z
the analyzer, cm per second (Fig. 1 8 )
μ = viscosity of the gas, poises
C = correction factor for Stokes law
Z> = length of charging region, cm (Fig. 1 )
E — field strength in charging region, stat-
volts per cm
E = field strength in precipitating region,
p
statvolts per cm
/ = c u r r e n t to electrode A of Fig. 1 , averaged
over the whole cycle, statamperes
Precipitation JULY 1 9 5 7
 1.2 1 d(ER)
= 4irNe (3)
R dR

ΙΕ, Ε; The solution of this equation is

IN, Ν, ~ E t l * Γ # ι Ί 2
Γ R M ' R i N i

-\-4ire
Λ

— —— ι ~4πβ
•JEJ L Ex (4)
Ο
Ζ 10
< in which the subscripts refer to two posi­
tions, taken here as the electrodes 5 and A.
Also

N2_Ri Ε ι
(5)
N t ~ R t E 2

or / Λ for fixed values of / and D (from equation

/ /
ι/ / 2). Fig. 17 shows t h e maximum variations
ι / / in Ε and Ν radially across t h e charging
// /
space, as calculated from equations 4 and 5.
Ο 10 20 30 The voltage gradient and ion concentra­
GRADIENT E,,STATVOLTS PER CM tion referred t o in Figs. 6 t o 16, are t h e
average gradient in space C and t h e ion
F i g . 1 7 . M a x i m u m variations o f Ε a n d Ν i n
concentration a t this average gradient
t h e particle charger. A , B , a n d C s h o w t h e assuming its position t o be midway be­
gradient v a r i a t i o n , a n d Α ' , B ' , a n d C s h o w tween 5 and A, as calculated from equation
t h e i o n d e n s i t y variation 2.

A, A'—Νι=0.7 (ΙΟ)»
Β, B ' — N i = 1 . 4 ( 1 0 ) 8

C, C — Ν χ = 2.8(10) 8
A p p e n d i x I V . Measurements of
M o b i l i t y and Charge Per U n i t
iC = Boltzrnann constant = 1.38 (10)~ 16 Mass
L = length of analyzing region, cm (Figs.
2 and 18) As the charged particle is carried with
M — particle mobility, cm per second per velocity v axially b y t h e air stream, it is
z

statvolt per cm forced by t h e electric field toward t h e inner

iV = ion concentration in charging region, cylinder a t radial velocity v , as indicated r

ions per cubic cm in Fig. 18. Both of these components

Q = charge per gram of particles, stat- vary with radial position. T h e radial
coulomb per gram distribution of field strength and therefore
R = radius of position in charger, cm (Fig. of v is known, and t h a t of v need not be
r z

17) known in order t o calculate the particle F i g . 1 8 . M o b i l i t y a n a l y z e r p a r t i c l e entrance

Τ=absolute temperature, degrees Kelvin mobility M. at A a n d s a m p l i n g g r o o v e a t Β
Z7=rms thermal velocity of ions, cm per
second
(6)
V=analyzer voltage, statvolts (Fig. 2) V Z
= T T tion 10 into equation 9,
DR
v = ~ - r = MEp (7)
A p p e n d i x I I I . A e r o s o l Charger
r
DT MV 2ττ n
η
and and
The square wave applied between S and
Λ actually is a clipped 60-cycle sine wave.
(8)
a ri a
By removing the upper portion of each half M= ^ -\n-==0A22~-
T (Π)
wave, flat tops extending 150 degrees with r In ft 2TTLV η LV

a voltage variation of 4 % , are obtained. The spread in t h e mobility curve is not

The 60-cycle frequency is satisfactory for where ζ and r are t h e axial and radial posi­ caused entirely b y actual particle mobility
the 0.72-cm spacing between S and A, and tions of t h e particle, E is t h e field strength distribution. Some broadening of the curve
for the ranges of particle mobility and field
p

at radius r, and V is t h e potential difference is caused by small eccentricity of the

strength under investigation. between cylinders. cylinders, finite widths of entrance and exit
Ions are present in C only during the openings, blending of streams a t the
half-cycle in which A is negative, so the dr dr__ entrance, and some degree of turbulence
effective charging time is 4 2 % of the total 3 = ~V Z = ~v r and diffusion in t h e air stream. The
ME, z
MV
time the particles spend in region C. If spread of the curve was found t o be affected
/ is the current to electrode A, averaged from equations 6, 7, and 8, and very little by L and q in t h e ranges used.
over time, then A measure of t h e spread is provided by
1 dividing t h e width of t h e curve a t t h e half
J = (0.42)(0.1)2ΤΓ#£> NemE (D
L = — in
MV - Γ- rdr peak ordinate, b y t h e average mobility,
'Jr, as indicated in Figs. 3 and 15. The
r

Substituting w = 480, the mobility of

positive ions, Ν is calculated as follows: But smallest spread obtained had a ratio of
0.14, so t h e spread caused b y conditions
dq = v 2irrdr z in the analyzer must be less than this.
iV=1.65(10) 7
(2)
RDE and In t h e charge-mass measurements on
dioctyl phthalate particles of uniform
The variation in Ε with radial position q= 27RFRL v rdr
z (10) size, the relation is simply
is easily obtained from Poisson's equation,
expressed in cylindrical co-ordinates, for a where q_ is t h e volume rate of air flow be­ SUE
= 1.17(10)- -
10
(12)
•2 =
long cylinder, in the following equation tween cylinders. Substituting from equa- 4ττ(0.984)α ;

JULY 1957 Hewitt—Small Particle Charging for Electrostatic Precipitation 305

 A p p e n d i x V . Theoretical Table I I . Correction Factor f o r Stokes L a w To take account of the two processes
operating simultaneous!}', the charging
Charging and M o b i l i t y rates, equations 19 and 22, should be
Particle Radius
Correction Factor, added. T h e resulting differential equation
The mobility M of a particle of radius a, Cm C is very difficult; for particular cases, t h e
carrying η elementary charges as obtained charge can be calculated by numerical
from Stokes Law is 0.07 (10) " 4 2.40
integration.
0.10 1.94
0.15 1.62
(13) 0.20 1.45
6ττμα
0.30 1.28 Refe rences
0.40 1.22
This is valid only when the particle radius 0.60 1.15
is large compared with the mean free path 0.80 1.11 1. A N E W E L E C T R O S T A T I C P R E C I P I T A T O R , G. W.
of the gas molecules. For small particles, 1.00 1.09 Penney. AIEE Transactions, vol. 56, 1937,
a correction factor must be used. The pp. 159-63.
mobility equation for air at 30 C (degrees 2. H A N D B O O K O N A E R O S O L S . United States
centigrade) and 730 millimeters pressure Atomic Energy Commission, Washington, D . C ,
1950, pp. 77-102.
then becomes 30 C and 730 millimeters pressure,
3. Proceedings, First National Air Pollution Sym­
dn _ Γ 2.22(10)- w"l 10 2
, posium, Los Angeles, Calif., 1949, pp. 1-13.
Jf=1.37(10)- C- (14) : 3,250 NEa
'^-Hb—J (18)
7
4. L I G H T S C A T T E R I N G AS A M E A S U R E O F P A R T I C L E
a dt ~ SIZE I N AEROSOLS, David Sinclair, Victor K.
La Mer. Chemical Reviews, Baltimore, Md., vol.
where Millikan's empirical correction factor and 44, 1949, pp. 245-67.
C is given in Table I I .1 1

Existing charging equations consider sep­ 7.24(10)- 5. PARTICLE CHARGING I N ELECTROSTATIC P R E ­

w = 45(10) £a CIPITATION, H. J. White. AIEE Transactions,
8 2
(19)
arately two methods by which a particle .1+7.24(10)) ~ Nt_ 7
vol. 70, pt. I I , 1951, pp. 1186-91.
can be charged by gas ions, field charging 6. CHARGING OF SPHERICAL PARTICLES I N A N
and thermal c h a r g i n g . In field charging,
5-8
In thermal charging, the ions reach the I O N I Z I N G F I E L D , M. M. Pauthenier, M. Moreau-
ions are driven onto the particle by the particles by virtue of their thermal energies. Hanot. Journal de Physique et Le Radium,
electric field. T h e equations for this Considering thermal charging alone, the Paris, France, vol. 3, Ser. 7, 1932, pp. 590-613.
mechanism alone are equations given by White are 7. RESEARCH O N THE PHYSICAL BASIS OF ELEC­
TRICAL G A S P U R I F I C A T I O N , R. Ladenburg. Annalen
dn -^1 der Physik, Leipzig, Germany, vol. 4, 1930, pp.
(15) — = 7Γ UNa*e aKt (20) 863-97.
dt
8. THE MECHANISM OF CHARGING OF MIST
PARTICLES, P. Arendt, H. Kallman. Zeitschrift
aKT / Tr Ue aNt\
2

fur Physik, Berlin, Germany, vol. 35, 1925, pp.

In (21)
421-41.
(16) 9. A D I R E C T T E S T OF K E L V I N ' S EQUATION CON­
where U is the rms thermal velocity of t h e NECTING VAPOR PRESSURE A N D R A D I U S OF C U R V A ­
ions. For air a t 30 C, assuming t h e ions T U R E , Victor K. La Mer, Ruth Gruen. Transactions,
Faraday Society, London, England, vol. 48, 1952,
to consist of single molecules in thermal pp. 410-15.
equilibrium with uncharged molecules,
10. A N ELECTROSTATIC D U S T W E I G H T SAMPLER,
G = l+2 (17) E. C. Barnes, G. W. Penney. Journal of Industrial
k+2 dn . 5.56(10) - · -
= 15.7(10) iVaV 4
(22) Hygiene and Toxicology, Chicago, 111., vol. 20,
dt no. 3, March 1938, pp. 259-65.
k is the dielectric constant of the particle
11. G E N E R A L L A W OF F A L L OF A SMALL SPHERICAL
and m is the mobility of the ions. For and B O D Y T H R O U G H A G A S , R. A. Millikan. Physical
£ = 5.1, the dielectric constant of dioctyl Review, New York, Ν . Y., vol. 22, Ser. 2, 1923,
phthalate, and for positive ions in air a t « = 1.8(10) a In 5
(l+0.9aNt) (23) pp. 1-23.

Discussion of 5 7 - 8 9 and
57-90
G. W. Hewitt and G. W. Penney : Measure­
ments of t h e charges on particles whose
diameters are smaller than t h e wavelength
of visible light, must be made indirectly.
There are various possible sources of error.
For this reason, there is some degree of
uncertainty in t h e results, and it is par­
ticularly important t o compare t h e results
of independent investigations.
The experimental results presented in
the two papers are rather difficult to com­
pare because there are several parameters;
different ranges are covered; and the data
are plotted in different ways. T h e Penney-
Lynch (57-89) investigation was directed
primarily toward the industrial precipitator
which involves large values of Nt, while the
Hewitt work was directed more toward the
low-voltage 2-stage precipitator with its
lower values of Nt. However, there are
regions of overlap in which comparisons can
be made. I t is difficult for t h e casual reader
to compare t h e results, so it seems desirable
to point out t h e degree of agreement.
Since, in t h e two papers, there are no
points for which all parameters are exactly
the same, points were chosen for which t h e
parameters are nearly t h e same a n d correc­
tions were made for t h e parameter differ­
ences. For example, in Fig. 11 of t h e
Penney-Lynch paper, t h e positive particle
charge a t iW = 6 (10) and 11 kv (for which7
Fig. 5 shows t h e field strength t o be 2.7 kv
per cm), for particles of 0.15-micron radius,
is 33. In comparison, Fig. 12 of t h e Hewitt
paper (57-90) shows 34 charges for a 0.14-
micron particle a t iV/ = 6 (10) and Ε = 12
statvolts per cm or 3.6 kv per cm. Correct­
ing for t h e size difference b y using Fig. 9 of
paper 57-90, and for difference in field
strength by using Fig. 7, t h e indicated
charge becomes 33, t h e same as t h e value
given by paper 57-89. Using this method
of comparison, values from various curves
of paper 57-89 compare with t h e Hewitt
data as follows.
Particle charge values shown in the lower
half of Fig. 8 and in t h e whole range of
Fig. 9, are about 1 0 % lower than values
obtained from paper 57-90. I n Fig. 10,
comparisons can be made in t h e lower por­
tion of t h e positive curve. T h e values
shown are lower than the Hewitt data by
less than 10%.
I t is only in Fig. 7 t h a t the two sets of
data differ by more than 10%. In this case,
the Penney-Lynch values of charge are
higher than t h e Hewitt values; t h e differ­
ence varies from 1 5 % a t 0.16-micron t o
2 7 % a t 0.28-micron radius. Except for
these discrepancies in Fig. 7, t h e agreement
7
between t h e two sets of d a t a is good, con­
sidering t h e difficulties in this type of experi­
ment.
Although there are small differences in t h e
results, both papers indicate t h a t t h e charge
acquired by small particles in t h e electric
fields commonly used in electrostatic pre­
cipitation, are considerably higher than those
predicted by existing charging equations,
especially a t large values of Nt.

306 Hewitt—Small Particle Charging for Electrostatic Precipitation JULY 1 9 5 7