materials

Review
Research Status of High-Purity Metals Prepared by
Zone Refining
Liang Yu 1,2,3 , Xiaoan Kang 1,2,3 , Luona Chen 1,2,3 , Kun Luo 4 , Yanli Jiang 1,2,3, * and Xiuling Cao 5, *

1 Key Laboratory of New Processing Technology for Nonferrous Metals & Materials, Ministry of Education,
Guilin University of Technology, Guilin 541004, China; [email protected] (L.Y.); [email protected] (X.K.);
[email protected] (L.C.)
2 Guangxi Scientific Experiment Center of Mining, Metallurgy and Environment, Guilin 541004, China
3 Collaborative Innovation Center for Exploration of Nonferrous Metal Deposits and Efficient Utilization of
Resources, Guilin University of Technology, Guilin 541004, China
4 School of Materials Science and Engineering, Changzhou University, Changzhou 213164, China;
[email protected]
5 School of Exploration Technology and Engineering, Hebei Geosciences University,
Shijiazhuang 050031, China
* Correspondence: [email protected] (Y.J.); [email protected] (X.C.); Tel.: +86-135-9163-3992 (Y.J.);
+86-135-1331-0032 (X.C.)

Abstract: The zone refining method is a physical method for effectively purifying metals. Increas-
ing yield and reducing impurity content have always been the focus of its research. This article
systematically summarizes the relevant research on the production of high-purity metals by zone
refining, including mechanisms, parameter optimization, zone refining types, analysis methods,




limitations, and future development directions, and it provides relevant theoretical foundations for

 the production of high-purity metals as well.
Citation: Yu, L.; Kang, X.; Chen, L.;
Luo, K.; Jiang, Y.; Cao, X. Research Keywords: zone refining; high-purity metals; purification
Status of High-Purity Metals
Prepared by Zone Refining. Materials
2021, 14, 2064. https://doi.org/
10.3390/ma14082064 1. Introduction
High-purity metals (99.999% (5N) or higher) are widely used in modern electronic
Academic Editor:
information, aerospace, defense, and military industries listed in Table 1. The development
Antonio Sánchez-Coronilla
trend of modern science and technology requires high-purity or ultra-high-purity of metals,
because some important characteristics of metals are affected by the type and amount of
Received: 9 December 2020
impurities in the matrix, and some characteristics are even masked by trace elements [1].
Accepted: 13 April 2021
Metals are often purified by vacuum distillation technology [2], ion exchange tech-
Published: 20 April 2021
nology [3], extraction technology [4], electrolytic refining technology [5], zone refining
technology [6], and other single or combined technologies.
Publisher’s Note: MDPI stays neutral
Among them, the zone refining technology has a wide range of applications; in
with regard to jurisdictional claims in
published maps and institutional affil-
addition, it is simple and easy to control, has no pollution, has a high product purity, and
iations.
is suitable for the final stage of preparing high-purity metals. Since Pfann [7] first used
the zone refining technology to purify germanium in 1955, it has been widely used in
the purification of high-purity metals and the preparation of high-quality, multi-variety
refractory single crystals. Zone refining is a technology of deeply purifying metals. Its
essence is to use the difference in solubility of impurity elements in the solid and molten
Copyright: © 2021 by the authors.
state of the main metal to precipitate the impurities or change the distribution of the
Licensee MDPI, Basel, Switzerland.
impurity elements. It provides an effective and easy method for preparing high-purity
This article is an open access article
distributed under the terms and
metals. Theoretically, high-purity metals up to 8N can be obtained.
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).

Materials 2021, 14, 2064. https://doi.org/10.3390/ma14082064 https://www.mdpi.com/journal/materials

Materials 2021, 14, 2064 2 of 21

Table 1. Application of high-purity metals.

Metal Application
In [8] ITO targets, CIGS solar cells, liquid crystal displays, etc.
Sn [9] Packaging materials, integrated circuits, refractory materials, etc.
Ni [10] Stainless steel, alloy steel, high-temperature structural materials, etc.
Cu [11] Audio products, integrated circuits, fatigue-resistant cables, etc.
Co [12] Magnetic materials, super alloys, electronic component targets, etc.
Ti [13] Large-scale integrated circuits, decorative materials, etc.
Ga [14] Semiconductor materials, solar cells, catalytic materials, etc.
Ge [15] Integrated circuits, photovoltaic cells, infrared optical materials, etc.
Te [16] Aerospace, atomic energy, electronics industry, etc.
Al [17] Target materials, integrated circuit wiring, optoelectronic storage media, etc.

2. Zone Refining Mechanism

2.1. Basic Principles
Zone refining is a technology of deeply purifying metals [18]. The distribution of
impurity elements in the solid and liquid phase of the bulk metal melt is determined by
the thermodynamic properties of the system. Impurities exist as solid solutions in the
main metal. Solid solution is generally due to the presence of metal B atoms in the crystal
lattice of metal A, thereby forming a solid solution. Due to the presence of trace impurities,
the melting point of the metal may decrease or increase. The extent to which the melting
point decreases or increases depends on the content of impurities. Decreasing the melting
point causes the solid solution to change from the molten state to the solid state, and
the impurities migrate from the solid phase to the liquid phase. When the melting point
increases, the opposite is true, and the impurities migrate from the liquid phase to the solid
phase. For the binary system composed of A and B (the metal A contains impurities B),
there is the following relationship [1,19,20]:

R( T 0f )2 x B2
∆T f = x B1 [1 − ]. (1)
∆H f A x B1

In the formula, “1” represents the molten phase, “2” represents the solid solution
phase; xA1 and xB1 are the concentrations of metal A and impurity B before the change,
while xA2 and xB2 are the concentrations of A and B after the change; ∆H f A is the heat of
dissolution when 1 mol A melts into the solution from a solid solution state, and its value
is positive. Due to the low impurity content, it can be regarded as a constant; T 0f is the
melting point of pure A; T f is the melting point of the solid solution at the mole fraction of
xA2 ; ∆T f = T 0f − T f is the melting point change value; R is the gas constant, and the value
is 8.314.
Assume
x
k0 = B2 . (2)
x B1
If k0 < 1, that is xB1 > xB2 ; then, xA1 < xA2 , and the concentration of A in the solid
phase is greater than the concentration of A in the liquid phase. When ∆T f > 0, the melting
point rises, as shown in Figure 1b. If k0 > 1, that is, xB1 < xB2 ; then, xA1 > xA2 , and the
concentration of A in the solid phase is smaller than the concentration of A in the liquid
phase. When ∆T f < 0, the melting point drops, as shown in Figure 1a. In Figure 1, the
upper part is the molten liquid zone; the middle is the solid–liquid two-phase equilibrium
zone; the lower part is the solid phase zone.
Materials 2021,14,
Materials2021, 14,2064
x FOR PEER REVIEW 33 of
of 21
21

Figure1.1.Binary
Figure Binarysystem
systemphase
phasediagram; (a)kk0 0>> 1;
diagram;(a) (b) kk00 < 1.
1; (b)

In
In the
the case of kk00 << 11 (see
case of (see Figure
Figure 1b),
1b), the
the solid
solid solution
solution containing
containing impurities
impurities at
at the
the
phase P the molten state S
phase point P is heated to the molten state S and then cooled to the temperature T; the
point is heated to and then cooled to the temperature T; the
first
first solid
solid phase
phase point
point to to condense
condense isis Q,
Q, and
and thethe impurity
impurity content
content isis reduced
reduced compared
compared to to
the phase point P. Then, the solid phase at the phase point Q is heated and
the phase point P. Then, the solid phase at the phase point Q is heated and melted until it melted until
it reachesthe
reaches thephase
phasepointpointR,R,after
afterwhich
whichititisis condensed,
condensed, and
and the
the impurity
impurity content
content in
in the
the
obtained solid at the phase point E is reduced again, and the heating is repeated with
obtained solid at the phase point E is reduced again, and the heating is repeated with
condensation, the impurity content in the solid phase is continuously reduced to achieve
condensation, the impurity content in the solid phase is continuously reduced to achieve
Materials 2021, 14, x FOR PEER REVIEW 4 of 21
the purpose of purifying metals, but the opposite is true for k0 > 1.
the purpose of purifying metals, but the opposite is true for k0 > 1.
2.2. Analysis of Changes in Impurities during Zone Refining
2.2. Analysis of Changes in Impurities during Zone Refining
The
The two sides
specific of Formula
process of zone (3)refining
respectivelyis shown derive x:
in Figure 2a. The metal materials to be
The specific process of
purified are placed in the tubular furnace; zone refining is shown in Figure
then, install a movable 2a. The metalring
heating materials
outsidetothe be
, C0 k0 (1 k0 ) ko x
purified are placed in the tubular
tube (high frequency heating ring canS be used). furnace;
C = then, install
e l >a 0. movable heating ring outside the
(4)
l
tube (high frequency heating ring can be used).
Let the initial concentration of an impurity in the metal ingot be C0, which is locally
melted into a melting zone; then, slowly move the heating ring to the right. The melting
zone also slowly moves from the left end to the right end. The melted metal at the left end
gradually solidifies again.
At this time, impurities with k0 < 1 are precipitated at the interface between the “reso-
lidification zone” and the melting zone, and the concentration of impurities distributed in
the liquid phase is greater than that in the solid phase.
Therefore, as the melting zone moves to the right, the impurities also move to the
right, and finally, they are concentrated at the tail end, and the impurities of k0 > 1 mainly
accumulate at the head end of the ingot. So, we repeat the above process many times; just
Figure
Figure 2.
2. Specific
likeofa zone
Specific process
process of
broom
zone refining;
sweeping
refining; (a)
the impurities
(a) Single-pass
Single-pass zone
to the two
zone refining;
refining; (b)
ends, themelting
(b) Multi-pass
Multi-pass melting
impuritieszone are
zone
removed from
refining.
refining.
both ends, purifying the middle of the ingot.
When
It
Let can thebeusing
initial multi-melting
obtained that the Czone
concentration moving
Sof an refining, theinadvantages
direction
impurity along
the metal of
theingot zone
melting be refining
Czone is can
0 , which be seen.
anisincreas-
locally
As function,
ing shown
melted intoinaand Figure
it can
melting 2b, bea seen
zone; series
then, ofslowly
from closely
Equationmovespaced(3)
the heaters
that whenare
heating used
x/l→∞,
ring to right.
melt
there
to the S/C0→1,
is Cinto
The multiple
and
melting
melting
the
zone alsozones
purificationslowly ineffect
the ingot;
moves is not
from after
themultiple
ideal at this
left endtime,
tozone
the sorefining, end.the
the value
right of
The impurity
x/lmelted concentration
is not suitable
metal atifthe distri-
theleft
width
end
bution
gradually reaches a
solidifiessteady
again. state or limits distribution.
of the melting zone is fixed, and thus the ingot should not be made too long. The distri-
bution Impurities
At ofthis time,
impuritiesdoimpurities
not
alongdiffuse in
thewith the
ingot k0solid
< 1 are
after phase, but the
the precipitated
molten zonemass transfer
atpasses
the interfacerate between
through in
thethe liquid
ingot theis
phase is invery
“resolidification
shown Figure fast,3.andand
zone” the concentration
the melting zone, of impurities in the liquidofphase
and the concentration is uniform.
impurities After
distributed
a zone
in the
For refining
liquid and
phase
the case purification,
ofiskgreater
0 > 1, the than the distribution
that
impurities in the theofmelting
in solid impurities
phase. zone along the lengthinofthe
are enriched the solid
ingot
can be
phase. Therefore,
expressed
Since theas the
by melting
Formula
impurities zone
(3)
cannot moves
[19,21]:
move to to thethe right,
right withthethe impurities
melting zonealso move in thetosolid
the
right, and finally, they are concentrated
phase, nor can they migrate to the left-side solidification∙k zone.at the tail end, and the impurities
It is equivalentof k > 1 mainly
0 to distrib-
(1 − /l
accumulate
uting at the head
the impurities in end
the of theCSingot.
length = Cthe
of 1rightmost
0 So, −we k0 )el (melting
repeat .
the0 above process
zone width)many of times; (3)
just
the ingot
like
evenlya broom sweeping the impurities to the two ends, the impurities are removed from
CSwithin
is the the range of
impurity the length ofatthe
concentration theleft x-l, as shown
distance x frominthe Figure
head3b. end, and l is the
bothItends,
can be purifying
seen from the themiddle of the ingot.
combination of Figure 3 that for impurities containing k0 < 1
length of the melting zone.
and k0 > 1 in the metal, in order to obtain a better purification effect, the middle part of the
bar ingot should be a high-purity metal zone. The ratio of the length of the ingot to the
width of the fixed melting zone (L/l) should not be too large, and the number of repetitions
of melting has the best value, which is about L/l, and the cut length at both ends should
 , C0 k0 1 k0
CS = e l > 0. (4)
l

Materials 2021, 14, 2064 4 of 21

When using multi-melting zone refining, the advantages of zone refining can be
seen. As shown in Figure 2b, a series of closely spaced heaters are used to melt into
multiple melting zones in the ingot; after multiple zone refining, the impurity concentration
distribution reaches a steady state or limits distribution.
Impurities do not diffuse in the solid phase, but the mass transfer rate in the liquid
Figure 2. Specific process of zone refining; (a) Single-pass zone refining; (b) Multi-pass melting zone refining.
phase is very fast, and the concentration of impurities in the liquid phase is uniform. After
a zoneItrefining and purification,
can be obtained that the the distribution
CS moving of impurities
direction along the along the length
melting zone isofanthe ingot
increas-
can be expressed by Formula (3) [19,21]:
ing function, and it can be seen from Equation (3) that when x/l→∞, there is CS/C0→1, and
the purification effect is not ideal at this
h time, so the valueiof x/l is not suitable if the width
of the melting zone is fixed, andC = C
S thus0 1 − (ingot
the )e− x·k0 /l
1 − k0should not. be made too long. The distri- (3)
bution of impurities along the ingot after the molten zone passes through the ingot is
CS in
shown is the impurity
Figure 3. concentration at the distance x from the head end, and l is the length
of theFor
melting zone.
the case of k0 > 1, the impurities in the melting zone are enriched in the solid
The two sides of Formula (3) respectively derive x:
phase. Since the impurities cannot move to the right with the melting zone in the solid
phase, nor can they migrate to the left-side C0 k0 (1 −solidification
k 0 ) −k o x zone. It is equivalent to distrib-
,
uting the impurities in the length CS = e l > 0.
of thel rightmost l (melting zone width) of the ingot (4)
evenly within the range of the length of the left x-l, as shown in Figure 3b.
ItItcan
canbebeobtained
seen from thatthe CS moving direction
thecombination of Figurealong
3 thatthe
formelting zonecontaining
impurities is an increasing
k0 < 1
function,
and k0 > 1and it can
in the be seen
metal, fromto
in order Equation
obtain a(3) when x/l→∞,
that purification
better therethe
effect, S /C0 →1,
is Cmiddle and
part ofthe
the
purification effect be
bar ingot should is not ideal at thismetal
a high-purity time, zone.
so theThevalue of x/l
ratio of is
thenot suitable
length if the
of the width
ingot of
to the
the melting zone is fixed, and thus the ingot should not be made too long.
width of the fixed melting zone (L/l) should not be too large, and the number of repetitionsThe distribution
of
ofimpurities
melting has along the ingot
the best value,after
whichtheismolten
about zone
L/l, andpasses through
the cut lengththe
atingot
both is shown
ends shouldin
Figure 3.
be greater than the length of the melting zone l.

Figure3.3.Distribution
Figure Distributionofofimpurities
impuritiesalong
alongthe
theingot
ingotafter the
after themolten
molten zone passes
zone through
passes throughthethe
ingot; (a) (a)
ingot; Distribution curve
Distribution of
curve
of multiple
multiple zone zone molten
molten impurities
impurities along
along thethe ingot
ingot when
when k0 k<0 1;
< 1;
(b)(b) Distributioncurve
Distribution curveofofmultiple
multiplezone
zonemolten
moltenimpurities
impurities
whenkk0 0>> 1.
when 1.

3. Influencing of k0 > and

For the caseFactors 1, theOptimization
impurities in the melting
of Zone zone are enriched in the solid phase.
Refining
SinceSince
the impurities
Pfann [22]cannot
publishedmove topioneering
the the right with
workthe
ofmelting zone indomestic
zone refining, the solidand
phase,
for-
nor can they migrate to the left-side solidification zone. It is equivalent to distributing
eign scholars have conducted a series of research and discussion on the influencing factors the
impurities in the length
of zone refining from aofpractical
the rightmost l (melting zone
and theoretical width) ofThese
perspective. the ingot evenly
factors are within
mainly
the range of the length of the left x-l, as shown in Figure 3b.
It can be seen from the combination of Figure 3 that for impurities containing k0 < 1
and k0 > 1 in the metal, in order to obtain a better purification effect, the middle part of the
bar ingot should be a high-purity metal zone. The ratio of the length of the ingot to the
width of the fixed melting zone (L/l) should not be too large, and the number of repetitions
of melting has the best value, which is about L/l, and the cut length at both ends should be
greater than the length of the melting zone l.
 equilibrium distribution coefficient, zone refining rate, melting zone width, diffusio
layer thickness, and zone refining times.
Materials 2021, 14, 2064 Yang [23] believes that the ingot length/area length ratio (L/l), the type5 and
of 21 purity
the container, the composition of the ambient gas, and the degree of vacuum all need
be carefully studied; otherwise, it will become a limitation to achieve the required perfo
3.mance.
Influencing Factors and Optimization of Zone Refining
Since Pfann [22] published the pioneering work of zone refining, domestic and foreign
3.1. Balanced
scholars Distribution
have conducted Coefficient
a series of research and discussion on the influencing factors
of zoneThe
refining from a practical
equilibrium and theoretical
distribution coefficientperspective. Theseisfactors
k0 of impurities arefactor
the key mainlythat dete
equilibrium
mines their migration, and it runs through the segregation purification. By layer
distribution coefficient, zone refining rate, melting zone width, diffusion observing th
thickness,
degree to and zone the
which refining times. distribution coefficient k0 deviates from 1, it is general
equilibrium
Yang [23] believes that the ingot length/area length ratio (L/l), the type and purity of
possible to predict the purification ability of the zone refining process to the correspon
the container, the composition of the ambient gas, and the degree of vacuum all need to be
ing impurities. However, in the actual operation process, usually, the impurities have n
carefully studied; otherwise, it will become a limitation to achieve the required performance.
had time to diffuse, and the temperature has changed greatly, causing the solid/liqu
interface
3.1. BalancedtoDistribution
advance greatly, forming a solid phase of new components; that is, non-equ
Coefficient
librium solidification
The equilibrium actually occurs,
distribution and
coefficient k0 impurities
of impurities areisconcentrated
the key factor at the
that solidificatio
deter-
interface [24].
mines their migration, and it runs through the segregation purification. By observing the
degreeTherefore,
to which thethe equilibrium distribution coefficient
equilibrium distribution coefficient k 0 deviates from 1, it is generally
k0 is not fixed, so the effective di
possible to predict the purification ability of the zone refining
tribution coefficient (keff) is introduced to improve the zone refining process to the corresponding
efficiency. The sim
impurities. However, in the actual operation process, usually, the impurities have not had
plest method is to optimize keff. The BPS equation is as follows [25,26]:
time to diffuse, and the temperature has changed greatly, causing the solid/liquid interface
to advance greatly, forming a solid phase keff = kof0 /new
k0 + components;
(1 − k0 )e fΔ/Dthat
. is, non-equilibrium (
solidification actually occurs, and impurities are concentrated at the solidification inter-
face [24].
where f is the zone melting rate, D is the impurity diffusion rate, and δ is the thickness
Therefore, the equilibrium distribution coefficient k is not fixed, so the effective
the boundary layer. As shown in Figure 4, when f ≫ 0D/δ, the rate of change of keff increas
distribution coefficient (keff ) is introduced to improve the zone refining efficiency. The
rapidlymethod
simplest until itisapproaches
to optimize k1; when f ≫ D/δ, keff tends to k0, and the melting efficiency
eff . The BPS equation is as follows [25,26]:
this time depends on the balance of specific impurities partition coefficient. In terms
h i
purification effect, keff is expected
k e f f = k0 / to
k0be
+ (close )e−k0f .∆/D .
1 − k0to (5)
It can be clearly seen from Equation (5) that if the δ value is large, the keff is also larg
which
where isthe
f is notzone
conducive
meltingto zone
rate, refining.
D is the impurity diffusion rate, and δ is the thickness
of the boundary
Therefore,layer.
the δAs shown
value in Figure
must 4, whenδf isdirectly
be reduced. D/δ, theaffected
rate of change keff
by theofmixing of th
increases rapidly until it approaches 1; when
melt. During the zone melting process, measures such f  D/δ, k tends to k
eff as rotation,, and the melting
0 external magnetic fiel
efficiency at this time
and oscillation depends
stirring areon the balance
often used. Ifof the
specific
meltimpurities
only haspartition coefficient.
convection, In
δ is about 1 mm
terms of purification effect, keff is expected to be close to k0 .
which can be reduced to 0.1 mm by stirring. Stirring can be reduced to 0.01 mm [27].

Figure
Figure 4. Relationship
4. Relationship between
between keff fδ/D.
keff and and fδ/D.
Materials 2021, 14, 2064 6 of 21

It can be clearly seen from Equation (5) that if the δ value is large, the keff is also large,
which is not conducive to zone refining.
Therefore, the δ value must be reduced. δ is directly affected by the mixing of the melt.
During the zone melting process, measures such as rotation, external magnetic field, and
oscillation stirring are often used. If the melt only has convection, δ is about 1 mm, which
can be reduced to 0.1 mm by stirring. Stirring can be reduced to 0.01 mm [27].

3.2. Zone Refining Rate

The selection of the moving speed of the melting zone (that is, the zone refining
rate) directly affects the purification efficiency, which is related to the production cost.
Therefore, in the purification process, it is necessary to formulate an appropriate rate
according to the characteristics of the material itself. When the melting zone movement
rate f is very large, the solid phase crystallization rate increases accordingly, making it
difficult to remove impurities in the unit cell, and the purification effect is not obvious;
when f is small, the component diffusion can be fully carried out, and the impurities
inside the crystal can be effectively removed, making the purification efficiency higher [28].
When multiple impurities are involved, the optimal zone melting speed is determined
by the impurity with keff closest to the host metal [29]. Prasad [28] optimized its process
parameters when purifying tellurium and found that when the zone melting speed was
30 mm/h, it produced better results and obtained 7N pure tellurium material. Zhang [30]
found that the melting rate has a significant effect on the zone melting process when
purifying metallic tin. When the zone refining rate was reduced from 1.4 to 0.6 mm/min,
the metal purity in the ingot increased from 99.99824% to 99.99906% after 10 times of
refining. Wan [31] proposed an improved vacuum zone refining process. By modifying the
effective distribution coefficient, the axial segregation of impurities was studied. When the
zone refining rate was 1 mm/min, the aluminum content was greater than that of the 5N
(99.9992%) sample.
When preparing single crystal materials, the use of a lower zone refining rate can
effectively remove impurity elements inside the crystal and reduce crystal defects, but
too low a zone refining rate will cause severe segregation of elements within the crystal
and increase the concentration gradient of the component, which is not conducive to
the preparation of single crystals; the use of a higher zone refining rate can improve the
segregation of impurities and improve production efficiency, but it is not conducive to the
elimination of gaseous impurities, which affects the stability of the solid/liquid interface,
and it also affects the preparation of single crystals. Therefore, the choice of the optimal
refining rate is to strike a balance between the formation of single crystals and the reduction
of the degree of segregation [32].

3.3. Length of Melting Zone

The length of the melting zone is affected by many factors, such as thermal field, zone
refining speed, crucible thermal conductivity, etc. In zone refining, the purity of the product
obtained in the narrow melting zone is higher than that in the wider melting zone, but
the impurity concentration of the narrow melting zone increases faster, making it difficult
to remove impurities in the melting zone, which must be compensated by extending the
purification time [33]. Therefore, in order to improve the purification efficiency, the actual
production process generally adopts the operation method of using a wide melting zone
and a higher zone refining rate at the beginning of purification and then changing to a
narrow melting zone and a lower zone refining rate. Jun [34] purified industrial cerium by
induction heating and found that the larger zone refining length in the early zone refining
and the shorter zone refining length in the later stage can improve the segregation efficiency
of solute during the zone refining process, especially for k0 far from the whole solute.
During the first zone refining, the impurity concentration of the melting zone will
increase for k0 < 1; for k0 > 1, the impurity concentration will decrease. In order to meet
the two situations at the same time, to obtain the best separation effect after the first zone
Materials 2021, 14, 2064 7 of 21

refining, the melting zone must be as large as possible; that is, the length of the melting
zone is the full ingot length. However, when the zone is melted a few times, see Figure 5.
Materials 2021, 14, x FOR PEER REVIEW 7 of 21
As the length of the melt zone increases, the “limit distribution” curve moves upwards,
and the final purity that can be achieved decreases.

Figure 5. Relationship between impurity concentration and number of zone refining passes [35].

Ho [36]
Ho [36] obtains
obtains the
the optimal
optimal melting
melting zone
zone length
length value
value for
for ten
ten passes
passes of of zone
zone refining.
refining.
It can be seen from Figure 6a that the length of the melting zone increases
It can be seen from Figure 6a that the length of the melting zone increases with the increase with the increase
of the
of the distribution
distributioncoefficient
coefficientandanddecreases
decreaseswithwiththetheincrease
increase ofof
the number
the number of of
zone refining.
zone refin-
Following this step for zone refining, a considerable separation effect can
ing. Following this step for zone refining, a considerable separation effect can be obtained. be obtained.
As shown
As shown in in Figure
Figure 6b, when kk00 <<1,
6b, when 1,the
themaximum
maximum solutesolute removal
removal raterate decreases
decreases with
with
the increase of the partition coefficient, and the opposite is true when k
the increase of the partition coefficient, and the opposite is true when k0 > 1. However, the
0 > 1. However, the
maximum solute
maximum soluteremoval
removalrate rateincreases
increases with thethe
with increase of zone
increase refining
of zone times.
refining Ghosh
times. [37]
Ghosh
calculated
[37] the zone
calculated melt melt
the zone length whenwhen
length meeting the maximum
meeting solute solute
the maximum removal rate according
removal rate ac-
to Ho’s numerical model. It can be obtained that as the number of smelting
cording to Ho’s numerical model. It can be obtained that as the number of smelting passes passes increases,
the shorter zone melt length can be used to obtain the maximum solute removal rate [38].
increases, the shorter zone melt length can be used to obtain the maximum solute removal
Table 2 lists the zone refining length when the maximum solute removal rate is obtained
rate [38]. Table 2 lists the zone refining length when the maximum solute removal rate is
per pass. Table 3 lists the purity grade of metal after zone refining.
obtained per pass. Table 3 lists the purity grade of metal after zone refining.
Table 2. Zone refining length when the maximum solute removal rate is obtained per pass.
Table 2. Zone refining length when the maximum solute removal rate is obtained per pass.
Zone Refining Times ≥20
Zone Refining Times 1 1 2 2 3 3
4 4
5–8 5–8
9–19 9–19
≥20
Normalized melting zone
Normalized melting zone length
length 1 1 0.350.35 0.25 0.25 0.2 0.2 0.15 0.150.1 0.1 0.05
0.05

Table 3. The purity grade of metal after zone refining.

Elements Al As Cd Cu Fe Mg Ni Pb Ga Si Sn Te Zn Ag
Purity grade 9N 9N 7N 10N 9N 8N 7N 7N 7N 9N 7N 7N 7N 8N
Elements Ge In Ce
Purity grade 10N 7N 7N
Materials 2021, 14, 2064 8 of 21

Table 3. The purity grade of metal after zone refining.

Elements Al As Cd Cu Fe Mg Ni Pb Ga Si Sn Te Zn Ag
Purity grade 9N 9N 7N 10N 9N 8N 7N 7N 7N 9N 7N 7N 7N 8N
Elements Ge In Ce
Materials
Purity2021, 14, x FOR PEER
grade 10N REVIEW
7N 7N 8 of 21

Figure 6. (a) Optimal zone

zone refining
refininglength
lengthof
of1–10
1–10passes;
passes;(b)
(b)Maximum
Maximumsolute
soluteremoval
removal rate
rate during
during multi-pass
multi-pass zone
zone refin-
refining.
ing.
3.4. Zone Refining Scans
3.4. Zone Refining Scans
The zone refining process needs to be repeated many times. As the number of times
The zone
increases, refining process
the purification effectneeds to be repeated
also increases, manya times.
but after certainAs thezone
pass number of times
melting, the
increases, the purification effect also increases, but after a certain pass zone
impurity concentration distribution is close to the “limit distribution”, as shown in Figure melting, the5.
impurity concentration
At this time, in the distribution
process ofis driving
close to the
the “limit distribution”,
impurities to both as shown
sides, in Figureon
impurities 5.
bothAt thiswill
sides time, in diffuse
also the process ofmiddle.
to the driving For
the impurities to both
the convenience ofsides, impurities
operation, on both
the empirical
sides
formula for the number of zone refining times n can be expressed as n = kL/l (where for-
will also diffuse to the middle. For the convenience of operation, the empirical k is
mula for the number
the empirical constant,ofthe
zone refining
value rangetimes
is 1 ton1.5;
canLbe expressed
is the length ofas the
n = ingot;
kL/l (where
l is theklength
is the
empirical constant,
of the melting zone);theinvalue
actualrange is 1 tothe
operation, 1.5;frequently
L is the length
usedof the ingot;islL/l
condition is the length
= 10, of
taking
the
k = melting
1.5; then,zone);
n = 15in[39].
actual operation, the frequently used condition is L/l = 10, taking k =
1.5; then, n = 15 [39].
3.5. Application of Current in Impurity Transmission
3.5. Application
Applying of Current field
a current in Impurity
duringTransmission
the refining process (see Figure 7) can improve the
segregation of impurities
Applying at theduring
a current field solidification interface
the refining through
process (see electromigration.
Figure 7) can improveDost [40]
the
conducted aof
segregation numerical simulation
impurities on the process
at the solidification of transporting
interface impurities in the cadmium
through electromigration. Dost [40]
zone refining
conducted system. The
a numerical results show
simulation thatprocess
on the the conductivity of cadmium
of transporting is higher,
impurities in thewhich
cad-
can enhance the migration rate of impurities in cadmium. The gas in Figure
mium zone refining system. The results show that the conductivity of cadmium is higher, 7 is argon,
which is used
can to prevent
enhance active metals
the migration rate from reacting with
of impurities other gases.
in cadmium. The gas in Figure 7 is
argon, which is used to prevent active metals from reacting with other gases.
 Applying a current field during the refining process (see Figure 7) can improve the
segregation of impurities at the solidification interface through electromigration. Dost [40]
conducted a numerical simulation on the process of transporting impurities in the cad-
mium zone refining system. The results show that the conductivity of cadmium is higher,
Materials 2021, 14, 2064 9 of 21
which can enhance the migration rate of impurities in cadmium. The gas in Figure 7 is
argon, which is used to prevent active metals from reacting with other gases.

Materials 2021, 14, x FOR PEER REVIEW 9 of 21

Figure
Figure 7. 7. Schematic
Schematic diagram
diagram ofof the
the zone
zone refining
refining system
system under
under applied
applied current.
current.
3.6. Inclination
3.6. Inclination
During the refining process in the horizontal area, the longitudinal section of the in-
gotDuring the refining
will become tapered process
due tointhe
themass
horizontal
transferarea,
andthe longitudinal
even cause the section
melt to of the ingotTo
overflow.
will become tapered due to the mass transfer and even cause the
avoid this phenomenon, the crucible should be inclined at an angle θ relative tomelt to overflow. Tothe
avoid
hori-
this phenomenon, the crucible should
zontal direction, as shown in Equation (6) [41]: be inclined at an angle θ relative to the horizontal
direction, as shown in Equation (6) [41]:
θ = tan 1 × 2h0 (1 − γ/l) (6)
θ = tan−1 × 2h0 (1 − γ/l ) (6)
where h0 is the original height of the ingot and γ is the ratio of solid and liquid metal
where h0 is the original height of the ingot and γ is the ratio of solid and liquid metal density.
density.
Magnetic
Magnetic field, gas
field, gasflow
flowrate, and
rate, andmelt
meltstability will
stability also
will alsohave
have ananimportant
important influence
influence
ononthe purification effect. In addition, the phenomenon of backflow
the purification effect. In addition, the phenomenon of backflow in the melting in the melting zone
zone
will also affect the effect of zone refining. This is because the cooling rate
will also affect the effect of zone refining. This is because the cooling rate of the upper of the upper and
and
lower
lower ingots
ingots differs
differsgreatly,
greatly, and
andthethe
ingot
ingotshrinks
shrinksunevenly
unevenly andandbends
bendsupwards
upwards [42].
[42].
The
Theexperimental
experimental optimization
optimization of the above-mentioned
of the above-mentioned zonezone
refining parameters
refining is not is
parameters
universal because they depend on the properties of the specific system, such
not universal because they depend on the properties of the specific system, such as equip- as equipment
specifications, ingot diameter,
ment specifications, melt viscosity,
ingot diameter, and impurity
melt viscosity, distribution
and impurity coefficient.
distribution coefficient.
4. Types of Zone Refining
4. Types of Zone Refining
Based on the same principle, there are various zone refining technologies. There are
Based on the same principle, there are various zone refining technologies. There are
two main types, refining in the suspension area and refining in the horizontal area, as
two main types, refining in the suspension area and refining in the horizontal area, as
shown in Figure 8.
shown in Figure 8.

Figure (a)(a)
8. 8.
Figure Horizontal zone
Horizontal refining
zone technology;
refining (b)(b)
technology; Suspended zone
Suspended refining
zone technology.
refining technology.

The
Theemergence
emergenceofofrefining
refiningininthe
the suspended
suspended area
area has greatly promoted
has greatly promoted the
thestudy
studyof
ofrefractory
refractorymetals.
metals.This
Thismethod
methodcan cannot
not only effectively remove volatile metals
only effectively remove volatile metals and andgas
gas impurities in refractory metals, as well as avoid secondary pollution of metals in the
impurities in refractory metals, as well as avoid secondary pollution of metals in the re-
refining process, but also effectively control the metal melt flow [43]. Table 4 lists the
fining process, but also effectively control the metal melt flow [43]. Table 4 lists the differ-
differences between the two methods.
ences between the two methods.

Table 4. Differences in zone refining technology [43].

Category Heating Method Advantage Disadvantage Scale

The melting zone is supported by
Electron beam heating, The ingot does not touch the
surface tension, so controlling the
Floating zone induction heating, container, so the product purity is Small
shape and stability of the melting
refining plasma heating, light high, and the equipment occupies batch
zone is the key, and the output is
heating a small space.
Materials 2021, 14, 2064 10 of 21

Table 4. Differences in zone refining technology [43].

Category Heating Method Advantage Disadvantage Scale

The melting zone is
supported by surface
Electron beam heating, The ingot does not touch the container,
Floating zone tension, so controlling the Small
induction heating, plasma so the product purity is high, and the
refining shape and stability of the batch
heating, light heating equipment occupies a small space.
melting zone is the key,
and the output is low.
Simple equipment, continuous
purification of multiple melting zones,
easy loading and unloading of
Horizontal zone Induction heating,
materials, easy identification of Large footprint Batch
refining resistance heating
interfaces, and the total length of
ingots can be increased or decreased
as needed

5. High-Purity Metal Analysis Method

The purity testing of high-purity metals should be based on the actual application
needs as the main standard, which is generally measured by the reduction method. From
the development and application of metal purity analysis at home and abroad in recent
years, it can be seen that in the determination of trace and ultra-trace metal elements,
a basic mode can be attributed to the effective sample decomposition method, efficient
separation and enrichment method, and simple, fast, and accurate instrument analysis
method [1].
The analysis methods to determine the purity of high-purity metal materials are
divided into chemical methods and physical methods, and they mainly consist of chemical
analysis methods. Table 5 compares various chemical analysis methods.
It can be seen that GDMS can be directly analyzed by solid sample injection, with a
wide measurement range, a variety of single detection elements, and high detection accu-
racy. It has obvious advantages in high-purity detection, but the equipment is expensive.
ICP-MS is more popular and has the advantages of speed, economy, and convenience.
NAA can be used for precious metal analysis and testing.
High-purity metal purity analysis is an important indicator to measure the production
level of industrial science and technology. It plays an inestimable role in the develop-
ment of high-purity metal production processes and the development of the materials
science industry.
At the same time, it also follows the progress of science and economic development
with increasing improvement, and the analysis of metal purity in the future will develop
toward the following aspects [1]:
(1) In terms of the number of measured elements, it will develop from single elements to
multiple elements simultaneously or continuously.
(2) In the analysis method, it will develop from offline/manual operation to online/
automatic mode.
(3) In terms of data collection and processing, the application of mathematical methods
such as chemo-metrics, pattern recognition, expert systems, artificial intelligence, and
neural networks will help to improve the integrity and accuracy of test data.
Materials 2021, 14, 2064 11 of 21

Table 5. Comparison of various analysis methods for measuring metallic elements.

Analytical
Sampling Method Measuring Range (g/g) Explanation
Method
The matrix effect is small, the sample processing is simple, it
can avoid pretreatment pollution and loss, it is fast and efficient,
the detection limit is low, and standard samples can be used,
GDMS [44] Solid Constant~10−12 but the price is expensive. The discharge is stable, and the
analysis sample can be peeled layer by layer for surface and
depth analysis. The self-absorption effect is small. Multiple
elements can be determined simultaneously.
High resolution and sensitivity, a wide measurement range, a
ICP-MS Liquid 10−6 ~10−12 need to try to eliminate the matrix effect and isotope
interference, and the sample processing process is longer.
The number of single analysis is small, the analysis range is
GF-AAS Solid 10−6 ~10−9
narrow, and it is not suitable for high melting point metals.
Fast speed, high sensitivity, small matrix effect, low detection
ICP-AES limit, and serious spectrum interference and matrix interference.
Liquid Constant~10−6
[45] Multi-element simultaneous analysis method with the widest
range of analysis elements and the largest content span.
Fast, simple, and economical non-destructive testing, which can
TXRF Solid Constant~10−12 perform multi-element analysis at the same time, but it can only
perform surface analysis, and the detection limit is low.
High sensitivity, high accuracy and precision, and sensitivity
varies from element to element. It can only measure the content
NAA Solid, liquid 10−6 ~10−13 of elements. It requires a small nuclear reactor and has the risk
of nuclear radiation. Less sampling, non-destructive analysis.
The detection equipment is expensive.
It has a lower detection limit than ICP-MS, can be used for
NTIMS Liquid 10−10 ~10−14
isotope age studies, and can accurately measure Re and Os.

6. Numerical Simulation
Zone refining experimental research takes a lot of time. Therefore, many numerical
models have been developed to effectively predict solute distribution or optimize exper-
imental parameters as well as provide guidance for empirical experiments. Reiss [46]
assumes that the solute partition coefficient k0 and the melting zone length L remain un-
changed, establishes a differential equation between the concentration of each substance,
and obtains a mathematical model of zone refining. Lee [47] proposed a model capable of
simulating zone refinement and drawing conclusions. This model can pass the repeated
pass method even when the zone travel speed and zone length are relatively large and
thus cannot be properly purified. Cheung [48] combined a numerical model and genetic
algorithm to establish an optimized melting zone length model to achieve maximum pu-
rification efficiency. Jun [34] uses induction heating to refine the industrial cerium in a cold
crucible and introduces the numerical model of solute redistribution in the process of zone
refinement into the genetic algorithm to search for the optimal zone length of different zone
channels to improve the solute redistribution efficiency. The experiment and calculation
results show that the genetic algorithm is a useful tool to obtain the optimal zone length
during the zone refining process, but the stability of the molten zone should be maintained
in the cold crucible.
Tan Yang [49] reported that the zone refining by combining finite element simulation
data and experimental results analysis concluded that the zone refining experiment on
the distribution of temperature field around the heat source with ellipsoid, semicircular
canals is more suitable than square grooves for the purification of zone refining, as shown
in Figure 9. The scope of the safety threshold value of indium under a given preheated
temperature will greatly boost zone refining efficiency and does not destroy the balance
 Materials 2021, 14, 2064 12 of 21

of zone refining. In addition, Chen Luona [50] applied the oxidation refining-zone refin-
Materials 2021, 14, x FOR PEER REVIEWing joint method to refining indium heat coupling finite element analyses, as shown in12 of 21
Figure 10. Here, argon gas flow in a horizontal transverse turn-back, along the long axis of
4, x FOR PEER REVIEW 12 of 21“up” in the
the zone refining equipment in the opposite direction from exports, presents
zone refining equipment, forming annular flow. The model more accurately reflects the
oxidation refining and zone refining temperature field as well as flow field distribution.

Figure 9. Temperature cross-section: (a) SiC thermal conductor, (b) graphite electrothermal assem-
bly.
9. Temperature cross-section: (a)
(a) SiC
SiC thermal
Figure
Figure 9. Temperature cross-section: thermalconductor,
conductor,(b)(b)
graphite electrothermal
graphite assembly.
electrothermal assem-
bly.

Figure
Figure 10.10. Flowdiagram
Flow diagram of
of argon
argon gas
gasflow
flowininthe zone
the refining
zone equipment.
refining equipment.

Figure 10. Flow diagram of argon gas flow in the zone refining equipment.
6.1. Iterative
6.1. IterativeModeling
Modeling of
of Constant
Constant KK
6.1. Iterative ModelingFor different
ofFor
Constant distributioncoefficients
differentKdistribution coefficients
andand melting
melting zone zone lengths,
lengths, Bochegov
Bochegov [51] con-
[51] consid-
sideredthethe limit distributionimpurities
of impurities after purification
zoneby zone technology
refining technology
For different ered limit
distribution distribution
coefficientsofand melting after purification
zone lengths, by Bochegovrefining
[51] con- and
and proposed analytical solutions. It divides the ingot into three regions in the length di-
sidered the limit distribution of impurities after purification by zone refining technology
rection, and the impurity distribution function is shown in Equation (7).
and proposed analytical solutions. It divides the ingot into three regions in the length di-
rection, and the impurity distribution function is, shown0in≤ xEquation
AeBx ≤ L − 2l (7).
⎧
Materials 2021, 14, 2064 13 of 21

proposed analytical solutions. It divides the ingot into three regions in the length direction,
and the impurity distribution function is shown in Equation (7).

Ae Bx , 0 ≤ x ≤ L − 2l


k( L−l −k)

AeB( L−2l ) Ae B( L−2l )
l k −1 e l

CS( x) = S( x ) + , L − 2l < x ≤ L – l (7)
S( L−2l )+l k−1 ek S( L−2l )+l k−1 ek
Ae B( L−2l )
( L − x ) k −1 , L − l

< x ≤ L


S( L−2l )+l k−1 ek
Materials 2021, 14, x FOR PEER REVIEW 13 of 21
A and B are constants, and B is derived from known conditions:

k  Bl 
B=e −1 . (8)
l k
B= ( − 1). (8)
l
A is determined according to non-normalized conditions:
A is determined according
Z L−2l to non-normalized
Z L−l conditions:
Z L
Cs0 L = C s1
L 2l
( x ) dx + l
Cs2 ( x ) dx + L
Cs3 ( x )dx. (9)
Cs0 L0 = ∫0 Cs1 (x)dx +L∫−L 2l2l Cs2 (x)dx+ ∫LL− Cl (x)dx.
l s3
(9)
x is the coordinate of the crystallization front relative to the starting point of the ingot, and
x is the coordinate of the crystallization front relative to the starting point of the ingot, and
cs(x) is the impurity concentration function on the coordinate x on the solid phase, where Cs0
cs(x) is the impurity concentration function on the coordinate x on the solid phase, where
is the initial impurity concentration. S is determined according to the following expression:
Cs0 is the initial impurity concentration. S is determined according to the following expres-
sion:
k ∞ k n ( l0 − x − l ) k + n
 
S( x ) = − ∑ n 1)( k(l+ x2)l)·k+n
(10)
l n =0 k l ∞ k ( k + 0 · · (k + n)
S(x)= − ∑n = 0 (10)
l )(k ) (k
k(k + 1 +2 ⋯ + n)
where n is the number of zone refining.
where n is the number of zone refining.
The vertical axis corresponds to the ratio C/C of the impurity concentration to its
The vertical axis corresponds to the ratio C/Cs0
s0 of the impurity concentration to its
initial concentration, and the horizontal axis corresponds to the coordinate relative to the
initial concentration, and the horizontal axis corresponds to the coordinate relative to the
origin of the ingot. Thinner lines correspond to the distribution with a finite amount of
origin of the ingot. Thinner lines correspond to the distribution with a finite amount of
passages with their numbers increasing as the lines approach the limiting distribution, as
passages with their numbers increasing as the lines approach the limiting distribution, as
shown in Figure 11.
shown in Figure 11.

Limiting distribution
Figure 11. Limiting distributionfor
forthe
thecase ofof
case molten zone
molten zone l = 0.2
sizesize andand
l = 0.2 impurity distribution
impurity (a) k =(a)
factorfactor
distribution 2 and
k = 2(b) k =(b)
and 0.5.k
= 0.5.
6.2. Iterative Modeling of Variable K
6.2. Iterative
Spim [52]Modeling of Variable
proposed K
a new iterative model that can be used to study the effect of zone
melt Spim
length[52]
andproposed a new
distribution iterativeon
coefficient model that efficiency,
refining can be used to studythe
including thefinal
effect of zone
impurity
melt length and
concentration distribution
distribution coefficient
and on refining
the minimum efficiency,
zone pass including
required to reachthe
thefinal
finalimpurity
state. In
concentration distribution and the minimum zone pass required to reach the final state.
In this model, by considering four different regions along the sample, the solute concen-
tration distribution in each subzone of the melting zone is quantified. The formula is as
follows:
dx 1 n 1
⎧ ki ∑M
q = 0 CS q dx ,X=0
Z
Materials 2021, 14, 2064 14 of 21

this model, by considering four different regions along the sample, the solute concentration
distribution in each subzone of the melting zone is quantified. The formula is as follows:
 !
  M −1
dx n −1
∑ CS(q dx) , X = 0



 ki Z
q =0



 h i
 Cn
 k i dx n −1 n

S( X −dx )
+ · · · + Z C S( X + Z −dx! )
− C S( X −dx )
, 0 < X ≤ 1−Z
CSn(X ) = . (11)
Materials 2021, 14, x FOR PEER REVIEW M −1 14 of 21
C0 dx n
∑ CS(q dx) , 1 − Z < X < 1



 Z − Z
q =0



∑qM=−01 CSn(q dx)



C0 − , X=1

M
Cheung [53] established a database containing the changes in solids concentration
(CS) andCheungliquid concentration
[53] established a(Cdatabase L) to obtain the variable
containing k, so during
the changes the simulation
in solids concentrationpro-
cess,
(C for liquid
S ) and each Cconcentration
L in any melting (CLzone, k can the
) to obtain be easily
variable accessed from the simulation
k, so during database. Compar-
process,
for
ingeach CL inmodel
the two any melting
simulations zone,with k canconstant
be easilyk accessed
and variable fromk the for database.
aluminumComparingrefining, it
the
cantwo model simulations
be concluded that using with constant
variable k inkthe variable k is
andsimulation forcloser
aluminumto the refining, it canim-
experimental be
concluded that using
purity distribution curve than kconsidering
variable in the simulation constant is k.closer
The to the experimental
results show that theimpurityvariable
distribution
solute distributioncurve thanmethodconsidering
is used to constant
simulate k. The results show
the impurity that the variable
distribution solute
in the refining
distribution method is used to simulate the impurity distribution
process in different regions, which is closer to the experimental result k than the usual in the refining process in
different
method. regions, which is closer to the experimental result k than the usual method.
Cheung
Cheung [48] interacted
interacted with the numerical numerical model through the AI (artificial (artificial intelli-
gence,
gence, AI) AI) search
search method
method to to quantitatively
quantitatively determinedetermine the the influence
influence of of the
the variable
variable solute
solute
distribution coefficientk konon
distributioncoefficient thethedistribution
distribution of impurities
of impurities in multi-pass
in multi-pass purification by zone
purification by
refining. The results show that the interaction between
zone refining. The results show that the interaction between the numerical model and the the numerical model and the AI
search
AI search technology
technology can can
provide
provide the best set of
the best setmelting
of melting zonezonelengths to achieve
lengths maximum
to achieve maxi-
purification efficiency. As shown in Figure 12, each node
mum purification efficiency. As shown in Figure 12, each node represents the length represents the length of a fusion
of a
zone
fusionand is and
zone linked to subsequent
is linked to subsequent nodesnodesof possible
of possiblesolutions,
solutions,resulting in aninexplosive
resulting an explo-
combination
sive combination of expansion.
of expansion.
Zhang
Zhang [54] carried out
[54] carried out aa zonezone melting
melting process
process to to purify
purify Sb Sb under
under the the conditions
conditions of of
inert
inert gas protection and heater movement rates of 2, 1, and 0.5 mm/min, respectively. By
gas protection and heater movement rates of 2, 1, and 0.5 mm/min, respectively. By
fitting
fittingthe theimpurity
impurity concentration
concentration curve curveobtained
obtained fromfrom the experiment
the experiment with the withSpim
themodel,
Spim
the
model,k values of Pb, As, and Fe impurities at each moving
eff the keff values of Pb, As, and Fe impurities at each moving rate are retrieved. rate are retrieved. This helps to
This
more accurately predict the time required for zone refining
helps to more accurately predict the time required for zone refining and achieve the effi- and achieve the efficiency of Sb
zone refining.
ciency of Sb zone refining.

Figure 12.Figure 12. Representative

Representative tree of possibilities.
tree of possibilities.

6.3. Iterative Modeling Considering Diffusion Area

Nakamura et al. [55] proposed a numerical simulation model that divides the melting
into two independent regions, namely the diffusion region and the stirring region, as
shown in Figure 13a, and the vertical axis C represents the solid and liquid. There is an
interface between them, and there are diffusion regions and stirring regions in the molten
region. This numerical model introduces a transfer ratio q to illustrate the amount of so-
 δ
( k)CL(i 1) × q + CL(i 1) × m j + C(x)
CL(i) = , 2 ≤ j ≤ m − 10. (13)
m-j

Burton [26] established two important equations related to the distribution of solutes
Materials 2021, 14, 2064 when there is a diffusion layer between the solid zone and the molten zone. 15 of 21

CL CS
δ = exp −fδ/D (14)
CL CS

6.3. Iterative
AssumeModeling
that the Considering
exponential Diffusion Area
factor in Equation (14) is the transmission ratio q.
Nakamura et al. [55] proposed a numerical simulation model that divides the melting
into two independent regions, namely −fδ/D
expthe = q region and the stirring region,
diffusion (15)as
shown in Figure 13a, and the vertical axis C represents the solid and liquid. There is an
The experimental values of f and D have been determined, and if the transfer ratio q
interface between them, and there are diffusion regions and stirring regions in the molten
is clear, the width of the diffusion region δ can be estimated. The advantage of this model
region. This numerical model introduces a transfer ratio q to illustrate the amount of solute
is that when the δ of one impurity is determined by parameter fitting, the same δ value
transferred from the diffusion zone to the stirring zone during the diffusion process.
can be used to derive the distribution curve of other impurities.

Figure 13. (a) Schematic diagram of the melting zone; (b) Schematic diagram of the computation model.
Figure 13. (a) Schematic diagram of the melting zone; (b) Schematic diagram of the computation model.

Figure 13b is a schematic diagram of the model calculation; the length of the melting
zone is equal to am (mm). When the first segment is the diffusion area, the concentration is
as shown in Equation (12):

CLδ (i) = C L(i−1) + (1 − k)CLδ (i−1) × (1 − q). (12)

The new mixing zone consists of three parts: the distribution of the diffusion zone,
the (m−j) segment with the concentration of CL(i−1) in the previous agitating zone, and
one segment of the new solid zone with the concentration of C(x) . Therefore, the average
concentration in the stirring area CL(i−1) becomes

(1 − k)CLδ (i−1) × q + C L(i−1) × (m − j)+C ( x )

C L (i ) = , 2 ≤ j ≤ m − 10. (13)
(m − j)
Materials 2021, 14, 2064 16 of 21

Burton [26] established two important equations related to the distribution of solutes
when there is a diffusion layer between the solid zone and the molten zone.

CL − CS
= exp(− f δ/D ) (14)
C δL − CS

Assume that the exponential factor in Equation (14) is the transmission ratio q.

exp(− f δ/D )= q (15)

The experimental values of f and D have been determined, and if the transfer ratio q
is clear,
Materials 2021, 14, x FOR PEER REVIEW the width of the diffusion region δ can be estimated. The advantage of this16model
of 21
is that when the δ of one impurity is determined by parameter fitting, the same δ value can
be used to derive the distribution curve of other impurities.
6.4. Evaluation Modeling and Experiment
6.4. Evaluation Modeling and Experiment
Since there are various experimental parameter combinations in the process of im-
Since there are various experimental parameter combinations in the process of im-
proving zone refining efficiency, it is very time-consuming and even unfeasible to study
proving zone refining efficiency, it is very time-consuming and even unfeasible to study
only through experiments.
only through experiments.
In addition, refining systems in different regions have the best experimental param-
In addition, refining systems in different regions have the best experimental param-
eters. Therefore, the application of the numerical model as described above is a good as-
eters. Therefore, the application of the numerical model as described above is a good
sistant to guide the experimental performance. In this way, the influence of each parame-
assistant to guide the experimental performance. In this way, the influence of each pa-
ter on the refining result can be predicted, and the theoretically optimal parameter com-
rameter on the refining result can be predicted, and the theoretically optimal parameter
bination can be obtained in a short time and at low cost.
combination can be obtained in a short time and at low cost.
Therefore, experimental research is necessary to prove the accuracy of simulation
Therefore, experimental research is necessary to prove the accuracy of simulation
prediction or to correct the model itself. It is recommended to combine simulation and
prediction or to correct the model itself. It is recommended to combine simulation and
experiment to study the refinement of metal regions.
experiment to study
In order theout
to carry refinement of metal
zone refining andregions.
prepare high-purity metal indium, corre-
In order to carry out zone refining and prepare
sponding process improvement and equipment research high-purity metal indium,
and development are correspond-
required.
ing
Forprocess improvement
this reason, our teamand equipment
Li Qing research
[56], Tan and development
Yang [49], and Li Yichengare[57]
required.
designed Forthe
this
reason, our team Li Qing [56], Tan Yang [49], and Li Yicheng [57] designed the
vacuum zone refining system for a multi-refining zone, whose structure is shown in Fig- vacuum
zone refining system for a multi-refining zone, whose structure is shown in Figure 14.
ure 14.

Figure
Figure 14.14. Schematic
Schematic diagram
diagram of vacuum
of vacuum refining-electromigration
refining-electromigration high-purity
high-purity indium
indium refining
refining plant
plant in multi-refining
in multi-refining zone.
zone.

The vacuum zone refining/electromigration refining device is composed of stepping

motor numerical control equipment, a high-purity argon gas purifier, a high-frequency
power supply, an induction heating coil, an electromagnetic shielding device, a quartz
Materials 2021, 14, 2064 17 of 21

The vacuum zone refining/electromigration refining device is composed of stepping

motor numerical control equipment, a high-purity argon gas purifier, a high-frequency
power supply, an induction heating coil, an electromagnetic shielding device, a quartz
furnace tube, and a vacuum pump. The quartz furnace tube is equipped with a cooling
jacket, and a quartz boat is placed inside to carry indium. In the process of zone refining,
the width of the melting zone is controlled by the output of high-frequency power, while
Materials 2021, 14, x FOR PEER REVIEW 17 of 21
the moving speed of the melting zone is precisely controlled by the numerical control table
of the stepping motor. The general speed is 3–30 mm/hour. An electromagnetic shielding
device is used to prevent electromagnetic interference between the melting zones. The
atmosphereduring
atmosphere duringthe therefining
refiningprocess
processisiscontrolled
controlledjointly
jointlybybythe thediaphragm
diaphragmvacuum vacuum
pumpand
pump andthethehigh-purity
high-purityargon
argongas gaspurifier.
purifier.
First,the
First, thediaphragm
diaphragmvacuumvacuum pump pump pumps the vacuum vacuum up uptotoabout
about0.06 0.06MPaMPafor for20
min. Then, the air valve of the vacuum
20 min. Then, the air valve of the vacuum pump pump path is closed, and the air valve leading toto
closed, and the air valve leading
ultra-high-purity argon gas is opened for argon
ultra-high-purity argon gas is opened for argon gas (the pressure gas (the pressureofofargon
argongas gasisisalways
always
slightlyhigher
slightly higherthan
than 1 atmospheric
1 atmospheric pressure).
pressure). The The above
above operation
operation waswas repeated
repeated 3 times
3 times to
to complete
complete theexchange
the air air exchange process,
process, and theandhigh-purity
the high-purityindium indium was prepared
was prepared in the inareathe
area melting/electromigration
melting/electromigration process process
underunder the protection
the protection of ultra-high-purity
of ultra-high-purity argon argon
gas. Thegas.
The vacuum
vacuum zone refining/electromigration
zone refining/electromigration joint joint
method method for high-purity
for high-purity indium indium extrac-
extraction
can
tionbecan
used bewithout high vacuum
used without under the
high vacuum undercondition of the preparation
the condition of high-purity
of the preparation of high-
6N indium,
purity and it can
6N indium, andsignificantly reduce reduce
it can significantly the technical threshold
the technical and gas
threshold andrawgasmaterial
raw ma-
costs,
terialindicating that thethat
costs, indicating technology has obvious
the technology industrial
has obvious application
industrial prospects.
application prospects.
InInorder
ordertotosolve
solvethetheproblem
problemofofsmallsmallprocessing
processingcapacity
capacityofofmetal
metalindium
indiumininthe the
zone
zonerefining
refiningprocess,
process,LiLiYicheng
Yichengetetal.al.[57]
[57]adopted
adoptedthe themethod
methodofofsimultaneous
simultaneousheating heating
ofofmultiple
multiplefurnace
furnace tubes, which
whichdoubled
doubledthe theprocessing
processing capacity
capacity of of
thethezonezone refining
refining pro-
process. Figure 15 shows the working picture of the vacuum refining/electromigration
cess. Figure 15 shows the working picture of the vacuum refining/electromigration high-
high-purity
purity indium indium refining
refining plant
plant in the
in the single
single refining
refining zonezoneof 4offurnace
4 furnace tubes.
tubes. At At present,
present, the
the
small unit of the laboratory has been able to complete the continuous treatment of 100ofkg
small unit of the laboratory has been able to complete the continuous treatment
100 kg indium
indium per year.
per year. The whole
The whole processprocess
has ahas a high
high energy
energy utilization
utilization rate,rate,
with with
no no waste
waste wa-
water, wastegas,
ter, waste gas,waste
wasteresidue
residue discharge,
discharge, andand environmental
environmental friendliness,
friendliness,and and6N 6Nindium
indium
can
canbebeobtained.
obtained.

Figure 15. Vacuum zone refining/electromigration high-purity indium refining device designed
Figure 15. Vacuum zone refining/electromigration high-purity indium refining device designed
with multiple furnace tubes.
with multiple furnace tubes.

LiLiQing
Qing [56]
[56] and
and Chen
Chen Luona
Luona [50][50] designed
designed a new
a new vacuum
vacuum zonezone refining/electromi-
refining/electromigration
gration high-purity indium refining device designed by Tan Yang [49]
high-purity indium refining device designed by Tan Yang [49] with multiplewith multiple
furnace fur-
tubes.
nace tubes. The heat of an alumina ceramic heating sheet was introduced into the graphite
sheet as a constant temperature heat source, as shown in Figure 16.
Materials 2021, 14, 2064 18 of 21

Materials 2021, 14, x FOR PEER REVIEW

The heat 18 of as
of an alumina ceramic heating sheet was introduced into the graphite sheet 21 a
constant temperature heat source, as shown in Figure 16.

Figure16.
Figure 16.Zone
Zonerefining
refiningequipment
equipment diagram.
diagram.

7.7.Existing
ExistingProblems
Problemsand
andProspects
Prospects
Afternearly
After nearly70
70years
yearsofofdevelopment,
development, thethe zone
zone refining
refiningtechnology
technologyhashasmade
madegreat
great
progressininboth
progress bothprocess
processresearch
researchandandapplication
applicationfield
fieldexpansion.
expansion.The
Theprocess
processisisbecoming
becom-
ing more
more mature,
mature, but there
but there are problems
are still still problems in following
in the the following aspects
aspects [1,27]:
[1,27]:
(1)(1) The
Therequirements
requirements for forraw
rawmaterials
materials areare
high. The The
high. purity of theofraw
purity thematerial rod must
raw material rod
meet certain requirements, and the gap impurities C, H, O, etc. contained
must meet certain requirements, and the gap impurities C, H, O, etc. contained in it in it must
be controlled
must within
be controlled a certain
within range,range,
a certain so as to
so avoid the power
as to avoid and temperature
the power and temperaturefluc-
tuations caused by gas impurities in the zone refining process, which
fluctuations caused by gas impurities in the zone refining process, which affects the affects the qual-
ity of the
quality of material.
the material.
(2) For impurities with a balanced distribution coefficient close to 1, the content of the
(2) For impurities with a balanced distribution coefficient close to 1, the content of the
raw material rods must be strictly controlled so as not to affect the final performance
raw material rods must be strictly controlled so as not to affect the final performance
of the materials: for example, magnesium, calcium, iron, and antimony in bismuth;
of the materials: for example, magnesium, calcium, iron, and antimony in bismuth;
lead, magnesium, silicon, and aluminum in indium.
lead, magnesium, silicon, and aluminum in indium.
(3) The influence of side effects. Side effects include the evaporation of high vapor pres-
(3) The influence of side effects. Side effects include the evaporation of high vapor
sure impurities in the melting zone caused by agitation, temperature increase, or low
pressure impurities in the melting zone caused by agitation, temperature increase,
pressure inert gas flow through the melting zone, due to chemical reactions between
or low pressure inert gas flow through the melting zone, due to chemical reactions
impurities (such as between carbon and sulfur, hydrogen, or oxygen), and the slag-
between impurities (such as between carbon and sulfur, hydrogen, or oxygen), and
ging process often brings complex problems that are difficult to predict. Sometimes,
the slagging process often brings complex problems that are difficult to predict.
in order to remove an impurity, it is necessary to add a second reaction element to
Sometimes, in order to remove an impurity, it is necessary to add a second reaction
complete, and the latter is selected to generate a compound with the former. During
element to complete, and the latter is selected to generate a compound with the former.
zone refining, it is easier than removing certain impurities alone.
During zone refining, it is easier than removing certain impurities alone.
(4) The bar size is limited. The emergence of refining in the suspension zone solves the
(4) The bar size
crucible is limited.
pollution existingTheinemergence of refining
the conventional zoneinrefining,
the suspension
making the zone solves
zone the
refin-
crucible pollution existing in the conventional zone refining, making
ing technology develop rapidly. However, due to gravity, the technique requires that the zone refining
technology
the size of thedevelop rapidly.rod
raw material However, due to gravity,
must be controlled withinthe technique
a certain rangerequires
to obtainthat a
the size of the raw
stable refining zone. material rod must be controlled within a certain range to obtain a
(5) stable refining zone.
Low production efficiency and high cost. The low speed and multiple refining re-
(5) Low production
quired by the zone efficiency and high
refining greatly cost.theThe
extend low speed
production timeandandmultiple
reduce the refining
pro-
required by the zone
duction efficiency; refining
after each zonegreatly extend
refining, the the
firstproduction
and last ends time andbar
of the reduce
need to the
production
be removed, efficiency; after each
which reduces the zone refining,
utilization ratethe
of first
raw and last ends
materials and of the bar need
increases the
toproduction
be removed, which reduces the utilization rate of raw materials and increases the
cost.
production cost.
(6) The development speed of product testing and analysis technology lags behind. With
(6) The the development
improvementspeed of producttechnology,
of purification testing anda analysis
variety oftechnology lags behind.
highly sensitive impurity With
the improvement of purification technology, a variety of highly
analysis methods came into being, but these methods have certain limitations, which sensitive impurity
restrict the development of zone refining technology.
Materials 2021, 14, 2064 19 of 21

analysis methods came into being, but these methods have certain limitations, which
restrict the development of zone refining technology.
(7) The large-scale application of advanced zone refining technology is limited. The
research on zone refining technology has made great progress, and it has been used
in industrial production in certain fields, but some of the deficiencies have hindered
its large-scale application: the suspension zone refining technology equipment in-
vestment is expensive, and raw material storage container selection is more difficult.
Moreover, it is susceptible to secondary pollution during operation, and the optimal
production process parameters are easily changed with the material.
In recent years, the market demand for high-purity metals has continued to expand,
and zone refining has become the preferred method for preparing such materials, with
broad application prospects. In view of the current market demand, the focus of research
on zone refining at this stage should be shifted to the following aspects [27,54].
(1) Combining zone refining technology with other purification technologies, developing
an ideal purification method combining multiple technologies, such as electromi-
gration/zone refining, vacuum degassing-zone refining, to effectively remove gas
impurities and ensure the stability of subsequent zone refining proceed to obtain
higher purity materials.
(2) Upgrade the zone refining equipment, increase the degree of automation and improve
the zone refining technology to obtain a stable and simple operation process, improve
production efficiency, and reduce production costs.
(3) Improve trace element analysis technology to make simultaneous or continuous de-
termination of trace and ultra-trace multi-elements. In addition, the industrialization
of zone refining technology is also the main development direction in the future. With
the deepening of research, zone refining technology will definitely develop toward
the industrialization of low cost, practicality, high efficiency, high reliability, and
high complexity.

Author Contributions: Conceptualization, Y.J. and L.Y.; software, X.K. and L.C.; writing—original
draft preparation, Y.J. and L.C.; writing—review and editing, L.Y., X.K. and L.C.; supervision, X.C.
and K.L.; project administration, X.C. and L.Y.; funding acquisition, X.C., L.Y. and K.L. All authors
have read and agreed to the published version of the manuscript.
Funding: The work was supported by the National Natural Science Foundation of China (51465014),
Guangxi Innovation Driven Development Project (Grant No. AA17204021), the foundation of
Guangxi Key Laboratory of Optical and Electronic Materials and Devices (No. 20KF-4) and Founda-
tion of introduction of senior talents in Hebei Province (H192003015).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Data sharing is not applicable for this article.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Guo, X.Y.; Tian, Q.H. High Purity Metal. Materials; Metallurgical Industry Press: Beijing, China, 2010.
2. Guo, X.; Zhou, Y.; Zha, G. A novel method for extracting metal Ag and Cu from high value-added secondary resources by
vacuum distillation. Sep. Purif. Technol. 2020, 242, 116787. [CrossRef]
3. Virolainen, S.; Suppula, I.; Sainio, T. Continuous ion exchange for hydrometallurgy: Purification of Ag(I)–NaCl from divalent
metals with aminomethylphosphonic resin using counter-current and cross-current operation. Hydrometallurgy 2014, 142, 84–93.
[CrossRef]
4. Itam, Z.; Beddu, S.; Mohammad, D. Extraction of metal oxides from coal bottom ash by carbon reduction and chemical leaching.
Mater. Today Proc. 2019, 17, 727–735. [CrossRef]
5. Lu, X.W.; Peng, J.B. Preparation of high purity indium by electrolytic refining method. Min. Metall. 2018, 27, 54–57.
6. Zhao, Q.S.; Niu, X.D. Purification of high purity germanium by zone refining. Guangzhou Chem. Ind. 2019, 47, 88–90.
Materials 2021, 14, 2064 20 of 21

7. Pfann, W.G.; Olsen, K.M. Purification and prevention of segregation in single crystals of germanium. Phys. Rev. 1953, 89, 322–323.
[CrossRef]
8. Ho, C.D.; Yeh, H.M.; Yeh, T.L. Optimal zone lengths in multi-pass zone-refining processes. Sep. Technol. 1996, 6, 227–233.
[CrossRef]
9. Liu, Y.B.; Yang, H.A. Determination on 15 trace impurity elements in high purity tin by glow-discharge mass spectrometry.
Yunnan Metall. 2018, 47, 72–77.
10. Reeks, W.; Davies, H.; Marchisio, S. A review: Interlayer Joining of Nickel Base Alloys. J. Adv. Join. Process. 2020. [CrossRef]
11. Ledford, C.; Rock, C.; Tung, M. Evaluation of electron beam powder bed fusion additive manufacturing of high purity copper for
overhang structures using in-situ real time backscatter electron monitoring. Procedia Manuf. 2020, 48, 828–838. [CrossRef]
12. Huang, Y.; Zhang, Z.; Cao, Y. Overview of cobalt resources and comprehensive analysis of cobalt recovery from zinc plant
purification residue-a review. Hydrometallurgy 2020, 193, 105327. [CrossRef]
13. Jiao, H.; Song, W.L.; Chen, H. Sustainable recycling of titanium scraps and purity titanium production via molten salt electrolysis.
J. Clean. Prod. 2020, 261, 121314. [CrossRef]
14. Xie, H.; Zhao, H.; Wang, J. High-performance bismuth-gallium positive electrode for liquid metal battery. J. Power Sources 2020,
472, 228634. [CrossRef]
15. Li, X.Y.; Lin, Z.L.; Mi, J.R. Research on preparation technology of ultra-high purity germanium polycrystalline material. Yunnan
Metall. 2020, 49, 56–60.
16. Xu, S.L.; Liu, Z.Q.; Chang, Y.C. Summary of the preparation method of high purity to tellurium and their oxides. Mar. Electr.
2019, 39, 46–50.
17. Mahinroosta, M.; Allahverdi, A. Production of high purity α-alumina and γ-alumina from aluminum dross. Encycl. Renew.
Sustain. Mater. 2020. [CrossRef]
18. Li, W.L.; Luo, Y.H. The review of high purity metal produced by zone refining. Min. Metall. 2010, 19, 57–62.
19. Zhou, Z.H.; Mo, H.B. Preparation of high purity indium by electrolytic refining-zone melting. Chin. J. Rare Met. 2004, 28, 807–810.
20. Hong, W.; Hong, Y.; Dan, W. Making the metal in high solidification by the way of zone refining. Chem. Eng. 2001, 3, 16–17.
21. Pfann, W.G. Techniques of zone melting and crystal growing. Solid State Phys. 1957, 4, 423–521.
22. Pfann, W.G. Principles of zone-melting. J. Miner. Met. Mater. Soc. 1952, 4, 747–753. [CrossRef]
23. Yang, G.; Govani, J.; Mei, H. Investigation of influential factors on the purification of zone-refined germanium ingot. Cryst. Res.
Technol. 2014, 49, 269–275. [CrossRef]
24. Zhai, X.J.; Fu, Y.; Zhang, X.M. The calculation and determination on distribution coefficient of impurities Fe and Si in aluminum.
Light Met. 2000, 12, 37–39.
25. Hu, G.X.; Cai, X. Materials Science; Shanghai Jiao Tong University Press: Shanghai, China, 2006.
26. Burton, J.A.; Prim, R.C.; Slichter, W.P. The distribution of solute in crystals grown from the melt. Part, I. Theoretical. J. Chem. Phys.
1953, 21, 1987–1991. [CrossRef]
27. Liu, W.S.; Liu, S.H.; Ma, Y.Z. The latest development of zone melting technology. Rare Met. Cem. Carbides 2013, 41, 66–71.
28. Prasad, D.S.; Munirathnam, N.R.; Rao, J.V. Effect of multi-pass, zone length and translation rate on impurity segregation during
zone refining of tellurium. Mater. Lett. 2006, 60, 1875–1879. [CrossRef]
29. Ho, C.D.; Yeh, H.M.; Yeh, T.L. The optimal variation of zone lengths in multi-pass zone refining processes. Sep. Purif. Technol.
1999, 15, 69–78. [CrossRef]
30. Zhang, H.; Zhao, J.; Xu, J. Preparation of high-purity tin by zone melting. Russ. J. Non Ferr. Met. 2020, 61, 9–20. [CrossRef]
31. Wan, H.; Xu, B.; Yang, B. The impurities distribution and purification efficiency of high-purity aluminum preparation by zone
melting in vacuum. Vacuum 2020, 171, 108839. [CrossRef]
32. Liu, J.; Zee, R.H. Growth of molybdenum-based alloy single crystals using electron beam zone melting. J. Cryst. Growth 1996, 163,
259–265. [CrossRef]
33. Su, H.; Shen, Z.; Ren, Q. Evolutions of rod diameter, molten zone and temperature gradient of oxide eutectic ceramics during
laser floating zone melting. Ceram. Int. 2020, 46, 18750–18757. [CrossRef]
34. Hu, J.; Ruo, Q.B. Application of a genetic algorithm to optimize redistribution process in zone refining of cerium. Rare Metal.
Mater. Eng. 2017, 46, 3633–3638.
35. Li, H.G. Principles of Metallurgy; Science Press: Beijing, China, 2005.
36. Ho, C.D.; Yeh, H.M.; Yeh, T.L. Numerical analysis on optimal zone lengths for each pass in multi-pass zone-refining processes.
Can. J. Chem. Eng. 1998, 76, 113–119. [CrossRef]
37. Ghosh, K.; Mani, V.N.; Dhar, S. Numerical study and experimental investigation of zone refining in ultra-high purification of
gallium and its use in the growth of GaAs epitaxial layers. J. Cryst. Growth 2009, 311, 1521–1528. [CrossRef]
38. Ghosh, K.; Mani, V.N.; Dhar, S.A. Computational study and experimental validation of zone refining process for ultra-purification
of gallium for opto-electronic applications. Trans. Indian Inst. Met. 2008, 61, 201–206. [CrossRef]
39. Bachmann, K.J.; Clark, L.; Buehler, E. Zone melting of indium phosphide. J. Electron. Mater. 1975, 4, 741–756. [CrossRef]
40. Dost, S.; Liu, Y.C.; Haas, J. Effect of applied electric current on impurity transport in zone refining. J. Cryst. Growth 2007, 307,
211–218. [CrossRef]
41. Yuan, G.; Shiming, Z. Preparation of ultra-pure indium. Non Ferr. Met. Extr. Metall. 2013, 7, 61–64.
42. Zhou, Z.H.; Mo, H.B. Preparation of high purity indium. Min. Metall. Eng. 2003, 23, 40–43.
Materials 2021, 14, 2064 21 of 21

43. Yin, W.H. New development of high purity and ultra high purity refractory metals. Rare Metal. Mater. Eng. 1991, 3, 1–9.
44. Zhang, X.; Friedrich, S.; Friedrich, B. Production of high purity metals: A review on zone refining process. J. Cryst. Process.
Technol. 2018, 8, 23. [CrossRef]
45. Liang, J.; Li, Y.C.; Lin, X.H. Application of high purity metal detection technology. China Molybdenum Ind. 2019, 43, 5–8.
46. Reiss, H. Mathematical methods for zone-melting processes. J. Miner. Met. Mater. Soc. 1954, 6, 1053–1059. [CrossRef]
47. Lee, H.Y.; Oh, J.K.; Lee, D.H. Purification of tin by zone refining with development of a new model. Metall. Trans. 1990, 21,
455–461. [CrossRef]
48. Cheung, T.; Cheung, N.; Tobar, C.M. Application of a genetic algorithm to optimize purification in the zone refining process.
Mater. Manuf. Process. 2011, 26, 493–500. [CrossRef]
49. Tan, Y. Contact Heating Metal. Indium Zone-Refining Device and the Analysis of Temperature Field Simulation; Guilin University of
Technology: Guilin, China, 2017.
50. Chen, L.N. Research on Purification of Indium by Oxidation Refining-Zone Refining Combined Method; Guilin University of Technology:
Guilin, China, 2021.
51. Bochegov, V.I.; Parakhin, A.S. Lmiting impurity distribution during zone refining. Tech. Phys. Lett. 2014, 40, 460–461. [CrossRef]
52. Spim, J.A.; Bernadou, M.J.S.; Garcia, A. Numerical modeling and optimization of zone refining. J. Alloy. Compd. 2000, 298,
299–305. [CrossRef]
53. Cheung, N.; Bertazzoli, R.; Garcia, A. Numerical and experimental analysis of an approach based on variable solute distribution
coefficients during purification by zone refining. Sep. Purif. Technol. 2007, 52, 504–511. [CrossRef]
54. Zhang, X.; Friedrich, S.; Liu, B. Computation-assisted analyzing and forecasting on impurities removal behavior during zone
refining of antimony. J. Mater. Res. Technol. 2020, 9, 1221–1230. [CrossRef]
55. Nakamura, M.; Watanabe, M.; Tanaka, K. Zone refining of aluminum and its simulation. Mater. Trans. 2014, 55, 664–670.
[CrossRef]
56. Li, Q. Purification of Indium by Zone Refining Method and Its Simulation; Guilin University of Technology: Guilin, China, 2018.
57. Li, Y.C.; Liu, Y.; Zhang, C.S.; Le, W.H.; Yu, L.; Luo, K. Research on the preparation of high-purity indium by zone refining. Min.
Metall. Eng. 2014, 34, 104–107.