You are on page 1of 24

Advanced Powder Technol., Vol. 14, No. 2, pp.

143– 166 (2003)


Ó VSP and Society of Powder Technology, Japan 2003.
Also available online - www.vsppub.com

Invited review paper

Electrostatics of particles

SHUJI MATSUSAKA ¤ and HIROAKI MASUDA


Department of Chemical Engineering, Kyoto University, Kyoto 606-8501 Japan

Received 2 October 2002; accepted 26 November 2002

Abstract—In powder handling, each particle collides with another particle or a wall, and conse-
quently becomes charged up to a certain value. Such contact charging is experienced in various elds.
In the present review, the basic concepts of contact charging are summarized; in particular, the effect
of the contact potential difference and the initial charge on the charge transfer is described in detail.
Furthermore, the variation of the particle charging caused by repeated impacts on a wall is formu-
lated. This theory is extended to the particle charging in gas– solids pipe ow, where each particle
has a different amount of charge; the distribution of the particle charge is also analyzed theoretically.
In addition, the method of measuring important electrostatic properties, the technique of detecting
particle charging and the application of particle charging are described.

Keywords: Electrostatics; particle charging; contact potential difference; charge distribution; measure-
ment.

NOMENCLATURE

a1 ; a2 constant in (36) and (37) (—)


b1 ; b2 constant in (36) and (37) (C / kg)
C0 capacitance between bodies at critical separation (F)
Cc Cunningham slip correction factor (—)
c constant in (39) (kg/ A)
Di inside diameter of pipe (m)
Dp particle diameter (m)
d constant in (39) (kg/ A)
dl thickness of powder layer (m)
E; E1 ; E2 Young’s modulus (Pa)
Ef strength of electric eld (V / m .D N/C))

¤
To whom correspondence should be addressed. E-mail: matsu@cheme.kyoto-u.ac.jp
144 S. Matsusaka and H. Masuda

e elementary charge (C)


f; f1 ; f2 ; f3 probability density function of particle charge per unit mass
(kg/ C)
fc frequency of particle collision (1/s)
fic probability density function of initial particle charge per unit
mass (kg/ C)
fk probability density function of impact charging factor (kg/ C)
g gravitational acceleration (m / s2 /
h Planck constant (J s)
I; I1 ; I2 electric current (A)
k impact charging factor (D qm1 =n0 / (C /kg)
kN mean value of impact charging factor (C /kg)
k0 constant in (6) (V / C)
k1 constant in (7) (charging efciency) (—)
k2 constant in (9) (1 / s)
k3 D Vb =q (V / C)
k4 constant in (35) (—)
ke elasticity parameter (1 / Pa)
m mass ow ratio of particles to gas (—)
n number of particle collisions (—)
nN mean number of particle collisions (—)
n0 relaxation number (—)
q net charge on particle (C)
qc charge accumulated by contact on particle (C)
qm particle charge per unit mass (charge-to-mass ratio) (C /kg)
qm0 ; qm1 particle charge per unit mass at a reference point (C /kg)
qm1 particle charge per unit mass at x D 1 (C /kg)
q0 initial charge (C)
qr leakage charge (C)
1qc charge transferred by contact (C)
S contact area (m2 /
t elapsed time (s)
uN average gas velocity (m / s)
V total potential difference (V)
V0 zero-point potential (V)
Vb potential difference arising from space charge (V)
Vc ; V1=2 contact potential difference based on surface work function (V)
Electrostatics of particles 145

Ve potential difference arising from image charge (V)


VM=Au contact potential difference of metal against gold (V)
VP=Au contact potential difference of powder against gold (V)
vN average particle velocity (m / s)
ve electrical migration velocity of charged particle (m /s)
vi impact velocity (m / s)
Wp mass ow rate of particles (kg/ s)
x distance from starting point (m / s)
1x length of pipe (m)
z0 critical gap between contact bodies (m)

Greek
® D Vb =Ve .D k0 =k3 / (—)
1; 11 ; 12 change of energy level caused by charge transfer (J)
"0 absolute permittivity of gas (F / m)
"p absolute permittivity of powder layer (F /m)
¹ viscosity of gas (Pa s)
º1 ; º2 Poisson’s ratio (—)
½ gas density (kg/ m3 /
½c charge density (C / m3 /
½p particle density (kg/ m3 /
¾0 standard deviation of initial charge per unit mass (C / kg)
¾k standard deviation of impact charging factor (C /kg)
Á; ÁI ; Á2 ; ÁM ; ÁI work function (J)
! frequency of radiation (1 / s)

1. INTRODUCTION

Electrostatic particle charging is an important phenomenon related to powder han-


dling. It is commonly a nuisance [1 – 3] and the source of explosion hazards [4].
Various applications have, however, been developed, e.g. electrophotography [5, 6],
dry powder coating [7 – 9], electrostatic precipitator [10], separation of powder
[11, 12], electromechanical particulate operation [13, 14], powder ow measure-
ment [15– 18], tomography [19] and many others. To improve the performance of
these applications and to reduce the risk of dust explosion, a correct understanding
of particle charging is required.
Contact charging has been known for a old times [20]. In spite of the long history,
there are still unsolved problems and inconsistent experimental results are also
reported. This is because many factors, such as physical, chemical and electrical
146 S. Matsusaka and H. Masuda

characteristics, and environmental conditions, affect particle charging, and thus it is


difcult to reproduce the same electrostatic phenomenon if there are some uncertain
elements. Although complete theory applicable to any kind of particle charging is
not established at present, there are several fundamentals to explain the particle
charging and electrostatic phenomena [21– 23].
In the present review, we summarize the basic concepts of contact charging and
then analyze impact charging of particles. We also describe several experimental
techniques and show the actual particle charging in gas–solids pipe ow; further-
more, we mention the application of the electrostatic phenomenon.

2. BASIC CONCEPTS AND THEORY

2.1. Contact charging


When two different materials are brought into contact and separated, an electric
charge is usually transferred from one to the other. This phenomenon is often
called ‘contact electrication’ or ‘contact charging’. When they are rubbed, it can
be called ‘frictional electrication’ or ‘tribo-charging’; as for short contact, it can
be called ‘impact charging’. Contact charging is also classied into three categories
according to the contacting materials, i.e. metal – metal contacts, metal – insulator
contacts and insulator– insulator contacts.

2.1.1. Metal– metal contacts. Charging of metals is usually unnoticeable be-


cause the charge transferred runs away from the contact point. In fact, charge trans-
fer occurs even for metal – metal contact when the metals are isolated electrically
before and after the contact. Figure 1 shows two metals with different work func-
tions ÁI and Á2 in contact. Assuming that electrons transfer by tunneling so that
thermodynamic equilibrium maintains, the contact potential difference (CPD) Vc is
given by:
Vc D V1=2 D ¡.Á1 ¡ Á2 /=e; (1)
where V1=2 is the CPD of metal 1 against metal 2 and e is the elementary charge. The
amount of transferred charge is equal to the product of the CPD and the capacitance
between the two bodies. The capacitance depends on the state of the contacting
part. Although the position of the electrons can vary as the metals are separated, the
charge after the separation 1qc is approximated by the following simple equation:
1qc D C0 Vc ; (2)
where C0 is the capacitance between the bodies at the critical separation dis-
tance. Figure 2 shows the theoretical and experimental results for the charge after
contact [21]. Although the experimental results are somewhat less than the theoret-
ical ones, the tendencies are in reasonable agreement. The difference is probably
caused by the surface roughness, impurities, oxidized layer, separation speed, etc.
Electrostatics of particles 147

Figure 1. Electron potential energy at a metal– metal contact.

Figure 2. Charge on a chromium sphere, 4 mm in diameter, in contact with another metal sphere,
13 mm in diameter, as a function of the CPD of chromium against each metal [21].

2.1.2. Metal– insulator contacts.


(i) Electron transfer. The concept of contact charging between metals can be
extended to charging for metal – insulator contacts [24]; however, experimental data
are often scattered widely [25] and inconsistent results are also reported [26, 27].
When a linear relationship between the transferred charge and the work function is
obtained experimentally, it is very probable that the charging will occur by means of
the electron transfer and the following relationship is applicable to the electrostatic
characterization:
1qc / .ÁM ¡ ÁI /; (3)
where ÁM is the work function of the metal and ÁI is the apparent work function
of the insulator. The proportionality constant in this relationship depends on the
characteristics of the insulator. Figure 3 shows the linear relationship obtained
148 S. Matsusaka and H. Masuda

Figure 3. Charge density of nylon 66 by contacting with various metals. The horizontal axis is the
CPD of each metal against a gold reference VM=Au [24], where the CPD is dened as VAu=M .

experimentally by Davies [24]. Murata and Kittaka also produced evidence of


electron transfer by means of the measurement of photoelectric emission [28].

(ii) Ion transfer. Several researchers have suggested that contact charging be-
tween a metal and an insulator may be due to the transfer of ions [20, 29, 30]. Insu-
lators may contain ions in the body or on the surface. These ions can be transferred
by diffusion, relative acidity, afnities and the kinetic effect based on shearing off.
When ions exist in excess, ion transfer affects contact charging; however, when the
ions are few, electron transfer will control the contact charging.

(iii) Material transfer. The impact or friction between two bodies can result in
the transfer of material from one to the other. When a metal slides over a polymer,
large amounts of polymer may transfer to the metal, and the transferred polymer
can carry charge and change the effective contact potential difference. Also, when
brittle particles impact on a metal wall, elements of particles can be transferred on
the metal [31].

(iv) Effect of the separation state on charging. Particle charging can be inu-
enced by the state of separation. If the charge transferred to the surface of the insu-
lator ows back to the metal, the net charge will be reduced. This reduction depends
on properties such as conductivity and the separation state such as the speed of slid-
ing or rolling. In addition, a gaseous discharge may occur during the separation
[32– 34].
Electrostatics of particles 149

Figure 4. Energy level diagram for insulator– insulator contact.

2.1.3. Insulator– insulator contacts. The tendency of the charging between


insulators can be evaluated using a ‘triboelectric series’, in which insulators are
ranked in an order such that a material higher up the series will always charge
positive when touched or rubbed with a material lower down. However, it is
not always correct and there is no theory in the triboelectric series. To analyze
the contact charging between insulators in more detail, several models have been
presented. Almost all of them are similar to those for metal – insulator contact, but
the movement of electrons or ions in the body was more restricted. Figure 4 shows a
contact charging model that includes the effect of an electric eld [35, 36]. When the
insulators come in contact, charges ow from the lled surface state of the insulator
1 to the empty surface state of insulator 2. It is assumed that the system remains
in equilibrium by tunneling until the surfaces are separated over the critical gap;
nally, the expression for the energy level including the effect of the electric eld
Ef is given as

Á1 C 11 C eEf z0 D Á2 ¡ 12 ; (4)
where 1 is the change of the energy level caused by the charge transfer.

2.2. Particle charging by repeated impacts on a metal wall

2.2.1. Charge transfer. When a particle impacts on a metal wall, each acquires
an equal and opposite charge. The amount of charge transferred depends on the total
potential difference V between the contact bodies, which is made up of two parts:
Vc based on the surface work function and Ve arising from image charge, which is
150 S. Matsusaka and H. Masuda

induced in the wall by an external charge. These are related as follows [37– 39]:
V D Vc ¡ Ve : (5)
The potential difference Ve is approximated by:
Ve D k0 q; (6)
where k0 is a constant and q is the charge on the particle before impact. The charge
transferred in a repeated impact process is analyzed by approximating the process to
an equivalent rate process, i.e. the charge transfer is treated as a continuous quantity
dqc =dn to obtain the enveloping curve of the remaining charge after a number of
collisions:
dqc
D k1 CV ; (7)
dn
where n is the number of collisions, k1 is the charging efciency and C is the
capacitance between the contact bodies, which is given by:
"0 S
CD ; (8)
z0
where "0 is the absolute permittivity of the gas, S is the contact area and z0 is the
critical gap including the surface roughness between the contact bodies.
Charge relaxation with elapsed time dqr =dt is approximated by [40]:
dqr
D ¡k2 q; (9)
dt
where k2 is a constant. When the particle collides at regular intervals, i.e. the
frequency is dened as fc , (9) is rewritten as:
dqr k2
D ¡ q: (10)
dn fc
From (5)– (8) and (10), the following equation is derived:
dq dqc dqr
D C
dn dn dn
"0 S k2
D k1 .V c ¡ k0 q/ ¡ q: (11)
z0 fc
Solving (11) with initial conditions (n D 0; q D q0 /, the following equation is
derived:
 ´   ´¼
n n
q D q0 exp ¡ C q1 1 ¡ exp ¡ ; (12)
n0 n0
where:
Vc
q1 D ; (13)
k2 z0
k0 C
k1 "0 Sfc
Electrostatics of particles 151

Figure 5. Charge on a rubber sphere accumulated by repeated impacts with a steel wall [39].

and:
1
n0 D : (14)
k0 k1 "0 S k2
C
z0 fc
Figure 5 shows the variation in the charge on a rubber sphere by repeated impacts
[39]. The transferred charge caused by an impact decreases with the number of
collisions and the charge approaches a limiting value. The limiting value tends to
decrease as the interval between collisions increases because the charge relaxation
increases with elapsed time. The broken lines in Fig. 5 are the results calculated
using (12). The experimental results are in agreement with the calculated values.

2.2.2. Effect of elasticity. When a particle impacts on a hard plate, the particle is
deformed and a contact area is produced. If the particle is a sphere with a smooth
surface and the contact deformation can be approximated by a Hertzian deformation
pattern [41], the maximum contact area S during the impact is represented by:
4=5
S D 1:36ke2=5 ½p2=5 Dp2 vi ; (15)
where ke is the elasticity parameter, ½p is the density of the sphere, Dp is the particle
diameter and vi is the impact (incident) velocity. If the particle is not spherical, the
particle shape should be taken into account [42]. The elasticity parameter ke is given
by
1 ¡ º12 1 ¡ º22
ke D C ; (16)
E1 E2
where º is Poisson’s ratio, E is Young’s modulus, and subscripts 1 and 2 represent
the sphere and the plate, respectively.
Figure 6 shows the relationship between the transferred charge and the maximum
contact area calculated using (15). The transferred charge is approximately
152 S. Matsusaka and H. Masuda

Figure 6. Relationship between transferred charge and maximum contact area during impact of a
rubber sphere of 0.032 m in diameter on a steel wall [39].

proportional to the contact area and, therefore, it is deduced that the maximum
contact area based on the elastic deformation controls the amount of the transferred
charge.

2.3. Charging in gas–solids pipe ow

In gas– solids pipe ow, particles are charged as a result of the collisions with the
wall; the charged particles form an electric eld and the electric eld inuences the
total potential difference V , i.e. [37, 38]:

V D Vc ¡ Ve ¡ Vb (17)

where Vb .D k3 q/ is the potential difference arising from the electric eld, which
is called the ‘space charge effect’. When a length of metal pipe that is isolated
electrically is grounded, the charge transferred from the particles to the wall ows
to earth. The charge transferred per unit time is detected as electric current. The
current I generated from the pipe of a length from x to x C 1x is expressed as:
  ´¼  ´¼
I n.x/ n.1x/
D fqm0 ¡ qm1 g exp ¡ 1 ¡ exp ¡ ; (18)
Wp n0 n0

where Wp is the mass ow rate of particles, qm0 and qm1 are the charge-to-mass
ratio at x D 0 and x D 1, respectively. n is the number of particle collisions and
n0 is the relaxation number. When x D 0, (18) becomes:
  ´¼
I n.1x/
D fqm0 ¡ qm1 g 1 ¡ exp ¡ : (19)
Wp n0
Electrostatics of particles 153

Furthermore, for n.1x/ ¿ n0 , (19) is simplied as:


I n.1x/
D fqm0 ¡ qm1 g ; (20)
Wp n0
where:
6Vc
qm1 D n k2 z0 o
; (21)
¼½p Dp3 .k0 C k3 / C
k1 "0 Sfc
and:
1
n0 D : (22)
.k0 C k3 /k1 "0 S k2
C
z0 fc
Substituting k0 D 2z0 =.¼ "0 Dp2 / (for the image charge effect), k2 D 0 (for low
electric relaxation) and k3 D 3z0 m½Di u=.2¼"
N 3
N (for the space charge effect)
0 ½p D p v/
into (21) and (22), gives the following equations, respectively [38, 43]:
3"0 Vc
qm1 D ; (23)
½p Dp z0 .1 C ®/
and:
¼Dp2
n0 D ; (24)
2k1 S.1 C ®/
in which ® is the ratio of the space charge to image charge effect, i.e.:
3 ½Di uN
®D m : (25)
4 ½p Dp vN
where m is the mass ow ratio of particles to gas, ½ is the gas density, ½p is the
particle density, Di is the inside diameter of the pipe, uN is the average gas velocity
and vN is the average particle velocity.

2.4. Charge distribution of particles in gas– solids pipe ow


Although particle charging depends on various factors, the main factors are con-
sidered to be the number of particle collisions, initial charge on the particles and
the state of the impact charging. To derive the charge distribution of particles, we
introduce the probability density functions of these factors [44].

2.4.1. Probability density function of particle collision with the wall inside.
Assuming that the probability density function of particle collision is expressed as
a normal distribution, which is derived from a binomial distribution, and also using
the relationship between the number of particle collision and the particle charge, the
154 S. Matsusaka and H. Masuda

probability density function of the charge to mass ratio f1 .qm / is represented by:
2 n ± o2 3
qm ¡ qm0 ²
6 q m1 ln 1 C ¡Nnk 7
qm1 1 qm1 ¡ qm
f1 .qm / D p exp 6
4 ¡ 7 ; (26)
5
qm1 ¡ qm 2¼ nk N N
2nk 2

where nN is the mean number of particle collisions and k is the impact charging factor
.D qm1 =n 0 /. When qm0 D 0 and jqm j ¿ jqm1 j in the early stages of the particle
charging, (26) is simplied as:
 ¼
1 .qm ¡ nk/
N 2
f1 .qm / ¼ p exp ¡ : (27)
N
2¼ nk N 2
2nk

2.4.2. Probability density function of initial charge on particles. Using a prob-


ability density function of initial charge fic .qm0 /, which is based on the form of
(26), the probability density function f2 .qm / after traveling through the pipe is rep-
resented by the following equation:

Z qm1
f2 .qm / D f1 .qm /fic .qm0 / dqm0
¡1
qm1 1
D q
qm1 ¡ qm 2¼.nk N 2 C ¾02 /
2 n ± ² o2 3
qm
6 q m1 ln 1 C ¡ N
nk 7
qm1 ¡ qm
£ exp 6
4 ¡ 7;
5 (28)
N C ¾0 /
2.nk 2 2

where ¾0 is the standard deviation of the initial charge. When jqm j ¿ jqm1 j, (28)
is simplied as:
( )
1 .qm ¡ nk/
N 2
f2 .qm / ¼ q exp ¡ : (29)
N 2 C ¾ 2/
2¼.nk N 2 C ¾02 /
2.nk
0

Substituting ¾0 D 0 in (29) gives the same form as (27).


Figure 7 shows the charge distributions calculated using (28). In this calculation, it
is assumed that the charge transferred from the particles to the wall is positive. As a
result, the amount of the particle charge increases negatively with the mean number
N
of particle collisions n.D 0; 2; 4; 8; 16; 32; 64/ and approaches an equilibrium value
(qm1 D ¡6 mC /kg). Since the transferred charge depends both on the particle
charge and the number of collisions, the variation of the charge distribution is
somewhat complicated.
Electrostatics of particles 155

Figure 7. Effect of the mean number of particle collisions nN on the charge distribution (qm1 D
¡6 mC/ kg, qm0 D 0 mC/ kg, ¾0 D 0:3 mC/ kg, k D ¡0:2 mC/ kg).

2.4.3. Probability density function of impact charging factor. The impact charg-
ing factor k is not always constant, i.e. it depends on electrostatic properties, particle
impact velocity, particle diameter, surface roughness, angle of incidence, etc. In the
same manner as (28), the probability density function of particle charge f3 .qm / is
represented by:
Z 1
f3 .qm / D f2 .qm /fk .k/ dk; (30)
¡1

where fk .k/ is the probability density function of the impact charging factor.
Assuming that fk .k/ is represented by a normal distribution, (30) becomes:

Z 1
qm1 1
f3 .qm / D q
¡1 qm1 ¡ qm 2¼.nk N 2 C ¾02 /
2 n ± qm ² o2 3
6 q m1 ln 1 C ¡ N
nk 7
qm1 ¡ qm 7
£ exp 6
4 ¡ 5
N 2 C ¾02 /
2.nk
 ¼
1 N 2
.k ¡ k/
£p exp ¡ dk
2¼ ¾k 2¾k2
156 S. Matsusaka and H. Masuda
Z 1
qm1 1 1
D q
qm1 ¡ qm 2¼ ¾k ¡1 nk N 2 C ¾02
2 n ± ² o2 3
qm
6 q m1 ln 1 C ¡ N
nk N 27
qm1 ¡ qm .k ¡ k/
£ exp 6
4 ¡ ¡ 7 dk; (31)
N 2 C ¾02 /
2.nk 2¾k2 5

Figure 8. Effect of the standard deviation of the impact charging factor ¾k on charge distribution
(¾ 0 D 0:3 mC/ kg, nN D 3:0; kN D ¡0:2 mC / kg).

Figure 9. Comparison between experimental data and theoretical curves (y ash: Dp50 D 12¹m;
(E) uN D 77 m/ s, 1x D 0:5 m; (n) uN D 58 m/ s, 1x D 2:0 m [44].
Electrostatics of particles 157

where kN and ¾k are the mean value and the standard deviation of the impact charging
factor, respectively. When jqm j ¿ jqm1 j, (31) is simplied as:
Z 1 ( )
1 1 .qm ¡ nk/
N 2 N 2
.k ¡ k/
f3 .qm / ¼ q exp ¡ ¡ dk: (32)
2¼ ¾k ¡1 .nk N 2 C ¾02 / N 2 C ¾02 /
2.nk 2¾k2

Figure 8 shows the effect of the standard deviation of the impact charging factor
¾k on the charge distribution. The charge distributions for ¾k D 0 and for ¾k 6D 0
are calculated using (29) and (32), respectively. The shape of the distribution curve
becomes skewed as the value of ¾k increases, i.e. the tail on the left-hand side of the
distribution curve becomes larger.
Figure 9 shows typical examples of the results for the charge distribution of
particles. The values of particle charge are widely distributed, including positive
charge as well as negative charge, and the distributions are skewed, i.e. there is
a longer tail on the left-hand side. Calculated lines are also added to Fig. 9. The
experimental data agree well with the theoretical curves taking into account the
distribution of the impact charging factor.

3. CHARACTERIZATION AND MEASUREMENT

3.1. Electrostatic properties


3.1.1. Work function. When materials are exposed to electromagnetic radiation,
electrons are librated. The number of electrons emitted depends on the intensity of
the radiation. The kinetic energy of the electrons emitted depends on the frequency
of the radiation. The radiation is regarded as a stream of photons, each having an
energy h!, where h is the Planck constant and ! is the frequency of the radiation.
A photon can only eject an electron if the photon energy exceeds the work function
Á of the solid, i.e. there is the minimum frequency (or threshold frequency) at
which ejection occurs. For many solids, the photoelectric effect occurs at ultraviolet
frequencies or above, but for some materials (having low work functions) it occurs
with light. The maximum kinetic energy of the photoelectron depends on the energy
of a photon and the work function. This is applied to a technique for determining
the work function or analyzing the surface properties of various materials [28, 45].

3.1.2. CPD. The CPD directly controls the contact charging as mentioned in
Section 2.1. In fact, the surface of materials is not pure and is usually covered with
an oxide lm, and thus the CPD of the materials used may differ from the values
shown in literature. To measure the CPD between a powder and a wall, a measuring
system based on the Kelvin – Zisman method was developed [40, 46, 47]. The
measuring system is shown in Fig. 10. This system has an electric circuit in which
there is a capacitor made up of two electrodes, a DC supply and an electrometer in
series. Powder is lled in the concavity of the lower electrode and upper electrode
158 S. Matsusaka and H. Masuda

Figure 10. Measurement of the CPD based on the Kelvin– Zisman method.

Figure 11. Characterizationof the coated particles by the Kelvin– Zisman method [47].

made of gold oscillates vertically. When the voltage applied on the upper electrode
is equal to the potential difference between the powder and the upper electrode, the
induced currents detected by the electrometer become zero. Hereafter, the applied
voltage is called a zero-point potential V0 , which is expressed by the following
equation:
½c dl2
V0 D VP=Au C ; (33)
2"p
where VP=Au is the CPD between the powder and gold reference, ½c is the charge
density, dl is the thickness of the powder layer, and "p is the absolute permittivity
of the powder layer. When the charge of the powder layer is sufciently small
.½c ¼ 0/; V0 is equal to VP=Au. Figure 11 shows an example of the measurements
Electrostatics of particles 159

for analyzing the CPD of coated particles. The value of the CPD varies according
to the thickness of the coat, i.e. the value approaches that of the coating material
as the thickness increases. In addition, various studies such as evaluation of toners
[40] and particles with different functional groups [48], and analysis of the effect of
temperature and humidity [49] have also been carried out using the method.

3.2. Particle charge


3.2.1. Charge-to-mass ratio. In general, the charge on particles is measured with
a ‘Faraday cage’ because the method is simple and reliable. A number of charged
particles are put into a metal enclosure that is isolated electrically and the induced
charge is measured. The charge-to-mass ratio is obtained by dividing the charge by
the mass. In gas– solids pipe ow, the charged particles are collected on the lter in
the Faraday cage.

3.2.2. Distribution of particle charges. Mazumder et al. [50] developed the


electrical-single particle aerodynamic relaxation time (E-SPART) analyzer, which
is based on the dynamics of the aerosol particles in an electrostatic and acoustic
eld. The motion of the particles can be analyzed with a laser Doppler velocimeter
and particle diameter Dp is determined from the phase lag of the particle motion
relative to the motion of the gas. The charge of each particle can be determined
from the electrical migration velocity Ve . When the motion is governed by Stokes’
low, the charge is given by:
3¼¹Dp ve
qD ; (34)
Ef Cc
where ¹ is the viscosity of the gas, Ef is the strength of the electric eld and Cc is
the Cunningham slip correction factor. Masuda et al. developed a simple method
to directly determine the charge-to-mass ratio of each aerosol particle qm [51]. The
value can be calculated from the two-dimensional trajectory of the particle moving
in an electrostatic and gravitational eld:
g
q m D k4 ; (35)
Ef
where k4 is a constant, which is the ratio of the vertical velocity to the horizontal
velocity controlled by electrostatic force, and g is the gravitational acceleration.
Since these methods enable us to collect data efciently, the distribution of the data
can be obtained easily.

4. PROCESS

In pneumatic transport, particles repeatedly collide with an internal wall, and


particles and the wall are charged. When a metal pipe is grounded, the charge
160 S. Matsusaka and H. Masuda

Figure 12. Electric current generated from a detecting pipe as a function of powder ow rate (quartz
sand: D p50 D 320 ¹m [43].

transferred from the particles to the wall ows to the ground, and can be detected
as electric currents. The currents are caused by the particle charging in the pipe
and thus the value of the current should be proportional to the particle ow rate.
As shown in Fig. 12, a linear relationship is obtained experimentally [43]. This
relationship is also explained by (18), where the right-hand side term must be a
constant. In general, this is satised in dilute and high-speed conveying systems.
As for dense-phase conveying systems, the surrounding particles prevent the free
contacts between the particles and the wall; as a result, the currents are less than
those expected under ideal conditions. Using smaller particles, the adhesiveness
also prevents free contacts. In addition, the initial charge on the particles affects the
charge transfer (see (18)). In fact, the particles collide with many different walls
before arriving at the pipe, e.g. hopper, feeder, chute, disperser, etc., and hence the
polarity and the amount of the particle charge change according to the operating
condition. To estimate the charge transferred from the particles to the wall, the
initial charge has to be known beforehand.

5. APPLICATIONS

5.1. Online measurement of particle ow rate in the dilute phase


The electric current detected from a pneumatic transport pipeline depends on the
particle ow rate and the initial charge per unit mass of particles. If the current
and the initial charge are known, the particle ow rate can be determined using
(19). The continuous measurement of the current is very easy, but that of the initial
Electrostatics of particles 161

Figure 13. A novel method for measuring the particle ow rate.

charge is not so easy. If the particle ow rate is determined from only the electric
currents, it is very convenient for the measurement. Figure 13 shows a method for
measuring the particle ow rate [52, 53]. The system has two different detecting
pipes connected in series with electrical isolation. The charge balances for the rst
and the second detecting pipes are given by the following equations, respectively:
I1
D a1 qm0 C b1 ; (36)
Wp

and:
I2
D a2 qm1 C b2 ; (37)
Wp
where I is the electric current, Wp is the particle ow rate, qm is the charge per
unit mass of particles at the inlet of each detecting pipe, a and b are constants, and
subscripts 1 and 2 refer to the rst and second detecting pipes, respectively. Since
the charge at the outlet of the rst pipe is equal to that at the inlet of the second pipe,
the charge balance is expressed as:

I1
qm0 ¡ qm1 D : (38)
Wp
From (36)– (38), the powder ow rate Wp is given by the following equation:

Wp D cI1 C dI2 ; (39)


162 S. Matsusaka and H. Masuda

here:
.a1 ¡ 1/a2
cD ; (40)
a1 b2 ¡ a2 b1

and:

a1
dD : (41)
a1 b2 ¡ a2 b1

In (40) and (41), the denominator of the right-hand side must not be zero. Therefore,
the electrostatic property of the rst detecting pipe must differ from that of the
second, i.e.:

b1 b2
6D : (42)
a1 a2

Figure 14 shows the electric currents generated from the two detecting pipes. The
polarity of the currents is determined by the relative electrostatic property between
particles and the wall. Since particles are fed continuously into the pipe, the currents
almost keep constant. The responsiveness in this system is so high that a small
uctuation of the feed rate can be detected as the variation of the current. Figure 15
shows the results on the measurement of the particle ow rate. The relative error
is less than 10%. This system can be applied to even polymer particles over wide
range [54].

Figure 14. Electric currents generated from the two detecting pipes [52].
Electrostatics of particles 163

Figure 15. Comparison between the online method and weighing method [52].

Figure 16. Comparison between the online method and Faraday cage method [52].

5.2. Online measurements of particle charge

The above system can be applied to the measurement of particle charge. From
(36)– (38), the following equation is derived:
.b2 ¡ a2 b1 /I1 ¡ b1 I2
qm0 D : (43)
a2 .a1 ¡ 1/I1 C a1 I2
164 S. Matsusaka and H. Masuda

Figure 16 shows the comparison between the values obtained by this online method
and those obtained by the Faraday cage method.

6. CONCLUSION

Contact charging is usually classied into three categories according to the contact-
ing materials. For metal – metal contacts, the charge transferred can be explained by
the CPD. As for metal – insulator contacts and insulator– insulator contacts, many
other factors affect the charge transfer. When a linear relationship between the
transferred charge and the work function is obtained experimentally, it is very prob-
able that the charging will occur by means of electron transfer and the CPD plays
an important role. When a particle impacts on a metal wall, the charge transfer
depends on the total potential difference, including the effect of the initial charge.
The charge transferred is approximately proportional to the contact area. When the
particle deforms elastically, the charge can be estimated theoretically. The particle
charging caused by repeated impacts on a metal wall can be expressed using an
equation with exponential functions. This theoretical analysis can be extended to
the particle charging in gas– solids pipe ow, where the charge transferred per unit
time, i.e. electric current, is an important evaluation factor. Since many particles
ow in the pipe, the charge on particles is widely scattered. The charge distribution
can be explained by introducing the probability density functions of (i) the num-
ber of particle collision, (ii) the initial charge on the particles and (iii) the amount
of charge transferred by an impact. In addition to the theoretical study, extensive
experimental studies have been carried out. Various measuring techniques to char-
acterize the electrostatic properties and charge transferred have been developed, and
some of them are used widely in the laboratory and industry.

REFERENCES
1. S. Joseph and G. E. Klinzing, Vertical gas– solid transition ow with electrostatics, Powder
Technol. 36, 79– 87 (1983).
2. M. Nifuku, T. Ishikawa and T. Sasaki, Static electrication phenomena in pneumatic transporta-
tion of coal, J. Electrostat. 23, 45– 54 (1989).
3. I. Adhiwidjaja, S. Matsusaka, S. Yabe and H. Masuda, Simultaneous phenomenon of particle
deposition and reentrainment in charged aerosol ow — effects of particle charge and external
electric eld on deposition layer, Advanced Powder Technol. 11, 221– 233 (1999).
4. T. B. Jones and J. L. King, Powder Handling and Electrostatics— Understanding and Preventing
Hazards. Lewis, London (1991).
5. L. B. Schein, Electrophotography and Development Physics, 2nd edn. Springer, Berlin (1992).
6. L. B. Schein, Recent advances in our understanding of toner charging, J. Electrostat. 46, 29– 36
(1999).
7. J. F. Hughes, Electrostatic Powder Coating. Research Studies Press/Wiley, New York (1984).
8. A. G. Bailey, The science and technology of electrostaticpowder spraying, transport and coating,
J. Electrostat. 45, 85– 120 (1998).
Electrostatics of particles 165

9. W. Kleber and B. Makin, Triboelectric powder coating: a practical approach for industrial use,
Particle Sci. Technol. 16, 43– 53 (1998).
10. D. A. Lloyd, Electrostatic Precipitator Handbook. Adam Hilger, Bristol (1988).
11. R. Gupta, D. Gidaspow and D. T. Wasan, Electrostatic separation of powder mixtures based on
the work functions of its constituents, Powder Technol. 75, 79– 87 (1993).
12. D. K. Yanar and B. A. Kwetkus, Electrostatic separation of polymer powders, J. Electrostat. 35,
257– 266 (1995).
13. M. Ghadiri, C. M. Martin, J. E. P. Morgan and R. Clift, An electromechanical valve for solids,
Powder Technol. 73, 21– 35 (1992).
14. W. Balachandran, D. Hu, M. Ghadiri, S. E. Law and S. A. Thompson, The study of the
performance of an electrostaticvalve used for bulk transport of particulate materials, IEEE Trans.
Indust. Applic. 33, 871– 878 (1997).
15. B. C. O’Neill and C. A. Willis, An electrostatic method for the measurement of powder ow
rates in pipes, in: Electrostatics ’97, pp. 303– 306. IOP Publishing, Oxford (1987).
16. H. Masuda, S. Matsusaka and S. Nagatani, Measurements of powder ow rate in gas– solids
pipe ow based on the static electrication of particles, Advanced Powder Technol. 5, 241– 254
(1994).
17. H. Masuda, S. Matsusaka and H. Shimomura, Measurement of mass ow rate of polymer powder
based on static electrication of particles, Advanced Powder Technol. 9, 169– 179 (1998).
18. J. B. Gajewski, Monitoring electrostatic ow noise for mass ow and mean velocity measure-
ment in pneumatic transport, J. Electrostat. 37, 261– 276 (1996).
19. M. Machida and B. Scarlett, Development of displacement current tomography system, Particle
Charact. 15, 36– 41 (1998).
20. W. R. Harper, Contact and Frictional Electrication. Clarendon Press, Oxford (1967).
21. W. R. Harper, The Volta effect as a cause of static electrication, Proc. R. Soc. A 205, 83– 103
(1951).
22. J. Lowell and A. C. Rose-Innes, Contact electrication, Adv. Phys. 29, 947– 1023 (1980).
23. T. B. Jones, Electromechanics of Particles. Cambridge University Press, Cambridge (1995).
24. D. K. Davies, Charge generation on dielectric surfaces, Br. J. Appl. Phys: J. Phys. D 2,
1533– 1537 (1969).
25. A. Chowdry and C. R. Westgate, The role of bulk traps in metal– insulator contact charging,
J. Phys. D: Appl. Phys. 7, 713– 725 (1974).
26. T. J. Fabish and C. B. Duke, Molecular charge states and contact charge in polymers, J. Appl.
Phys. 48, 4256– 4266 (1977).
27. G. A. Cottrell, J. Lowell and A. C. Rose-Innes, Charge transfer in metal–polymer contacts and
the validity of ‘contact charge spectroscopy’, J. Appl. Phys. 50, 374– 376 (1979).
28. Y. Murata and S. Kittaka, Evidence of electron transfer as the mechanism of static charge
generation by contact of polymers with metal, Jpn. J. Appl. Phys. 18, 421 (1979).
29. E. S. Robins, J. Lowell and A. C. Rose-Innes, The role of surface ions in the contact
electrication of insulators, J. Electrostat. 8, 153– 160 (1980).
30. J. Guay, J. E. Ayala, A. F. Diaz and L. H. Dao, The question of solid-state electron transfer in
contact charging between metal and organic materials, Chem. Mater. 3, 1068– 1073 (1991).
31. K. Tanoue, A. Ema and H. Masuda, Effect of material transfer and work hardening of metal
surface on the current generated by impact of particles, J. Chem. Engng, Japan 32, 544– 548
(1999).
32. G. S. Rose and S. G. Word, Contact electrication across metal-dielectric and dielectric-
dielectric interfaces, Br. J. Appl. Phys. 8, 121– 126 (1957).
33. R. Elsdon and F. R. G. Mitchell, Contact electrications of polymers, J. Phys. D: Appl. Phys. 9,
1445– 1460 (1976).
34. T. Matsuyama and H. Yamamoto, Charge relaxation process dominates contact charging of a
particle in atmospheric conditions, J. Phys. D: Appl. Phys. 28, 2418– 2423 (1995).
166 S. Matsusaka and H. Masuda

35. E. J. Gutman and G. C. Hartmann, Triboelectric properties of two-component developers for


xerography, J. Imag. Sci. Technol. 36, 335– 349 (1992).
36. G. S. P. Castle and L. B. Schein, General model of sphere-sphereinsulator contact electrication,
J. Electrostat. 36, 165– 173 (1995).
37. B. N. Cole, M. R. Baum and F. R. Mobbs, An investigation of electrostatic charging effects in
high speed gas– solids pipe ows, Proc. Inst. Mech. Eng. 184, 77– 83 (1969– 1970).
38. H. Masuda, T. Komatsu and K. Iinoya, The static electrication of particles in gas–solids pipe
ow, AIChE J. 22, 558– 564 (1976).
39. S. Matsusaka, M. Ghadiri and H. Masuda, Electrication of an elastic sphere by repeated impacts
on a metal plate, J. Phys. D: Appl. Phys. 33, 2311– 2319 (2000).
40. T. Itakura, H. Masuda, C. Ohtsuka and S. Matsusaka, The contact potential difference of powder
and the tribo-charge, J. Electrostat. 38, 213– 226 (1996).
41. S. P. Timoshenko and J. N. Goodier, Theory of Elasticity, 3rd edn. McGraw-Hill, New York
(1970).
42. H. Masuda and K. Iinoya, Electrication of particles by impact on inclined metal plates, AIChE J.
24, 950– 956 (1978).
43. H. Masuda, S. Matsusaka, S. Akiba and H. Shimomura, Electrication of ne particles in
gas–solids pipe ow, KONA Powder and Particle 16, 216– 222 (1998).
44. S. Matsusaka, H. Umemoto, M. Nishitani and H. Masuda, Electrostatic charge distribution of
particles in gas–solids pipe ow, J. Electrostat. 55, 81– 96 (2002).
45. Y. Murata, Photoelectric emission and contact charging of some synthetic high polymers, Jpn. J.
Appl. Phys. 18, 1– 8 (1979).
46. H. Yoshida, T. Fukuzono and H. Masuda, Measurement of contact potential difference between
a powder bed and a metal, J. Soc. Powder Technol., Japan 28, 226– 231 (1991).
47. H. Yoshida, T. Fukuzono, H. Ami, Y. Iguchi and H. Masuda, The effect of the surface
modication of particles on the contact potential difference between a powder bed and metal,
J. Soc. Powder Technol., Japan 29, 504– 510 (1992).
48. K. Tanoue, K. Morita, H. Maruyama and H. Masuda, Inuence of functional group on the
electrication of organic pigments, AIChE J. 47, 2419– 2424 (2001).
49. T. Nomura, N. Taniguhi and H. Masuda, Inuence of the atmospheric condition for toribo-
charging of powder, J. Soc. Powder Technol., Japan 36, 168– 173 (1999).
50. M. K. Mazumder, R. E. Ware, T. Yokoyama, B. J. Rubin and D. Kamp, Measurement of particles
size and electrostaticcharge distributionson toners using E-SPART analyzer, IEEE Trans. Indust.
Applic. 27, 611– 619 (1991).
51. H. Masuda, K. Gotoh and N. Orita, Charge distribution measurement of aerosol particles,
J. Aerosol Res., Japan 8, 325– 332 (1993).
52. H. Masuda, S. Matsusaka and S. Nagatani, Measurements of powder ow rate in gas– solids
pipe ow based on the static electrication of particles, Advanced Powder Technol. 5, 241– 254
(1994).
53. H. Masuda, S. Matsusaka and H. Fujimoto, On-line simultaneous measurement of powder ow
rate and electric charge of particles in gas solids pipe ow — effect of air velocity on accuracy
of measurement, Kagaku Kogaku Ronbunshu 21, 573– 580 (1995).
54. H. Masuda, S. Matsusaka and H. Shimomura, Measurement of mass ow rate of polymer powder
based on static electrication of particles, Advanced Powder Technol. 9, 169– 179 (1998).

You might also like