Chemical Engineering Journal 311 (2017) 217–226

Contents lists available at ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier.com/locate/cej

Study on recovery of valuable metals from waste mobile phone PCB

particles using liquid-solid fluidization technique
Chunhu Zhao a, Xiaoping Zhang a,b,⇑, Jiangling Ding a, Yaru Zhu a
a
College of Environment and Energy, South China University of Technology, Guangzhou 510006, Guangdong, China
b
The Key Lab of Pollution Control and Ecosystem Restoration in Industry Clusters, Ministry of Education, South China University of Technology, Guangzhou 510006, Guangdong, China

h i g h l i g h t s

The liquid-solid fluidization technique is used for separation of the WPCB particles.

The metal composition of the WPCB particles was chemically analyzed.

Theoretical fluidization velocities of single components were calculated.

Particle shape factor and water velocity distribution were considered.

The appropriate operating velocity range of the WPCB particles was determined.

a r t i c l e i n f o a b s t r a c t

Article history: In this paper, an effort was made to separate the valuable metals from waste mobile phone printed circuit
Received 26 August 2016 board (PCB) particles using liquid-solid fluidization technique. Based on the mechanical liberation char-
Received in revised form 24 October 2016 acteristics of waste mobile phone PCBs, four closely sieved particles having a granularity, respectively, of
Accepted 13 November 2016
0.075–0.106, 0.106–0.18, 0.18–0.25 and 0.25–0.35 mm were chemically analyzed and charged into the
Available online 14 November 2016
fluidized bed to study the fluidization separation behavior. The minimum fluidization velocities and ter-
minal settling velocities of single components were served as an important tool for the evaluations of the
Keywords:
experimental data. The results showed that, in the process configuration, the fluidization separation was
Mobile phone
PCBs
effective within each limited range of particle size with proper control on water velocity. When the water
Liquid-solid fluidization velocity was in the appropriate range, higher metal yield and grade could be obtained. The calculated
Minimum fluidization velocity minimum fluidization velocities and terminal settling velocities of waste mobile phone PCB particles
Terminal settling velocity were in reasonable agreement with the actual operating water velocities. Additionally, the influence of
the porosity of perforated plate distributor and the height of the fluidized bed on the metal grade and
yield was also discussed. The application of liquid-solid fluidization technique in beneficiation of waste
mobile phone PCB particles is expected to be one of the important developmental directions.
Ó 2016 Elsevier B.V. All rights reserved.

1. Introduction WPCBs recycling techniques, as this is determined by the complex-

ity of the WPCBs compositions and the limitation of the recycling
Recently, because of increasingly new technological innovation techniques. It is well known that WPCBs contain more than 60 ele-
and rapidly short product life cycles, WPCBs have received consid- ments including plenty of toxic heavy metals and toxic organic
erable attention for the resources recycling and environment pro- substances [5–7]. If WPCBs could not be disposed and recycled
tection. In China, tremendous progress in WPCBs recycling properly, these toxic materials may cause serious environmental
technology has been achieved, such as dismantling, shredding, pollution [8]. In addition, the technologies as previously mentioned
physical beneficiation, magnetic separation, hydrometallurgy, and show relatively low separation efficiency, limited treating capacity
pyrometallurgy [1–4]. Despite the achievements being remarkable, and high economic cost, which are difficult to be widely applied in
there are still some obstacles that limit the industrial application of industry. Therefore, the solution lies in developing and adopting
the most effective and relevant technology and in this context, flu-
idization techniques appear promising.
⇑ Corresponding author at: College of Environment and Energy, South China
The fluidization technique has been widely applied in many
University of Technology, Guangzhou 510006, Guangdong, China.
areas of engineering applications due to the advantages of less
E-mail address: [email protected] (X. Zhang).

http://dx.doi.org/10.1016/j.cej.2016.11.091
1385-8947/Ó 2016 Elsevier B.V. All rights reserved.
218 C. Zhao et al. / Chemical Engineering Journal 311 (2017) 217–226

environment pollution, higher material recovery, and lower eco- velocity of single components can be served as an important tool
nomic cost. In the mineral industry, this technique is used in grav- in understanding the fluidization behavior.
ity separation process to segregate the mineral particles having In this paper, liquid-solid fluidization technique is used to sep-
different physical properties [9–11]. Much work has been devoted arate the valuable metals from waste mobile phone PCB particles
to investigating the segregation of particles, and different models with proper control on size and water velocity. And the minimum
have been proposed to illustrate the importance of process param- fluidization velocities and terminal settling velocities of single
eters on segregation. It was found that the efficiency on the segre- components are presented in order to determine suitable operating
gation of the solid particles depends largely on the key process conditions. Additionally, the influence of the porosity of perforated
parameters such as size [12,13], density [12,14], and shape [13]. plate distributor and the height of the fluidized bed on the metal
Dong et al. [15] studied the fluidization characteristics of fine grade and yield is also studied. Overall, the fluidization technique
anthracite using a pulsing dense-phase gas-solid fluidized bed shows the advantages of large treating capacity, low treating cost,
and explained the effects of the pulsating air flow on the minimum high separation efficiency, and little environmental pollution. It
fluidization velocity, stability of the bed density, and motion of the has great application potential to be developed and popularized
heavy medium. Barghi et al. carried out an experimental work and in the field of waste mobile phone PCBs processing.
reported that density differences between components play an
important role in the segregating fluidization process of two- 2. Theoretical consideration
solid beds [16].
From an experimental point of view, it has been shown that In this study, the formulas of minimum fluidization velocity and
when the particles differ in density, they have tendency to segre- terminal settling velocity were used to analyze the experimental
gate during the fluidization process. The particles of lower density results for the fluidization separation of waste mobile phone PCB
will fluidize in the upper part of the bed and particles of higher particles. It was based on the knowledge of the effective forces
density will fluidize at the column bottom. Additionally, particle (gravity, buoyancy, drag force, and particle contact forces) acting
mixing and segregation phenomenon may take place in the flu- upon an individual particle in suspension. In liquid-solid fluidiza-
idized bed. Fluidized beds consisting of single-sized particles show tion process, the drag force is the predominant force for mixing
smooth expansion as the liquid superficial velocity is increased or segregation of particles, and it can be used to predict the mini-
from the minimum fluidization velocity to the single particle ter- mum fluidization velocity. Additionally, when the sum of the drag
minal settling velocity [17]. According to Rasul [18], in a binary flu- force and buoyancy force is equal to the gravity force, the particle
idized bed of particles with different densities, mixing and attains the terminal settling velocity.
segregation phenomenon occurs during the fluidization process
and the transitional behavior through a change in the bulk density 2.1. Minimum fluidization velocity
of different components comprising the particle system was inves-
tigated. It was observed that if choosing appropriate operating con- When the bed is in the critical fluidization state, it can be con-
ditions, any type of particle mixtures might be significantly sidered to be the end of the fixed bed. Therefore, based on the the-
separated in the fluidized beds. oretical analysis of pressure drop in fixed bed, Ergun equation can
For waste mobile phone PCB particulate system, because it con- be used to predict the minimum fluidization velocity, and it is
sists of a variety of irregular particles with different densities, it is described as follows:
difficult to obtain a model which can accurately predict its fluidiza-
ð1  emf Þ2 lumf 1  emf qf umf
2
tion separation process. Therefore, in terms of the prediction for Dp
¼ 150  þ 1:75  ð1Þ
the fluidization behavior, it requires more consideration on the Lmf emf
3
ðus dp Þ
2
emf
3 us dp
effective forces acting on a single particle. As a matter of fact, the
where, Dp is the pressure drop, Lmf is the bed height in the critical
forces acting on a single spherical particle, such as weight, drag
fluidization state, emf is the void fraction in the critical fluidization
and buoyancy, play a critical role in the fluidization process [19–
state, umf is the minimum fluidization velocity, dp is the diameter
21]. It can be concluded that the particle segregation state is deter-
of the particle, us is the shape factor of the particle and, qf is the
mined by the net force acting on a particle and the particle agglom-
fluid density.
eration behavior. Whenever an assembly of particles with more
Rearranging Eq. (1), the equation of pressure drop in fixed bed
than one density or size undergoes fluidization, equilibrium of
can be expressed as the following formula:
the forces can be rapidly reached [22–25].

2

As is known, the force balance on the particle leads to the con- 1:75 dp umf qf 150ð1  emf Þ dp umf qf
cept of terminal settling velocity, which is fundamental to optimiz- þ
e3mf us l e3mf u2s l
ing the operation conditions of fluidized bed composed of waste
PCB particles. Additionally, the drag force is the predominant force
3
dp qf ðqs  qf Þg
¼ ð2Þ
for mixing or segregation of particles, and it can be used to predict l2
the minimum fluidization velocity [13]. Many researchers studied
dp umf qf d3p qf ðqs qf Þg
the separation and recovery of materials from waste printed circuit where, ¼ Rep;mf , ¼ Ar
l l2
boards [26–30]. They reported that appropriate fluidization condi- For simplification, Eq. (2) can be rewritten as follows:
tion can be optimized by knowing the physical properties of the
WPCBs particles. Nevertheless, considering the minimum fluidiza- 1:75 150ð1  emf Þ
Re2p;mf þ Rep;mf ¼ Ar ð3Þ
tion velocity and terminal settling velocity of different components e3mf us e3mf u2s
of waste mobile phone PCB particulate system, there is not still
where, Rep,mf is the Reynolds number, and Ar is the Archimedes
much detailed work done. When the fluid velocity reaches to the
number.
minimum fluidization velocity, the particle will be in the mixing
According to Wen and Yu [31], for the different particulate sys-
state. And when the fluid velocity is higher than the terminal set-
tems, the approximate expressions were established:
tling velocity of a single particle, the particle will be separated out
of the fluidized bed. Therefore, in waste mobile phone PCB partic- 1
 14 ð4Þ
ulate system, due to the difference in the densities of various com- us e3mf
ponents, the minimum fluidization velocity and terminal settling
 C. Zhao et al. / Chemical Engineering Journal 311 (2017) 217–226 219


1  emf 2F D dp uqf
 11 ð5Þ CD ¼ ¼f ¼ f ðReÞ ð13Þ
u2s e3mf Ap qu2 l
Then, Eq. (3) can be rewritten as follows: where, Ap is the cross-sectional area of the particle in suspension,
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi and Ap ¼ p4 dp .
2

Rep;mf ¼ dp umf qf =l ¼ 33:72 þ 0:0408Ar  33:7 ð6Þ For the non-spherical particles, the particle shape factor may
have a certain impact on the drag coefficient. Wang et al. [35]
where Ar ¼ dp qf ðqs  qf Þg=l2 .
3
summed up the shape factors of irregularly shaped particles, which
Using Eq. (6), the minimum fluidization velocities of waste were shown in Table 1.
mobile phone PCB particles can be calculated. In addition, the drag coefficient is a function of Reynolds num-
ber, and the relationship between the shape factor and the Rey-
2.2. Terminal settling velocity nolds number was shown in Table 2.
Especially noteworthy is that there exists boundary effect
When the sum of the drag force and buoyancy force is equal to between the particles and the fluidized bed wall. In order to elim-
the gravity force, the particle attains the terminal settling velocity. inate the impact of boundary effect, the value of dt/dp should be
For a spherical particle, the resulting correlation can be written as, greater than 100. Thus, when the value of dt/dp is smaller than
p p qf u2t p 100, the drag coefficient of the particle should be modified as
dp qs g ¼ C D d p qf g
3 2 3
dp þ ð7Þ below:
6 4 2 6
where ut is the terminal settling velocity of the particle, CD is the ðC D Þav ¼ C W ðC D Þcv ð14Þ
drag coefficient, dp is the diameter of the particle, g is the accelera-
tion due to gravity and, qf and qs, are the densities of the fluid and C W ¼ ð1 þ 2:104dp =dt Þ dt =dp < 10 ð15Þ
particle, respectively.
2:5
In the multiple particle system with a varying particulate C W ¼ ð1  dp =dt Þ dt =dp > 10 ð16Þ
assemblage, the mathematical expressions for the system become
difficult due to the complex interaction between particles and where, (CD)av is the actual value of the drag coefficient, (CD)cv is the
fluid. The drag force is the predominant force for mixing or segre- calculated value of the drag coefficient.
gation of particles, and it should be modified when taking into con- In this study, Eq. (6) was used to calculate the minimum flu-
sideration the effect of neighboring particles. R.D. Felice [32] idization velocities of waste mobile phone PCB particles, and Eq.
proposed a drag force as a function of bed voidage as (10) was used to calculate the metallic and nonmetallic particle
terminal settling velocities. The discrepancies between the experi-
f D ¼ F D f ðeÞ ð8Þ mental data and the calculated results are analyzed, and it can act
Therefore, Eq. (7) can be expressed as following formula. as a guide for the fluidization separation of waste mobile phone
PCB particles.
p p qf u2t
dp gðqs  qf Þ ¼ C D f ðeÞ
3 2
dp ð9Þ
6 4 2 3. Materials and methods
where, f(e) is a function of bed voidage in the multi-particle flu-
idized system. Rearranging Eq. (9), the terminal settling velocity 3.1. Materials
of the particle has the following form:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi The materials used in this study were printed circuit boards of
4 dp ðqs  qf Þg mobile phone, and were of the type that was currently being dis-
ut ¼ ð10Þ
3 C D qf f ðeÞ carded. The electronic components were removed from the printed
circuit boards, and then the boards were cut into small 20-mm-
The function of voidage according to Di Felice is given as f(e)
square pieces. The pieces were crushed using a micro high-speed
= eb, where
universal pulverizer and the pulverized particles were filtered.
Vb  Vp Vp Four closely sieved samples of waste mobile phone PCB particles
e¼ ¼1 ð11Þ
having a granularity, respectively, of 0.075–0.106, 0.106–0.18,
Vb Vb
0.18–0.25 and 0.25–0.35 mm were analyzed by scanning electron
" !#
log ðReÞ2 microscopy (SEM), atomic absorption spectrophotometer (AAS)
b ¼ 3:7  0:65 exp 1:5  ð12Þ and X-ray diffraction spectrum (XRD).
2

The Eqs. (10) and (11) indicate that the terminal settling veloc- 3.2. Experimental apparatus
ity of a spherical particle in the multiple particle system depends
not only on the particle size, particle density, fluid density, and All the experiments of this study were conducted in a cylindri-
the Reynolds number, but also depends on the voidage around cal Plexiglas fluidization column with an inner diameter of 50 mm,
the particle. It can be assumed that the voidage is directly propor- and the schematic diagram of the experimental apparatus was
tional to the number of contacts associated with a particle [33]. shown in Fig. 1. A perforated plate distributor containing holes of
Therefore, for an individual particle, a value of 1.0 can be used 1 mm diameter is fitted at the bottom of the fluidization column.
for the voidage around the particle. If the number of contacts were The water was pumped into the fluidization column from the bot-
more than 12, the voidage corresponding to a randomly packed tom through a calibrated flow-meter by centrifugal pump and
bed was considered. The predicted value of the voidage around a recycled from the column top to the reservoir. A pair of ball values
particle was used to calculate the drag force according to R.D. connected to the calibrated flow-meter and the bypass line con-
Felice’s correlation [32]. Kafui et al. [34] has successfully used trols the flow water into the fluidization column. The flotsam is
the method to model their fluidized bed systems. separated from the upper end of the fluidization column to the
According to the dimensional analysis, the equation of the drag water channel, and is filtered. Subsequently, at the bottom of the
coefficient can be obtained. fluidization column, the jetsam is collected and is chemically ana-
220 C. Zhao et al. / Chemical Engineering Journal 311 (2017) 217–226

Table 1
The shape factors of irregularly shaped particles.

Shape factor Flat Rectangular Polygon Approximate spherical

us Range 0.47–0.58 0.58–0.71 0.67–0.82 0.75–0.91
Average 0.53 0.65 0.75 0.85

Table 2
Non-spherical particle drag coefficient.

us Re
1 10 100 400 1000
0.670 28 6 2.2 2.0 2.0
0.806 27 5 1.3 1.0 1.1
0.846 27 4.5 1.2 0.9 1.0
0.946 27.5 4.5 1.1 0.8 0.8
1.000 26.5 4.1 1.07 0.6 0.46

Metal yield:
According to the equilibrium of mass, the total mass of the sam-
ple is the sum of the flotsam and jetsam.

F ¼W þM ð18Þ

where, F is the total mass of the sample, W is the mass of the flot-
sam, and M is the mass of the jetsam.
In another way,

F  x f ¼ M  xm þ W  xw ð19Þ

where, xf is the concentration of the metal in the sample, xm is the

metal concentration in the jetsam, and xw is the metal concentra-
tion in the flotsam.
Rearranging Eqs. (18) and (19), the following formula can be
obtained:
xf  xw
M¼ F ð20Þ
xm  xw
Using Eq. (20), the equation of the metal yield can be expressed
as follows:

M  xm
Fig. 1. Schematic diagram of the experimental apparatus: a- feeder; b - fluidized
g¼  100% ð21Þ
bed; c - distributor; d - calibrated flow-meter; e - ball value; f - centrifugal pump; g F  xf
- circulating water channel; h - mesh translates.

4. Results and discussions

lyzed. Using this apparatus, the fluidization experiments were car-
ried out to explore the effectiveness of liquid-solid fluidization for
4.1. The metal composition of waste mobile phone PCBs
the separation of waste mobile phone PCB particles.

Fig. 2 shows the X-ray diffraction spectrum analysis results of

3.3. Methods waste mobile phone PCB particles. The intensity of diffraction peak
increases with the increase of metal concentration. Contrasted
For preliminary study, the fluidization separation of waste with the standard XRD patterns of copper, the diffraction peaks
mobile phone PCB particles was carried out. All the feed samples of 43.44°, 50.56°, 74.12°, 89.91° correspond to the crystal faces of
were fluidized at different levels of water velocity. As before, each copper. The results show that most of the metal elements exist
time 50 g of feed was charged into the fluidized bed and at steady as a simple substance in the waste mobile phone PCB particles,
state, one set of samples was collected from the bottommost sam- and the copper content is the highest. This is in good agreement
ple collection port. Then the samples were filtered and dried. These with the AAS analysis results.
products were chemically analyzed. After that, the metal yield and The metal composition of waste mobile phone PCBs was shown
grade were plotted against the respective water velocity. in Table 3. To be clear, the values represent the analysis results for
The metal grade and yield were defined as follows. particles that were sufficiently mixed after pulverization of waste
Metal grade: printed circuit boards obtained from 100 or more mobile phones,
and it is believed that these values represent the average values.
C  V  N  103
x¼  100% ð17Þ From Table 3, waste mobile phone PCBs contains many metallic
m
elements such as Cu, Sn, Pb, Fe, Al, Zn, and Ni, including a small
where, C is the measured value of the metal, V is the volume of the amount of precious metal Au and Ag. It can also be seen that in dif-
sample, N is the dilution ratio of the sample, and m is the total mass ferent particle size range, the relative enrichment degree of these
of the filtered products. metals is quite different. This is mainly attributable to the differ-
 C. Zhao et al. / Chemical Engineering Journal 311 (2017) 217–226 221

From Fig. 3, it can be clearly seen that the metal particles of

waste mobile phone PCBs appear basically near-spherical
(us = 0.85), while the nonmetal particles are mostly acicular or floc-
culent (us = 0.65), which have higher drag coefficients. Conse-
quently, by Eq. (10), it can be inferred that the actual terminal
settling velocities of the metal and nonmetal particles are less than
the calculated values. This effectively means that at smaller veloc-
ities than the calculated values, the nonmetal particles will flow
out of the fluidized bed, resulting in an increase of metal grade.
Under the experimental condition, the Reynolds number can be
calculated, and then according to Table 2, the drag coefficient can
be obtained. From Eq. (6) and Eq. (10), the corresponding mini-
mum fluidization velocities and terminal settling velocities of dif-
ferent components are calculated, and the results are shown in
Tables 4 and 5, respectively. Based on these calculated data, the
fluidization separation process of waste mobile phone PCB parti-
cles will be analyzed.
As shown in Table 4 and Table 5, there has a significant differ-
ence in the minimum fluidization velocities and terminal settling
Fig. 2. X-ray diffraction spectrogram of waste mobile phone PCB particles.
velocities between the metallic particles and the nonmetallic par-
ticles, and the minimum fluidization velocities and terminal set-
ences in the grinding properties between the metallic particles and tling velocities of the non-metallic particles are obviously less
the nonmetallic particles. The metals, which were difficult to be than that of the metallic particles. Therefore, the fluidization sepa-
crushed because of its high strength and ductility, are mostly dis- ration of the WPCBs particles can be achieved at a certain range of
tributed in the larger particle size range. Conversely, the metals, the water velocity. In the fluidization separation process, the parti-
which were easy to be crushed, are mainly concentrated in the cles can be affected by various influence factors such as the inter-
smaller particle size range. action of the particles, the fluidized device and other additional
In addition, it can be observed that, in terms of the metal com- factors interference [36]. Therefore, in the interference conditions,
position, Cu, which is used in the circuitry, has the highest content when the particles are segregated according to the minimum flu-
in the particle size range of 0.25–0.35 mm. Sn and Pb are mainly idization velocities and terminal settling velocities, the water
concentrated in the range of 0.075–0.106 mm and 0.25–0.35 mm. velocity should be determined by the properties of the various
And Al is mainly concentrated in the range of 0.075–0.106 mm, components of the WPCBs particles.
0.106–0.18 mm and 0.18–0.25 mm. Furthermore, in terms of the
precious metal composition, Au and Ag, which are used as contact 4.3. Liquid-solid fluidization separation of the WPCBs particles at
materials or as plating layers due to their high conductivity and various particle sizes
chemical stability, are mainly enriched in the particle size range
of less than 0.106 mm. Additionally, the total metal content of As shown in Table 3, compared with the other metals, the con-
52.83% are distributed in the range of 0.25–0.35 mm. Based on tent of copper is the highest in the waste mobile phone PCB parti-
the above analysis results, it provides a basis for the fluidization cles, which is the major enrichment object. Thus, all the
separation process of waste mobile phone PCB particles. experiments are designed towards the beneficiation of copper
through fluidization. In the experiments, in each case 50 g of waste
mobile phone PCB particles with different particle size (0.075–
0.106, 0.106–0.18, 0.18–0.25 and 0.25–0.35 mm) were charged
4.2. Mathematical calculation of the minimum fluidization velocity into the fluidized bed to study the effect of water velocity on the
and terminal settling velocity metal grade and yield, the height of the fluidized bed was 50 cm,
and the porosity of the perforated plate distributor was 1%. The
Under ideal conditions, the spherical particles are used for experimental results were shown in Figs. 4–7.
investigating the fluidization separation behavior and calculating From Figs. 4–7, it can be observed that the trend of the copper
their minimum fluidization velocities and terminal settling veloci- grade and yield changing with the water velocity was similar to
ties. However, in the actual fluidization separation process, the that of the total metal. With the increase of the water velocity,
particles are mostly non-spherical, especially for waste mobile the metal grade increased, while the metal yield generally
phone PCB particles. Therefore, it is necessary to observe the decreased. It indicated that the nonmetallic particles were firstly
shapes of the metallic and non-metallic particles. In this study, separated, and the metallic particles with smaller densities were
the morphology of the WPCBs particles was analyzed by scanning separated afterwards. Based on the calculated terminal settling
electron microscopy, and the SEM image was shown in Fig. 3. velocities of different components (see Table 5), it can be inter-

Table 3
The metal composition of the printed circuit boards from waste mobile phone.

Particle size, mm Metallic element content, wt%

Cu Sn Pb Al Fe Zn Ni Ag Au, ppm Total
<0.075 2.16 0.14 0.44 0.86 0.7 0.03 0.13 0.0038 7.82 4.46
0.075–0.106 8.98 3.29 2.44 2.35 0.26 0.038 0.086 0.0025 9.75 17.45
0.106–0.18 21 2.32 1.5 1.71 0.15 0.023 0.037 0.001 4.5 26.74
0.18–0.25 28.5 2.15 1.56 1.90 0.14 0.022 0.027 0.0013 4.62 34.30
0.25–0.35 43.89 4.45 3.15 1.1 0.15 0.033 0.05 0.0035 0.72 52.83
222 C. Zhao et al. / Chemical Engineering Journal 311 (2017) 217–226

Fig. 3. SEM image of waste mobile phone PCB particles.

Table 4
Properties of waste mobile phone PCB particles: particle sieve size range, values of the average sieve diameter (d p) and the minimum fluidization velocities (u mf).

Fraction dp, mm umf, m/s

(sieve size, mm)
Epoxy resin Glass fiber Copper Tin Lead Aluminum
0.075–0.106 0.0905 6.36  106 7.61  105 4.29  104 3.41  104 5.62  104 9.24  105
0.106–0.18 0.143 1.59  105 1.90  104 1.07  103 8.51  104 1.40  103 2.31  104
0.18–0.25 0.215 2.87  105 4.29  104 2.40  103 1.91  103 3.14  103 5.21  104
0.25–0.35 0.3 6.99  105 8.33  104 4.61  103 3.68  103 6.00  103 1.01  103

Table 5
Properties of waste mobile phone PCB particles: particle sieve size range, values of the average sieve diameter (d p) and the terminal settling velocities (ut).

Fraction dp, mm ut, m/s

(sieve size, mm)
Epoxy resin Glass fiber Copper Tin Lead Aluminum
0.075–0.106 0.0905 0.0022 0.0077 0.0186 0.0166 0.0213 0.0086
0.106–0.18 0.143 0.0028 0.0097 0.0234 0.0208 0.0268 0.0108
0.18–0.25 0.215 0.0074 0.0256 0.0702 0.0626 0.0803 0.0326
0.25–0.35 0.3 0.0087 0.0302 0.0830 0. 0740 0.0949 0.0385

Fig. 4. Fluidization separation results for 0.075–0.106 mm waste mobile phone PCB Fig. 5. Fluidization separation results for 0.106–0.18 mm waste mobile phone PCB
particles. particles.

preted that, for particles with the same diameter, the terminal set- It can also be found that, at the primary stage of fluidization
tling velocities of the nonmetallic particles are smaller than that of separation process, the copper yields remain unchanged. When
the metallic ones. Therefore, when the water velocity was at a cer- the water velocity increases to a certain value, there has been a
tain value higher than the terminal settling velocity of the non- clearly declining trend in the copper yield. Compared with the cal-
metallic particles, the metal grade increased. Additionally, the culated terminal settling velocity of copper, the actual separation
densities of some metallic particles, such as aluminum, were velocity of copper is smaller, but it is still in the reasonable range.
almost the same as that of the nonmetallic ones. In the fluidization As is known, the nonmetallic particles do not have a neat edge but
process, it will flow out of the fluidized bed along with the non- are covered with many tiny burrs. Therefore, in the actual separa-
metallic particles, resulting in the decrease of the total metal yield. tion process, the nonmetallic particles can easily intertwine with
 C. Zhao et al. / Chemical Engineering Journal 311 (2017) 217–226 223

poor separation results were achieved with the particles of

0.075–0.106 mm and 0.106–0.18 mm, when the metal yield was
about 95%, the metal grade was less than 80%. It might be caused
by two reasons: (1) with the decrease of particle size, the particle
surface adhesion force increased, resulting in a subsequent
increase of particle aggregation, and (2) the materials in these
ranges mainly consist of glass fiber and resin, the electrostatic
effect between the particles and the fluidized bed wall make it dif-
ficult to be separated. Thus, for finer waste mobile phone PCB par-
ticles (dp < 0.18 mm), it not only consumed more energy during the
grinding process, but also showed a lower separation efficiency in
the fluidization process. Consequently, the liquid-solid fluidization
technique is suitable for waste mobile phone PCB particles with
particle size above 0.2 mm, where better metal grade and yield
can be obtained.
Based on the above analysis, it can be concluded that high metal
grade and yield can be obtained when the appropriate water veloc-
Fig. 6. Fluidization separation results for 0.18–0.25 mm waste mobile phone PCB ity is adjusted, and the summary of the appropriate operating
particles. water velocity range was shown in Table 6. From Table 6, for
0.075–0.106 mm particles, at the water velocity range of 0.014–
0.016 m/s, the copper grade was about 40%, but the copper yield
was more than 83.7%. For 0.106–0.18 mm particles, at the water
velocity range of 0.020–0.023 m/s, the copper yield was more than
83.6% and the copper grade was about 60%. In addition, for 0.18–
0.25 mm sized particles, the copper grade was high at the water
velocity range of 0.033–0.044 m/s, where the value was 69.79–
75.76%, and the yield was more than 83.1%. For 0.25–0.35 mm par-
ticles, the copper grade was about 75% and the copper yield was
also more than 82.4% at the water velocity range of 0.050–
0.075 m/s. To summarize, for waste mobile phone PCB particles
with different particle size, the nonmetallic particles and the
metallic ones can be segregated effectively in the respective better
operating water velocity range.
It is also necessary to compare the experimental results
obtained in this work with the literatures’. M. Sarvar et al. [37]
using wet jigging and froth flotation methods to recycle the metal
from WPCBs. In their work, 95.6%, 97.5% and 85% of metal content
of coarse size, middle size and small size fraction are recovered. S.
Fig. 7. Fluidization separation results for 0.25–0.35 mm waste mobile phone PCB
K. Dey et al. [38] studied the processing of electronic waste in a
particles.
counter current teeter-bed separator. It shows that a recovery of
94% is achievable at nearly 60% mass yield. Additionally, a number
the metallic ones, which will be entrained out of the fluidized bed. of articles have been published in recent years pertaining to recycle
In addition, the metal and nonmetallic particles are a collection of the copper using the degradation technique [39,40]. In these stud-
particles with diameters within a certain range. Therefore, the cor- ies, the copper recovery rate was above 95%. Compared with these
responding separation velocity is subsequently more diverse, literatures, the liquid-solid fluidization technique that used in this
resulting in some particles being separated from the fluidized work shows good separation efficiency, large treatment capacity
bed at lower velocities. This experimental phenomenon was also and low economic cost. It can be believed that it would be of
observed by Han [29]. importance for the engineering practice.
From Fig. 4, on contrast to the other WPCBs particulate systems,
for 0.075–0.106 mm and 0.18–0.25 mm particles, there has a 4.4. Liquid-solid fluidization separation of the WPCBs particles at
greater descending range of the total metal yield. This may be various porosities
because there has a relatively higher percentage of aluminum in
this particulate system (see Table 3). Additionally, for 0.106– The porosity and the distribution of the pores are the key
0.18 mm particles, it is noteworthy that the calculated terminal parameters of liquid distributor, which are particularly significant
settling velocity of tin is 0.0208 m/s, which means that tin is com- in the expansion of the fluidized bed and the flow state of the par-
pletely separated out of the fluidized bed when the water velocity ticles. The equation of the porosity can be described as follows:
reaches to the value of 0.021 m/s. This phenomenon is unique for
this particulate system, and according to the AAS analysis, the con- pd2 d
2

tent of tin in the jetsam is rare, which giving indirect evidence to a¼  ¼ 0:907 ð22Þ
4s2 sin 60 s2
the applicability of the terminal settling velocity.
It can be seen obviously from Figs. 4–7 that, under the experi- where, d is the inner diameter of the distributed holes, s is the pore
mental conditions, the separation efficiency decreased with the spacing.
decrease of the particle size. The better separation results were In this experiment, four fluid distributors were used to investi-
with 0.18–0.25 mm and 0.25–0.35 mm sized particles, where the gate the effect of porosity on the copper grade and yield. And the
metal yield was above 95% at about 90% of the metal grade. The porosity for the four fluid distributors can be calculated by Eq.
(22), which was 2%, 1.5%, 1%, and 0.5%.
224 C. Zhao et al. / Chemical Engineering Journal 311 (2017) 217–226

Table 6
The appropriate operating water velocity range of waste mobile phone PCB particles with different particle size.

Particle size, Water velocity, Copper Total metal

mm m/s
Grade, % Yield, % Grade, % Yield, %
0.075–0.106 0.014–0.016 38.93–41.54 83.7–96.2 66.41–70.67 73.1–84.5
0.106–0.18 0.020–0.023 57.84–66.26 83.6–95.4 70.70–78.80 86.0–93.1
0.18–0.25 0.033–0.044 69.79–75.76 83.1–99.9 80.01–87.92 80.0–94.6
0.25–0.35 0.050–0.075 72.15–75.87 82.4–98.9 86.79–91.09 82.1–98.3

The pressure drop of the fluidized bed is given by:

DPB ¼ Hqs ¼ 0:925 kPa ð23Þ

where, H = 50 mm, qs = 1.85 kg/m3

The pressure drop of the perforated plate distributor can be cal-
culated using the standard relationship:

qf u2
DP D ¼ c ð24Þ
2g a2

where, c is the resistance coefficient of the perforated plate distrib-

utor, qf is the fluid density, u is the operating fluid velocity.
In the experiments, in each case 50 g of waste mobile phone
PCB particles with particle size of 0.18–0.25 mm were charged into
Fig. 9. Variation in the copper grade at different water velocity for the WPCBs
the fluidized bed to study the effect of water velocity on the copper
particles with different porosity.
grade and yield. The height of the fluidized bed was 50 cm, and the
porosities of the perforated plate distributor were 2%, 1.5%, 1%, and
0.5%, respectively. The experimental results were shown in Figs. 8 4.5. Liquid-solid fluidization separation of the WPCBs particles at
and 9. various fluidized bed heights
As shown in Figs. 8 and 9, it indicated that, at the same water
velocity, the variation of porosity had no significant effect on the In the experiments, the WPCBs samples with particle size of
grade and yield of copper. This is mainly because the porosities 0.25–0.35 mm have been used to study the influence of bed height
can meet the requirements of the stability of the pressure drop on the grade and yield of copper with the increase of the water
of the fluidized bed. Based on the different porosity (2%, 1.5%, 1%, velocity, and the experimental results were shown in Figs. 10
and 0.5%), the pressure drop of the perforated plate distributor and 11.
can be calculated by Eq. (13), and which are all greater than As shown in Figs. 10 and 11, with the increase of the fluidized
0.1DPB. Therefore, at the series of the perforated plate distributor, bed height, the copper grade decreased, while the copper yield
a good and stable fluidized bed can be established and maintained, increased. The experimental results were in good agreements with
so that the fluidization process of waste mobile phone PCB parti-
cles can be operated stably and effectively.

Fig. 8. Variation in the copper yield at different water velocity for the WPCBs Fig. 10. Variation in the copper yield at different water velocity for the WPCBs
particles with different porosity. particles with different bed height.
 C. Zhao et al. / Chemical Engineering Journal 311 (2017) 217–226 225

5. Conclusions

In this study, an attempt was made to separate the valuable

metals from waste mobile phone PCB particles using liquid-solid
fluidization technique. Based on this approach, the experimental
investigation of this paper leads to conclusions that can be summa-
rized as follows: Waste mobile phone PCB particles contain many
metallic elements such as Cu, Sn, Pb, Fe, Al, Zn, and Ni, including
a small amount of precious metal Au and Ag. Compared with the
other metals, the content of copper is the highest in waste mobile
phone PCB particles, which is the major enrichment object. With
regard to the fluidization separation of waste mobile phone PCB
particles, the minimum fluidization velocities and terminal settling
velocities of different particles were successfully used for evaluat-
ing the fluidization separation process. It shows that the experi-
mental results were in good agreements with the theoretical
analysis. According to the experimental results and the theoretical
analysis, for waste mobile phone PCB particles with different par-
ticle size, the nonmetallic particles and the metallic particles can
Fig. 11. Variation in the copper grade at different water velocity for the WPCBs
particles with different bed height.
be segregated effectively in the respective better operating water
velocity ranges. At the same water velocity, the variation of poros-
ity has no significant effect on the grade and yield of copper. Addi-
the theoretical analysis. From Eq. (10), it can be seen that the ter- tionally, proper fluidized bed height is conducive to the
minal settling velocity will increase when the voidage of waste beneficiation of copper. Therefore, this study can be served as a
mobile phone PCB particles increases. Therefore, higher fluidized theoretical guide for engineering applications, especially for the
bed height is conducive to the beneficiation of copper. e-waste recycling.
It can also be found from the figures that the copper grade at the
bed height of 30 cm was higher than that of 50 cm and 80 cm when Acknowledgements
corresponding to the same water velocity, while the copper yield at
the bed height of 30 cm was obviously lower than that of 50 cm This study was supported by Natural Science Foundation of
and 80 cm. The main reason for this phenomenon lies in the inter- China (21377041), the Science and Technology Project of Guang-
face between the dense-phase zone and the dilute-phase zone. As dong Province (2013B021300022), State Key Laboratory of Sub-
is known, at a particular water velocity, when the height of feed tropical Building Science Foundation (2013KB28), and the High-
inlet is varying above this interface, the voidage is high and the tech Modern Industry Public Service Platform Construction of
particle density is low, leading to a very minimal impact on the Meizhou (2011B010600029). We thank our visiting professor Dr.
grade and yield of copper. On the contrary, when the height of feed Donald Barnes for providing valuable comments on the
inlet is varying below this interface, there will be a significant manuscript.
impact on that of copper. Therefore, some heavy metal particles
were entrained out of the fluidized bed along with the non- References
metallic particles, leading to an obvious decrease in the copper
grade and yield. [1] C. Eswaraiah, T. Kavitha, S. Vidyasagar, S.S. Narayanan, Classification of metals
and plastics from printed circuit boards (PCB) using air classifier, Chem. Eng.
From Fig. 11, it can also be found that, at the velocity lower than Process. 47 (2008) 565–576.
0.0625 m/s, the copper grade at the height of 30 cm was higher [2] J. Li, H. Lu, J. Guo, Z. Xu, Y. Zhou, Recycle technology for recovering resources
than those of 50 cm and 80 cm. However, when the velocity was and products from waste printed circuit boards, Environ. Sci. Technol. 41
(2007) 1995–2000.
higher than 0.0625 m/s, the copper grade at the height of 30 cm [3] Z. Wang, N.J. Miles, T. Wu, F. Gu, P. Hall, Recycling oriented vertical vibratory
decreased to be lower than those of the other two heights. The separation of copper and polypropylene particles, Powder Technol. 301 (2016)
cause of this phenomenon is that at a particular water velocity, 694–700.
[4] J.B. Wang, J. Guo, Z.M. Xu, An environmentally friendly technology of
the jet velocity of the distribution pores plays an important role disassembling electronic components from waste printed circuit boards,
on the particle flow behavior. It is known that the jet velocity is Waste Manage. 53 (2016) 218–224.
usually 2–10 times of the operating water velocity. When the [5] J.B. Wang, Z.M. Xu, Disposing and recycling waste printed circuit boards:
disconnecting, resource recovery, and pollution control, Environ. Sci. Technol.
bed height is lower, the influence of the jet velocity on the particles
49 (2015) 721–733.
is relatively larger, resulting in a lower copper grade. [6] L. Zhan, X.S. Xiang, B. Xie, J. Sun, A novel method of preparing highly dispersed
In addition, it can be observed that there is little difference in spherical lead nanoparticles from solders of waste printed circuit boards,
Chem. Eng. J. 303 (2016) 261–267.
the copper grade and yield between the bed height of 50 cm and
[7] S.B. Wath, M.N. Katariya, S.K. Singh, G.S. Kanade, A.N. Vaidya, Separation of
that of 80 cm. This may be due to the samples were in a stable flu- WPCBs by dissolution of brominated epoxy resins using DMSO and NMP: A
idization state at the bed height of 50 cm and 80 cm, the effect of comparative study, Chem. Eng. J. 280 (2015) 391–398.
the interface between the dense-phase zone and the dilute-phase [8] R.R. Liu, J.Y. Chen, G.Y. Li, T.C. An, Using an integrated decontamination
technique to remove VOCs and attenuate health risks from an e-waste
zone was relatively smaller, and the water velocity determined dismantling workshop, Chem. Eng. J. (2016), http://dx.doi.org/10.1016/j.
the separation efficiency. Obviously, there is a minimum separa- cej.2016.05.004.
tion bed height of the fluidization column. When the actual sepa- [9] N. Epstein, Liquid solids fluidization, in: W.C. Yang (Ed.), Handbook of
Fluidization and Fluid-Particle Systems, Marcel Dekker, New York, 2003, pp.
ration height is greater than it, the further increases of the bed 705–764.
height have no effect on the separation efficiency. Considering [10] P.V. Chavan, J.B. Joshi, Analysis of particle segregation and intermixing in solid-
the reasons for the pump energy consumption, the author chose liquid fluidized bed, Int. Eng. Chem. Res. 47 (2008) 8458–8470.
[11] R.D. Felice, Liquid suspensions of single and binary component solid particles-
50 cm for the final separation bed height. an overview, China Particuol. 5 (2007) 312–320.
[12] V.S. Patwardhan, C. Tien, Sedimentation and liquid fluidization of solid
particles of different sizes and density, Chem. Eng. Sci. 40 (1985) 1051–1060.
226 C. Zhao et al. / Chemical Engineering Journal 311 (2017) 217–226

[13] H.P. Liu, L.Y. Zhang, T.P. Chen, S.W. Wang, Z.J. Han, S.H. Wu, Experimental Environmental Science and Technology, Prog. Environ. Sci. Technol. 1 (2007)
study on the fluidization behaviors of the superfine particles, Chem. Eng. J. 262 1410–1414.
(2015) 579–587. [27] K. Huang, J. Guo, Z.M. Xu, Recycling of waste printed circuit boards: a review of
[14] J.F. He, M.B. Tan, R. Zhu, Y.M. Zhao, P. Hong, Y. Jiang, Density-based separation current technologies and treatment status in china, J. Hazard. Mater. 164
performance of a secondary air-distribution fluidized bed separator (SADFBs) (2009) 399–408.
for producing ultra-low-ash clean coal, Fuel 172 (2016) 178–186. [28] Y.H. Zhou, K.Q. Qiu, A new technology for recycling materials from scrap
[15] L. Dong, E.H. Zhou, L.H. Cai, C.L. Duan, Y.M. Zhao, Z.F. Luo, Fluidization printed circuit boards, J. Hazard. Mater. 175 (2010) 823–828.
characteristics of a pulsing dense-phase gas-solid fluidized bed for high- [29] Y.B. Han, W.Z. He, L.D. Li, G.M. Li, J.W. Huang, Mathematical analysis of the gas-
density separation of fine anthracite, Energy Fuels 30 (2016) 7180–7186. solid fluidized bed separation of metals and nonmetals from waste PCB
[16] S. Barghi, C.L. Briens, M.A. Bergougnou, Mixing and segregation of binary powders, Powder Technol. 295 (2016) 142–151.
mixtures of particles in liquid-solid fluidized beds, Powder Technol. 131 [30] W.H. Li, Y. Zhi, Q.Y. Dong, L.L. Liu, J.H. Li, S.L. Liu, H.H. Xie, Research progress on
(2003) 223–233. the recycling technology for nonmetallic materials from waste printed circuit
[17] D. Mihal, G. Radmila, A. Zorana, Segregation in water fluidized beds of sand board, Procedia Environ. Sci. 16 (2012) 569–575.
particles, Powder Technol. 235 (2013) 173–179. [31] C.Y. Wen, Y.H. Yu, Chem. Eng. Prog. Symp. Ser. 62 (1966) 110–111.
[18] M.G. Rasul, V. Rudolph, F.Y. Wang, Particles separation using fluidization [32] R.D. Felice, The voidage function for fluid-particle interaction systems, Int. J.
techniques, Int. J. Miner. Process. 60 (2000) 163–179. Multiphase Flow. 20 (1994) 153–159.
[19] B. Formisani, Packing and fluidization properties of binary mixtures of [33] A.K. Mukherjee, B.K. Mishra, Experimental and simulation studies on the role
spherical particles, Powder Technol. 66 (1991) 259–264. of fluid velocity during particle separation in a liquid-solid fluidized bed, Int. J.
[20] R.D. Felice, Hydrodynamics of liquid fluidization, Chem. Eng. Sci. 50 (1995) Miner. Process. 82 (2007) 211–221.
1213–1245. [34] K.D. Kafui, C. Thornton, M.J. Adams, Discrete particle-continuum modeling of
[21] J. Yang, A. Renken, A generalized correlation for equilibrium of forces in liquid- gas-solid fluidized beds, Chem. Eng. Sci. 57 (2002) 2395–2410.
solid fluidized beds, Chem. Eng. J. 92 (2003) 7–14. [35] D.Z. Wang, G.Z. Qiu, Y.H. Hu, Resources Processing, Science Press, Beijing, 2005.
[22] A.K. Mukherjee, B.K. Mishra, R.V. Kumar, Application of liquid-solid [36] M. Schmidt, A. Buck, E. Tsotsas, Environmental investigation of process
fluidization technique in beneficiation of fines, Int. J. Miner. Process. 92 stability of continuous spray fluidized bed layering with internal separation,
(2009) 67–73. Chem. Eng. Sci. 126 (2015) 55–66.
[23] N. Mostoufi, J. Chaouki, Prediction of effective drag coefficient in fluidized [37] M. Sarvar, M.M. Salarirad, M.A. Shabani, Characterization and mechanical
beds, Chem. Eng. Sci. 54 (1999) 851–858. separation of metals from computer Printed Circuit Boards (PCBs) based on
[24] T. Renganathan, K. Krishnaiah, Voidage characteristics and prediction of bed mineral processing methods, Waste Manage. 45 (2015) 246–257.
expansion in liquid-solid inverse fluidized bed, Chem. Eng. Sci. 60 (2005) [38] S.K. Dey, V. Ari, A. Das, Processing of electronic waste in a counter current
2545–2555. teeter-bed separator, J. Environ. Manage. 107 (2012) 45–51.
[25] B. Formisani, G.D. Cristofaro, R. Girimonte, A fundamental approach to the [39] M.F. Xing, F.S. Zhang, Degradation of brominated epoxy resin and metal
phenomenology of fluidization of size segregation binary mixtures of solids, recovery from waste printed circuit boards through batch sub/supercritical
Chem. Eng. Sci. 56 (2001) 109–119. water treatments, Chem. Eng. J. 219 (2013) 131–136.
[26] M. Xu, G.M. Li, W.Z. He, H. Li, Metals recovery from waste printed circuit [40] N. Li, X.B. Lu, S.T. Zhang, A novel reuse method for waste printed circuit boards
boards using fluidization technique, International Symposium on as catalyst for wastewater bearing pyridine degradation, Chem. Eng. J. 257
(2014) 253–261.