Tri Bo Electric

Chemical Engineering Science 65 (2010) 57815807
Contents lists available at ScienceDirect
Chemical Engineering Science

journal homepage: www.elsevier.com/locate/ces
Review
Triboelectric charging of powders: A review

S. Matsusaka a,n, H. Maruyama a, T. Matsuyama b, M. Ghadiri c
a
Department of Chemical Engineering, Kyoto University, Kyoto 615-8510, Japan
b
Department of Environmental Engineering, Soka University, Tokyo 192-8577, Japan
c
Institute of Particle Science and Engineering, University of Leeds, Leeds LS2 9JT, UK
a r t i c l e in fo abstract
Article history: Particles are often electrostatically charged by frictional contact during powder-handling operations.
Received 23 January 2010 This phenomenon is called triboelectric charging or contact electrication. The charged particles
Received in revised form cause problems such as particle deposition and adhesion. In addition, if particles are excessively
4 July 2010
charged, an electrostatic discharge may occur, which can pose a risk of re and explosion hazards; thus,
Accepted 8 July 2010
to mitigate the adverse effects, it is important to elucidate the underlying triboelectric charging
Available online 14 July 2010
mechanisms. The electrostatics is, on the other hand, very useful in a number of applications that have
Keywords: been developed using the principles. In this review, the basic concepts and theories of charge transfer
Powder technology between solid surfaces are summarized, and chemical factors depending on materials and environ-
Particulate processes
mental effects are described. To theoretically analyze the process of particle charging, relevant models
Triboelectric charging
are discussed. Using the models, particle charging by repeated impacts on a wall is formulated. To
Charge transfer
Characterization experimentally evaluate particle charging, measurement and characterization methods are outlined.
Electrostatic application Furthermore, important applications and computer simulations are described.
& 2010 Elsevier Ltd. All rights reserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5782
2. Basic concepts of contact charging and charge transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5782
2.1. Electron transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5782
2.1.1. Work function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5782
2.1.2. Effective work function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5783
2.1.3. Surface state model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5783
2.1.4. Molecular-ion-state model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5784
2.1.5. Quantum chemical calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5784
2.2. Ion transfer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5785
2.3. Material transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5785
3. Chemical concept and environmental conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5786
3.1. Inorganic compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5786
3.2. Organic compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5786
3.3. Surface treatment and charge control agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5787
3.4. Moisture and temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5787
4. Mechanism of particle charging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5787
4.1. Condenser model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5787
4.2. Charge relaxation model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5788
4.3. Impact on a wall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5789
4.4. Repeated impacts of a single particle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5790
4.5. Particle charging in gassolids pipe ow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5791
4.6. Control of triboelectric charging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5793
5. Characterization of triboelectric charging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5795
5.1. Work function and contact potential difference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5795
n
Corresponding author.
E-mail address: [email protected] (S. Matsusaka).
0009-2509/$ - see front matter & 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ces.2010.07.005
5782 S. Matsusaka et al. / Chemical Engineering Science 65 (2010) 57815807
5.2. Specic charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5795

5.3. Charge distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5795
5.4. General techniques for characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5797
5.5. Single particle impact test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5798
5.6. Gassolids pipe ow test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5798
5.7. Atomic force microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5799
6. Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5799
6.1. Separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5799
6.2. Powder coating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5799
6.3. Electrophotography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5800
6.4. Self-assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5800
6.5. Measurements in pneumatic transport. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5801
6.5.1. Particle ow rate and specic charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5801
6.5.2. Signal analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5801
7. Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5802
7.1. Triboelectric charging by repeated impacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5802
7.2. Behavior of charged particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5803
8. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5803
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5804
Appendix AEquations of image charge effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5804
Appendix BEquations of space charge effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5804
Appendix CEquations of probability density function of specic charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5805
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5805
1. Introduction particle charging caused by repeated impacts on a wall is

formulated, which is expanded to analyze the charge distribution
In industry, powders and particulate solids are widely used as of particles. Also, measurement of particle charge and characteriza-
raw materials, intermediates, or nal products. When handled in tion of electrostatic properties are outlined. Furthermore, important
air, the surfaces become triboelectrically charged and various applications and computer simulations are described.
phenomena occur; for instance, the charged particles in pneu-
matic transport lines or in uidized beds experience electrostatic
forces, and tend to adhere to the walls (Joseph and Klinzing, 1983; 2. Basic concepts of contact charging and charge transfer
Nifuku et al., 1989; Adhiwidjaja et al., 2000; Guardiola et al.,
1996; Yao et al., 2004). If the particles are excessively charged, an When two different materials are brought into contact and
electrostatic discharge will occur, which can pose a risk of re and separated, an electric charge is transferred from one to the other.
explosion hazards (Jones and King, 1991; Ohsawa, 2003; Nifuku This phenomenon is often called contact electrication or
and Katoh, 2003). On the other hand, electrostatic forces can contact charging. When they are rubbed against each other, it
control the motion of charged particles; thus, many applications can be called frictional electrication, triboelectric charging, or
have been developed (Mazumder, 1999), e.g. electrophotography simply tribo-charging; as for short contact during collision, it can
(Schein, 1992/1996, 1999), electrostatic powder coating (Hughes, be called impact charging (Matsusaka and Masuda, 2003). In
1984; Bailey, 1998; Kleber and Makin, 1998), electrostatic practice, it is not easy to classify the contacting process for charge
precipitation (Lawless, 1999), particle separation (Gupta et al., transfer purposes into groups, such as sliding, rolling, and impact,
1993; Yanar and Kwetkus, 1995), and electromechanical valve for and thus the term triboelectric charging is used in such a broad
solids (Ghadiri et al., 1992; Balachandran et al., 1997). In addition, sense.
the charge on the particles can provide useful information on the
state of the process, e.g. powder ow rate (ONeill and Willis, 2.1. Electron transfer
1987; Matsusaka and Masuda, 2006; Gajewski, 2006, 2008),
concentration distribution (Machida and Scarlett, 2005) and 2.1.1. Work function
others (Matsusaka et al., 2008a). We start with the well-established concept of electron transfer
Contact charging and electromechanics of particles have been based on the contact potential difference (CPD) due to a difference
studied for many years (Schnurmann, 1941; Harper, 1951, 1967; of work functions for metal-to-metal contact. The triboelectric
Lowell and Rose-Innes, 1980; Jones, 1995); however, there are charging of metals is usually unnoticeable since the charge
still many unknowns, and in some cases, inconsistent results have transferred moves away from the contact point due to its
been reported. This is because there are many factors, such as conductivity. However, when the metals are isolated electrically
chemical, physical and electrical properties and environmental after the contact, the transferred charge can be measured. The
conditions, which affect the process. Moreover, the amount of charge transfer is explained in terms of electron transfer arising
charge on particles is distributed, which makes the estimation of from the difference in work function between the surfaces. Two
the electrostatic charge and the control of the process more metals with different work functions f1 and f2 in contact are
difcult. To analyze and control particle charging, the measure- schematically shown in Fig. 1. Assuming that electron transfer takes
ment of electrostatic charge and the evaluation of electrostatic place by tunneling so that thermodynamic equilibrium prevails, the
characteristics are important (Matsusaka and Masuda, 2003, contact potential difference Vc is given by (Harper, 1951)
2006). Also, to improve existing processes and to develop new
applications, it is necessary to obtain an in-depth understanding f1 f2
Vc V1=2 , 1
based on theoretical analyses. e
In the present review, the basic concepts and theories of where V1/2 is the contact potential difference of metal 1 against
charge transfer between solid surfaces are summarized and the metal 2 (Matsusaka and Masuda, 2003), e is the elementary charge.
S. Matsusaka et al. / Chemical Engineering Science 65 (2010) 57815807 5783
Fig. 1. Electron potential energy for metalmetal contact.
The amount of the transferred charge is equal to the product of the 0.8
contact potential difference and the capacitance between the two Theoretical
bodies. The capacitance depends on the state of the contacting
surfaces. Although the position of the electrons can vary after the 0.6 Au
metals are separated, the net charge transferred Dqc is Rh
Pt
Charge (pC)
approximated by the following equation:
0.4 Ag Cu
Dqc C0 Vc , 2
Ni Steel Experimental
where C0 is the capacitance between the bodies at the critical
separation distance where the charge transfer is cut off. The 0.2
theoretical and experimental results for the charge after contact Co
and separation are shown in Fig. 2 (Harper, 1951). Although the
experimental results are somewhat less than the theoretical 0
ones, the tendencies are in reasonable agreement. The 0 0.2 0.4 0.6 0.8 1
difference is probably caused by uncertain factors, such as Contact potential difference (V)
surface roughness, impurities, oxidized layer, separation speed,
Fig. 2. Charge on a chromium sphere, 4 mm in diameter, in contact with another
and others.
metal sphere, 13 mm in diameter, as a function of the CPD of chromium against
each metal (Harper, 1951).
2.1.2. Effective work function
The charge transfer for insulatormetal contact can be
explained by a simulated concept of the metal-to-metal electron
1
transfer. This assumes an apparent or effective work function to
be assigned to the insulator. The amount of the transferred charge
is determined so as to equalize the energy levels of the two
Charge density (mCm2)
materials, and the following relationship is applicable to the 0.5

electrostatic characterization:
fI fM
Dqc C0 , 3 0
e
where fI is the effective work function of insulator and fM is the

work function of metal. The effective capacitance C0 depends on 0.5
the characteristics of the insulator. A linear relationship obtained
experimentally is shown in Fig. 3 (Davies, 1969). Murata and
Kittaka (1979), also, produced the evidence of the electron
1
transfer by comparing contact charging and photoelectric 1 0.5 0
emission experiments. Contact potential difference (V)
Fig. 3. Charge density of Nylon 66 by contact with various metals. The horizontal
2.1.3. Surface state model axis is the CPD of gold against each metal VAu/M (Davies, 1969).
The main criticism of the effective work function model is
that there is no available free electron in an insulator. To cover
this point and to explain the charge transfer for insulator
insulator contacts, several modied models have been presented empty surface states of insulator 2. The driving force for the
(Lowell and Rose-Innes, 1980; Lee, 1994; Bailey, 2001). Some of charge transfer between the surfaces is the difference in
them are similar to those for insulatormetal contact, but the the effective work functions of the two surfaces. The charge
movement of electrons in the body is more restricted. In one of transfer will cease when the Fermi levels of the two materials
the models it is assumed that available energy levels of electron coincide with each other. The charge transfer causes the Fermi
are only on surface, not in bulk, and the available level is called energy of the insulator 1 and insulator 2 to change with D1 and
surface state (Fig. 4) (Gutman and Hartmann, 1992; Anderson, D2, respectively. The charge transfer also produces a potential
1994; Cotler et al., 1995). When the insulators come to contact, difference between the surfaces Ef z0. The net effect is that the
electrons move from the lled surface states of insulator 1 to the Fermi levels of the surfaces are the same at contact and the
eEf z 0 Electron qIM A(EF) E

energy
1 f (E)
1
1
M(E)
2 2
E Acceptor
1
Fermi e
2
level E f A(E)
Z0 <E I>
Insulator 1 Insulator 2 Insulator 1 Insulator 2
D(E)
(+) ()
f (E)
Before contact After contact Donor
Fig. 4. Energy level diagram for insulatorinsulator contact. f (E)
Metal Window Insulator

expression for the energy level is given as
Fig. 5. Molecular-ion-state model for a metalinsulator contact (electron injection
f1 D1 eEf z0 f2 D2 : 4 into acceptor states of polymer).
For the charging of toner particles used in electrophotography,

it was reported that the data agree with the high density limit of <E 2 >
the surface state density (Cotler et al., 1995), which directly q2,1 = min [ D1 ( E ), A2 ( E )] f1 ( E ) {1 f 2 ( E )}d E
<E >
corresponds to the effective work function model. However, there
are cases where the high density limit does not stand (Ikezaki Electron energy
et al., 2005). (E) 1 f (E)
A1 1 1 f2 (E)
2.1.4. Molecular-ion-state model A2
(E)
Since the physicochemical structure of the surface states was Acceptor
difcult to strictly dene, Fabish and Duke (1977) proposed the
Acceptor
molecular-ion-state model assuming that polymers have donor
and acceptor states and charge is carried by electrons. Thus, <E1>
despite the inclusion of ion in the name of the model, this is
actually one of the electron transfer models. <E2>
Donor
A schematic illustration of a metalinsulator contact for
electron injection into the acceptor of the polymer insulator is Donor
(E)
shown in Fig. 5. In this model, electrons can only tunnel into the D1
polymer whose energy is in a narrow range, so-called window, f1 (E)
around the Fermi level of the metal. The density of the charge D2 (E) f2 (E)
acquired from the metal of a given Fermi level EF is therefore
Insulator 1 Insulator 2 Charge-transfer
supposed to be the density of the insulator states at energy EF. The
general equation for the charge transfer qIM is expressed as
integrated
Z EF DE Z EF
Fig. 6. Molecular-ion-state model for an insulatorinsulator contact.
qIM rD Ef E dE rA Ef1f Eg dE, 5
EF EF DE
Z /E2S
where DE is the window ( E0.4 eV) for the charge transfer, rD(E)
minrD1 E, rA2 Ef1 Ef1f2 Eg dE, 6
and rA(E) are the density of the insulator states for the donor and /E1 S
acceptor, respectively, f(E) is the probability of the charge states
occupied by electrons, and 1 f(E) is the probability of the charge
in which, min (rDi(E),rAj(E)) is the smaller state density of the
states unoccupied. As shown in Fig. 5, for f(EF) 0 in Eq. (5), the
donor and acceptor. The polarity of the insulators after contact
charge density qIM is approximately equal to rA(EF) DE.
depends on the values of the centroid energies.
A schematic illustration of the states for an insulatorinsulator
contact is shown in Fig. 6 (Duke and Fabish, 1978). Donor states of
insulator 1 align in energy with acceptor states of insulator 2
within the energy interval between the distribution centroids 2.1.5. Quantum chemical calculation
/E1S and /E2S. Hence, the charge transfer can take place to an Instead of using the assumed Gaussian distribution of the
extent limited by the smaller state distribution within this energy molecular-ion-state, an evaluation of the electronic state of
interval. The charge transfer q2,1, i.e. the variation of the charge polymer was tried by computational molecular orbital calcula-
density of the insulator 2 is obtained by summing up the number tion. Yanagida et al. (1993) calculated the level of the highest
of aligned donoracceptor charge states. The general equation is occupied molecular orbital (HOMO) of an oligomer, using a semi-
expressed as empirical molecular orbital method. The calculated values were
Z /E2S almost proportional to the measured values of the threshold
q2,1 minrD2 E, rA1 Ef2 Ef1f1 Eg dE energy of photoemission, which corresponded to the effective
/E1 S work function of polymers. This result showed that quantum
a
Fig. 7. Model of interface between polytetrauoroethylene (PTFE) and aluminum
(Al) for quantum chemical calculation (Yoshida et al., 2006). Fig. 9. (a) Electron distribution and (b) variation of the distribution (the solid and
broken lines mean, respectively, increase and decrease of electrons by contact
charging) (Yoshida et al., 2006).
20
chemical calculation can be used to understand the charge
transfer between surfaces.
10
Relative energy (eV)
2.2. Ion transfer
0 HOMO level
In the early 1960s, ion transfer caused by the adsorption of
water on a solid surface was believed to be the major mechanism
of charge transfer. However, water vapor affects other surface
10
properties, e.g. increasing the surface conductivity and lowering
the electrical breakdown strength of air. Therefore, the charge
transfer and the maximum potential are lowered in moist
20
atmospheres (Lee, 1994).
0 2 6 4 8
The mechanism of ion transfer has regained attention in recent
Density of states (eV1) years. In the electrophotographic industry, the external addition
Fig. 8. Electron density of surface states of PTFE contacting with aluminum of ionic charge control agents (CCAs) is widely used to accelerate
(Shirakawa et al., 2008). and control the charging process. These additives generally
consist of a mobile ion and an immobile ion. For example,
substances with cationic group have a positive charging property,
while ones with anionic groups have a negative charging
chemical calculations are applicable to the evaluation of the
property. The performance of these additives has been examined
triboelectric charging of polymers.
(see Section 3.3); however, their roles in triboelectric charging
Yoshida et al. (2006) and Shirakawa et al. (2008) studied the
remain obscure. As an experimental approach to observe ion
charge transfers for a polymermetal contact system using
transfer, secondary ion mass spectrometry (SIMS) was used
another molecular orbital method, paying attention to surface
(Mizes et al., 1990). Also, the scanning probe microscopy
defects. When an atom is missing a neighbor to which it would be
(SPM) was applied to study triboelectric charging, showing
able to bind, a dangling bond occurs. Such defects can be made
that the ionic species in the insulators can inuence the
during frictional contact. A model of surface contact used for the
charging (Saurenbach et al., 1992). To make clear the mechanism
calculations is shown in Fig. 7. The molecular chains of
of the ion transfer, further experimental studies should be
polytetrauoroethylene (PTFE), are arranged perpendicular to
conducted.
the surface of aluminum Al, and dangling bonds are placed at the
interface. The densities of the states (DOS) of the PTFE with
the dangling bond obtained by the quantum chemical calculation 2.3. Material transfer
are shown in Fig. 8 (Shirakawa et al., 2008). A zero value of the
vertical-axis is equal to the HOMO level of the PTFE. Compared The impact and friction between two bodies can result in a
with the work function of metal, the charging tendency of the transfer of materials from one to the other, which can be
PTFEmetal contact can be analyzed. The electron distribution of fragments of the bodies as well as contaminated small dusts or
the interface of the PTFE obtained by the calculation is shown in impurities on the surfaces. For instance, when a metal object
Fig. 9(a) and the variation of the distribution by contact charging slides over a polymer surface, a certain amount of polymer
is shown in (b). The solid and broken lines mean, respectively, will transfer to the metal surface and the metal can also transfer
increase and decrease of electrons. This gure indicates that the to the polymer. If the transferred material carries charge,
electrons of the carbon atom (C) increase. In general, it is believed charge transfer will occur (Lowell and Rose-Innes, 1980).
that PTFE easily accepts negative charge because of the high When brittle particles impact on a metal wall, elements of
electronegativity of uorine atom (F); however, this result shows particles are easily transferred on the metal (Tanoue et al., 1999).
that the C atom with dangling bond is more effective to transfer However, there is currently no theoretical model of charge
electrons than the uorine atom. Although the quantitative transfer and its equilibrium state with such mass transfer
analysis is still limited, it is expected that the quantum mechanism.
3. Chemical concept and environmental conditions 300

MgO
3.1. Inorganic compounds Eu2O3
ZnO
200 Nd2O3 PbO
The work function or energy band concept has been widely Tm2O3 HgO
applied to explain the contact charging between metals Al2O3
Specific charge (mC kg1)

D 2O3
Dy
(Harper, 1967). For insulators, it is, however, difcult to predict 100 Y2O3
In2O3
the charging tendency from this concept because of the lack CeO2
of the data for band structures. For a practical approach to CuO Ga2O3
predict the charging tendency of insulators, several methods 0 Sc2O3 ZrO2
have been developed (Oguchi and Tamatani, 1986, 1993). As2O3
Cr2O3
The amount of charge on particles after contact with other kinds
Fe2O3
of particles was measured using a blow-off method (see 100
Section 5.2). The results for metal oxidized particles are shown
in Fig. 10. The specic charge, i.e. charge to mass ratio of particles, TiO2
shifts linearly to a negative direction with an increase in the 200 Nb2O3
GeO2
generalized electronegativity wi for constituent metal ions, which
SiO2
is dened as V2O5
300
wi l 2Zw0 , 7 0 10 20
Electronegativity i
where Z is the valency of the metal ion and w0 is the Paulings Fig. 10. Relationship between specic charge obtained by blow-off method and
electronegativity for the metal element. The same tendency is generalized electronegativity wi for metal ions (oxidized ne particles: 0.30.5 mm
observed for metal uorides and metal suldes. in diameter; reference particles for contact charging: surface oxidized iron,
4474 mm in sieve diameter) (Oguchi and Tamatani, 1986).
3.2. Organic compounds

Specific charge of metal beads (C kg1)
10
Shinohara et al. (1976) approximated the electronic states of Salicylaldehyde anils
polymer with the chemical constants such as e value in
copolymerization,1 Hammett substituent constant,2 and the HO
ionization potential, and investigated the relationship between N
the triboelectric charging and the electronic state. Gibson (1975,
1984) investigated the substituent effect on triboelectric charging Substituted
X polystyrenes
using a cascade method, where metal beads (steel or zinc) were
+CHCH 2+ n
fed at the top of an inclined plate coated with a lm made of 1
salicylaldehyde anils or substituted polystyrenes, and the specic
charges of the metal beads were measured with a Faraday cage. X
The logarithms of the absolute values of the specic charge
as a function of the Hammett substituent constant are
shown in Fig. 11. Good linear correlations exist both for the
anils and the polystyrenes; thus, the triboelectric charging is
likely to be related to the molecular structure. The triboelectric
charging is regarded as a solid state electrochemistry, where there
0.1
is no transport medium (electrolyte) and also the reaction
0.5 0 0.5 1
depends on physical contact. In both gas and liquid phases,
Hammett substituent constant
1 Fig. 11. Triboelectric charging as a function of substituent constant for

e values of the Q e concept of copolymerization proposed by Alfrey and
salicylaldehyde anils (1) using 250 mm steel beads at 0% relative humidity,
Price (1947) represents the electron density on the double bond, and accounts for
organic lms were positively charged, and for substituted polystyrenes and
the polar effect of substituents in monomers; its value is negative when
(2) using zinc powder at 56% relative humidity, organic lms were negatively
substituents are electron-repelling and positive when they are electron-attracting.
2 charged. X is a substituent (Gibson, 1975).
s values of Hammetts rule is called the substituent constant and is
determined by the type and position (m- or p-) of the substituent regardless of the
kind of reaction; its value tends to be negative for electron-repelling groups and
positive for electron-attractive groups.
molecular orbital energy levels can be correlated with Hammett
Substituent constants (s values). substituent constants.
Substituent para effect meta effect Gibson (1975, 1984) also explained the triboelectric charging
by a model based on electron transfer. According to the model,
NH2 0.66 0.16 when the energy level of the highest occupied molecular orbital
OH 0.37 0.12
(HOMO) of an organic solid is higher than the Fermi level of the
CH3 0.17 0.07
H 0 0 metal, the organic solid is positively charged; while, when that of
Cl 0.23 0.37 the lowest unoccupied molecular orbital (LUMO) is lower than
COCH3 0.50 0.38 the Fermi level of the metal, it is negatively charged. The direction
CN 0.66 0.56 of electron transfer for metalorganic contact is shown in Fig. 12
NO2 0.78 0.71
(Diaz and Guay, 1993).
Energy control agents (CCAs) either on toner surfaces or to bulk.

LUMO
Examples of surface charge control agents are fumed silica and
HOMO highly uorinated polymeric material such as polyvinylidene
Fermi level uoride. Bulk charge control agents are blended into the polymer.
LUMO For positive bulk CCAs, amine and quaternary ammonium salts
are used. The counterions can be halogens (Cl , Br , I ),
HOMO uoroborates (BF4 ), sulfates (RSO4 ) or sulfonates (RSO3 ). For
Organicmolecules Metal Organic molecules negative bulk CCAs, metal complex dyes are used (Schein,
(negatively charged) (positively charged) 1992/1996; Mazumder, 1999). Although little information ap-
pears available on how these charge control agents affect the
Fig. 12. Direction of electron transfer for metalorganic contact.
triboelectric charging, the charging process can involve transfer of
the counterion on the toner surface to the carrier surface upon
contact. Some negative CCAs such as chromium or cobalt complex
azo dyes have a tendency to remove a proton, depending on the
Alumina content ()
molecular structure. Moreover, if they depend on the acidic nature
0 0.05 0.1 and a water layer is present on the surface, H + can move in the
200
aqueous layer between the contact surfaces (Schein, 1992/1996).
A class of compounds such as a polyester salt has the
advantage to stabilize triboelectric charge. These agents are called
100 charge stabilizers. In comparison with the charge control agents,
the charge stabilizers show lower charging magnitude but long-

term charge stability. When used in combination with a CCA, the
0 controllability of the triboelectric charging is further improved
(Michel et al., 2001).
100 3.4. Moisture and temperature
Environmental conditions such as relative humidity and

200 temperature affect triboelectric charging. Greason (2000) mea-
sured the charge on a metal sphere with a Faraday cage after
contact with an insulating material under various conditions. The
300 relationships between the charge and relative humidity for three
0 0.05 0.01 temperatures are shown in Fig. 14. At a given temperature, the
Silica content () charge decreases with increasing relative humidity. This can be
Fig. 13. (a) Charge as a function of alumina content for alumina-coated titania and due to increased leakage caused by a decrease in the electric
(b) contribution of silica for aluminasilica-coated titania (Oguchi and Tamatani, resistance on the surface. Also, it is true that the charge decreases
1993). more rapidly where the charged particles are kept at a high relative
humidity (Nomura et al., 2003). The effects of temperature on the
charge on the sphere are shown in Fig. 15. The charge tends to
decrease with increasing temperature. This tendency is remarkable
3.3. Surface treatment and charge control agents when the relative humidity is low (Greason, 2000).
Surface modication of particles is of great importance for

industrial applications, since surfaces affect various properties such 4. Mechanism of particle charging
as triboelectric charging, adhesion, friction, and owability. One of
the simple and effective methods for the surface modication is to 4.1. Condenser model
coat it with different materials. To investigate the effect of the
coating materials and their contents on contact charging, titania A schematic illustration of the condenser model of particle
particles of 0.150.3 mm in diameter coated with alumina and silica charging by contact is shown in Fig. 16. The contact region
were brought into contact with iron. The experimental results are
shown in Fig. 13 (Oguchi and Tamatani, 1993). The contact charge 25
shifted toward the positive direction with an increase in the alumina
content, while the silica contributed to a shift toward the negative 10C
direction. 20C
20
Charge (nC)
For polymeric materials, the molecular structure affects the 30C

charging characteristics (Shinohara et al., 1976; Lowell and
Akande, 1988); thus, the contact charge can be controlled by
changing the molecular structure of the surfaces. Irradiation with 15
ultraviolet rays (Uyama and Ikada, 1990) and plasma processing
(Kodama et al., 1993) are useful methods to change the charging
characteristics of polymers.
In electrophotography, pigmented polymer particles called 10
0 0.2 0.4 0.6 0.8
toner are used to reproduce an image on paper. The toner
particles are strongly required to have a denite amount and Relative humidity ()
polarity of charge since the particles have to be properly moved Fig. 14. Triboelectric charge of a metal sphere,12.7 mm in diameter, as a function
by the electrostatic force. This can be achieved by adding charge of relative humidity (Greason, 2000).
25
Relative humidity
before contact
0.1
0.3
20
Charge (nC)
0.5
0.7 contact and
charge transfer
15
separation
10
10 20 30
Temperature (C)
Fig. 15. Triboelectric charge of a metal sphere, 12.7 mm in diameter, as a function charge relaxation
of temperature (Greason, 2000).
Fig. 17. Schematic illustration of contact process with charge relaxation.
Critical z0
gap z0 Particle S Paschen curve
Contact potential
Paschen's break down potential

or potential of particle surface
difference Vc
Contact Potential due to
area S remaining charge
Beginning point
of charge
Fig. 16. Condenser model. relaxation
Terminal point of
charge relaxation
between a particle and a wall is regarded as a capacitor. When a
particle impacts and rebounds on a wall, the contact time is short,
Potential curve of an
but still long enough for the charge transfer; thus, the transferred approaching particle
charge Dq caused by impact can be represented by the condenser
model, i.e. (Matsusaka et al., 2000): Contact gap
Dq kc CV , 8 Fig. 18. Schematic illustration of the charge relaxation model.
where kc is the charging efciency, C is the capacitance, and V is

the total potential difference. The capacitance C is given by
4.2. Charge relaxation model
e0 S
C , 9
z0 Matsuyama and Yamamoto (1995a, b) proposed another
charging model, named charge relaxation model. The concept
where e0 is the absolute permittivity of gas, S is the contact area
of the charge relaxation process is sequentially shown in Fig. 17;
and z0 is the critical gap including the geometrical factors
i.e. (a) there are two neutral bodies separated; (b) when the
between the contact bodies. The total potential difference V at
bodies are brought into contact with each other, charge is
the contact gap is given by
transferred through the contact gap; (c) the bodies are in
V Vc Ve Vb Vex , 10 separation; (d) relaxation of the transferred charge occurs due
to gas discharge in the separation process (Matsusyama and
where Vc is the potential difference based on the surface Yamamoto, 2006a).
work functions, Ve is that arising from the image charge, which A quantitative scheme of this model is indicated in Fig. 18. To
is given by determine the breakdown voltage in the gap, the Paschen curve is
Ve ke q, 11 applied, which is widely used in the air insulating technology to
give the gas breakdown limit voltage between two parallel
where q is the particle charge held on the particle before contact. electrodes as a function of pressure and gap distance (Paschen,
Vb is that arising from the space charge caused by surrounding 1889). Now, it is assumed that a particle with a small initial
charged particles, which is given by charge approaches a metal wall, and a charge is transferred by a
contact with the wall. When the charge transferred on the particle
Vb kb q: 12
is high enough to make the potential curve intersect the Paschen
Vex is the potential difference arising from other electric elds. For curve in the separation process, the charge relaxation occurs due
instance, an external electric eld may be applied in this system. to gas breakdown, and the potential curve with the residual
When the wall is insulator, the wall surface can have charge and charge leaves the Paschen curve downwards after the process. As
form an electric eld, which affects the total potential difference. a result, the total charge remaining on the particle becomes larger
If the charge is accumulated by contact charging, the total than the initial charge. This difference of the charge is the net
potential difference will decrease with increasing surface charge transferred charge to be observed. On the other hand, when the
(Matsusaka et al., 2003a). particle approaching the wall has a large amount of charge which
gives the potential curve exceeding the Paschen curve, the particle Particle feeding
releases the surplus charge by the gas breakdown before impact.
Compressed air
The charge remaining on the particle is kept after impact since the
potential curve does not exceed the Paschen curve. Therefore, the Sintered plate
released charge is equivalent to the net transferred char-
ge. The remaining charge depends on the dielectric constant Glass tube
and the diameter of the particle as well as the breakdown voltage
of the gas.
Through-type
4.3. Impact on a wall Digital storage
Faraday cage oscilloscope
(FC1) Pre. Amp.1
In powder handling operations, individual particles acquire
charge during collision with walls. An understanding of the Photodetector Transducer
charging process of a single particle is a basic requirement for the PC
development of a theory of triboelectric charging of particles PI
(Bailey, 1993). Several studies have been reported, in which a
single particle of a few millimeters in diameter was made to Faraday cage
collide with a metal target, and the transferred charge was (FC2) Target Pre. Amp.2
measured (Masui and Murata, 1983, 1984; Yamamoto and Insulator Sieve mesh
Scarlett, 1986; Matsuyama and Yamamoto, 1989, 1994,
1995ac, 1997). Also, single-particle experiments with a larger Sintered plate
sphere, 31 mm in diameter (Matsusaka et al., 2000) and with a
particle as small as 100300 mm (Matsuyama et al., 2003) were Vacuum line
performed. These methods have several advantages, i.e. the
Fig. 19. Impact charging test rig for single particles.
contact state during the particle collision can be reproducible by
controlling the impact velocity and angle. An example of the
impact charging test rig for single particles is shown in Fig. 19
(Watanabe et al., 2006, 2007a, b). A particle is fed into the funnel Impact charge q
at the top of the rig and is accelerated by air owing through a Equilibrium charge qe
glass tube. The particle then collides with a metal target in a Initial particle charge qi
collection chamber. To measure the particle charge before impact,
an open ended (through-type) Faraday cage (FC1) is installed. The 16 pC q0
second Faraday cage (FC2) is installed in the collection chamber to 0.5 pC
measure the particle charge after impact. The amount of charge v = 7 ms1
transferred during the impact, i.e. impact charge, is quantied by 0.9 pC
Characteristic line
subtracting the initial charge (measured by FC1) from the particle v = 9 ms1
charge after impact (FC2). The relationship between the impact
1.6 pC
charge and initial charge has a linear trend, as shown in Fig. 20.
The impact charge at zero initial charge Dq0 is a characteristic v = 13 ms1
charge, and it increases with the impact velocity. The intercept on
the horizontal axis is the equilibrium charge qe, where there is no
net charge transfer. Dq0 and qe are important characteristics of the Fig. 20. Linear relationships between impact charge and initial charge (spherical
sugar granules, 500600 mm in sieve diameter) (Watanabe et al., 2006).
charging tendency of particles. The characteristic line with these
values is expressed as the following equation:

q
Dq Dq0 1 i : 13 20
qe
EC: +10 pC
Equilibrium charge (pC)
The equilibrium charges of particles made of different materials

0
as a function of impact velocity is shown in Fig. 21. Although there SG: 16 pC
are some uctuations in the data, the equilibrium charge is
independent of the impact velocity but depends on the material. 20
The effect of impact angle on the triboelectric charging was LM: 18 pC
investigated by Ema et al. (2003). The particle impact tests using
an inclined target and a rotating target are illustrated in Fig. 22. As 40
particles impact on the metal target in a stream, the value of
ASP: 40 pC
electric current generated from the target is represented by the
following equation: 60
0 10 20 30
W Impact velocity (ms1)
I Dq, 14
mp
Fig. 21. Equilibrium charge as a function of impact velocity (EC: ethylcellulose
where W is the mass ow rate, mp is the mass of a particle, and Dq granules, SG: sugar granules, aLM: a-lactose monohydrate, ASP: aspirin)
is the impact charge of a particle. The relationship between (Watanabe et al., 2006).
normalized current I/Imax and impact angle y is shown in Fig. 23,
in which Imax is the maximum current at a constant normal tendency can be explained by a rollingslipping model. For
component of impact velocity. The electric current increases with y r60o, the effective contact area increases with the angle
the impact angle up to 60o and then decreases. This charging because of the increase in the rotation of the particle on the
h P P
h
r
vt0 P
vt0 P
vn0 v0
v0 vn0 P P
Fig. 22. Particle impact tests: (a) Inclined target system and (b) Rotating target Fig. 24. Effective contact area: (a) Rolling and (b) Rolling & slipping.
system.
1.0
Normalized electric current ()
0.8 Electrometer 2
Electrometer 1
0.6 Target plate 2
Target plate 1
1
vn = 2.5 m s
0.4
vn = 1.8 m s1
vn = 1.1 m s1
0.2 0 < vt < 31m s1 Through-type Digital storage
Faraday cage oscilloscope
0
0 15 30 45 60 75 90
Impact angle (deg)
Fig. 23. Relationship between normalized electric current I/Imax and impact angle Particle
y (vn and vt: normal and tangential component of impact velocity) (Ema et al.,
2003). Fig. 25. Electrostatic charging test for successive particle impacts.
target (Fig. 24(a)). As for y 460o, the effect of the slip on the target
increases with the angle; thus, the effective contact area 10
decreases (Fig. 24(b)). Polystyrene particle
Dp = 200 m
5 P = 100 Torr
Impact charge (pC)
4.4. Repeated impacts of a single particle v = 30 ms-1
When particles repeatedly collide with a wall, the charge on 0

the particles varies according to the electrostatic properties
and the state of the collisions; thus, it is important to trace the
variation of the charge on a particle. As a rst step to analyze 5
successive impact charging, single-particle experiments were
carried out using two metal targets (Fig. 25) (Matsuyama
and Yamamoto, 1995c). The charge generated by the rst 10
impact affects the next impact charging. For polymer particles, 10 5 0 5 10
the surface charge caused by impact is non-uniform, and Initial charge (pC)
the localization on the surface affects the successive particle
Fig. 26. Impact charging with 300 mm polymer particles (Matsuyama et al., 2003).
charging.
The relationship between the impact charge and the initial
charge of 200 mm polymer particles is shown in Fig. 26
(Matsuyama et al., 2003). The wide scatter on the data is rest point using a ne string, and then was dropped onto a metal
thought to be caused by the non-uniform surface charge. The plate. After impact on the plate, the string was then pulled up to
substantial effect of the localization of the surface charge on the prevent the sphere from falling again. This impact test was repeated
impact charge can be explained (Fig. 27). If charged area contacts many times, and the initial charge on the sphere and the transferred
with a wall, the impact charge will strongly depend on the initial charge were measured with a through-type Faraday cage.
charge; however, if the charged area is at the remote side, there is The variation in the charge on the sphere is shown in Fig. 29.
no effect. The transferred charge caused by an impact decreases with the
Repeated impact tests to study the charge accumulation were number of impacts and the accumulated charge approaches a
carried out by Matsusaka et al. (2000) (Fig. 28). To control the limiting value, which tends to decrease as the time interval
contact area easily, a larger sphere (31 mm in diameter), made of between impacts increases. This is because the leakage of the
synthetic rubber was used. The sphere was supported at its initial electrostatic charge increases with increasing elapsed time.
Initial charge at the contact 20

surface (overall relaxation) Impact charge
Theoretical
Charge of sphere (nC)

Relaxation of excess charge 10
(over equilibrium)
0
Theoretical line for uniform Initial charge Interval of impact
charged particle 10 s
20 s
Initial charge at the 30 s
opposite side (no effect) 10
0 5 10 15 20
Fig. 27. Concept of localization of initial charge on a particle. Number of impacts ()
Fig. 29. Charge of synthetic rubber sphere by repeated impacts (Matsusaka et al.,
2000).
String
Sphere
and
Metal foil for Through-type Vc
electric shielding q1 : 21
Faraday cage ke kr z0 =kc e0 Sf
High-speed
Impact It should be noted that the same form as Eq. (19) can also be
Digital oscilloscope camera
plate derived from the charge relaxation model in a phenomenological
Flash
Window level. By replacing qe and qi in Eq. (13) with qN and q, respectively,
the transferred charge per collision dq/dn is given by
dq Dq0
q Dq0 : 22
Trigger cable
dn q1
Data transfer cable Solving Eq. (22) with the initial conditions, n 0 and qq0 gives
Computer
the same form as Eq. (19). Here, n0 in Eq. (19) is given by
Fig. 28. Test to investigate the charge accumulation process by repeated impacts q1
of an elastic sphere. n0 : 23
Dq0
Comparing the charge relaxation model with the condenser model,
It is possible to formulate the particle charge generated by
one can notice a difference. For the condenser model, n0 is
repeated impacts. First, the condenser model (see Section 4.1) is
independent of the contact potential difference Vc (see Eq. (20)),
applied to the formulation. In order to obtain the charge qc as a
but qN is proportional to Vc (see Eq. (21)). On the other hand, for the
function of the number of collisions n, a continuous quantity
charge relaxation model, both n0 and qN are independent of Vc.
dqc/dn is used, i.e.
The exponential equation represented by Eq. (19) can be used
dqc for almost all the repeated impacts of a single particle. The effect
kc CV : 15
dn of the differences in experimental conditions including particle
diameter should be considered in the two terms, i.e. qN and n0.
The leakage of electrostatic charge dqr /dt is approximated by
(Itakura et al., 1996)
4.5. Particle charging in gassolids pipe ow
dqr
kr q, 16
dt In gassolids pipe ow, particles repeatedly collide with the
where kr is a constant. When the frequency of particle collisions is inner wall, as a result of which charge transfer takes place. When
dened as f, Eq. (16) is rewritten as a metal pipe is grounded, the charge transferred from the particles
to the wall ows to earth, and can be detected as electric currents
dqr kr (Masuda et al. 1976, 1994; Cartwright et al., 1985; Nieh and
q: 17
dn f Nguyen, 1988; Gajewski, 1989). A system to analyze the particle
charging in gassolids pipe ow is shown in Fig. 30 (Matsusaka
For simple repeated impacts of a single particle, the values of Vb
et al. 2008a). Particles that are continuously fed are dispersed into
and Vex in Eq. (10) are zero; therefore, from Eqs. (9)(11), (15) and
airow through an ejector. To increase the efciency of the
(17), the net charge transfer is derived as
particle charging, spiral pipes can be used instead of straight

dq dqc dqr kc ke e0 S kr kc e0 SVc pipes. The electric currents owing from the metal pipes to earth
q : 18
dn dn dn z0 f z0 are measured with an electrometer. To conrm the validity of the
charge balance in this system, the charge-to-mass ratio, i.e.
Solving Eq. (18) with the initial conditions, n 0 and qq0, gives specic charge, of particles at the inlet and outlet of the metal
the following exponential equation: pipe must be measured by connecting them to a Faraday cage.
Also, a digital oscilloscope can be used to analyze the pulsating
n n
q q0 exp q1 1exp , 19 electric signals of the charge transfer (see Section 6.5.2).
n0 n0
The electric currents generated from a metal pipe are shown in
where Fig. 31(a) (Masuda et al., 1994; Matsusaka and Masuda, 2006).
During the operation, the measured values are almost constant
1 (I 2.3 nA for Wp 7.0 10 7 kg s 1). The amount of charge of
n0 20
kc ke e0 S=z0 kr =f the particles that is being collected in the Faraday cage under the
Mist Electromagnetic shield

Compressor separator Pre-charge Detection pipe 10
pipe I1 = 2.3 nA
Insulator
Condenser I/Wp = 3.3mC/kg
Bag filter
Current (nA)
5
Ejector
A C B 0
Table feeder
Feed (fly-ash)
Faraday cage Average air velocity: 30 ms1
A -5
Blower 0 200 400 600
B Time, t (s)
Computer 60
Electrometer
C
qmIN = 3.9 mCkg1
qmOUT = 0.6mCkg1
Inlet
Charge (nC)
40 qmIN qmOUT = 3.3mCkg1
Digital oscilloscope
Fig. 30. System to analyze particle charging in gassolids pipe ow.

Outlet
20
same conditions as above is shown in Fig. 31(b). The values of the Flyash
specic charge at the inlet and outlet of the detector 0
0 20 40 60
are qmIN 3.9 10 3 and qmOUT 0.6 10 3 C kg 1, respectively.
Time (s)
The difference between these values qmIN qmOUT is equal to the
current generated per unit mass ow rate, I/Wp 3.3 Fig. 31. (a) Electric currents generated from a stainless steel detector and
10 3 C kg 1. This means that the charge balance is satised and (b) charges on particles at the inlet and outlet of the detector, which were
the system is working correctly. measured with a Faraday cage (Matsusaka and Masuda, 2006).
When the effect of particleparticle interactions on the particle

charging is negligible in dilute phase gassolids pipe ow, each
particle can freely collide with the inner wall. Under such Mass flow ratio ()
conditions, the electric current is proportional to the mass ow
10-3 10 -2 10-1 1 10
rate of particles (Fig. 32) (Masuda et al., 1998a). For dense-phase 10 -9
gassolids pipe ows, the surrounding particles prevent the
qm0 = 3.7 ~ 1.0 mCkg1
free particlewall contacts, and consequently the efciency of
Generated current (A)
the charge transfer is reduced. Also, for smaller particles, the

efciency decreases because of the agglomeration. In addition, the 1
10-10
initial charge on particles affects the electric currents as shown in 1
Fig. 33 (Masuda et al., 1994). In powder handling operations,
particles will collide with different walls before arriving at the
current detection pipe, e.g. hopper, feeder, chute, disperser, etc., 10-11 Quartz sand: 320 m
and hence the polarity and amount of charge on particles vary Average air velocity: 30 ms1
according to the conditions. To estimate the charge transferred
from the particles to the wall, the initial charge has to be known
10-12
beforehand. 10-6 10-5 10-4 10-3 10-2
The particle charging in gassolids pipe ow can be formu-
Particle flow rate (kgs1)
lated. When a particle moves from x to x+ Dx along the pipe axis,
the variation of the charge is derived from Eq. (19) as follows: Fig. 32. Relationship between generated current and particle ow rate (Masuda
et al., 1998a).
nx nDx
Dq qx Dxqx q1 q0 exp 1exp :
n0 n0 s redened as zero, Eq (25) becomes
24
I nDx
qm0 qm1 1exp : 26
The charges transferred from the particles to the pipe wall can be Wp n0
analyzed by the electric currents. When a length of Dx is isolated
electrically and grounded, the electric current I owing to earth is When the value of Dx is quite small and n(Dx)5n0, Eq. (26) is
expressed as (Matsusaka and Masuda, 2006): simplied as

I Dq nx nDx I nDx
qm0 qm1 exp 1exp qm0 qm1 , 27
Wp mp n0 n0 Wp n0
25 where
where Wp is the mass ow rate of particles, mp is the mass of the
6Vc
particle, qm0 and qmN are the specic charge at x 0 and x N, qm1 , 28
respectively. When the point of x at the inlet of the detection pipe prp D3p ke kb kr z0 =kc e0 Sf
6 The above theoretical approach can be used for analyzing the

Flyash charge distribution (Matsusaka et al., 2002). Although the particle
Stainless steel pipe
Inner diameter: 0.006 m charge distribution depends on various factors, the main factors
Length: 0.8 m are considered to be the number of particle collisions, initial
charge on the particles, and the state of the impact. Introducing
4 the probability density functions of these factors, one can obtain
= 0.3 mC kg1
I/Wp (mC . kg-1)
the equation of particle charge distribution (Appendix C).

Furthermore, the maximum (or the equilibrium) charge of
particles in gassolids pipe ow was studied by Matsuyama and
+
Yamamoto (2008). They conducted a theoretical calculation based on
2 - the charge relaxation model, taking into account the space charge
effect and compared with the experimental data cited from literature.
Average air velocity: 30 m s1

Particle flow rate: 0.33.6 mg s1 4.6. Control of triboelectric charging
0
0 2 4 6
qm0 (mC kg1) In general, the reproducibility of the triboelectric charging of
particles is poor; however, the control of the charge on particles is
Fig. 33. Effect of the initial charge per unit mass of particles (q/mp)0 on the current
possible using the triboelectric charging principles. In this section,
generated per unit mass ow rate of particles I/Wp (Masuda et al., 1994).
typical triboelectric characteristics in dilute phase gassolids pipe
ow is shown rst and then useful methods to control tribo-
where rp is the particle density and Dp is the particle diameter.
electric charging are explained.
The constants of ke regarding image charge and kb regarding the
Matsusaka et al. (2007) conducted experiments for triboelectric
space charge in Eq. (28) are, respectively, approximated by
charging of micrometer-sized particles using ve different kinds of
(Appendices A and B)
pipes. The relationships between the specic charge of particles and
the pipe length are shown in Fig. 34. The particles were charged
2z0
ke 29 positively by contact with the walls of the two kinds of stainless
pe0 D2p
steel. As for aluminum, copper, and brass pipes, the particles were
charged negatively. Among them, the charging level for the
and aluminum pipe was rather small. Although the absolute value of
3z0 mrg Di u the specic charge increased with pipe length, the rate of increase
kb , 30 gradually decreased and the specic charge approached an
2pe0 rp D3p v
equilibrium value depending on the wall material. In order to
where m is the mass ow ratio of particles to gas, rg is the density apply the theoretical model to the experimental results, it can be
of the gas, Di is the inner diameter of the pipe, u is the average gas assumed that the frequency of the particlewall impacts per unit
velocity and v is the average particle velocity. When the leakage of pipe length is constant, and that the number of impacts n is
the electrostatic charge is sufciently small compared to the proportional to the pipe length L; therefore, Eq. (19) is rewritten as
particle charging, kr can be neglected. Using Eqs. (29) and (30), n0
and qmN are expressed as L L
qm L qm0 exp qm1 1exp , 37
L0 L0
pD2p
n0 31
2kc S1 a where L0 is the characteristic length of the particle charging. The
and experimental results were in good agreement with the calculated
ones (Fig. 34). Also, Eq. (37) can be used to evaluate the particle
3e0 Vc charging efciency gq, i.e.
qm1 , 32
rp Dp z0 1 a
q q L
where a is the ratio of the space charge effect to the image charge gq m m0 1exp : 38
qm1 qm0 L0
effect, i.e.
3 rg Di u 15
a m : 33
4 rp Dp v Pipe material: Particles: Alumina
Stainless steel (SUS 316)
10
For dilute phase ow (a 51), Eq. (27) is rewritten as Stainless steel (SUS 304)
Aluminum
I 5
aqm0 b, 34
Wp
0
where
2nDxkc S 5
a 35
pD2p
10 Copper
and Brass
15
6e0 Vc nDxkc S
b : 36 0 1 2 3
pD3p rp z0 Pipe length (m)
From Eq. (34), it is found that the transferred charge is Fig. 34. Relationships between specic charge of micrometer-sized particles and
proportional to the initial charge of particles. This relationship pipe length (dilute phase gassolids pipe ow systems, average air velocity:
was conrmed experimentally (see Fig. 33). 40 m s 1; solids lines: using Eq. (37)) (Matsusaka et al., 2007).
10 Pipe material: stainless steel


10
Experimental Theoretical
5
5
Brass Brass
0 0
Theoretical
5 5 Stainless Stainless
steel steel
10
10 Pipe material: brass
0 1 2 3 4 5
0 1 2 3 Total pipe length, L (m)
Length of second pipe, L (m)

Fig. 35. Effect of initial charge on particle charging (Matsusaka et al., 2007). 10
Experimental Theoretical
5
Brass Brass
The particle charging efciency for 3-m pipe in this experiment is
0
found to be in a range from 93% to 99%.
The effect of the initial charge on particle charging is shown in 5 Stainless Stainless
Fig. 35. The experimental results were obtained using two steel steel
different pipes made of stainless steel and brass. Although the 10
particle charging depends on the initial charge on particles and 0 0.5 1 1.5 2 2.5
pipe materials, all the experimental results agree well with the Total pipe length, L (m)
results calculated using Eq. (37). These results imply that the
Fig. 36. Control of particle charging by a system combining two different pipe
triboelectric charging of particles can be controlled by arranging
materials (Matsusaka et al., 2007).
pipes made of different materials.
The experimental result for the specic charge obtained by
connecting 1-m brass pipe and 1-m stainless steel (SUS316) pipe
alternately is shown in Fig. 36(a). The particles are charged 10
negatively in the brass pipes but positively in the stainless steel
pipes. As a result, the values of the specic charge are within a 5

certain range. The experimental results are in good agreement
with the solid lines calculated using Eq. (37). The result obtained
by 0.5-m pipes instead of 1-m pipes is shown in Fig. 36(b). It is 0
clear that the specic charge in Fig. 36(b) is within a narrower
range compared with that in Fig. 36(a). 5
Examples of the general calculation to control the triboelectric
charging in gassolids pipe ow using two different pipe 10
materials A and B are shown in Fig. 37. Although the charge is 0 1 2 3 4 5
uctuated positively and negatively, the uctuation level
decreases with decreasing each pipe length. The upper limit Pipe length (m)
10
qmA for pipe A and the lower limit qmB for pipe B are given by
n
o
n
o
LA LA
qmA1 1exp DLA0 qmB1 exp DLA0 LB
1exp DLB0 5
qmA *

LA
1exp DLA0 LB
DLB0
39 0
and
n
o
n
o 5
LA
LB
qmB1 1exp DLB0 qmA1 exp DLB0 LB
1exp DLA0
qmB *
,
LA
1exp DLA0 LB
DLB0 10
0 1 2 3 4 5
40 Pipe length (m)
where DLA and DLB are the pipe lengths, LA0 and LB0 are the Fig. 37. General calculations of the triboelectric charging in gassolids pipe ow
characteristic lengths, and the subscripts A and B denote the two using two different pipe materials.
different materials. Also, the polarity and amount of charge
can be controlled by changing the pipe length ratio of A to B
(r DLA/DLB). As the pipe length DLA, DLB-0, the limiting values
of the charge satisfy the relation of qm * qmA * qmB *. If two materials are arranged in parallel instead of in series, the
Substituting DLA rDLB into Eq. (39) and setting DLA, DLB-0, charge on particles will approach the limiting value without the
the limiting value qm * is given by uctuations mentioned above. In addition, various shapes and
structures are available for a particle charging control device.
qmA1 rLB0 qmB1 LA0
qm * 41 Matsusaka et al. (2008b) developed a high-efciency particle charger
LA0 rLB0 having an inverted truncated cone. Micrometer-sized particles are
Therefore, the charge on particles can be controlled positively, introduced into the charger from the tangential direction at the top,
negatively, or neutrally using two materials. carried spirally downward, and discharged to the tangential
direction at the bottom. The particles are triboelectrically charged by Power voltage supply
contact with the side wall of the charger due to the centrifugal force. Thermohygrostat
Since two different metal sheets are attached in the side wall,
Electromagnetic
particles can contact with these metals alternately.
shield
5. Characterization of triboelectric charging Vibrating

upper electrode
5.1. Work function and contact potential difference
Powder
When materials are exposed to electromagnetic radiation, Lower electrode

electrons are emitted (photoelectric effect). The number of the
emitted electrons depends on the intensity of the radiation. Also, Computer
the kinetic energy of the electrons depends on the frequency or
wave length of the radiation. The radiation is regarded as a stream
of photons, each having an energy hn, where h is the Planck
constant and n is the frequency of the photon. If the photon Electrometer
energy exceeds the work function of the material f, a photon can
Fig. 38. Measurement of contact potential difference based on the KelvinZisman
eject an electron out of the surface; thus, there is the minimum method.
frequency, i.e. threshold frequency, at which ejection occurs. For
many solid materials, the photoelectric effect occurs at ultraviolet
frequencies or above, and for some materials having low work Oscillating
functions it occurs with visible light. The maximum kinetic energy upper electrode (Au)
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 Capacitor
materials (Murata, 1979; Murata and Kittaka, 1979).
When two dissimilar electrically conductive materials are in Powder
contact, a potential difference is generated between them. This is Lower electrode
called the contact potential difference (CPD). Its origin can be A
described in terms of the process necessary to bring the two
materials into thermal equilibrium, and it is expected that
electrons migrate through the surfaces when they make contact. Fig. 39. Equivalent electric circuit of the KelvinZisman method.
If there is a net ow of electrons, they are electrostatically
charged. Therefore, the measurement of CPD is important to
evaluate the contact charging. In fact, the surface of materials is enclosure isolated electrically, charge is induced in the inner
not pure and is usually covered with an oxide layer, and thus the wall of the metal enclosure. The specic charge of particles is
CPD of the materials may signicantly differ from the values obtained by dividing the charge by the mass of particles. For dust
shown in literature. To measure the CPD between a powder and a in air or particles in gassolids pipe ows, a Faraday cage with a
wall, a measuring system based on the KelvinZisman method lter is used to collect the particles (Fig. 41). Also, an open ended
was developed (Fig. 38) (Yoshida et al., 1991, 1992; Itakura et al., Faraday cage, so-called through-type Faraday cage is available
1996; Tanoue et al., 2001a). The CPD measurement system has an (Matsuyama and Yamamoto, 1989, 1994; Matsusaka et al., 2000;
electric circuit consisting of a vibrating capacitor made up of two Watanabe et al., 2006, 2007a, b). Since this type of Faraday cage
electrodes, a DC bias voltage supply, and an electrometer (Fig. 39). does not obstruct the particle movement, the charge can be
Powder is lled in the concavity of the lower electrode and the measured in situ.
upper electrode made of gold oscillates vertically. When the In electrophotography, to evaluate the triboelectric charging of
voltage applied on the upper electrode is equal to the potential a toner-carrier system called developer, the blow-off method is
difference between the powder and the upper electrode, the used (Fig. 42). In this system, the toner particles are charged by
induced currents detected by the electrometer become zero. The contact with a carrier (larger bead). In the Faraday cage, a mixed
applied voltage is called zero-point potential V0, which is powder of the toner and the carrier is placed on a stainless steel
expressed by the following equation: lter whose mesh size is chosen to be between the sizes of the
toner particles and carrier beads. Nitrogen gas spurts from the
rc dl 2 upper nozzle of the Faraday cage and separates the mixture into
V0 VP=Au , 42
2ep toner and carrier. Only the toner is blown off through the lter
and the carrier with a charge equivalent but opposite in polarity
where VP/Au is the CPD between the powder and the gold
to that of the blown-off toner is left in the cage. The variation of
reference, rc, dl, and ep are, respectively, the volume charge
the charge in the cage is measured by an electrometer (Oguchi
density, the thickness of the powder layer, and the absolute
and Tamatani, 1993).
permittivity of the powder layer. When the charge of the powder
layer is sufciently small (rc E0), V0 is equal to VP/Au.
5.3. Charge distribution
5.2. Specic charge
An aerodynamic technology is widely used to measure particle
In general, the charge on particles is measured with a Faraday charge distribution for ne particles. A typical example of the
cage (Fig. 40) because of the advantages of simplicity and system is shown in Fig. 43. Particles are injected into a laminar
reliability. When charged particles are put into a metal ow, where a perpendicular electric eld is applied, and deposit
Positively charged particles

Laminar flow
Metal y
enclosure Electric field
Electro magnetic
shield x
Positively charged particles
Laminar flow
Powder Electrometer Electric field
V Negatively charged particles
Fig. 43. Measurement of electrostatic charge distribution.
Insulator
Fig. 40. Faraday cage. the particle, Dp is the particle diameter and Cc is the Cunningham
correction factor. Therefore, the charge-to-diameter ratio
qd (q/Dp) is derived as
Electro magnetic
3pm 3pm yd
shield qd vE , 44
Cc E Cc E t
Metal
enclosure where yd is the displacement and t is the time (Schein, 1992/
Insulator Filter 1996).
Mazumder et al. (1991) developed a laser-based instrument
called the electrical-single particle aerodynamic relaxation time
Electrometer
Mesh (E-SPART) analyzer, which measures the aerodynamic diameter
Dp and charge q of individual particles. In the measurement cell,
V electric eld and acoustic eld are formed horizontally (Fig. 44).
When a charged particle passes vertically through an acoustic
eld, the particle oscillates with a phase lag depending on
the particle diameter. Therefore, the particle diameter Dp is
determined by measuring the phase lag y, i.e.
s
Fig. 41. Faraday cage to measure charge on aerosol particles.
18m tan y
Dp , 45
Cc rp o
Electro magnetic where rp is the particle density and o is the angular velocity of
shield the acoustic wave. As a DC electric eld is simultaneously
superimposed in the cell, the charged particle experiences
Metal another velocity component. From the balance between the
enclosure electrostatic force and aerodynamic drag, the charge-to-
diameter ratio qd is obtained using Eq. (44). The particle velocity
Insulator vE is measured by a laser Doppler method. This analyzer is
available for analyzing bipolar charge distribution (Mazumder
et al., 1991; Matsusaka et al., 2003b).
For measuring particle charge, the particle motion analysis
V system (PMAS) is also used (Fig. 45). In the electrostatic
separation chamber, there are two parallel electrodes and an
electric eld is applied horizontally. Particles that are injected into
the chamber travel in the electric eld. The PMAS, which consists
Mesh of a short duration light source, a CCD camera with a zoom lens,
and the data processor, captures pairs of multiple particle images
Fig. 42. Blow-off method. using the double-spark light source with very short pulse interval.
The particle velocity can be obtained from the tracking images.
The amount of charge on particles can be calculated with both the
on the walls of the channel by the electrostatic force. From the velocity in the given electrical eld and the diameter of particles
position of the deposited particles, the amount of the charge can in the images (Ahn et al., 2004).
be determined. A lter placed at the end of the apparatus can be As shown in Fig. 43, charge spectrometers for ne particles use
used to collect the particles with small or no charge. The drag a laminar ow to carry the particles; however, it is not easy to
force on small particles immediately becomes equal to the control the airow and the injection of particles. The disturbance
Coulomb force; thus the force balance is given by of the airow causes a signicant measurement error. To simplify
the construction of the apparatus, and to increase the accuracy of
3 pmvE Dp
qE , 43 the measurement, a free air beam is used for injecting particles in
Cc
an electric eld (Fig. 46). The air in the chamber is sucked to pull
where q is the charge of the particle, E is the electric eld strength, the aerosol particles from an activation cell. The charged particles
m is the viscosity of air, vE is the electrostatic terminal velocity of in the measurement cell are deected onto the electrodes
Particle inlet Pump
Acoustic transducer
(1kHz)
Electrode Measurement
cell
Laser beam Measuring spot Activation

cell Air inlet
Particle outlet
Fig. 46. The q-test device for two-component toners.
Fig. 44. Electrical-single particle aerodynamic relaxation time (E-SPART) analyzer.
Short duration
light source
Particle
generator
Power
supply
Image capturing Data

device Monitor
processer
Fig. 45. Particle motion analysis system (PMAS).
according to the charge-to-diameter ratio. Both polarities can be

determined by one measurement. However, the electric eld
caused by the inhomogeneous geometry leads to a more complex Fig. 47. Charge distribution measurement system consisting of dispersion feeder,
calculation to obtain the particle path through the measurement sampling section, and measurement cell.
cell (Epping and Kuettner, 2002).

Fig. 47 shows a charge distribution measurement system
consisting of dispersion feeder, sampling section, and where xd and yd are, respectively, the x- and y-components of the
measurement cell. The dispersion of particles is necessary to displacement. The particle diameter and the displacement can be
feed single particles into the measurement cell. However, highly measured with an automatic digital image processing system. An
charged aerosol particles readily deposit on the wall. To prevent example of two-dimensional distribution of specic charge and
the charged particles from depositing in the sampling section, an particle diameter measured by this system is shown in Fig. 48. The
AC eld called the electric curtain (Masuda et al., 1972) can be cumulative distribution, dimensionless [0, 1] is given by the
used. In the measurement cell, the particles fall under gravity multiple integral of the frequency distribution.
with moving horizontally by the electrostatic force. The
horizontal and vertical velocities of the particles vx and vy in the
cell are, respectively, given by (Masuda et al., 1993; Matsusaka 5.4. General techniques for characterization
et al., 2008c)
Various techniques have been proposed to characterize the
qECc
vx 46 triboelectric charging of particles. Although these techniques
3pmDp
were developed from practical needs, the high performance of the
and measurement methods made them the object of scientic
interest. A representative method, namely the cascade method
mp gCc is shown in Fig. 49 (Oguchi and Tamatani, 1993). Particles are fed
vy , 47
3pmDp from the top of the reference plate that is held at a certain angle.
where g is the gravitational acceleration. Therefore, the specic The triboelectric charging occurs while the particles cascade
charge q/mp is derived as down the slope. The total amount of charge produced on the
particles is measured with a Faraday cage. Although the polarity
q g vx g xd of the charge on the particles is opposite from that on the plate,
, 48
mp E vy E yd the absolute values are the same as each other.
Hopper
10 9
0.08 Triboelectric
Frequency (kgC -1 m-1)
charging plate
0.06
0.04
0.02 2
0
0 -2 Faraday
2 -4
4
6
cage
8 -6
10
12
14 -8 Fig. 49. Cascade method.
Fig. 48. Two-dimensional frequency distribution of specic charge and particle

diameter.
Vibrating plate
Another useful characterization method is shown in Fig. 50.
The device consists of a vibrating feeder and a charging plate
(Higashiyama et al., 1997). The particles move from one end
toward the other while contacting with the plate. The charge of
the particles falling from the edge is measured with a Faraday
cage.
A uidized bed can be used for a triboelectric charging device Faraday
(Fig. 51) (Iuga et al., 2005). When different kinds of particles are cage
mixed, the triboelectric charging is complicated, i.e. the charge
Fig. 50. Vibration method.
transfer occurs during collision with different types of particle,
with the same type of particle, and with the inner wall, depending
on the initial surface charge on the contact area. To measure the
charge distribution of the particles, they are classied in a charge
spectrometer, such as free fall separator with a horizontal electric
eld. Air output screen
Fig. 52 shows a vertical array of Faraday cage sensors, which is
combined with a uidized bed (Zhao et al., 2002). This
measurement system consists of special Faraday cages mounted
vertically in cascade and a normal one located at the bottom. The
series of Faraday cages have open holes on the upper and lower
covers. The measurements are performed during the free falling of
particles. Each particle experiences gravity and space charge Objects input
repulsion; thus, large particles drop faster in the vertical direction
and highly charged particles move faster in the radial direction. As Objects output
a result, the particles are separated in the Faraday cages. The
charge and mass of particles are measured with each Faraday
Charge
cage. In addition, various analyses such as particle size Clean air
spectrometer
distribution in the Faraday cages are useful for total evaluation
of the particle charging. Fig. 51. Fluidization method.
The characteristics of the triboelectric charging can be
evaluated by various methods; however, the particles may not
be sufciently charged. Kittaka and Murata (1976), Rowley impacting on the target at different incident angles with a high
(2001), and Matsusaka et al. (2008b) proposed a method using a velocity. The analysis of the charge transfer as a function of the
cylinder in which particles can be charged by a circular airow initial charge of the particle determines an equilibrium charge, i.e.
(Fig. 53). As strong centrifugal forces act on the particles in an initial charge for no net charge transfer (see Fig. 20).
motion, the particles continue to rotate in the cylinder until the
airow is stopped. After rotation, the charged particles are taken
out rapidly from the cylinder through a hole using a vacuum 5.6. Gassolids pipe ow test
pump and are led into a Faraday cage.
Charge transfer in dilute phase gassolids pipe ow can be
analyzed from electric currents generated in a length of detection
5.5. Single particle impact test pipe, as mentioned in Section 4.5; i.e. the transferred charge per
unit mass of particles in the pipe is obtained as the value of the
Watanabe et al. (2006, 2007a, b) developed an impact charging electric current divided by the mass ow rate of the particles I/Wp.
test rig, as mentioned in Section 4.3. The device consists of an From the relationship between I/Wp and the specic charge qm0 at
impact target and two sets of Faraday cages. The device allows the inlet of the detection pipe, a characteristic of particle charging
charge measurements of single particles as small as 100 mm is obtained, and the equilibrium specic charge is determined (see
They performed two different experiments, i.e. the generation of

Fluidized bed charge by repeated contacts and by scanning across the surface of
the substrate, showing that the particle charging by scanning was
more effective than that by simple contacts.
#1
6. Applications
#2
#3 6.1. Separation
Vertical array
#4 of Faraday The development of a process to separate solid materials can
cage sensors
improve the possibilities for recycling wastes. If particles or small
#5
solids have largely different electrostatic properties, they can be
#6 separated. In this process, the particles are rst triboelectrically
charged and then fed into an electric eld separator. The particle
#7 trajectories are deected in the electric eld according to the
polarity and the amount of charge. Although the method to
triboelectrically charge the particles is simple and easy, the charge
Fig. 52. Vertical array of Faraday cage sensors combined with a uidized bed. transfer depends on the environmental conditions; thus tem-
perature and humidity should be controlled in a certain range. A
typical electrostatic separator is shown in Fig. 54 (Yanar and
Air Kwetkus, 1995). Particles that are triboelectrically charged in a
cyclone separator enter the electrostatic separator where a DC
electric eld is applied. The trajectory of the particles depends on
the electrostatic charge and mass of particles. The inhomogeneous
electric eld is effective to achieve a large deection of the
Vacuum charged particles in the upper part of the separator. The bottom of
the separator consists of a series of collection trays.
Faraday
Another electric separator is shown in Fig. 55 (Soong et al.,
cage
2007). Particles are charged in a venturi by contact with the wall
Fig. 53. Centrifugal method for measuring maximum triboelectric charge of and other particles, and sent to an electric eld between two
particles. series of louvered plates. The charged particles are collected on
the louvered plates by adjusting the applied voltage according to
the charge on particles.
Fig. 33). This method is available for micrometer-sized particles Saeki (2006) proposed a different type of separator consisting of a
(Matsusaka and Masuda, 2006). vibratory conveyor equipped with two plate electrodes (Fig. 56). The
advantage of the vibratory separator is to avoid the inuence of the
adhesion between oppositely charged particles and also particlewall
5.7. Atomic force microscopy
adhesion. The two electrodes are inclined toward the horizontal at an
angle y and are xed to the vibrating plate. The upper electrode has
Force measurement with a micrometer-sized particle attached negative polarity and the lower electrode is grounded. The separation
to the cantilever of atomic force microscope (AFM) has received unit vibrates in the direction of angle a to the x-axis. The charged
increasing interest since 1991 (Ducker et al., 1991). One distinct particles are fed onto the lower electrode through a feeder. Positively
advantage of the so-called colloid probe technique is that a charged particles are attracted toward the upper electrode by the
quantitative comparison between experimental force curves and electrostatic force and are repelled from the lower electrode, and
theoretical expectations becomes possible using a probe of known thus the tangential vibration force acting on the particles is reduced.
geometry, while the disadvantage is that it is unsuitable for On the other hand, negatively charged particles are attracted toward
statistical analysis of a number of particles since the preparation the lower electrode and strongly experience the vibration force from
of the colloid probes is not easy. In an attempt to better the lower electrode and consequently conveyed on the x-direction.
understand the origin of adhesion forces, quantitative studies of The difference in the particle trajectory enables the separation.
the surface force interaction as a function of the distance between
particle and substrate have been carried out. The measurements
have revealed two dominant forces, i.e. the electrostatic force due 6.2. Powder coating
to localized contact charging on the particles and the van der
Waals force. The sensitivity in the measurement of the charge Electrostatic powder coating is a dry nishing process. The
transfer is 10 electrons (Gady et al., 1998). powder used in the process, which is a mixture of nely ground
To study the contact charging of glass particles, Nishitani et al. particles of pigment and resin, is electrostatically charged and
(2000) used this method. Single particles attached to cantilevers sprayed onto an electrically conductive work piece to be coated.
were charged by repeated contact with a at stainless steel. The The main difference from a conventional liquid paint is that the
charge accumulated on the particle was analyzed by considering a powder coating does not require a solvent to keep the binder and
virtual point charge in the particle. Matsuyama et al. (2006) also ller parts in a liquid suspension form. Therefore, the powder
performed similar measurements, and discussed how to deter- coating offers several advantages. It does not use volatile organic
mine the surface charge density and the contact area from the compounds (VOCs), which are hazardous to human health, and
force curve. Furthermore, the consistency between the measured over sprayed powder can be collected and recycled.
results and their charge relaxation model was discussed. Bunker The general process is as follows; coating powder is uidized
et al. (2007) studied the charging of pharmaceutical particles. by air and then blown through a pipe to a charging gun. Two
Air out space-charge eld, which is directed toward the grounded

conductive workpiece. As a result, the charged particles drift
Air filter
toward the workpiece and deposit not only onto the surface of
Gas-solids
the workpiece facing the spray gun but also in hidden regions due
flow
Cyclone to the space-charge eld. The particles that deposit onto the
separator workpiece adhere by electrostatic image forces. After spraying,
the workpiece is transferred to a high-temperature curing
region in which the powder melts and ows to produce a smooth
coating layer.
Electrode Electrode
() (+)
6.3. Electrophotography
Electrostatic
separator
Electrophotography is a photocopying technique developed by
Chester Carlson in 1938. In this process, a visible image is created
using electrostatic latent images in the form of surface charge
Collection trays
patterns on a photoconductive surface. The visible images consist
of ne particles called toner. Toner is triboelectrically charged by
Air filter mixing it with larger carrier beads. The mixture is called
developer. In some cases the toner is triboelectrically charged
against walls without the use of carrier beads.
Air out
A typical process of the electrophotography consists of six
Fig. 54. Typical electrostatic separator. steps: (i) the surface of photoreceptor, which is an insulator in the
absence of light, is uniformly charged. (ii) Light reected from the
image discharges the normally insulating photoreceptor produ-
cing a latent imagea charge pattern on the photoreceptor that
Air
mirrors the information to be transformed into the real image. (iii)
Feeder
Toner particles charged spread over the surface adhere to the
latent image, transforming it into a real image. (iv) The developed
toner on the photoreceptor is transferred to paper charged on the
Triboelectric backside. (v) The image is permanently xed to the paper by
charging pipe melting the toner. (iv) The photoreceptor is discharged and
cleaned of any excess toner (Schein, 1992/1996).
For black and white copiers, the photoreceptor can transfer the
Electrode Electrode image directly to paper. However, for most color copiers the
() (+) image is formed from four colors (cyan, magenta, yellow and
black) of toner and the image is built up rst on an intermediate
surface; then, another transfer takes place where the full-color
image is transferred to paper.
6.4. Self-assembly
Vacuum
Self-assembly of components is a useful method for preparing
Fig. 55. Electrostatic separator with louvered plates. microstructured materials with interesting mechanical or optical
properties. Although crystallization of particles with different
sizes or shapes can be achieved, the methods to assemble binary
lattices of particles of the same sizes but with different properties
Positively charged particles are limited. One of the available methods is the electrostatic self-
assembly (ESA) of macroscopic components using interactions
Negatively charged particles
caused by triboelectric charging. The systems comprise two kinds
of particles (usually spheres) that are charged oppositely when
agitated on a at, metallic surface. The interplay of repulsive
Upper electrode Feeder
interactions between like-charged particles and attractive inter-
(negative polarity)
actions between unlike-charged ones results in the self-assembly
Lower electrode of these particles into highly ordered arrays.
(grounded) The type of the lattice formed by self-assembly depends on the
x number ratio of the two different particles. For example, when the
Vibrating number ratio, Nylon-6, 6 (NNyl)/Teon (NTef), is close to unity, a
plate y
square lattice array is formed (Fig. 57(a)). When particles of one
kind are present in excess, they arrange themselves into either a
Fig. 56. Electrostatic separator using vibration. pentagonal lattice array or a hexagonal lattice array. When NNyl/
NTef 41, each Teon particle is surrounded by ve or six Nylon-
6,6 particles (Fig. 57(b) and (c)). Conversely, when NNyl/NTef o1,
types of guns are available, i.e. the triboelectric and corona guns. each Nylon-6,6 particle is surrounded by ve or six Teon
For the triboelectric gun, powder is charged by frictional contact particles. These observations suggest that the lattices are
with the walls in the gun. The sprayed, charged particles form a steady-state products of self-assembly (Grzybowski et al., 2003).
6.5. Measurements in pneumatic transport Insulator

First detector Second detector
6.5.1. Particle ow rate and specic charge Gas solids flow
In gassolids pipe ow, the charges transferred from the
particles to the wall are detected as electric currents as mentioned
qm0 qm1
in Section 4.5. For dilute phase ow, the electric current I I2
I1
generated from a pipe is formulated with the particle ow rate
Wp and the specic charge of particles qm0 at the inlet of the pipe Electric shield
(see Eq. (34)). When two different detection pipes are connected
Autoswitch
in series with electrical isolation (Fig. 58), the following equations
are held for the rst and second pipes, respectively:
Electrometer
I1
a1 qm0 b1 49 Computer
Wp
and Fig. 58. Current detection system for measuring particle ow rate and specic charge.
I2
a2 qm1 b2 , 50
Wp
(49)(51), the following equation is derived:
where subscripts 1 and 2 refer to the rst and second detection
b2 a2 b1 I1 b1 I2
pipes, respectively. Since the charge at the outlet of the rst pipe qm0 : 56
a2 a1 1I1 a1 I2
is equal to that at the inlet of the second pipe, the charge balance
is expressed as The particle charge can also be calculated using Eq. (56).
I1
qm0 qm1 , 51 6.5.2. Signal analysis
Wp
The current detected in gassolids pipe ow consists of many
From Eqs. (49)(51), the particle ow rate Wp is expressed as pulsating electric signals, which vary in a wide range from
Wp cI1 dI2 , 52 positive to negative. In fact, these electric signals contain useful
information on the state of particles owing in the pipe. An
here example of the signals detected with a digital oscilloscope is
a1 1a2 shown in Fig. 59(a). The variation of the signals on a magnied
c 53 time scale is shown in Fig. 59(b). The pulsating signals are caused
a1 b2 a2 b1
by the uctuation of the particle ow rate. Even though particles
and continuously ow in the pipe, the particle concentration
a1 uctuates on a time scale of milliseconds; as a result, a number
d : 54
a1 b2 a2 b1 of clouds of particles ow in the pipe.
The denominator of the right-hand side in Eqs. (53) and (54) must For instance, when a cloud of negatively charged particles
not be zero; this means that the electrostatic property of the rst approaches the detection pipe, positive charge is induced on the
detection pipe must differ from that of the second one, i.e. inner surface of the detection pipe. This occurs by the fact that the
electrons on the inner surface ow to earth, and consequently a
b1 b2 negative peak is detected with the digital oscilloscope. If the
a : 55
a1 a2 particles impact on the wall of the detection pipe, charge transfer
Therefore, on-line measurement of particle ow rate is possible will occur. Since the transferred charges remain there while the
by measuring the two electric currents (Masuda et al., 1994, particles are in the pipe, no electric signal is detected. When the
1998b; Matsusaka and Masuda, 2006). In addition, from Eqs. particles pass out of the pipe, the whole charge held on the wall
surface ows to earth, and a peak signal is detected.
The shape and intensity of the electric signals depend on the
condition such as the polarity and amount of the charge of the
cloud at the inlet and outlet of the detection pipe. Typical particle
charging in gassolids pipe ow and the shapes of the electric
signals are illustrated in Fig. 60 (Matsusaka et al., 2008a). Here, it
Nylon-6,6 is supposed that the particle charging is represented by an
exponential equation (Cole et al., 19691970; Masuda et al., 1976;
Teflon Matsusaka et al., 2007) and that the detection pipe length is larger
than the cloud; thus, the tails of the two peaks do not overlap
with each other.
These electric signals can be applied to the correlation method
for measuring particle velocity (Yan et al., 1995; Gajewski, 1996).
Two detection pipes are used for the measurement system.
Three sets of electric signals are shown in Fig. 61 (Matsusaka
et al., 2008a). Since particles that are negatively charged
obtain negative charge by contact with the wall of the rst
detection pipe, a larger positive signal is detected at the outlet.
By passing through the intermediate pipe, the particles obtain
positive charge by contact with the wall, material of which is
Fig. 57. Electrostatic self-assembly: (a) Square lattice array; (b) Pentagonal lattice different from that of the detection pipes, and the polarity of
array and (c) Hexagonal lattice array. the particle changes from negative to positive. The polarity
80
Electric signal (mV)
Intermediate pipe
40 Detection pipe 1 Detection pipe 2
0 a
20
0
-40 -20 Detection pipe 1
0 2 4 6 8 10
Time, t (s) 20 Detection pipe 2
0
-20
80

40 20
0
-20 Detection pipe 1
0
20 Detection pipe 2
0
-40 -20
0 0.2 0.4 0.6 0.8 1
Time (s) c
20
Fig. 59. (a) Signals detected with a digital oscilloscope and (b) signals on a 0
-20 Detection pipe 1
magnied time scale (Matsusaka et al., 2008a).
20 Detection pipe 2
0
-20
Detection 0 50 100 150

pipe Transport pipe Time (ms)
Specific charge, qm
(1) Fig. 61. Measurement of particle velocity by the correlation method (Matsusaka
et al., 2008a): (a) Intermediate pipe: L0.5 m; (b) Intermediate pipe: L1 m; and
(2) (c) Intermediate pipe: L2 m.
(3) Pipe length L
were localized randomly on the particle surface (Fig. 62(a)), and

the charges were considered to work as the initial charge at the
next impact. To reduce the load of the calculation, the patch-like
Inlet Outlet charges were redistributed as an axisymmetric charge
distribution as shown in Fig. 62(b). The charge transfer can be
calculated based on the charge relaxation model (see Section 4.2).
The calculated results i.e. impact charge and the accumulated
0 charge on the particle are shown as a function of the number of
(1) impacts (Fig. 63). Both the impact charge and the accumulated
Electric signal
charge vary exponentially.

0 Yoshida et al. (2003) calculated multi-impact charging in a
(2) vibrating vessel using three-dimensional distinct element method
(DEM). The surface of the particles is divided into many charging
sites (Fig. 64). In this calculation, it is assumed that (i) particles
0 are insulators, (ii) the charge transfer occurs by impact between
(3) Time, t particles as well as particle and wall, depending on the contact
potential difference and the local electric eld, and (iii) the
Fig. 60. Effect of particle charging on the shape of electric signal: (a1) Particle
charges transferred on the particle remain at the contact sites. The
charging and (a2) Variations of electric signal.
specic charge of the particles as a function of time elapsed is
shown in Fig. 65. The particle charge approaches an equilibrium
value, and the effect of the number of the sites on the particle
again changes to negative by passing through the second charge is not so signicant if the number of the sites is sufciently
detection pipe. The time interval between the rst and second large.
detection pipes increases proportionally with the length of the Matsuyama et al. (2009) developed another DEM simulation to
intermediate pipe. The particle velocity can be calculated from analyze the development of charge on particles in a shaker. In
the time interval. their analysis, particle charging model, space charge effect,
electrical boundary condition of a container, and long range force
were comprehensively involved.
7. Simulation In gassolids pipe ow, particles are charged to an equilibrium
value by collisions with the pipe wall and other particles. Tanoue
7.1. Triboelectric charging by repeated impacts et al. (2001b) carried out numerical simulations of particle
charging based on the condenser model (see Section 4.1). They
Matsuyama and Yamamoto (2006b) proposed a Monte Carlo used the standard ke model to express the turbulent ow in the
simulation for a charge accumulating process of a single particle pipe and determined the number of collisions of particles on the
during successive impacts. In the calculation, patch-like charges pipe wall by Lagrangian particle tracking, showing that the effect
+q
+q
q
E E q
Fig. 62. (a) Schematic illustration of randomly given patch-like charge and
(b) schematic illustration of charge redistribution for axisymmetric calculation
(Matsuyama and Yamamoto, 2006b).
Fig. 64. Model particles used in the DEM simulation (the number of charging sites
is (a) 32, (b) 200, and (c) 392) (Yoshida et al., 2003).
100
Impact charge (pC)
50 6
Specific charge ( C kg 1)
Negative
0 charge
4
50 Number of charging
cites of a particle
100 2 32
0 20 40 60 200
1000 392
Accumulated charge (pC)
0
750 0 5 10 15 20 25 30
Vibration time (s)
500 Fig. 65. Specic charge of particles in vibrating vessel (Yoshida et al., 2003).
=6
250 Dp = 3.2 mm
8. Summary
0
0 20 40 60
The research on triboelectric charging of particles has been
Number of impact () reviewed from theoretical and experimental viewpoints to
Fig. 63. Calculated results by a Monte Carlo simulation: (a) impact charge and evaluate the current state of understanding and applications.
(b) accumulated charge (Matsuyama and Yamamoto, 2006b). From this work, the following points are concluded:
of the particle size, air velocity, and pipe diameter on the particle (1) The triboelectric charging can be caused by electron transfer,
charging. Cangialosi et al. (2006) used a statistical model to ion transfer, and material transfer. Among them, the electron
determine the particlewall collisions and used the charge transfer is considered to be the base of the charge transfer. For
relaxation model (see Section 4.2) to predict the equilibrium metalmetal contacts, the difference in work function is used
charge and the impact charge caused by multiple impact events. to explain the charge transfer. For inorganic compounds such
Watano (2006) used three-dimensional discrete element method as oxides, the charging tendency can be explained by the
(DEM) to calculate the contact area for particle charging in generalized electronegativity of metal ion. For organic
pneumatic conveying systems. compounds, the charging tendency depends on the functional
groups contained in them. For polymermetal contacts, the
electric polarity after charge transfer can be determined by
the energy level of HOMO and LUMO of the organic solids and
7.2. Behavior of charged particles the Fermi level of the metal. In addition, environmental
conditions such as temperature and humidity have to be
Lim et al. (2006) studied the ow behaviors of charged taken into account.
particles in gassolids pipe ow using DEM coupled with (2) To analyze the particle charging, two basic models, i.e. the
computational uid dynamics (CFD), showing that the charged condenser model and the charge relaxation model exist. The
particles are spread by the electrostatic diffusion. Also, Mio et al. former is based on the effective contact potential difference
(2006) simulated the developing behavior of toners in a two- depending on the work functions and initial charge. The latter
component development system of electrophotography using the is based on the gas breakdown predicted by the Paschen
large-scale DEM (150,000 particles) and discussed the effect of the curve. Irrespective of the difference in the concept of particle
bias voltage and mechanical operation conditions on the behavior charging, these models derive the same type of exponential
of toners. equations for the charge accumulation by repeated impacts.
These equations enable us to analyze actual complicated represented by

particle charging processes.
(3) The values of work function or the contact potential difference Ve ke q, 11
characterizing the triboelectric charging can be obtained
experimentally. For the measurement of the charge on where
particles, the Faraday cage method is available; in fact, there
1 1 1
are several types of Faraday cages, which are used for each ke , A:1
4pe0 r1 r2
purpose. In recent years, there is increasing interest in the
measurement of the charge distribution of particles to analyze in which
processes in detail. To precisely measure the charge distribu-
tions, dispersion and sampling of the charged particles are Dp
r1 A:2
important since positively and negatively charged particles 2
agglomerate and highly charged particles easily deposit before
measurement. To characterize the particle charging tendency, and
several practical methods are proposed. In particular, the single
Dp
particle impact test and the gassolids pipe ow test are useful r2 2z0 : A:3
2
to evaluate the equilibrium charge and transferred charge. The
atomic force microscopy is also useful for analyzing the Substituting Eqs. (A.2) and (A.3) into Eq. (A.1) gives
electrostatic force of a single particle in detail.
(4) Triboelectric charging of particles can be applied for various 1 1 1
ke
purposes, e.g. electrostatic separation, dry powder coating and 2pe0 Dp Dp 4z0
electrophotography. The electrostatic self-assembly is ex-
pected as a technique to produce new functional materials.
2z0
Also, the measurement techniques using triboelectric char- 29
ging of particles are essential to evaluate and control particle
pe0 D2p
handling processes.
(5) Numerical simulation is a powerful tool to analyze particle
charging processes. Monte Carlo simulation is available to
study the charge accumulation by repeated impacts. DEM Appendix B. Equations of space charge effect
simulation is also useful for the calculation since the
information of all the particle collisions are taken into account Charged particles owing through a vertical pipe are illu-
for the charge transfer. In gassolids pipe ow, Lagrangian strated in Fig. B.1 (Matsusaka and Masuda, 2006). When the
particle tracking method is available to study the charge charged particles are dispersed uniformly, the electric eld E is
accumulation process. In recent years, the analysis of the expressed as a function of radius r, i.e.
behavior of charged particles in a complicated system such as
np rq
two-component development system of electrophotography Er B:1
2e0
comes to attract attention.
where np is the particle number concentration, which is given by
Acknowledgements 6m rg u
np , B:2
p D3p rp v
The authors acknowledge nancial support from the Interna-
tional Fine Particles Research Institute (IFPRI). This research was also where m is the mass ow ratio of particles to gas, rg is the density
supported by Core-to-Core Program for Advanced Particle Handling of the gas, u is the average gas velocity, rp is the particle density,
Science, JSPS, and a Grant no. S0901039 from MEXT, Japan. and v is the average particle velocity.
Appendix A. Equations of image charge effect
An image charge is illustrated in Fig. A.1 (Matsusaka and

Masuda, 2006). The potential difference Ve in the gap is
Uniformly charged Grounded

Spherical particles conductive wall
E(r)
q q
Dp
Di
z0
r1 r2
Gas-solids pipe flow
Fig. A.1. Potential difference caused by image charge. Fig. B.1. Electric eld generated by space charge.
The potential difference Vb in the gap z0 close to the wall is Cartwright, P., Singh, S., Bailey, A.G., Rose, L.J., 1985. Electrostatic charging
represented by characteristics of polyethylene powder during pneumatic conveying. IEEE
Transactions on Industry Applications IA-21, 541546.
Z Di =2
3z0 mrg Di u Cole, B.N., Baum, M.R., Mobbs, F.R., 19691970. An investigation of electrostatic
Vb Er dr q, B:3 charging effects in high-speed gassolids pipe ows. In: Proceedings of the
Di =2z0 2pe0 rp D3p v Institution of Mechanical Engineers, vol. 184, pp. 77-83.
Cotler, P., Castle, G.S.P., Schein, L.B., 1995. General model of spheresphere
From Eq. (B.3) and Eq. (12) in the text, the following equation is insulator contact electrication. Journal of Electrostatics 36, 165173.
derived: Davies, D.K., 1969. Charge generation on dielectric surfaces. British Journal of
Applied Physics: Journal of Physics D 2, 15331537.
3z0 mrg Di u Diaz, A.F., Guay, J., 1993. Contact charging of organic materials: ion vs. electron
kb : 30 transfer. IBM Journal of Research and Development 37, 249259.
2pe0 rp D3p v
Duke, C.B., Fabish, T.J., 1978. Contact electrication of polymers: a quantitative
model. Journal of Applied Physics 49, 315321.
Ducker, W.A., Senden, T.J., Pashley, R.M., 1991. Direct measurement of colloidal
forces using an atomic force microscope. Nature 353, 239241.
Appendix C. Equations of probability density function of Ema, A., Yasuda, D., Tanoue, K., Masuda, H., 2003. Tribo-charge and rebound
specic charge characteristics of particles impact on inclined or rotating metal target. Powder
Technology 135/136, 213.
Epping, R.H., Kuettner, A., 2002. Free air beam in an electric eld for determination
Assuming that the probability density function of particle of the electrostatic charge of powders between 1 and 200 mm. Journal of
collision is expressed as a normal distribution, which is derived Electrostatics 55, 279288.
Fabish, T.J., Duke, C.B., 1977. Molecular charge states and contact charge exchange
from a binomial distribution, and also using the relationship in polymers. Journal of Applied Physics 48, 42564266.
between the number of particle collision and the particle charge, Gady, B., Reifenberger, R., Rimai, D.S., 1998. Contact electrication studies using
the probability density function of the specic charge f1(qm) is atomic force microscope techniques. Journal of Applied Physics 84, 319322.
Gajewski, A., 1989. Measuring the charging tendency of polystyrene particles in
represented by (Matsusaka et al., 2002)
pneumatic conveyance. Journal of Electrostatics 23, 5566.
" 2 # Gajewski, J.B., 1996. Monitoring electrostatic ow noise for mass ow and mean
qm1 1 qm1 ln 1 qm qm0 =qm1 qm nk
f1 qm p exp , velocity measurement in pneumatic transport. Journal of Electrostatics 37,
qm1 qm 2pnk 2nk 2
261276.
Gajewski, J.B., 2006. Non-contact electrostatic ow probes for measuring the ow
C:1
rate and charge in the two-phase gassolids ows. Chemical Engineering
where n is the mean number of particle collisions and k is the Science 61, 22622270.
Gajewski, J.B., 2008. Electrostatic nonintrusive method for measuring the electric
impact charging factor qm1 =n0 . charge, mass ow rate, and velocity of particulates in the two-phase gassolid
Using a probability density function of initial charge fic (qm0), pipe owsits only or as many as 50 years of historical evolution. IEEE
the probability density function f2(qm) after traveling through the Transactions on Industry Applications 44, 14181430.
Ghadiri, M., Martin, C.M., Morgan, J.E.P., Clift, R., 1992. An electromechanical valve
pipe is represented by
for solids. Powder Technology 73, 2135.
Z qm1 Gibson, H.W., 1975. Linear free energy relationshipstriboelectric charging of
f2 qm f1 qm fic qm0 dqm0 : C:2 organic solids. Journal of the American Chemical Society 97, 38323833.
1 Gibson, H.W., 1984. Control of electrical properties of polymers by chemical
modication. Polymer 25, 327.
Furthermore, using a probability density function of the impact Grzybowski, B.A., Winkleman, A., Wiles, J.A., Brumer, Y., Whitesides, G.M., 2003.
charging factor fk(k), the probability density function f3(qm) after Electrostatic self-assembly of macroscopic crystals using contact electrica-
traveling through the pipe is represented by tion. Nature Materials 2, 241245.
Z 1 Greason, W.D., 2000. Investigation of a test methodology for Triboelectrication.
Journal of Electrostatics 49, 245256.
f3 qm f2 qm fk k dk: C:3 Guardiola, J., Rojo, V., Ramos, G., 1996. Inuence of particle size, uidization
1
velocity and relative humidity on uidized bed electrostatics. Journal of
Substituting the probability density functions gives theoretical Electrostatics 37, 120.
Gupta, R., Gidaspow, D., Wasan, D.T., 1993. Electrostatic separation of powder
equations. mixtures based on the work functions of its constituents. Powder Technology
75, 7987.
Gutman, E.J., Hartmann, G.C., 1992. Triboelectric properties of two-component
References developers for xerography. Journal of Imaging Science and Technology 36,
335349.
Harper, W.R., 1951. The Volta effect as a cause of static electrication. Proceedings
Adhiwidjaja, I., Matsusaka, S., Yabe, S., Masuda, H., 2000. Simultaneous
phenomenon of particle deposition and reentrainment in charged aerosol of the Royal Society of London Series A 205, 83103.
oweffects of particle charge and external electric eld on the deposition Harper, W.R., 1967. Contact and Frictional Electrication. Clarendon Press, Oxford.
layer. Advanced Powder Technology 11, 221233. Higashiyama, Y., Ujiie, Y., Asano, K., 1997. Triboelectrication of plastic particles on
Ahn, Y.C., Kim, D.H., Kim, S.C., Hwang, E.G., Lee, C.G., Kim, G.T., Shin, H.S., Lee, J.K., a vibrating feeder laminated with a plastic lm. Journal of Electrostatics 42,
Lee, S.H., 2004. Measurement and control of triboelectrically charged silica and 6368.
glassy carbon particles. Particulate Science and Technology 22, 305320. Hughes, J.F., 1984. Electrostatic Powder Coating. Research Studies Press, Wiley,
Alfrey, T., Price, C.C., 1947. Relative reactivities in vinyl copolymerization. Journal New York.
of Polymer Science 2, 101106. Ikezaki, K., Teraoka, S., Namekawa, M., 2005. Ultrasound aided tribocharging of
Anderson, J.H., 1994. A comparison of experimental data and model predictions for polymeric powders. Journal of the Institute of Electrostatics Japan 29,
tribocharging of two-component electrophotographic developers. Journal of 173178.
Imaging Science and Technology 38, 378382. Itakura, T., Masuda, H., Ohtsuka, C., Matsusaka, S., 1996. The contact potential
Bailey, A.G., 1993. Charging of solids and powders. Journal of Electrostatics 30, difference of powder and the tribo-charge. Journal of Electrostatics 38,
167180. 213226.
Bailey, A.G., 1998. The science and technology of electrostatic powder spraying, Iuga, A., Calin, L., Neamtu, V., Mihalcioiu, A., Dascalescu, L., 2005. Tribocharging of
transport and coating. Journal of Electrostatics 45, 85120. plastics granulates in a uidized bed device. Journal of Electrostatics 63,
Bailey, A.G., 2001. The charging of insulator surfaces. Journal of Electrostatics 937942.
51/52, 8290. Jones, T.B., 1995. Electromechanics of Particles. Cambridge University Press,
Balachandran, W., Hu, D., Ghadiri, M., Law, S.E., Thompson, S.A., 1997. The study of Cambridge.
the performance of an electrostatic valve used for bulk transport of particulate Jones, T.B., King, J.L., 1991. Powder Handling and Electrostatics: Understanding and
materials. IEEE Transactions on Industry Applications 33, 871878. Preventing Hazards. Lewis, Chelsea.
Bunker, M.J., Davies, M.C., James, M.B., Roberts, C.J., 2007. Direct observation of Joseph, S., Klinzing, G.E., 1983. Vertical gassolid transition ow with electro-
single particle electrostatic charging by atomic force microscopy. Pharmaceu- statics. Powder Technology 36, 7987.
tical Research 24, 11651169. Kittaka, S., Murata, Y., 1976. A new system for measuring the charging tendency of
Cangialosi, F., Liberti, L., Notarnicola, M., Stencel, J., 2006. Monte Carlo simulation solid particles. Journal of Electrostatics 2, 111119.
of pneumatic tribocharging in two-phase ow for high-inertia particles. Kleber, W., Makin, B., 1998. Triboelectric powder coating: a practical approach for
Powder Technology 165, 3951. industrial use. Particulate Science and Technology 16, 4353.
Kodama, J., Foerch, R., McIntyre, N.S., Castle, G.S.P., 1993. Effect of plasma Matsuyama, T., Yamamoto, H., 2006b. Montecarlo simulation of charge accumu-
treatment on the triboelectric properties of polymer powders. Journal of lating process of a single particle due to successive impacts. Journal of the
Applied Physics 74, 40264033. Society of Powder Technology, Japan 43, 169173.
Lawless, P.A., 1999. Electrostatic precipitators. In: Webster, J. (Ed.), Wiley Matsuyama, T., Yamamoto, H., 2008. Maximum electrostatic charge of powder in
Encyclopedia of Electrical and Electronics Engineering, vol. 7. John Wiley & pipe ow. Journal of the Society of Powder Technology, Japan 45, 373379.
Sons, Inc., pp. 115. Matsuyama, T., Ogu, M., Yamamoto, H., Marijnissen, J.C.M., Scarlett, B., 2003.
Lee, L.-H., 1994. Dual mechanism for metalpolymer contact electrication. Impact charging experiments with single particles of hundred micrometre
Journal of Electrostatics 32, 129. size. Powder Technology 135/136, 1422.
Lim, E.W.C., Zhang, Y., Wang, C.-H., 2006. Effects of an electrostatic eld in Matsuyama, T., Ohtsuka, M., Yamamoto, H., 2006. Measurement of force curve due
pneumatic conveying of granular materials through inclined and vertical to electrostatic charge on a single particle using atomic force microscope.
pipes. Chemical Engineering Science 61, 78897908. Journal of the Society of Powder Technology, Japan 43, 174180.
Lowell, J., Rose-Innes, A.C., 1980. Contact electrication. Advances in Physics 29, Matsuyama, T., Supuk, E., Ahmadian, H., Hassanpour, A., Ghadiri, M., 2009. Analysis
9471023. of tribo-electric charging of spherical beads using distinct element method.
Lowell, J., Akande, A.R., 1988. Contact electricationwhy is it variable? Journal of Powders and Grains 2009. AIP, New York, pp. 127130.
Physics D: Applied Physics 21 125137. Mazumder, M.K., Ware, R.E., Yokoyama, T., Rubin, B.J., Kamp, D., 1991. Measure-
Machida, M., Scarlett, B., 2005. Process tomography system by electrostatic charge ment of particle size and electrostatic charge distributions on toners using E-
carried by particles. IEEE Sensors Journal 5, 251259. SPART analyzer. IEEE Transactions on Industry Applications 27, 611619.
Mazumder, M.K., 1999. Electrostatic processes. In: Webster, J. (Ed.), Wiley
Masui, N., Murata, Y., 1983. Electrication of polymer particles by impact on a
Encyclopedia of Electrical and Electronics Engineering, vol. 7. John Wiley &
metal plate. Japanese Journal of Applied Physics 22, 10571062.
Sons, Inc., pp. 1539.
Masui, N., Murata, Y., 1984. Mechanisms of charge build-up on a polymer particle
Michel, E., Baur, R., Macholdt, H.-T., 2001. Charge stabilizers: properties and
by repeated impact. Japanese Journal of Applied Physics 23, 550555.
applications. Journal of Electrostatics 51/52, 9196.
Masuda, S., Fujibayashi, K., Ishida, K., Inaba, H., 1972. Connement and
Mio, H., Matsuoka, Y., Shimosaka, A., Shirakawa, Y., Hidaka, J., 2006. Analysis of
transportation of charged aerosol clouds via electric curtain. Electrical
developing behavior in two-component development system by large-scale
Engineering in Japan 92, 4352.
discrete element method. Journal of Chemical Engineering of Japan 39,
Masuda, H., Komatsu, T., Iinoya, K., 1976. The static electrication of particles in
11371144.
gassolids pipe ow. AIChE Journal 22, 558564. Mizes, H.A., Conwell, E.M., Salamida, D.P., 1990. Direct observation of ion transfer
Masuda, H., Gotoh, K., Orita, N., 1993. Charge distribution measurement of aerosol in contact charging between a metal and a polymer. Applied Physics Letters
particles. Journal of Aerosol Research, Japan 8, 325332. 56, 15971599.
Masuda, H., Matsusaka, S., Nagatani, S., 1994. Measurements of powder ow rate Murata, Y., 1979. Photoelectric emission and contact charging of some synthetic
in gassolids pipe ow based on the static electrication of particles. Advanced high polymers. Japanese Journal of Applied Physics 18, 18.
Powder Technology 5, 241254. Murata, Y., Kittaka, S., 1979. Evidence of electron transfer as the mechanism of
Masuda, H., Matsusaka, S., Akiba, S., Shimomura, H., 1998a. Electrication static charge generation by contact of polymers with metals. Japanese Journal
of ne particles in gassolids pipe ow. KONA Powder and Particle 16, of Applied Physics 18, 421.
216222. Nieh, S., Nguyen, T., 1988. Effects of humidity, conveying velocity, and particle size
Masuda, H., Matsusaka, S., Shimomura, H., 1998b. Measurement of mass ow rate on electrostatic charges of glass beads in a gaseous suspension ow. Journal of
of polymer powder based on static electrication of particles. Advanced Electrostatics 21, 99114.
Powder Technology 9, 169179. Nifuku, M., Katoh, H., 2003. A study on the static electrication of powders during
Matsusaka, S., Masuda, H., 2003. Electrostatics of particles. Advanced Powder pneumatic transportation and the ignition of dust cloud. Powder Technology
Technology 14, 143166. 135/136, 234242.
Matsusaka, S., Masuda, H., 2006. Simultaneous measurement of mass ow rate and Nifuku, M., Ishikawa, T., Sasaki, T., 1989. Static electrication phenomena in
charge-to-mass ratio of particles in gassolids pipe ow. Chemical Engineering pneumatic transportation of coal. Journal of Electrostatics 23, 4554.
Science 61, 22542261. Nishitani, Y., Muramatsu, H., Maruyama, H., Tanoue, K., Matsusaka, S., Masuda, H.,
Matsusaka, S., Ghadiri, M, Masuda, H., 2000. Electrication of an elastic sphere by 2000. Measurement of interaction forces between particle and surface in air
repeated impacts on a metal plate. Journal of Physics D: Applied Physics 33, using AFM. In: Proceedings of the Annual Meeting of Journal of the Society of
23112319. Powder Technology, Japan, May, pp. 6769.
Matsusaka, S., Umemoto, H., Nishitani, M., Masuda, H., 2002. Electrostatic charge Nomura, T., Satoh, T., Masuda, H., 2003. The environment humidity effect on the
distribution of particles in gassolids pipe ow. Journal of Electrostatics 55, tribo-charge of powder. Powder Technology 135/136, 4349.
8196. Oguchi, T., Tamatani, M., 1986. Contact electrication in inorganic binary
Matsusaka, S., Nishida, T., Gotoh, Y., Masuda, H., 2003a. Electrication of ne compounds. Journal of the Electrochemical Society 133, 841847.
particles by impact on a polymer lm target. Advanced Powder Technology 14, Oguchi, T., Tamatani, M., 1993. Contact electrication phenomena and powder
127138. surface treatments. Wear 168, 9198.
Matsusaka, S., Oki, M., Masuda, H., 2003b. Bipolar charge distribution of a mixture Ohsawa, A., 2003. Computer simulation for assessment of electrostatic hazards in
of particles with different electrostatic characteristics in gassolids pipe ow. lling operations with powder. Powder Technology 135/136, 216222.
Powder Technology 135/136, 150155. ONeill, B.C., Willis, C.A., 1987. An electrostatic method for the measurement of
Matsusaka, S., Oki, M., Masuda, H., 2007. Control of electrostatic charge on powder ow rates in pipes, Electrostatics 87. IOP Publishing, Oxford
particles by impact charging. Advanced Powder Technology 18, 229244. pp. 303306.
Matsusaka, S., Fukuda, H., Sakura, Y., Masuda, H., Ghadiri, M., 2008a. Analysis of Paschen, F., 1889. Ueber die zum Funkenubergang in Luft, Wasserstoff und
pulsating electric signals generated in gassolids pipe ow. Chemical Kohlensaure bei verschiedenen Drucken erforderliche Potentialdifferenz.
Annalen der Physik 273, 6996.
Engineering Science 63, 13531360.
Rowley, G., 2001. Quantifying electrostatic interactions in pharmaceutical solid
Matsusaka, S., Ando, K., Tanaka, Y., 2008b. Development of electrostatic charge
systems. International Journal of Pharmaceutics 227, 4755.
controller for particles in gases using centrifugal contact. Journal of the Society
Saeki, M., 2006. Vibratory separation of plastic mixtures using triboelectric
of Powder Technology, Japan 45, 380386.
charging. Particulate Science and Technology 24, 153164.
Matsusaka, S., Yoshitani, K., Tago, H., Nii, T., Masuda, H., Iwamatsu, T., 2008c.
Saurenbach, F., Wollmann, D., Terris, B.D., Diaz, A.F., 1992. Force microscopy of ion-
Sampling of charged ne particles by motion control under ac eld. Journal of
containing polymer surfaces: morphology and charge structure. Langmuir 8,
the Society of Powder Technology, Japan 45, 387394.
11991203.
Matsuyama, T., Yamamoto, H., 1989. Charge transfer between a single polymer
Schein, L.B., 1992/1996. Electrophotography and Development Physics, 2nd ed.
particle and a metal plate due to impact. KONA Powder and Particle 7, 1521.
Springer-Verlag/Laplacian Press, Berlin.
Matsuyama, T., Yamamoto, H., 1994. Charge transfer between a polymer particle Schein, L.B., 1999. Recent advances in our understanding of toner charging. Journal
and a metal plate due to impact. IEEE Transactions on Industry Applications of Electrostatics 46, 2936.
30, 602607. Schnurmann, R., 1941. Contact electrication of solid particles. Proceedings of the
Matsuyama, T., Yamamoto, H., 1995a. Charge relaxation process dominates Physical Society 53, 547553.
contact charging of a particle in atmospheric conditions. Journal of Physics Shinohara, I., Yamamoto, F., Anzai, H., Endo, S., 1976. Chemical structure and
D: Applied Physics 28, 24182423. electrostatic properties of polymers. Journal of Electrostatics 2, 99110.
Matsuyama, T., Yamamoto, H., 1995b. Characterizing the electrostatic charging of Shirakawa, Y., Ii, N., Yoshida, M., Takashima, R., Shimosaka, A., Hidaka, J., 2008.
polymer particles by impact charging experiments. Advanced Powder Quantum chemical calculation of electron transfer at metal/polymer inter-
Technology 6, 211220. faces. Journal of the Society of Powder Technology, Japan 45, 366372.
Matsuyama, T., Yamamoto, H., 1995c. Electrication of single polymer particles by Soong, Y., Irdi, G.A., McLendon, T.R., Hedges, S.W., Dilmore, R.M., Grifth, C.,
successive impacts with metal targets. IEEE Transactions on Industry Romanov, V., Haljasmaa, I., 2007. Triboelectrostatic separation of y ash with
Applications 31, 14411445. different charging materials. Chemical Engineering & Technology 30, 214219.
Matsuyama, T., Yamamoto, H., 1997. Charge-relaxation process dominates contact Tanoue, K., Ema, A., Masuda, H., 1999. Effect of material transfer and work
charging of a particle in atmospheric condition: II. The general model. Journal hardening of metal surface on the current generated by impact of particles.
of Physics D: Applied Physics 30, 21702175. Journal of Chemical Engineering of Japan 32, 544548.
Matsusyama, T., Yamamoto, H., 2006a. Impact charging of particulate materials. Tanoue, K., Morita, K., Maruyama, H., Masuda, H., 2001a. Inuence of functional
Chemical Engineering Science 61, 22302238. group on the electrication of organic pigments. AIChE Journal 47, 24192424.
Tanoue, K., Tanaka, H., Kitano, H., Masuda, H., 2001b. Numerical simulation of potential difference measurement. Japanese Journal of Applied Physics 32,
tribo-electrication of particles in a gassolids two-phase ow. Powder 56035610.
Technology 118, 121129. Yanar, D.K., Kwetkus, B.A., 1995. Electrostatic separation of polymer powders.
Uyama, Y., Ikada, Y., 1990. Electrostatic properties of UV-irradiated and surface- Journal of Electrostatics 35, 257266.
grafted polymers. Journal of Applied Polymer Science 41, 619629. Yao, J., Zhang, Y., Wang, C.-H., Matsusaka, S., Masuda, H., 2004. Electrostatics of the
Watanabe, H., Samimi, A., Ding, Y.L., Ghadiri, M., Matsuyama, T., Pitt, K.G., 2006. granular ow in a pneumatic conveying system. Industrial & Engineering
Measurement of charge transfer due to single particle impact. Particle & Chemistry Research 43, 71817199.
Particle Systems Characterization 23, 133137. Yoshida, H., Fukuzono, T., Masuda, H., 1991. Measurement of contact potential
Watanabe, H., Ghadiri, M., Matsuyama, T., Ding, Y.L., Pitt, K.G., Maruyama, H., difference between a powder bed and a metal. Journal of the Society of Powder
Matsusaka, S., Masuda, H., 2007a. Triboelectrication of pharmaceutical powders Technology, Japan 28, 226231.
by particle impact. International Journal of Pharmaceutics 334, 149155. Yoshida, H., Fukuzono, T., Ami, H., Iguchi, Y., Masuda, H., 1992. The effect of the
Watanabe, H., Ghadiri, M., Matsuyama, T., Ding, Y.L., Pitt, K.G., 2007b. New surface modication of particles on the contact potential difference between a
instrument for tribocharge measurement due to single particle impacts. powder bed and metal. Journal of the Society of Powder Technology, Japan 29,
Review of Scientic Instruments 78, 024706. 504510.
Watano, S., 2006. Mechanism and control of electrication in pneumatic Yoshida, M., Shimosaka, A., Shirakawa, Y., Hidaka, J., Matsuyama, T., Yamamoto, H.,
conveying of powders. Chemical Engineering Science 61, 22712278. 2003. Estimation of electrostatic charge distribution of owing toner particles
Yamamoto, H., Scarlett, B., 1986. Triboelectric charging of polymer particles by in contact with metals. Powder Technology 135/136, 2334.
impact. Particle Characterization 3, 117121. Yoshida, M., Ii, N., Shimosaka, A., Shirakawa, Y., Hidaka, J., 2006. Experimental and
Yan, Y., Byrne, B., Woodhead, S., Coulthard, J., 1995. Velocity measurement of theoretical approaches to charging behavior of polymer particles. Chemical
pneumatically conveyed solids using electrodynamic sensors. Measurement Engineering Science 61, 22392248.
Science and Technology 6, 515537. Zhao, H., Castle, G.S.P., Inculet, I.I., 2002. The measurement of bipolar charge in
Yanagida, K., Okada, O., Oka, K., 1993. Low-energy electronic states related to polydisperse powders using a vertical array of Faraday pail sensors. Journal of
contact electrication of pendant-group polymers: photoemission and contact Electrostatics 55, 261278.

Tri Bo Electric

Uploaded by

Copyright:

Available Formats

Tri Bo Electric

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tri Bo Electric

Uploaded by

Copyright:

Available Formats

Chemical Engineering Science 65 (2010) 57815807

Contents lists available at ScienceDirect

Chemical Engineering Science

Triboelectric charging of powders: A review

5.2. Specic charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5795

1. Introduction particle charging caused by repeated impacts on a wall is

Fig. 1. Electron potential energy for metalmetal contact.

materials, and the following relationship is applicable to the 0.5

where fI is the effective work function of insulator and fM is the

eEf z 0 Electron qIM A(EF) E

Metal Window Insulator

For the charging of toner particles used in electrophotography,

2.2. Ion transfer

3. Chemical concept and environmental conditions 300

Specific charge (mC kg1)

3.2. Organic compounds

1 Fig. 11. Triboelectric charging as a function of substituent constant for

Energy control agents (CCAs) either on toner surfaces or to bulk.

the charge stabilizers show lower charging magnitude but long-

100 3.4. Moisture and temperature

Environmental conditions such as relative humidity and

Surface modication of particles is of great importance for

For polymeric materials, the molecular structure affects the 30C

Fig. 17. Schematic illustration of contact process with charge relaxation.

Paschen's break down potential

Dq kc CV , 8 Fig. 18. Schematic illustration of the charge relaxation model.

where kc is the charging efciency, C is the capacitance, and V is

The equilibrium charges of particles made of different materials

4.4. Repeated impacts of a single particle v = 30 ms-1

When particles repeatedly collide with a wall, the charge on 0

Initial charge at the contact 20

Charge of sphere (nC)

Mist Electromagnetic shield

Fig. 30. System to analyze particle charging in gassolids pipe ow.

When the effect of particleparticle interactions on the particle

the charge transfer is reduced. Also, for smaller particles, the

6 The above theoretical approach can be used for analyzing the

the equation of particle charge distribution (Appendix C).

Average air velocity: 30 m s1

10 Pipe material: stainless steel

Specific charge (mC kg1)

Specific charge (mC kg1)

pipes. As a result, the values of the specic charge are within a 5

5. Characterization of triboelectric charging Vibrating

When materials are exposed to electromagnetic radiation, Lower electrode

Positively charged particles

V Negatively charged particles

Fig. 43. Measurement of electrostatic charge distribution.

Particle inlet Pump

Laser beam Measuring spot Activation

Image capturing Data

Fig. 45. Particle motion analysis system (PMAS).

according to the charge-to-diameter ratio. Both polarities can be

cell (Epping and Kuettner, 2002).

Fig. 48. Two-dimensional frequency distribution of specic charge and particle

They performed two different experiments, i.e. the generation of

Air out space-charge eld, which is directed toward the grounded

6.5. Measurements in pneumatic transport Insulator

40 Detection pipe 1 Detection pipe 2