Materials Transactions, Vol. 50, No. 6 (2009) pp. 1370 to 1379
#2009 The Japan Institute of Metals
Sulfide Precipitation in Titanium-Added Steel with Residual Level of Copper (2)
—Precipitation in Ferrite Region—
Yasuhide Ishiguro1; *1 , Takashi Murayama2; *2 , Takeshi Fujita3 and Kotaro Kuroda4
1
Steel Research Laboratory, JFE Steel Corporation, Handa 475-8611, Japan
Steel Research Lab., JFE Steel Corp., Kawasaki 210-0855, Japan
3
Steel Research Lab., JFE Steel Corp., Fukuyama 721-8510, Japan
4
Department of Materials Science and Engineering, Nagoya University, Nagoya 464-8603, Japan
2
This report is the second paper in three consecutive papers to know the whole sulfide precipitation in Ti-added steel by adding a new
knowledge that even a trace level of copper (0.01%Cu) acts as a sulfide-former to form copper sulfide (Cu-S). The second paper focuses on the
sulfide precipitation in ferrite region in Ti-added steel, especially in real-produced hot-rolled and annealed steel sheets.
Although it has been believed that the sulfide precipitation is completed only in austenite region except for a few limited papers, Ti4 C2 S2 ,
MnS and Cu-S are really formed in ferrite region. When as-hot-rolled Ti-added steel is annealed, the sulfide precipitation transforms
morphologically and compositionally. The phenomenon can be explained through the same mechanism in the first paper that TiSdecomposition-induced re-solute sulfur is re-precipitated. When annealed in ferrite region, already-precipitated TiS is changed into Ti4 C2 S2 and
re-solute sulfur comes out along the estimated reaction ‘‘4TiS þ 2[C] ! Ti4 C2 S2 þ 2[S]’’, and it is re-precipitated in ferrite region as other
precipitates (Ti4 C2 S2 , MnS and Cu-S). The precipitation of Cu-S is estimated to occur at the lowest temperature among re-precipitated sulfides,
and therefore is affected by the precedent sulfide precipitation phenomena. [doi:10.2320/matertrans.MRA2008299]
(Received August 29, 2008; Accepted March 6, 2009; Published May 25, 2009)
Keywords: sulfide precipitation, copper sulfide, titanium carbo-sulfide, titanium sulfide, manganese sulfide, Ti-stabilized interstitial-free steel,
interstitial-atom-free steel (IF) steel, transmission electron microscopy, quantitative chemical analysis, extracted residue, residual
element, tramp element, ferrite region
1.
Introduction
Through three consecutive papers,1,2) the authors show that
sulfide precipitation in Ti-added steel should be interpreted
through one integrated concept to apply to both austenite and
ferrite regions This unified theory includes a new knowledge
that even a trace level of copper (0.01%Cu) acts as a sulfideformer to form copper sulfide (Cu-S) and that the sulfide
precipitations occurs even in ferrite region, that is, titanium
carbo-sulfide (Ti4 C2 S2 ) and manganese sulfide (MnS) and
Cu-S are formed in ferrite region.
This paper is the second part of three consecutive
papers,1,2) and focuses on sulfide precipitation in ferrite
region, especially at real-produced hot-rolled and annealed
steel sheets. Our standpoint3–20) and reviews of past studies
have already been stated in details in the first report.1)
Although the following two items are repeated topics in the
first report,1) we dare to make remarks on disputed items of
sulfide precipitation in ferrite region in Ti-added steel.
(1) It has not been reported that a small amount of Cu acts
as s sulfide-former in Ti-added steel.
(2) At some limited papers, it has been shown that the
experimental facts that sulfides are formed in ferrite region in
Ti-added steel without touching on any Cu-S precipitations.21–24) K. Yamada et al. say that TiS continuously
changes into Ti4 C2 S2 ,21,23) and H. Murakami et al. say that
the sulfide precipitation in ferrite region is additive-Tidependent phenomenon:22,24) the same behavior to Yamada’s
occurs in low-Ti-added steel but Ostwald’s ripening of TiS
*1Corresponding
*2Present
Japan
author, E-mail: [email protected]
address: Nippon Magnetic Dressing Co., Kurashiki 712-8074,
occurs in high-Ti-added steel. In the paper, the authors show
what really occurs in sulfide precipitations in ferrite region
and give mechanism to explain the phenomena in ferrite
region by using as-hot-rolled and as-annealed Ti-added steel
sheets with various conditions of coiling temperature and
annealing temperature, which is supposed to be within the
range of real production.
2.
Samples and Experimental Procedure
In understanding sulfide precipitation in Ti-added steel, the
authors selected Ti-stabilized Interstitial-Free (Ti-IF) steel,
because it is so simple a system that we are able to obtain a
precise result of sulfide characterization.
As shown in Table 1 and Fig. 1, the steels with 0.04% Ti
addition were prepared to carry out research on the sulfide
precipitation phenomena in ferrite region. The main characteristic is that all Ti-added samples include 0.01%Cu, which
is inevitably present even if degraded scraps are not used for
steelmaking. The samples are from commercial-based steel
which was produced through a trial operation from blastfurnace & LD-converter-based steelmaking and continuouscasting, and then produced by hot-rolling and the subsequent
annealing in continuous-annealing-line (CAL). It should be
noted that the sample of ‘‘0.04%Ti-stabilized IF steel’’ in this
paper is not exactly the same as ‘‘0.04%Ti-IF’’ in the first
paper1) and the content of sulfur differs by only 2 ppm, but
both are referred to as ‘‘0.04%Ti-IF’’, because steel product
has an allowable range of chemical compositions in real
production and these two samples can be regarded as same
steel sheets within an allowance.
As shown in Fig. 1 and Table 1, sulfide precipitation is
reported by using the samples prepared in the following
Sulfide Precipitation in Titanium-Added Steel with Residual Level of Copper (2) —Precipitation in Ferrite Region—
1371
Table 1 Chemical compositions of Ti-added steel (mass%).
Sample
Heat Treatment
C
Si
Mn
S
Cu
Ti
0.04%Ti-IF
(ii) Hot-rolling & annealing
range ()
0.0014
0.01
0.17
0.0084
0.01
0.039
As-Hot
(9)830°C
(5)CT: 640°C
(6)CT: 680°C (8) 770°C
(7)CT: 720°C
(a) As–hot
(b) As–annealed
Cu–S
Ti4C2S2
TiS
Fig. 1
TiS
Schematic diagram of heat treatment of (ii).
Cu–S
Ti4C2S2
conditions: (ii)a comparison of as-hot-rolled samples (referred to as ‘‘as-hot’’) among three levels of Coiling Temperature (CT) conditions ((5)640 C, (6)680 C and (7)720 C) as
well as a comparison of as-annealed samples among two
levels of annealing temperature conditions ((8)770 C and
(9)830 C). It should be noted that the annealing time
corresponds to about 90 sec., and the samples tend to cool
down rapidly because the heat capacity of steel sheet is small
and the ratio of surface to volume is large enough to remove
heat quickly. In short, nine conditions are compared each
other (= 1steel 3CT (as-hot + 2annealing)) to know
sulfide precipitation in ferrite region in Ti-added steel. Please
note that the above numbering intentionally starts with (ii) in
Table 1 and (5) in Fig. 1. They are consecutive numbers from
the first paper.1)
Experiments were carried out mainly through two methods: (a) microanalysis with the use of a Transmission
Electron Microscope (TEM) and its analytical spectrometric
gears and (b) quantitative chemical analysis of each sulfide
(referred to as ‘‘Tai-Betsu analysis’’) with the application of
extracted residue and its dissolution technique. These two
methods are shown in detail in the first report.1)
3.
3.1
Results
Sulfide precipitation in ferrite region by analyzing
as-hot and as-annealed samples of 0.04%Ti-IF
Samples with three CT conditions ((5)640 C, (6)680 C
and (7)720 C) in hot-rolling are prepared, and these as-hot
samples and their as-annealed ones at either (8)770 C or
(9)830 C are compared.
First of all, from a microscopic point of view, we briefly
explain the distinction between as-hot and as-annealed
samples. Figure 2 shows typical TEM micrographs of
sulfides which are picked out from the as-hot (CT:
(6)680 C) and its corresponding as-annealed ((9)830 C)
sample. Sulfide in as-hot sample mainly consists of roundshaped or oval-shaped TiS. In contrast, sulfide in as-annealed
sample is a multi-precipitate-combined shape, in which TiS
is in the center and Ti4 C2 S2 grows outside partially in touch
with TiS. Cu-S is confirmable in some of the overlapping
area between TiS and Ti4 C2 S2 . The important point is that
annealing process in ferrite region changes TiS in as-hot
Fig. 2 Typical example of morphological change of sulfides from (a) ashot-rolled (CT: 680 C) to (b) as-annealed (830 C) in 0.04%Ti-IF.
electron beam
Ideal (but unfeasible)
Ti
S
(b)TiS
(Ti : S = 2 : 1)
(a)Ti4C2S2
(c)Al supporting film
(d)Al mesh
(a)Ti4C2S2
Electron
· divergence
Real (and probable)
· intrusion
Ti Ti : S≠2 : 1
Al S
(ghost)
Mixture of main
target & effect of
neighboring area
S
Ti
(b)TiS
(Ti : S = 1 : 1)
Al(ghost)
(c)Al supporting film
(d)Al mesh
Fig. 3 Schematic diagram to show that EDX spectrum in combined-sulfide
inevitably includes signals other than targeted area.
sample into a multi-precipitate-combined sulfide. It means
that the sulfide precipitation is reconfigured in ferrite region
during annealing.
Before explaining in details, the authors want to call
attention to the interpretation of the sulfide in TEM micrographs by using EDX spectra. In the following part, a multiprecipitate-combined sulfide is a focus of discussion. Such a
precipitate has a three-dimensional morphology, but TEM
micrographs are based on a two-dimensional-projected
view of a three-dimensional precipitate. Therefore, no
matter how fine electron beam is focused on a certain
small area of a multi-precipitate-combined sulfide, characteristic X-ray will be emitted not only from the targeted
area but also from the neighboring area, the supporting film
and the mesh. As shown in Fig. 3, when overlapping area
between (a) Ti4 C2 S2 and (b) TiS are analyzed in TEM-EDX
spectrum, the spectrum is mainly composed of (a) Ti4 C2 S2 ,
but it inevitably includes the effect of (b) TiS, (c) Al
supporting film and (d) Al mesh, because electron beam is
somewhat divergent and intrudes inside. In interpreting the
EDX spectrum, every individual signal should be separated
1372
Y. Ishiguro, T. Murayama, T. Fujita and K. Kuroda
(2)center
S Ti (3)
MnCu
(Al)
0
4
S Ti
(Al)
8 (keV)
0
(Al) (3)
(1)whole area
(keV)
4
8
S Ti
(Al)
Mn
0
8 (keV)
4
Ti
(magnified)
(Al) S
Mn Cu
0
4
8 (keV)
S Ti (5)
Mn
(Al)
Cu
Ti (4)
S Mn
(Al)
Cu
0
Fig. 4
4
8 (keV)
8 (keV)
4
0
S
0
8
0
Ti
TiS–side at overlapping area
(10)
0
(Al)
0
Fig. 5
(9)
8
Cu (9)
(Al)
0
S
Ti
Ti4C2S2
4
4
Remaining
Sulfur
(Al)
Solute sulfur does not exist in steel
Remaining
Sulfur
0
4
(8)(Al) S Ti Mn
(keV)
8
0
4
8 (keV)
Cu
8 (keV)
TiS
0
8 (keV)
4
8 (keV)
4
Ti
S
(Al)
Cu
8 (keV)
The possible case of only Ti4C2S2
at targeted area
Solute sulfur does not exist in steel
Ti4C2S2–side near overlapping area
Ti
Ti
S Ti Cu Sulfide–
former
(Al)
(keV)
(Al) S
4
S
8 (keV)
Cu
Detailed EDX spectra in as-annealed (830 C) in 0.04%Ti-IF.
Remaining
Sulfur
Ti
4
0
Ti
(Al) S (7)
8 (keV)
4
4
(Al)
8 (keV)
4
Cu
(Al) S Ti
0
(6)
0
(10)
The possible case of only TiS
at targeted area
Solute sulfur does not exist in steel
Cu
S
8
4
(Al) S Mn
Cu
8 (keV)
0
4
(5)
Cu
(Al) S Ti
0
Detailed EDX spectra in as-hot-rolled (CT: 680 C) in 0.04%Ti-IF.
0
(keV)
4
(4)
(2)center (1)whole area
S Ti
S Ti Cu
(Al)
(keV) (Al)
Ti
Sulfide–
former
S Ti Cu
Sulfide–
former
Ti4C2S2
0
4
8 (keV)
Cu
8 (keV)
Fig. 6 Evidence of Cu-S and MnS around overlapping area in multi-combined sulfide of (9) and (10) in Fig. 5.
into the effect of each precipitate and we should comprehend
what precipitate exists in the targeted spot including the
neighboring area.
Figures 4 and 5 show the details of Fig. 2, which are EDX
spectra with TEM micrographs in as-hot and as-annealed
0.04%Ti-IF samples, respectively. It should be noted that the
detailed EDX spectrum at each spot is obtained from the
small area of below 10 nm. Figure 4 shows that the sulfide in
the as-hot sample is mainly round-shaped or oval-shaped TiS,
but it is not composed only of TiS. The surface is not smooth
and is covered with partially-attached tiny-sized precipitate
or a thinly-coated frilled structure of MnS, Cu-S and
Ti4 C2 S2 . The reason is based on the judgment that Mn peak
and Cu peak come from MnS and Cu-S, although mentioned
in details later by using Figs. 5 and 6.
In contrast, Fig. 5 shows that the sulfide in the as-annealed
sample is a multi-precipitate-combined shape, in which TiS
is in the center, Ti4 C2 S2 is formed partially in touch with TiS,
and Cu-S and MnS are confirmable in some of the overlapping area. The morphological change of sulfide precipitation occurs in ferrite region during annealing.
From Cu peaks in EDX spectra, the authors judge the
existence of Cu-S in a multi-precipitate-combined sulfide in
annealed Ti-IF steel, because Cu peak is not from metallic
Cu. The proof exists indirectly in the EDX spectra shown in
Fig. 5. Roughly speaking, EDX spectra of TiS and Ti4 C2 S2
are ‘‘Ti-K : S-K = 1:1’’ and ‘‘Ti-K : S-K = 2:1’’ in peak
height, respectively, because these k-factors for EDX
quantitative analysis are almost the same25) and the ratio of
the peak heights approximately corresponds to the atomic
Sulfide Precipitation in Titanium-Added Steel with Residual Level of Copper (2) —Precipitation in Ferrite Region—
(a)Zeroloss
(b) S–L edge
1373
(c) Ti–L edge
100nm
100nm
100nm
(d) Mn–L edge
(e) Cu–L edge
100nm
100nm
Fig. 7 Energy-filtered EELS mappings of sulfide in as-annealed sample: (a) Zeroloss, (b) S, (c) Ti, (d) Mn and (e) Cu.
ratio. When EDX spectra in Fig. 5 are reviewed through such
a simple criterion, the existence of Cu-S comes out clearly.
The EDX spectrum of (2) indicates that TiS is in the center.
The spectra of the edge points of (3) and (7) show that bulgeshaped precipitates are Ti4 C2 S2 . The EDX spectra of (4) to
(6) and (8) to (10) are from the overlapping areas and they
also have Mn-peak and Cu-peak. The point of focus is S-peak
and Ti-peak and is discussed by using Figs. 5 and 6 to prove
that Cu peak is from Cu-S.
In the sulfur-richer case, especially in (6) and (10), where
are rather TiS-side at overlapping area, if Cu is ignored,
sulfur cannot be fully stabilized as Ti-related sulfide, whether
it is TiS or Ti4 C2 S2 . Since sulfur does not exist as solid
solution in steel, Cu can be regarded as a sulfide-former based
on the Cu-peak in EDX spectra. Another example of (9),
where is a bulge-typed Ti4 C2 S2 near overlapping area, is
presumed to be mainly based on Ti4 C2 S2 by analogy of the
similar cases of (3) and (7). Therefore, according to ‘‘Ti-K :
S-K = 2:1’’, a half height equivalent of the Ti-peak is
subtracted from the S-peak and the remaining S-peak should
be stabilized as other kinds of sulfide. In this case, the most
probable sulfide is Cu-S, which is stabilized by a trace level
of 0.01%Cu. The identification of MnS is also possible in the
same way, especially by using the EDX spectra of (4) and (8).
Consequently, the Cu-peak and Mn-peak in the EDX
spectrum can be regarded as ‘‘Cu from Cu-S’’ and ‘‘Mn from
MnS’’, respectively.
Figure 7 shows elemental mappings of a typical sulfide in
the as-annealed sample through energy-filtered EELS. Cu
and Mn are localized at overlapping area in a multiprecipitate-combined sulfide and these two elements do not
always exist in the same points. The localized Mn and Cu
indicate that MnS and Cu-S are formed in different places
from each other. These results also support the above EDXbased analyses in Figs. 5 and 6.
The next topic is the difference caused by the coiling
temperature in hot-rolling and the annealing temperature in
annealing. Figure 8 shows TEM micrographs of as-hot
samples (CT: 640, 680 and 720 C) and their corresponding
as-annealed samples (770 and 830 C). It should be noted that
‘‘CuS’’ in Fig. 8 means ‘‘Cu-S’’ in the context of this paper
and is not identified as ‘‘CuS’’. Figure 9 shows Tai-Betsu
analysis of each sulfide in all conditions of hot-rolling and
annealing.
As shown in Fig. 8, it is difficult to tell the difference
in morphology between CT conditions and also between
annealing temperature conditions, although the change from
as-hot to as-annealed, which is mentioned in Fig. 2, can be
confirmed in any conditions. The morphological characteristics of sulfides among CT conditions and the ones among
annealing conditions are regarded as almost the same except
for as-hot of ‘‘CT: 720 C’’. The size of the sulfide is 150 to
400 nm in both as-hot and as-annealed samples.
In contrast, as shown in Fig. 9, Tai-Betsu analysis highlights a clear distinction of sulfides among the three coiling
temperature (CT) conditions in hot-rolling as well as between
the two annealing temperatures. First of all, Sulfur is fully
stabilized as sulfides in all samples. ‘‘Sulfur in steel’’ almost
the same as the summation of ‘‘calculated sulfur’’ by using
the metallic elements of each sulfide-former, that is, the sum
total of ‘‘(b)S as MnS’’+‘‘(c)S as FeS’’+‘‘(e)S as CuS’’+‘‘(f)S as Ti4 C2 S2 ’’+‘‘(i)S as TiS’’, whose parenthetical
1374
Y. Ishiguro, T. Murayama, T. Fujita and K. Kuroda
As–Annealed
0.04%Ti–IF
(7)CT: 720°C (6)CT: 680°C (5)CT: 640°C
As–Hot
(8)
770°C
(9)
830°C
(8)
770°C
(9)
830°C
(8)
770°C
(9)
830°C
100
84 ppm Sulfur in steel
80
Cu8S5
MnS
60
Ti4C2S2
40
TiS
20
Annealed
Annealed
(9)830°C
(8)770°C
(7)CT : 720°C
As-Hot
(9)830°C
(8)770°C
(6)CT : 680°C
As-Hot
(9)830°C
(8)770°C
(5)CT : 640°C
0
As-Hot
Precipitated Sulfur as sulfides (ppm)
Fig. 8 TEM micrographs of as-hot and as-annealed of 0.04%Ti-IF. (Note that ‘‘CuS’’ means ‘‘Cu-S’’ in the context.)
Annealed
Fig. 9 Tai-Betsu analysis of each kind of sulfide of as-hot and as-annealed
in 0.04%Ti-IF.
alphabet is from Fig. 5 in the first paper.1) The gap, or
‘‘unextracted sulfur’’ and ‘‘missing sulfur’’, which is mentioned in the first paper (Fig. 2-Method-A),1) corresponds to
‘‘(e)S as Cu-S’’, which is stabilized by a trace level of
0.01%Cu. In regard to the general behavior of the change
form as-hot to as-annealed Ti-added steel, as-hot samples are
mainly composed of TiS and the small amounts of other
sulfides, such as Ti4 C2 S2 , MnS and Cu-S, are also confirmed.
In as-annealed samples, TiS content reduces, while Ti4 C2 S2 ,
MnS and Cu-S contents increase. As for the comparison
between different CT and annealing temperatures, the sulfide
precipitation can be regarded as almost the same among three
CT conditions, and, in contrast, the annealing-temperaturebased difference of sulfide precipitation lies in the ratio of
each sulfide, that is, the 830 C-annealed sample contains less
TiS and more Ti4 C2 S2 , MnS and Cu-S than the 770 Cannealed one. Considering the effect of CT in hot-rolling
together, the above annealing-related tendency occurs more
markedly at high CT than at low CT.
Sulfide Precipitation in Titanium-Added Steel with Residual Level of Copper (2) —Precipitation in Ferrite Region—
1375
Through the above experimental facts, the following
statements can be derived about sulfide precipitation in
ferrite region, together with the new knowledge of Cu-S
precipitation caused by a trace level of 0.01%Cu.
(1) All sulfur is stabilized as sulfides in as-hot and asannealed samples.
(2) The sulfide precipitation in Ti-added steel is reconfigured morphologically and compositionally in ferrite
region, although there have been mentioned except for
limited papers.21–24) It is true that Ti4 C2 S2 , MnS and
Cu-S are precipitated in ferrite region and that TiS is
dissolved (reacted) in ferrite region.
(3) Sulfur is stabilized mainly as TiS until hot-rolling.
(4) During annealing, already-precipitated TiS is morphologically and compositionally changed into a multiprecipitate-combined sulfide, in which TiS is in the
center, Ti4 C2 S2 grows as a bulge-shaped sulfide
partially attaching to TiS, and MnS and Cu-S are
located at some of overlapping area between TiS and
Ti4 C2 S2 , but these two sulfides are not always observed
at the same place.
(5) Such a morphological change proceeds more markedly
when Ti-added sample is annealed at higher temperature.
(6) The ‘‘gap’’ in the first paper1) (Fig. 2-Method-A)
corresponds to Cu-S precipitation induced by a trace
level of Cu. The strange phenomenon of ‘‘S-rich TiS’’ in
the first paper1) (Fig. 2-Method-B) is misinterpreted by
the traditional knowledge that all sulfides in Ti-added
steel are only composed of TiS, Ti4 C2 S2 and MnS and
that a residual level of Cu has been regarded as solute in
steel.
in relation to the whole sulfide precipitation along the above
our reported concepts, although the data in the third report
will be actually conclusive evidence.2) The focus of attention
is whether the existence of Cu-S in Fig. 9 is regarded as a
precipitation at annealing temperature (direct cause) or as an
eventual precipitation during cooling associated with the
increase of Ti4 C2 S2 (consequence).
4.
Eventually, the newly-proposed mechanism is based on a
simple idea of the existence of (re-)solute sulfur and its reprecipitation, which corresponds to a common concept in the
first paper1) and in non-Ti-added steel.3–20)
The first reaction occurs during heat treatment, which is
annealing in this paper. When Ti-added steel is heat-treated,
already-precipitated TiS is changed into ‘‘first Ti4 C2 S2 ’’ and
simultaneously ‘‘re-solute sulfur’’ comes out near ‘‘reacted
TiS’’, especially near overlapping area between TiS and
Ti4 C2 S2 in a multi-combined sulfide. The parameters of
the first reaction are temperature and time-length of heattreatment and the content of solute carbon. Based on Le
Chatelier’s law, the amount of each sulfide-former element,
which works in the second reaction, also acts as a parameter.
The second reaction is that ‘‘re-solute sulfur’’ is reprecipitated during heat treatment. There are three candidates
to stabilize ‘‘re-solute sulfur’’ as ‘‘second Ti4 C2 S2 ’’, MnS and
Cu-S. The parameters are the amount of ‘‘re-solute sulfur’’,
temperature and time-length of heat-treatment and the
amount of each sulfide-former element. Strictly saying, it
corresponds to the amount of an effective sulfide-precipitatable element.
It should be noted that ‘‘re-solute sulfur’’ is virtual
existence to explain the phenomena for the sake of
convenience. When ‘‘re-solute sulfur’’ comes out along the
above (1st) reaction, it will be immediately precipitated as
the most energetically-stable sulfide in accordance with
4.1
Discussions
Precipitation of copper sulfide in annealed Ti-added
steel
As mentioned above, it is true that Cu-S precipitation
occurs in ferrite region, but we cannot judge that the
precipitation of Cu-S proceeds more actively at 830 C than at
770 C. The amounts of Ti4 C2 S2 and Cu-S are larger in
830 C-annealed samples than in 770 C-annealed ones, both
of which are analyzed at room temperature (RT) after
annealed and then cooled to RT, but are not made in-situ
analyses at annealing temperature.
The important thing for interpretation is that every Cu-S
precipitation behavior should maintain consistency with a
common metallurgical principle regardless of whether Ti is
added to steel or not. It seems strange that each Cu-S
phenomenon is explained by using each different logic. Our
group has already reported in the first paper that the Cu-S
precipitation is formed not in austenite region but during
cooling in austenite-heat-treated Ti-added steel and the
mechanism is based on a simple idea that solute sulfur and
re-solute sulfur are precipitated as Cu-S during cooling.1)
Furthermore, a trace level of Cu acts as a sulfide-former in
non-Ti-added steel and it is evident that Cu-S can be formed
at below 750 C in ferrite region, even during WQ.3–20)
In the paper, the Cu-S behavior in ferrite-annealed Tiadded steel will be discussed in details in the following part
4.2
Mechanism of sulfide precipitation in Ti-added
steel
The following three important facts are picked out from
Figs. 2 and 4 through 9, and then the mechanism is derived.
(a) TiS-related reconfiguration occurs during annealing in
ferrite region, and is deeply connected to the precipitation of Ti4 C2 S2 , MnS and Cu-S.
(b) Almost all of Ti4 C2 S2 is formed partially in touch with
TiS.
(c) MnS and Cu-S are formed at some of overlapping area
in TiS-Ti4 C2 S2 combined sulfide. Each precipitated
place is different rather than the same points.
For the sake of convenience, the mechanism is divided into
the following two steps on the base of the morphological
changes of (a) to (c). The schematic diagram is shown in
Fig. 10.
[First reaction]
(1st)
4TiS þ 2[C] ! Ti4 C2 S2 (first) þ 2[S (re-solute)]
[Second reaction] (three possible candidates)
(2nd-1) [re-solute S] þ 4[Ti] þ 2[C] ! Ti4 C2 S2 (second)
(2nd-2) [re-solute S] þ [Mn] ! MnS
(2nd-3) [re-solute S] þ [Cu] ! Cu-S
1376
Y. Ishiguro, T. Murayama, T. Fujita and K. Kuroda
First reaction
first re–solute
first Ti4C2S2
4TiS 2[C]
Ti4C2S2
2[S]
“Re–solute sulfur” is virtual existence.
When it comes out, it is re–precipitated.
(It will not be observed at RT)
TiS
first
Ti4C2S2
first
Ti4C2S2
(2nd–1) 2[S]
(2nd–2) [S]
(2nd–3) [S]
Second reaction
4[Ti]
[Mn]
[Cu]
2[C]
[Ti], [C]
[Mn]
[Cu]
Ti4C2S2
MnS
Cu–S
TiS
[Ti], [C]
TiS
[Ti], [C]
Ti4C2S2
TiS
[Mn]
[Cu]
[Cu]
[Mn]
MnS
Cu–S
Cu–S
MnS
Ti4C2S2
Ti4C2S2 second
MnS
Cu–S
Fig. 10 Estimated mechanism of sulfide precipitation in ferrite region in Ti-added steel.
(1st)
[Sulfur]
already-precipitated
at hot-rolled
(5-1)Cu8S5
(4-1)MnS
(3-1)Ti4C2S2
Precipitated Sulfur as sulfides
(ppm)
thermal condition. Therefore, in our opinion, it is difficult to
find ‘‘re-solute sulfur’’ in the sample even with TEM as far as
we observe in room temperature.
The phenomena in Figs. 2 to 9 can be understood along
both the above mechanism and the theory of Cu-S precipitation at low temperature in non-Ti-added steel.3–20)
When heat-treatment is taken at higher temperature and/or
for longer time, (1) the first reaction will proceed more
rightward, that is, ‘‘first Ti4 C2 S2 ’’ and simultaneously ‘‘resolute sulfur’’ will increase. (2) The ‘‘re-solute sulfur’’ is reprecipitated as one of three candidates of sulfides among
which are most energetically-stable in accordance with
thermal conditions. ‘‘Second Ti4 C2 S2 ’’ or MnS are likely
precipitates. (3) Cu-S can be precipitated during cooing if
‘‘re-solute sulfur’’ still remains after the precipitation of
‘‘second Ti4 C2 S2 ’’ and MnS.
The reason why Cu-S increases in 830 C-annealed sample
can be interpreted by the following idea: The first reaction
proceeds more rightward at 830 C but annealing time is not
long enough to stabilize all of ‘‘re-solute sulfur’’ as either
‘‘first Ti4 C2 S2 ’’ or MnS. Therefore, much of ‘‘re-solute
sulfur’’ still remains and then re-precipitated Cu-S is high in
content. In the case of 780 C-annealed sample, the amount of
‘‘re-solute sulfur’’ is small and so Cu-S is low in content.
Also, the reason why MnS and Cu-S are localized at some
of overlapping area can be also explained along the above
mechanism: TiS-decomposion-based ‘‘re-solute sulfur’’ at
the first reaction in our proposed mechanism is estimated to
exist near overlapping area. The reason why MnS and Cu-S
rarely coexist can be also explained along the above
mechanism: The place where Mn already stabilizes ‘‘resolute sulfur’’ does not have any re-solute sulfur, and
therefore, Cu-S cannot be formed there.
Moreover, the authors apply the above mechanism to more
specific examples by citing examples of as-hot (CT: 680 C)
Re-precipitated
through annealing
(5-2)Cu8S5
60
(4-2)MnS
(2)Reacted TiS
(3-2)Ti4C2S2
40
20
(1)Un-reacted TiS
0
As-Hot
Estimation:
(a)Most of Ti4C2S2 comes from
“first Ti4C2S2”.
(b)MnS and Cu-S are based on
second reaction
830°C-annealed
: already-precipitated sulfide
Fig. 11 Detailed analysis of as-hot (CT: 680 C) and 830 C-annealed
0.04%Ti-IF steel through newly-proposed mechanism.
and its 830 C-annealed samples. Figure 11 is a picked-out
comparison between the above two conditions from Fig. 9.
First of all, TiS in annealed sample can be regarded as (1) unreacted TiS, and therefore, ‘‘TiS in as-hot’’ minus ‘‘(1)TiS in
annealed sample’’ corresponds to (2) reacted TiS at the first
reaction in our mechanism. It can be estimated that the
already-precipitated (3-1)Ti4 C2 S2 and (4-1)MnS keep intact
during annealing but that (5-1)Cu-S may be dissolved away
during annealing, because it is thought that Cu-S is formed at
low temperature in ferrite region and cannot exist at
830 C.1–20)
By annealing, re-precipitated sulfides correspond to
(3-2)Ti4 C2 S2 , (4-2)MnS and (5-2)Cu-S. Along the first
reaction in our mechanism, the behavior of sulfur shows
the following features.
4TiS þ 2[C] ! Ti4 C2 S2 (first) þ 2[S (re-solute)]
‘‘4S as TiS’’
‘‘2S as Ti4 C2 S2 ’’ ‘‘2S as solid solution’’
Sulfide Precipitation in Titanium-Added Steel with Residual Level of Copper (2) —Precipitation in Ferrite Region—
The half amount of ‘‘(2)sulfur as reacted TiS’’ is changed
into ‘‘(3-2)the sulfur as first Ti4 C2 S2 ’’ and the same amount
of sulfur comes out as ‘‘re-solute sulfur’’.
In this case, the half amount of ‘‘(2)sulfur as reacted TiS’’
corresponds roughly to (3-2)Ti4 C2 S2 . It means that the
second reaction of (2nd-1)Ti4 C2 S2 does not occur. After the
second reaction of (2nd-2)MnS proceeds slightly, much of
‘‘re-solute sulfur’’ still remain and it is finally stabilized as
(2nd-3)Cu-S. After Ti4 C2 S2 and MnS are formed, Cu-S is
estimated to be form during cooling process. This mechanism
will be reinforced again in the next third paper by using longtime-annealing Ti-IF steel,2) although the Cu-S precipitation
during cooling is not completely supported by any experimental evidence so far. However, the Cu-S precipitation in
ferrite-annealed Ti-added steel keeps consistency with the
one in non-Ti-added steel.3–20)
Additionally, if this phenomenon is reviewed from a
different viewpoint, the authors can point out that sulfide
precipitation in real-produced Ti-added steel (annealed Tiadded steel product) can be regarded as non-equilibrium,
because TiS still remains and can react into Ti4 C2 S2 . If it is
kept at high temperature and/or for long time, the first and
second reactions in our newly-proposed mechanism will
proceed up to such equilibrium state that TiS is completely
changed into other sulfides. In other words, recent steel
process, which tends to focus on high efficient process of
rapid heating, short-time-heat-treatment and rapid cooling,
can be regarded as non-equilibrated process (intermediary
stage) from the viewpoint of sulfide reconfiguration, although
advanced steel products can be surely produced with good
quality.
Solute carbon is required for (1st) and (2nd-1) reactions,
because the sample is from Ti-IF steel, whose interstitial
elements, such as carbon and nitrogen, should be completely
stabilized as carbides and nitrides in principle. Please note
that ‘‘annealed Ti-IF steel’’ is ‘‘interstitial-element-free’’ to
stabilize such elements, but ‘‘Ti-IF steel before annealing’’,
or as-hot-rolled sample, is not always ‘‘interstitial-elementfree’’ and some as-hot Ti-IF sample, especially recent steel,
really includes solute carbon. The detailed information will
be also mentioned in the third paper from a mechanical
viewpoint of aging index.2)
4.3
Consistency of mechanism with past studies of Cu-S
and Cu diffusion
The proposed theory includes a new viewpoint of sulfide
precipitation. If it is a perfect theory, the new theory has to
explain consistently all the phenomena that the authors are
facing, and also it has to be able to explain all the phenomena
in the past papers which have been interpreted through a
traditional concept. In addition, it will be able to predict
some related phenomena.
As mentioned above, our mechanism is based on the existence of (re-)solute sulfur and its re-precipitation. Re-solute
sulfur comes out to matrix in the first reaction and then it is restabilized as other sulfides. The Cu-S precipitation occurs at
the last stage after the precipitations of Ti4 C2 S2 and MnS are
finished. Only when re-solute sulfur exists, Cu-S can be precipitated. In other words, if all of re-solute sulfur is completely
stabilized, Cu-S cannot be formed. In regard to the temper-
1377
ature of Cu-S precipitation, it can be predicted that Cu-S is
formed at low temperature in ferrite region, even during WQ,
by analogy with the data in non-Ti-added steel.3–20)
However, our theory includes two delicate points: (1) the
consistency with past studies on Cu-S and (2) the diffusion of
Cu in steel.
In regard to the first consistency problem, most of past
studies26–43,48) have reported that Cu-S is formed at temperatures as high as steel is melted in non-Ti-added steel,
although the precipitation of Cu-S has long been mentioned
in hot-shortness,26–30) welding,31–35) HSLA steel,36–39) magnetic steel sheet,40–43) steel codes in tires44–47) and strip-cast
sheet steel.48–53) In contrast, our past papers have already
reported consistently that the Cu-S precipitation occurs at
low temperature in ferrite region in non-Ti-added steel, even
during WQ.3,9–15) Also, it is reported in this paper that Cu-S in
Ti-added steel is formed in ferrite region after Ti4 C2 S2 and
MnS are precipitated in ferrite region. It is true that these past
concepts and our theory seem contradictory to each other, but
we think that these behaviors have the same root and that the
discrepancy is based on the interpretation of Cu-S in WQ
samples.
In our opinion, the WQ-based Cu-S phenomenon is a kind
of ‘‘booby-trap’’, leading to an incorrect interpretation that
Cu-S can be really precipitated at the temperature prior
to rapid cooling. The WQ method has been used for a
convenient technique to ‘‘freeze’’ precipitation prior to WQ.
However, the authors want to present a different viewpoint
that Cu-S can be formed even during WQ.1–20) This means
that WQ cannot ‘‘freeze’’ the Cu-S precipitation and it is
formed at a certain temperature during WQ. We think that the
data on Cu-S precipitations in past studies should be reevaluated and it would be possible to interpret Cu-S in the
past data as formed at low temperature in ferrite region.
Such a new challenging concept can be also supported
experimentally in various aspects by our past reports.3–20)
Therefore, we think that the first consistency problem has
gone away.
As for the second problem, the authors have not yet had a
consistent answer. In our opinion, the experimental data in
three consecutive papers1,2) would be suggestive evidence
that the diffusion of Cu is much faster than widely accepted
data.54) Through a series of our experimental results, the
precipitation of Cu-S can be interpreted as formed at low
temperature. The diffusion of Cu and sulfur is required for the
Cu-S precipitation at the range of the temperature to form it.
Since Cu is a substitutional element, the Cu diffusion should
be almost the same as Fe self-diffusion, and it should be
predominantly governed by a vacancy mechanism such that a
substitutional element can jump to a neighboring site with a
high probability when it is a vacancy site and the substitutional element has an internal energy greater than the
activation energy. Although the experimental facts are
completely different from the traditional concept of diffusion,
the authors cannot think of anything but the mechanism that
Cu diffuse very fast. In the different scientific field of
irradiation-induced embrittlement in material for atomic
reactors, some papers have reported that Cu diffusion is about
two orders of magnitude faster than Fe self-diffusion,55–59)
although these experiments are carried out by using Fe-1%Cu
1378
Y. Ishiguro, T. Murayama, T. Fujita and K. Kuroda
or Fe/Cu/Fe diffusion pairs heat-treated at over 1000 C.
They interpret the diffusion through Self-Interstitial-Atom
(SIA) mechanism. The SIA diffusion of Fe itself has been
often mentioned as well.60)
In addition, another strange phenomenon lies in the
estimated virtual existence of ‘‘re-solute sulfur’’. According
to our mechanism, as shown in Fig. 10, Cu moves to the ‘‘resolute sulfur’’ to be stabilized as Cu-S. Although sulfur is an
interstitial element and it should diffuse more rapidly than
substitutional Cu, ‘‘re-solute sulfur’’ stays near ‘‘reacted TiS’’
and is waiting for Cu there. As soon as re-solute sulfur comes
out to matrix by TiS-based first reaction in Fig. 10, it is
thought that Cu stabilizes it as Cu-S and the re-solute sulfur
does not enough time to diffuse out of there.
5.
Summary
The authors provide the following new concepts of sulfide
precipitation in ferrite region in Ti-added steel, especially Tistabilized IF steel.
(1) Even a trace level of 0.01%Cu can act as a sulfide-former
in ferrite region in Ti-added steel. It is presumed that Cu-S is
formed at temperatures lower than Ti4 C2 S2 and MnS.
(2) As in austenite region is, the sulfide precipitation in
ferrite region in Ti-added steel is interpreted through a simple
idea of the existence of re-solute sulfur and its re-precipitation.
(3) It is thought that the reconfiguration of sulfide precipitation occurs in ferrite region in Ti-added steel along the
following our proposed mechanism.
[First reaction]
(1st)
4TiS þ 2[C] ! Ti4 C2 S2 (first) þ 2[S (re-solute)]
[Second reaction] (three possible candidates)
(2nd-1) [re-solute S] þ 4[Ti] þ 2[C] ! Ti4 C2 S2 (second)
(2nd-2) [re-solute S] þ [Mn] ! MnS
(2nd-3) [re-solute S] þ [Cu] ! Cu-S
(4) Through annealing in ferrite region, already-precipitated
TiS, which is formed in hot-rolling, is changed into Ti4 C2 S2
and ‘‘re-solute sulfur’’ comes out to matrix. The ‘‘re-solute
sulfur’’ is re-precipitated as Ti4 C2 S2 , MnS and Cu-S. The reprecipitation depends on annealing conditions (temperature
and time) and the amounts of solute carbon and sulfideformer elements. ‘‘Second Ti4 C2 S2 ’’ and/or MnS are reprecipitated ahead of Cu-S. When ‘‘re-solute sulfur’’ still
remains after the precipitation of these two sulfides is
finished, it can be stabilized as Cu-S.
Acknowledgements
The authors would like to thank Dr. Kei Sakata (JFE Steel
Corporation) for the opportunity to conduct research on such
a pure-metallurgy-truth-seeking theme and for valuable
discussion. We would also like to thank Dr. Yoshihiro
Hosoya (JFE Steel Corp.) for long support of this research
and helpful communications.
We must thank Mr. Yoshikazu Kawabata, Dr. Mitsuo
Kimura, Mr. Takatoshi Okabe, Mr. Shunsuke Toyoda and
other members in Tubular Products & Casting Research
Department, Steel Research Laboratory, JFE Steel Corp. for
vigorous discussions at meetings.
Special thanks are given to Mr. Naruhito Shinbu, Mr.
Toshiaki Noguchi (JFE Techno-Research Co.), Mr. Kaoru
Murata (formerly Kokan Keisoku Co., one of the predecessors of JFE Tech., presently working at FEI Japan), Mr. Eiji
Okunishi (formerly Kokan Keisoku Co., presently working at
JEOL) for preparing TEM replica samples.
REFERENCES
1) Y. Ishiguro, T. Murayama, T. Fujita and K. Kuroda: Mater. Trans. 50
(2009) 1360–1369.
2) Y. Ishiguro, T. Murayama, T. Fujita and K. Kuroda: Mater. Trans. 50
(2009) 1380–1389.
3) Y. Ishiguro, T. Murayama, K. Sato and K. Araki: CAMP-ISIJ. 10
(1997) 703.
4) Y. Ishiguro, K. Yamada and K. Sato: CAMP-ISIJ. 10 (1997) 1466.
5) Y. Ishiguro, K. Yamada and K. Sato: Materia Jpn. 38 (1999) 383.
6) Y. Ishiguro, T. Murayama, K. Sato and K. Yamada: Chugoku-ShikokuBranch Joint Workshop of ISIJ No. 43 & JIM No. 40, (2000) pp. 37.
7) Y. Ishiguro, K. Yamada and K. Sato: Materia Japan 39 (2000) 946.
8) Y. Ishiguro, K. Yamada and K. Sato: Practical Electron Microscopy
and Its Application to Materials ed. by K. Maruyama, (The Iron and
Steel Institute of Japan & The Japan Institute of Metals, Sendai, 2002)
pp. 115.
9) Y. Ishiguro, T. Murayama and K. Sato: Mater. Trans. 46 (2005) 769–
778.
10) Y. Ishiguro, T. Murayama and K. Sato: Mater. Sci. Forum 561–565
(2007) 2029–2032.
11) Y. Ishiguro and K. Kuroda: CAMP-ISIJ. 21 (2008) 571.
12) Y. Ishiguro and K. Kuroda: AMTC Lett. 1 (2008) 128–129.
13) Y. Ishiguro and K. Kuroda: CAMP-ISIJ. 21 (2008) 1534–1537.
14) Y. Ishiguro and K. Kuroda: CAMP-ISIJ. 21 (2008) 863–866.
15) Y. Ishiguro and K. Kuroda: The 9th Asia-Pacific Microscopy Conf.,
(APMC-9, 2008) pp. 558–559. (Korean J. Microscopy 38 supplement
(2008) 558–559)
16) Y. Ishiguro, T. Murayama, A. Chino, K. Sato, Y. Shima, A. Kido and
M. Morita: Mech. Work. Steel Process. 35 (1997) 255–264.
17) Y. Ishiguro and T. Murayama: CAMP-ISIJ. 20 (2007) 436.
18) Y. Ishiguro, T. Fujita and K. Kuroda: CAMP-ISIJ. 20 (2007).
19) Y. Ishiguro and K. Kuroda: CAMP-ISIJ. 21 (2008) 572.
20) Y. Ishiguro and K. Kuroda: CAMP-ISIJ. 21 (2008) 213.
21) K. Yamada and M. Niikura: CAMP-ISIJ. 6 (1993) 1710.
22) H. Murakami, S. Sanagi and T. Senuma: CAMP-ISIJ. 7 (1994) 1835.
23) K. Yamada and M. Niikura: Tetsu-to-Hagane 81 (1995) 197–202.
24) H. Murakami: Tetsu-to-Hagane 89 (2003) 544–551.
25) G. Cliff and G. W. Lorimer: J. Microscopy 103 (1975) 203–207.
26) H. Morrogh: J. Iron Steel Inst. 154 (1946) 399–414.
27) P. A. Melford: J. Iron Steel Inst. 200 (1962) 290–299.
28) P. A. Melford: Phil. Trans. R. Soc. Lond. A 295 (1980) 89–103.
29) C. Nagasaki, A. Aizawa and J. Kihara: Trans. ISIJ 27 (1987) 506–512.
30) B. Mintz, R. Abushosha and D. V. Crowther: Mater. Sci. Technol. 11
(1995) 474–481.
31) J. E. Harbottle and S. B. Fisher: Nature 299 (1982) 139–140.
32) E. S. Kayali, J. M. Corbett and H. W. Kerr: J. Mater. Sci. Lett. 2 (1983)
123–128.
33) S. G. Court and G. Pollard: J. Mater. Sci. Lett. 4 (1985) 427–430.
34) S. B. Roberts and R. Fenn: Weld. Rev. 5 (1986) 6–7.
35) J. M. Dowling, J. M. Corbett and H. W. Kerr: Metall. Trans. A 17
(1986) 1611–1623.
36) E. M. Cox, M. C. Bachelder, N. H. Nachtrieb and A. S. Skapski: J. Met.
1 (1949) 27–31.
37) Z. Chen, R. Cochrane and M. H. Loretto: Proc. Conf. Anal. Electron.
Microscopy (1984) pp. 243–246.
38) H. Kejian and T. N. Baker: Mater. Sci. Technol. 8 (1992) 1082–1089.
39) S. Okada, T. Obara and K. Tsunoyama: Tetsu-to-Hagane 73 (1987)
S516.
Sulfide Precipitation in Titanium-Added Steel with Residual Level of Copper (2) —Precipitation in Ferrite Region—
40) N. F. Dubrov, I. A. Gorlach and M. V. Lyasko: Stal. Eng. 7 (1960) 517–
519.
41) V. F. Nosach and V. YA. Gol’dshteyn: Fiz. Metall. Metalloved. 44
(1977) 288–293.
42) T. Shimazu, K. Chikuma, T. Sakai and M. Tanino: Tetsu-to-Hagane 70
(1984) S568.
43) T. Sakai, T. Sakai, K. Chikuma, M. Tanino and M. Matsuo: Tetsu-toHagane 70 (1984) 2049–2056.
44) H. Hirakawa and Y. Ishikawa: Nihon Gomu Kyoukaishi 45 (1972)
920–931.
45) M. Ashida: Nihon Gomu Kyoukaishi 57 (1984) 501–508.
46) G. E. Hammer: J. Vac. Sci. Technol. A 19 (2001) 2846–2850.
47) T. Yasunaga, T. Nakayama, K. Okumura, T. Yamauchi, T. Shimizu and
M. Doi: Kobe Steel Eng. Rep. 54 (2004) 2–6.
48) A. Guillet E. Es-Sadiqi, G. L’Espérance and F. G. Hamel: ISIJ Int. 36
(1996) 1190–1198.
49) Y. Kobayahashi and K. Nagai: Materia Japan 43 (2004) 730–736.
50) Z. Liu, Y. Kobayahashi and K. Nagai: Mater. Trans. 45 (2004) 479–
487.
1379
51) Y. Kobayahashi and K. Nagai: Mater. Trans. 46 (2005) 26–33.
52) Y. Kobayahashi, Z. Liu and K. Nagai: J. Jpn. Inst. Met. 70 (2006) 440–
446.
53) Z. Liu, Y. Kobayahashi, J. Yang, K. Nagai and M. Kuwabara: Mater.
Trans. 47 (2006) 2312–2320.
54) Metal Databook rev.4, ed. by The Japan Institute of Metals, (Maruzen,
Tokyo, Sendai, 2004) Chap. 1.2.14, pp. 20–25.
55) S. Yoshida, N. Takuchi, Y. Kawanami, M. Kajiwara and T. Mori:
CAMP-ISIJ. 8 (1995) 1382.
56) Y. Kawanami, M. Kajiwara and T. Mori: CAMP-ISIJ. 8 (1995) 1383.
57) M. Kajiwara and T. Mori: CAMP-ISIJ. 10 (1997) 441.
58) J. Marian, B. D. Wirth, J. M. Perlado, G. R. Odette and T. Diaz de la
Rubia: Phys. Rev. B 64 (2001) 094303.
59) F. Sasaki, Z. Tang, Y. Nagai and M. Hasegawa: The 108th Conf.
Institute for Materials Research, Tohoku University, (2004) poster #34.
(http://www.imr.tohoku.ac.jp/jpn/event/kouen/pdf/2004/posterses10806.
pdf).
60) K. Kusunoki: Mater. Trans. 47 (2006) 1906–1909.