Metall. Res. Technol.

114, 113 (2017)


c EDP Sciences, 2016 Metallurgical Research
DOI: 10.1051/metal/2016056
&Technology
www.metallurgical-research.org

Comparison of 2D and 3D morphology of non-

metallic inclusions in steel using different
methods
Xuewei Zhang1 , Lifeng Zhang1 , Wen Yang1, Ying Zhang1,3 , Ying Ren1
and Yuanchi Dong2
1
School of Metallurgical and Ecological Engineering, University of Science and Technology Beijing,
Beijing 100083, P.R. China
e-mail: [email protected]
2
School of Metallurgy and Materials, Anhui University of Technology, Ma anshan 243002, P.R. China
3
Beijing Key Laboratory of Green Recycling and Extraction of Metals (GREM), University of Science
and Technology Beijing, No. 30 Xueyuan Rd, Hadian District, Beijing 100083, P.R. China

Key words: Abstract – In the current work, the two-dimensional (2D) morphology of MnS inclusions
MnS inclusions; 3D morphology; with different shapes, including spherical, rod-like, flower-like, polyhedron and aggregated
electrolysis; non-aqueous cluster, as well as spherical oxides were observed. The three-dimensional (3D) morphology
electrolyte of inclusions extracted from steel using acid dissolution, but the MnS was reacted away
with acid during the extraction of the inclusion from the steel matrix. It is shown that the
2D method is hardly accurate to detect the morphology and size of inclusions comparison
3D morphology of inclusions. Therefore, a method of separation, extraction and analysis
of non-metallic inclusions in steel was presented. Electrolytic extraction using non-aqueous
electrolytes were performed to dissolve the conductive steel matrix and remain nondestruc-
tive inclusions. Scanning electron microscopy (SEM) with an energy-dispersive spectrometer
(EDS) was used to obtain the integrated 3D morphology, size and chemical composition of
inclusions. The morphology of MnS inclusions was reclassified on the basis of the integrated
3D morphology under the following categories: elliptical, spherical, rod-like, plate-like, ir-
regular or polyhedron, dendrite and patch MnS precipitated on the surface of oxides. Non-
aqueous electrolysis extraction with the advantage of nondestructive to inclusions, low cost,
and high efficiency, and the subsequent analysis of morphology, size and chemical compo-
sition is more reliable and accurate than the traditional method. What is more, it allows
Received 23 February 2016
a better understanding of the origin and formation of inclusions, which is helpful for the
Accepted 20 September 2016
control and removed of inclusions in steels.

on-metallic inclusions have a sig- that the size distribution of inclusions ob-

N nificant impact on mechanical

properties of steel [1–3]. The fast
and accurate detection of inclusions in steel
tained from 3D detection was more accu-
rate than that by 2D detection, especially
for Al2 O3 clusters and aggregations. Suito
have been a difficulty for a long time. Zhang et al. [13–15] studied the size and amount
et al. [4] summarized nearly thirty detec- of inclusions on the cross section of steel
tion methods and described their advan- samples and concluded that 3D detection
tages and disadvantages. There have been was more reliable than 2D observation, es-
other reported detection methods, but the pecially for small inclusions. Zhang [16]
lack of details [5–9] Garcia et al. [10] stud- studied the 3D morphology of acid-resistant
ied the extraction process of inclusions from inclusions by partial acid extraction reveal-
steel samples using the method of acid dis- ing that different Al2 O3 inclusions under
solution that was limited to high Al2 O3 and 2D observation were actually a single large
SiO2 -based inclusions since other basic com- Al2 O3 cluster in 3D scale. Dekkers et al. [17]
ponents react with acid. Jonsson et al. [11,12] studied inclusions in Al-killed steel by acid
had studied several methods to reveal the 3D dissolution and focused on large inclusions
morphology of inclusions. It was reported Liu et al. [18] studied the 3D morphology of

Article published by EDP Sciences

 Xuewei Zhang et al.: Metall. Res. Technol. 114, 113 (2017)

<3 µm oxides inclusions by acid dissolution

method in order to investigate the effect of
small inclusions on the performance of the
steel
Non-metallic inclusions in steel are al-
ways expected to be removed. However,
it has been recognized that some inclu-
sions smaller than 3 µm might be benefi-
cial for grain refinement. It is important to
detect such small inclusions, especially to
observe their three-dimensional morpholo-
gies. For the extraction of inclusions from Fig. 1. Schematic diagram of tiny galvanic cells
steel [12, 13], the chosen reagent should generated on the surface of the steel matrix.
not too acid, and the generated complexes
should have low viscosity and high solubil-
ity in the organic solvent. The characteristics boundary is generally lower than that of
of various reagents are summarized in Ta- grains of the steel matrix. Thus, the grain
ble 1 [5, 7, 8, 11–13, 19, 20]. boundaries will be preferentially dissolved.
As a kind of nucleophilic reagent with The potential existed between non-metallic
strong complexing ability, triethanolamine inclusions and the steel matrix is also the
can be easily combined with Fe2+ near the reason for the electrochemical unevenness,
surface of the steel sample into diffluent inducing the separation between them. Dur-
complex (CH2 CH2 OH)3 NFe. Acetylacetone ing the dissolution process of the steel in
also has a strong complexing ability due to the electrolyte solution, complex multi-cell
the existing organic functional groups- car- reactions occur [22]. The tiny galvanic cells
bonyl (C = O) which is easy to combine with generated on the surface of steel matrix is
Fe2+ to form a large number of reddish or- schematically shown in Figure 1.
ange complexes, named acetylacetone iron,
The redox reactions are [23]:
which is soluble in the anhydrous alcohol
(ethanol) solvent, but generating very sticky Negative (carbon or carbide): 2H+ + 2e =
solution. The complexing reactions are as H2 ↑ (oxidation)
follow: Positive (Fe): Fe = 2e + Fe2+ (reduction)
(CH2 CH2 OH)3 N̈ + Fe2+ Redox reaction: 2H+ + 2Fe = 2Fe2+ + H2 ↑
→ (CH2 CH2 OH)3 NFe (1) Since every detection method has disad-
2CH3 CÖCHCOCH3 + Fe 2+ vantages, several methods are sometimes
combined with each other to comprehen-
→ Fe (CH3 COCHCOCH3 )2 (2)
sively and accurately investigate inclusions
The suitable electrolyte is used if the steel in steel [24, 25], which are summarized as
sample is uniformly dissolved, with small follows and compared in Table 3. It is im-
electrolyte change, stable current density, portant to develop a method to accurately
high electrolysis efficiency and a relatively and efficiently detect the amount, composi-
flat surface of the sample after electroly- tion, morphology and size of non-metallic
sis, and generates few colloids. The optimal inclusions in steel [26]. Inclusions can be
composition of the electrolyte was proposed extracted from Al deoxidized steel or Si-
as shown in Table 2. Mn deoxidized steels by the electrolytic ex-
Electrolytic extraction is an electrochem- traction method. The inclusions include ox-
ical process in electrolytes. Different regions ides such as MnO, SiO2 , Al2 O3 , CaO, MgO,
of the surface of steel matrix have different etc., sulfides such as MnS, CaS, CuS, TiS
electrode potential, forming a large num- etc., nitrides and carbonitrides like AlN,
ber of micro-cells. The dissolution occurs TiN, Ti(CN), V(CN), Nb(CN) and phosphide
in the anodic regions with lower potential, like FeP3 Electrolytic extraction using weak-
and precipitation occurs in the higher poten- acid electrolyte (Slimes) was reported in
tial cathodic region. The potential in grain literatures [10, 16, 27].
113-page 2
 Xuewei Zhang et al.: Metall. Res. Technol. 114, 113 (2017)

Table 1. Reagents used in the electrolysis process to extract inclusions from steel.

No. Name and Chemical Function Concentration Ref.

formula properties (wt.%)

– Low toxicity;
– Weak alkalinity;
– A control of PH stabilizer;
Triethanolamine – Reacting with inorganic acid or organic
1 solution into salt due to the presence of Complexing 10–15 [19]
(N(CH2 CH2 OH)3 )
lone pair of electrons on nitrogen atom;
– Generating a stable chelate with variety
of heavy metal ions.

– Toxic, colorless, transparent

Acetylacetone and ester aroma liquid; Complexing
2 10-15 [5, 7, 8, 11]
(CH3 COCH2 COCH3 ) – Used for metal complexing agent reagents
and also as a desiccant.

– Low toxicity, colorless and oily;

1,4-Butyrolactone – Soluble into methanol or ethanol; Complexing
3 – With organic functional groups 6–12 [20]
(C4 H6 O2 ) reagents
(C = O) in the molecular structure.

Glycerol Colorless, sweet, odorless and Reducing

4 2–5 [19]
(C3 H8 O3 ) viscous with clarity of liquid. reagents
Tetramethylammonium
Conductive
5 chloride Toxic. 1–4 [12]
reagents
(C4 H12 NCl)

– White crystals with low toxicity;

– Dissolvable in many organic solvents,
Lithium chloride such as ethanol or methanol; Conductive
6 1–3 [13, 19, 20]
(LiCl) – Forming adduct of compositions, reagents
due to its non-ionic bond.

Ethanol Flammable, volatile, colorless and

7 Solvent Remain [20]
(CH3 CH2 OH) transparent.
Methyl alcohol
8 Colorless and volatile with alcohol odor. Solvent Remain [19]
(CH3 OH)

Non-metallic inclusions have some im- tion of MnS particles were discussed. How
portant effects on mechanical properties of to analyze MnS particles and other kind of
steel. The qualified detection of inclusions inclusion accuracy and efficiency is yet a key
is significant to ensure the quality of steel problem for inclusion study even though
However, it is difficult to determine how there are many methods to evaluate inclu-
many metallographic surfaces must be cut sions in steel. So the experimental procedure
and grind resulting from randomness distri- for the extraction of inclusions from steel
bution of inclusions in the sample. Therefore samples using non-aqueous electrolyte was
there are serious problems and cannot ac- described and introduced.
commodate to the development of steel met-
allurgy using the traditional methods due to
hardly prove all kinds of inclusion were fully 1 Experimental procedure
observed in the sample.
The three electrolytes have been tested by
In the current work, the pros and cons means as shown in Table 2. The experimental
of the traditional method for the observa- results show that Nos. 1, 2, and 3 electrolytes
113-page 3
 Xuewei Zhang et al.: Metall. Res. Technol. 114, 113 (2017)

Table 2. Different electrolytes used for the extraction of inclusions from steel.

No. Composition (wt%) Ref.

1 1 pct 4-methyl ammonium chloride + 5 pct triethanolamine [19]
+ 5 pct glycerin + 89 pct methyl alcohol
2 2 pct 4-methyl ammonium chloride + 8 pct acetylacetone [21]
+ 5 pct glycerin + 6 pct triethanolamine + 79 pct methyl alcohol
3 1 pct lithium chloride + 6 pct 1,4-butyrolactone + 5 pct glycerin [20]
+ 88 pct ethanol or 88 pct methyl alcohol

Table 3. Comparison of different extraction methods for inclusions in steel.

Different advantages disadvantages Types

extraction
of inclusions
methods

– Inclusion amount, size distribution

– Incomplete 3D morphology
Partial acid estimate SiO2 , Al2 O3 ,
– Incomplete composition data
extraction and morphology possible TiN and VN
– Alkaline inclusions easily react
– Inexpensive and fast

– Only acid-proof inclusion extracted

Acid – 3D morphology possible – No size distribution
dissolution – Inclusion composition measured – The sample matrix must first be dis-
method – Inexpensive and fast solved
– Incomplete composition data

– Accurate inclusion amount mea-

Electrolytic sured
extraction – Time consuming to dissolve entire
– Captures rare large inclusions
using weak-acid large sample
– With subsequent analysis, can get
electrolyte – Incomplete composition data
3D
(Slimes) size distribution and morphology

– 3D morphology good revealed MnO, SiO2 ,

Electrolytic – Inclusions size and composition Al2 O3 , CaO,
extraction using measured MgO, MnS,
– No size distribution
non-aqueous – Can measure inclusions smaller CaS, CuS, TiS,
electrolyte than 1 µm AlN, TiN, VN
– Fast and so on.

gave no colloidal precipitates due to the fact Fe3 C) remained. Owing to the nonexis-
that the electrolyte contains a complexing tence of H+ , OH and O2− , FeO was hardly
agent. The colloidal precipitates with metal produced during the electrolysis process.
ions were formed in solution if the elec- Hence, after the separation of carbide (Fe3 C)
trolyte does not contain a complexing agent. by the powerful magnet, the remaining were
This experiment was repeated, using Nos. non-metallic inclusions whose original mor-
1, 2 and No. 3 electrolytes, with almost the phology and composition were kept due to
same results. Therefore each of them may be the fact of the low-temperature process with-
used; however, in the remainder of this work out acid or alkali solutions.
only electrolyte No. 1 was used. Steel samples were taken from a bloom
In the current study, since the organic so- of a Si-killed heavy rail steel They were
lution was used as an electrolyte, the steel dissolved in the electrolytic solution, and
matrix was dissolved after energization, in- the residues were inclusions. After magnetic
clusions and carbide (Mainly the magnetic separation, washing, filtration and drying,
113-page 4
 Xuewei Zhang et al.: Metall. Res. Technol. 114, 113 (2017)

the remaining inclusions were then sepa-

rated. The procedure is illustrated in Fig-
ures 2 and 3. By this method, the original
chemical composition, size and morphology
of inclusions in steel will be hardly destroyed
during extraction. Materials and equipment
were used as following:
– Constant current power supply (Do not
exceed 20 V for safety).
– Beaker with a sealed cap (400 ml).
– Refrigerator (or ice bath of the bigger
than 400 ml beaker).
– Ultrasound equipment.
– Electrolyte.
– Watch-glass (15 cm in diameter).
– Copper wires.
– Bracket.
– Powerful magnet.
– Filtration device.
Procedures of electrolytic extraction were
summarized as follows:
1. Cylinder pieces with 8 mm in diame-
ter and 110 mm in height from heavy
rail steel bloom of U75V were machined.
These samples were polished with SiC
papers and diamond suspensions. Then, Fig. 2. Extraction identification of nonmetallic inclu-
removing the small particles on the sur- sions in the current study.
face of samples using ultrasound equip-
ment.
2. The electrolyte in a 400 ml beaker was
prepared and the argon gas was injected
into the electrolyte in order to exclude the
air and have the effect of stirring.
3. During electrolysis, the anode is steel
specimens, while the cathode is stain-
less steel. The current density was set
at 0.04 A/cm2 and the electrolysis time
was 8 h (The total time should not be
exceeded 12 h). The temperature of elec-
trolyte is kept between –5 and 5 ◦ C.
4. After electrolysis, the anode with inclu-
sion and carbide particles was put into
ethyl alcohol by ultrasonic equipment. Fig. 3. Schematic of the electrolytic extraction method
5. The watch-glass was employed to pu- used in the current study (1-anode; 2-cathode; 3-
thermometer; 4-electrolytes; 5-beaker).
rify the solution which including parti-
cles and electrolyte. The new anhydrous
ethanol was added into the watch-glass
by the wash bottle, next, the solution was particles in the ethyl alcohol. Another
sucked away from the edge of watch- method, Filtration device was used to in-
glass using the glass pipette, and that tercept inclusion to the membranes.
cycle repeats until the solution becomes 6. Finally, all inclusion particles in ethyl
clear. Then, carbide (Fe3 C) was magnet- alcohol were elutriated, dried and col-
ically separated with magnet from all lected.
113-page 5
 Xuewei Zhang et al.: Metall. Res. Technol. 114, 113 (2017)

Fig. 4. 2D morphology of inclusions analyzed using the traditional 2D method, (a)–(e): pure MnS,
(f): 25.83%Al2 O3 -36.76%CaO-30.24%SiO2 -3.86%MgO-3.31%MnS sulfide-oxide inclusion.

7. The extracted inclusions were placed flat spherical oxides were observed. Since the in-
on the conductive tape and further ana- clusion observed here is a 2D cross section.
lyzed using the SEM and their composi- The size of the spherical MnS in Figure 4a
tion was detected using the EDS. may be smaller than its real size, the rod-like
MnS in 2D in Figure 4b may be a plate-like
one in 3D, the flower-like MnS in Figure 4c
2 Results and discussion may be a dendritic one in 3D, the polyhedral
MnS in Figure 4d may not be quadrangu-
lar in 3D, and the aggregated cluster MnS in
2.1 Analysis of inclusions using
the traditional method 2D in Figure 4e may be the accumulation of
many rod-like, spherical and plate-like MnS
particles, the patch MnS were damaged on
Figure 4 shows the 2D morphology of inclu- surface of spherical oxides in 2D in Figure 4f
sions observed by the traditional 2D method, and it had not been represented the real one.
MnS inclusions with the different shapes, in-
cluding spherical, rod-like, flower-like, poly- Figure 5 reveals the 3D morphology of
hedron and aggregated cluster, as well as inclusions extracted from steel using the
113-page 6
 Xuewei Zhang et al.: Metall. Res. Technol. 114, 113 (2017)

Fig. 5. 3D morphology of inclusions extracted by partial acid extraction (a) 24.67%MgO-

69.55%Al2 O3 -4.45%SiO2 -1.32%CaO; and by acid dissolution method (b) 41.53%Al2 O3 -58.47%SiO2
& (c) 100%Al2 O3 ; (d) 100% SiO2 by slime electrolysis using HCl+FeSO4 +CuSO4 as electrolyte.

traditional 3D method. Figure 5a shows a 2.2 Extraction of inclusion

rough 3D morphology of a MgO-Al2 O3 SiO2 using non-aqueous electrolysis
inclusion extracted by partial acid extrac-
tion Figure 5b shows a SiO2 -Al2 O3 inclusion The morphology of different inclusions in
extracted by acid dissolution method with steel can be distinguished by different col-
many encaved pots on its surface. These pots ors as shown in Figure 6, Employing the
were the root sites of precipitated MnS dur- new method, the characteristics, formation
ing solidification and the MnS was reacted mechanism, size chemical composition ad
away with acid during the extraction of the well as the effects on steel of nanoinclusions
and some special inclusions such as TiVN,
inclusion from the steel. The acid dissolu-
tion method is suitable for the extraction of which are difficult to be found by the tradi-
the irregular Al2 O3 clusters as shown in Fig- tional 2D detection.
ure 5c since the inclusion is not easily broken The critical factor for the better elec-
or damaged and lost its original morphol- trolytic extraction is mainly the electrolyte.
Triethanolamine coordinates with iron by
ogy, on the contrary it is not suitable if the
compound Al2 O3 -based inclusion contains N and O, and the coordination of N will
alkaline constituents may easily react with be preferred due to the low non-metallic
the acid solution. In the current study, poly- nature. The ability of coordination of OH−
carbonate membranes filters were used to group will be enhanced if H+ is lost. The
stability of complexes formed by Fe2+ is de-
separate inclusions from the acid solution.
Figure 5d is a 100% SiO2 inclusion with a rived from the low valence orbital energy
big crack on its surface that was extracted of Fe2+ which has strong coordination abil-
from the steel using electrolysis method with ity due to the formation of strong coordina-
HCl+FeSO4 +CuSO4 as an electrolyte. The tion bonds. Fe reacts with triethanolamine
preferentially, which will possibly mask Fe3+
cracks might be covered with MnS and the
MnS reacted away with the weak solution of by the formation of more stable tetrahedral
HCl+FeSO4 +CuSO4 . complexes with [N(C2 H2 O)3 ]3− but do noth-
ing with Fe2+ .
113-page 7
 Xuewei Zhang et al.: Metall. Res. Technol. 114, 113 (2017)

Fig. 6. Classification of inclusions by different colors.

Consequently, the new electrolytic ex- tated on the surface of oxides (Fig. 7f). The
traction method can be the substitution of advantage of this method can be easily as-
traditional 3D detection for inclusion re- sessed compared with the traditional 3D de-
search, it can reduce the dosage of acid, tection method using which the MnS on
the water contamination, the environmen- the oxide would be dissolved as shown in
tal pollution and the harm to human health, Figure 5b.
and will draw more attention and utilization The corresponding composition of the in-
of metallurgical workers. clusions in Figure 7 is listed in Table 4. It is
MnS inclusions were divided by indicated that the elliptical inclusion in Fig-
Sims [28] in 1938 into three types on ure 7a was an oxide wrapped by MnS, which
the basis of 2D morphology: spherical, can reduce the detrimental effect of the oxide
chain-shaped and angular irregular In 1981, inclusion on steel [30, 31].
Ito [29] reclassified MnS inclusions into
spherical, fan-shaped and chainshaped,
polyhedron and irregular based on the 2.3 Comparison 3D and 2D
2D morphology Oikawa [5] modified and morphology of inclusions
supplemented the classification based on
the rough and incomplete 3D morphol- Figure 8 shows schematic comparison dia-
ogy in 1995, as spherical, dendritic, and gram 3D and 2D morphology of the inclu-
polyhedral. The results obtained by the sions. It indicates the dashed line in Figure 8a
current authors were in accordance with the was random polishing metallographic sur-
previous researches mentioned above. face, and its actual morphology is a rectangle
The new electrolytic extraction method in 3D, rather than long strips in the case of 2D
was applied to extract MnS inclusions in as shown in Figure 8b. Furthermore, this in-
this paper The extracted inclusions were clusion may harm to the performance of steel
tiled on the carbon conductive resin and due to the size of 3D is larger than 2D for the
then detected using SEM + EDS to obtain same one. The dendritic inclusion is detri-
the chemical composition size and morphol- mental effects on the mechanical properties
ogy. Based on the results, the morphology of of the steel as shown in Figure 8c. However,
MnS inclusions were reclassified as shown only a few small spherical and elliptical in-
in Figure 7, including elliptical and spher- clusions were observed in 2D as shown in
ical (Fig. 7a), rod-like (Fig. 7b), plate-like Figures 8d and 8e shows patch MnS oxide
(Fig. 7c), irregular or polyhedron (Fig. 7d), in 3D, but only few spot were obtained on
dendrite (Fig. 7e), and patch MnS precipi- edge of inclusion in 2D as shown in Figure 8f.
113-page 8
 Xuewei Zhang et al.: Metall. Res. Technol. 114, 113 (2017)

Fig. 7. Morphology of inclusions attained by electrolytic extraction method (Inclusions are adhered
to the conductive tape).

Table 4. Chemical composition, size and morphology of inclusions by a new electrolytic extraction
method.
Inclusions observed by new electrolytic extraction method in heavy rail steel slab
Chemical Inclusions Inclusions Inclusions Inclusions Inclusions Inclusions
composition (a) Elliptical (b) Rod-like (c) Plate-like (d) Polyhedron (e) Dendrite (f) Patch
(wt.%) 17 µm 53 µm 58 µm 10 µm 25 µm 9 µm
MnS 83.79 100 100 100 100 10.58
MgO – – – – – 12.2
Al2 O3 1.56 – – – – 49.18
SiO2 – – – – – 13.12
CaO 2.28 – – – – 14.31
MnO 12.37 – – – – –

113-page 9
 Xuewei Zhang et al.: Metall. Res. Technol. 114, 113 (2017)

Fig. 8. Schematic comparison diagram 3D and 2D morphology of the inclusions.

The ratio of precipitation and area fraction of sis with an organic electrolyte solution was
statistics will be inaccurate by 2D. introduced, and then their complex com-
As mention above, there is serious unre- position, various morphologies, distribution
liability using the 2D method with advan- of size, physical and chemical were charac-
tages of this is convenient and commonly terized using scanning electron microscope
used in industrial practice due to haphaz- with energy dispersive spectrometer (SEM
ard distribution and rare number density of + EDS), On this basis, the thermodynam-
inclusions. Therefore, it is hardly accurate ics and kinetics of inclusions formation and
to detect the morphology and size of inclu- modification were calculated and predicted
sions and it is no significant to determine the using the thermodynamic software in order
inclusions using the traditional 2D method, to eliminate the effect of some disadvan-
no matter how many inclusions were ob- tages on steel quality. (b) The behavior of
served in steel, hardly found a qualified met- inclusion was conduct only by sampling in-
allographic surface to represent all of the spection, rather than make frequent to detect
inclusion. under the condition of the optimization dur-
As we all know, nonmetallic inclusions ing steelmaking process which needs to be
have harmful effects on mechanical proper- guided.
ties of steels because of their characteristics.
The development direction of inclusions can 3 Conclusions
be carried out as following: (a) A nonde-
structive method to extract inclusions from Non-aqueous electrolysis extraction was uti-
steel using non-aqueous solution electroly- lized to extract the inclusions. The SEM
113-page 10
 Xuewei Zhang et al.: Metall. Res. Technol. 114, 113 (2017)

+ EDS were applied to detect the morphol- [2] R. Farrar, P. Harrison, Mater. Sci. 22 (1987)
ogy, size and type of non-metallic inclusions. 3812-3820
The present of results can be summarized as [3] Z. Liu, Y. Kobayashi, F. Yin, M. Kuwabara,
ISIJ Int. 47 (2007) 1781-1788
follows:
[4] Lifeng Zhang, B. Thomas, ISIJ Int. 43 (2003)
1. Inclusions can be extracted by non- 271-291
aqueous electrolysis extraction with the [5] K. Oikawa, H. Ohtani, K. Ishida, T.
advantage of nondestructive to inclu- Nishizawa, ISIJ Int. 35 (1995) 402-408
sions, low cost high efficiency and so [6] M. Wakoh, T. Sawai, S. Mizoguchi, ISIJ Int.
36 (1996) 1014-1021
on, and the subsequent analysis of mor-
[7] K. Oikawa, K. Ishida, T. Nishizawa, ISIJ Int.
phology, size and chemical composition 37 (1997) 332-338
is more reliable and accurate than the tra- [8] J.Y. Park, J.K. Park, W.Y. Choo, ISIJ Int. 40
ditional method. The organic electrolyte (2000) 1253-1259
without any water was used, no reaction [9] N. Tsunekage, H. Tsubakino, ISIJ Int. 41
for forming OH− would occur in the pro- (2001) 498-505
cess of electrolysis. Therefore, in theory, [10] M. Fernandes, N. Cheung, A. Garcia, Mater.
Characterization 48 (2002) 255-261
the extracted inclusion particles would [11] H. Doostmohammadi, A. Karasev, P.
not be corroded by the electrolyte in the Jönsson, Steel Res. Int. 81 (2010) 398-406
whole process of electrolysis. [12] D. Janis, R. Inoue, A. Karasev, P. Jonsson,
2. A new classification of MnS inclusions Adv. Mater. Sci. Eng. 2014 (2014) 7
was proposed on the basis of the inte- [13] R. Inoue, S. Ueda, T. Ariyama, H. Suito, ISIJ
grated 3D morphology obtained using Int. (2011) 6
[14] J. Takahashi, H. Suito, Acta Mater. 49 (2001)
the non-aqueous electrolysis extraction
711-719
method, including spherical or elliptical, [15] J. Takahashi, H. Suito, Metall. Mater. Trans. A
plate-like, rod-like, irregular or polyhe- 34 (2003) 171-181
dral and dendritic. [16] L. Zhang, S. Li, J. Wang, X. Zuo, Iron Steel 44
3. The 2D method is hardly accurate to (2009) 75
detect the morphology and size of in- [17] R. Dekkers, B. Blanpain, P. Wollants, F.
clusions comparison 3D morphology of Haers, Ironmak. Steelmak. 29 (2002) 437-444
[18] L. Liu, B. Zheng, Q. Wu, Q. Zhang, J. Iron
inclusions. The good morphology of Steel Res. 19 (2007) 85-89
MnS was obtained through the non- [19] K. M .Fang, R.M. Ni, Metall. Trans. A 17
aqueous electrolysis extraction method (1986) 315-323
which could be used to evaluate the [20] F.F. Shi, H.G. Li, C.F. Wu, S.B. Zheng, Phys.
cleanliness of steel accurately, as well as Examination Testing 28 (2010) 42-46
to provide effectively guidance for the [21] H.G. Li, S.B. Zheng, Q. Zheng, L.W. Wang,
Steel Res. Int. 78 (2007) 898-902
control of chemical composition, size and
[22] R. Avci, B. Davis, M. Wolfenden, I. Beech,
amount of inclusions. Corros. Sci. 76 (2013) 267-274
[23] G. Wang, S. Li, X. Ai, C. Zhang, J. Iron Steel
Res. Int. 22 (2015) 566-572
Acknowledgements [24] M. Burty, C. Louis, P. Dunand, P. Osmont, La
The authors are grateful for support from the Revue de Metallurgie-CIT 97 (2000) 775-782
National Science Foundation China (Grant Nos. [25] L. Zhang, K. Cai, Project report: Cleanliness
51274034, 51334002, and No. 51404019), State Key investigation of low carbon Al-killed steel in
Laboratory of Advanced Metallurgy, Beijing Key Bao Steel, Report, Bao Steel, 1997
Laboratory of Green Recycling and Extraction of [26] K. Fang, R. Ni., Metall. Trans. A 17 (1986)
Metals (GREM), the Laboratory of Green Process 315-323
Metallurgy and Modeling (GPM2 ) and the High [27] R. Gurry, J. Christakos, C. Stricker, Trans.
Quality Steel Consortium (HQSC) at the School of ASM 50 (1958) 105-128
Metallurgical and Ecological Engineering at Uni- [28] C. Sims, F. Dahle, Trans. AFS 46 (1938) 65-132
versity of Science and Technology Beijing (USTB), [29] Y. Ito, N. Masumitsu, K. Matsubara, Trans.
Iron Steel Inst. Japan 21 (1981) 477-484
China.
[30] F. Hengerer, B. Lilljekvist, G. Lucas, Hart.-
Tech. Mit 30 (1975) 91-98
References [31] K. Tsubota, T. Sato, Y. Kato, K. Hiraoka,
ASTM Special Technical Publication 1327
[1] P. Harrison, R. Farrar, J. Mater. Sci. 16 (1981) (1998) 202-215
2218-2226
113-page 11