2019 Volume 60 Issue 10 Pages 2086-2095
Interphase precipitation is a phenomenon in which precipitates are generated on interphase boundaries during phase transformation, creating carbides having a form like bands or fibers in micro-alloyed steels. In isothermal transformation, a lower transformation temperature reduces the carbide diameter and band spacing, and a higher cooling rate reduces them in the continuous cooling process.
Various interphase precipitation mechanisms have been suggested, and some models for the ledge mechanism attempted to explain the regular band spacing in steels containing carbide-forming elements quantitatively. Recently, as the orientation relationship in the grains of the interphase boundary is not consistent with the ledge mechanism, three-dimensional interface structures have been suggested to explain the experimental results of observation by transmission electron microscopy in low carbon steel. The newly-proposed interphase structure model may explain both the ledge and quasi-ledge mechanisms.
Steels with tensile strength of more than 590 MPa, which are manufactured by using interphase precipitated carbides, have been developed and used practically not only in plate and sheet products but also in forged products to improve formability. In steels consisting of ferrite and a second hard phase, interphase precipitated carbides are used to realize high local ductility and to reduce the difference of hardness between the two phases. Ferritic steels strengthened by nanometer-sized carbides are developed to achieve excellent formability, realizing precipitation-strengthening of more than 300 MPa.
This Paper was Originally Published in Japanese in J. Japan Inst. Met. Mater. 81 (2017) 447–457. In order to provide a more precise explanation of the interphase precipitation phenomenon, the entire article was revised and recently published literature was cited. Figures 1 and 2 in the original paper were omitted to simplify the article.
Interphase precipitation is a well-known phenomenon that often appears in microalloyed steels. The arrangement of interphase precipitated carbides is observed to be regularly spaced bands or fibers aligned in one direction1) when thin foils are tiled to a certain angle in transmission electron microscopy. In carbon steels containing stronger carbide-forming elements, the carbide band spacing decreases with lowering of the transformation temperature. The addition of alloy elements which retard the transformation also affects the carbide band spacing.2)
The characteristic carbide bands of interphase precipitated NbC in Nb-bearing steel were first reported by Morrison3) in 1963, which was followed by the discovery of interphase precipitation of VC, TiC, (Ti,V)C and other carbides. Morrison proposed that carbides might precipitate on dislocations in ferrite introduced by the austenite-ferrite transformation. After Davenport et al.4) proposed that the regularly spaced bands of VC in ferrite were originated by precipitation on austenite-ferrite boundary during a transformation, the phenomenon came to be widely known as interphase precipitation. The carbide bands parallel to the boundary indicated that carbide nucleation had taken place periodically at the boundary. In 1974, Campbell et al.5) succeeded in taking micrographs showing that regularly spaced planar bands occurred along the planar boundary between austenite and ferrite in 0.2%C–12%Cr steel, while risers were free from carbide. This provided strong evidence that the regularly spaced bands were generated on the terrace in the austenite-ferrite interfaces. After the success of the experiment by Davenport et al., many kinds of carbide bands were observed in this manner, and several mechanisms of interphase precipitation were suggested. Moreover, since regularly spaced copper bands were also observed,6) interphase precipitation came to be considered a general precipitation phenomenon. Today, interphase precipitation in metals with phase transformation is widely known, and application to the production of high performance materials is strongly expected.7,8)
At the end of the 1980s, a ferritic-pearlitic plate steel, in which ferrite was strengthened by interphase precipitated carbides, was developed.9) In the 1990s, a ferritic-pearlitic hot-rolled high strength steel sheet strengthened by TiC was also developed in the tensile strength 590 MPa grade.10) Subsequently, a ferritic-martensitic steel sheet strengthened by interphase precipitated carbides was practically applied in the tensile strength 780 MPa grade.11) Until 1995, in the ferritic-martensitic and ferritic-bainitic steel sheets, the interphase precipitated carbides in ferrite were used to reduce the difference of hardness between the ferrite phase and the second hard phase (martensite or bainite) in order to improve stretch-flange formability.12)
In 2004, a ferritic steel sheet with interphase precipitated carbides having a diameter of 3 nm was successfully developed to realize high stretch-flange formability.13) In this steel sheet, nanometer-sized carbides were applied to strengthen ferritic steel with tensile strength of more than 780 MPa.
The following overview explains the phenomenon of interphase precipitation and its application to practical steel products, as many experimental results have been reported in this connection. The unit of mass% is used for the compositions in this article.
Figure 1 shows a transmission electron micrograph of interphase precipitated (Ti,Mo)C in Ti, Mo-bearing low carbon ferritic steel. Regularly spaced bands of fine carbides are one of the attractive microstructures in steels, and interface precipitated carbides are often observed in practical steel products. Table 1 shows a list of the interphase precipitated compounds reported in the literature.

Transmission electron micrograph showing interphase precipitated (Ti,Mo)C in low carbon steel.

Both simple carbides such as NbC, TiC, VC, Mo2C, Fe3C and complex carbides such as (Ti,Mo)C, (Ti, Nb)C are able to form interphase precipitates. Interphase precipitated NbC seems to have been discovered earlier than the other carbides since Nb was often added to TMCP (Thermomechanical Control Process) steels. Subsequently, interphase precipitated carbides were discovered one after another. Interphase precipitation frequently occurs at temperatures near or just above the ferrite bay in the TTT diagram.22) Interphase precipitated Cr23C6 was also easily observed since a slow interface migration rate causes carbides to coarsen in stainless steels.5) Interphase precipitated carbides have often been observed in steels containing strong carbide formers such as Nb, Ti, V or Cr, whereas fiber-like carbides are often observed in steels containing weak carbide formers such as Mo.23) However, the experimental result that the carbide morphology can change abruptly in one ferrite grain5,24) allows the mobility of the interphase between austenite and ferrite to affect the carbide morphology.
2.2 Band spacingAccording to the previous literature, band spacing becomes smaller when transformation takes place at a lower temperature. Figure 2 shows the rearranged data from the previous literature. Although these data were collected in different research organizations, an increased carbon content resulted in a smaller band spacing independently of the kind of carbide formers. This suggests that the state of the austenite-ferrite interface during transformation is strongly associated with interphase precipitation. Since the super ledge height is reduced by rapid migration of the interface,28) low temperature transformation causes a short band spacing. A high carbon content also seems to reduce band spacing. These reports suggested that the high transformation driving force calculated under the NPLE mode could provide fine and high density dispersion of VC in non-KS ferrite oriented to prior austenite grains.29) Recently, measurement of the ferrite grain orientation against prior austenite and VC detection by 3D-AP were carried out for V-bearing medium carbon steel. The results indicated that the ferrite grain had a non-KS orientation to the prior austenite grains.30) In addition, band spacing changed as a function of the distance from the ferrite grain boundary.31) These results suggested that interphase precipitation was strongly affected by transformation conditions. However, a further analysis of interphase precipitation seems to be necessary for quantitative prediction of band spacing.

Carbide band spacing of isothermally transformed low carbon steels.
Carbide band spacing data for continuous cooling after hot rolling were collected by Gray et al.32) Figure 3 shows the relationship between the cooling rate and band spacing. Rapid cooling reduces the mean of band spacing and increases its dispersion. These results can be attributed to the lowering of the transformation temperature by supercooling or the temperature distribution. The data collected from original papers are listed in Table 2 and are shown in the meter-kilogram-second system.

Carbide band spacing in continuously cooled low carbon steels.32)

Figure 4 shows the diameter of TiC or NbC as a function of the isothermal transformation temperature.37) The carbide diameter becomes finer as the temperature decreases and converges to around 3 nm under 650°C. Actually, the carbide shape is a thin cuboid since anisotropy depends on coherency of the interface between the carbide and ferrite matrix. Figure 5 shows a transmission electron inverse FFT image of a nanometer-sized (Ti, Mo)C carbide in the ferrite matrix. The larger boundaries are coherent with the ferrite matrix, having the Baker-Nutting relationship.38) The carbide diameters obtained in the continuous cooling process are also listed in Table 2 as a reference.

Carbide diameter of isothermally transformed microalloyed low carbon steels.37)

Transmission electron micrograph showing inverse FFT image of (Ti,Mo)C in low carbon ferritic steel.
The typical mechanisms of interphase precipitation which have been suggested are listed in Table 3.

The carbides and interface between austenite and ferrite are shown in Fig. 6.32) Carbon is scavenged on the austenite side in the interface, retarding interface migration. When carbides precipitate on the interface as a result of diffusion of a carbide-forming element, migration of the interface restarts. Repetition of the process of migration and precipitation creates straight carbide bands parallel to the interface.

Gray model explaining interface precipitation.32)
Figure 74) shows a schematic illustration of the carbon concentration at the interface in the carbide band formation process. Consumption of concentrated carbon by carbide precipitation restarts the migration of the interface that had been stopped by carbon solute drag. Recently, a new model combining this model with super ledge formation has been suggested to explain carbon band formation quantitatively.39,40) The interface between carbides migrates, forming super ledges, which then combine with each other, leaving the carbides.

Movement of austenite-ferrite interface controlled by carbon concentration after Davenport;4) Cγ; Equilibrium carbon concentration in γ, Cα; Equilibrium carbon concentration in α, $C_{\gamma }^{\gamma - \alpha }$; Carbon concentration on γ side interface when interface stops, $C_{\gamma }^{\gamma \textit{-precipitate}}$; Equilibrium carbon concentration on γ-carbide interface.
Although these mechanisms can explain carbide band formation, they do not agree with the experimental results for 0.2%C–12%Cr steel.5) On the other hand, the ledge mechanism described in the next section can explain the carbide band morphology and the results of observation of interphase precipitation.
3.2 Ledge mechanismFigure 841) shows a schematic illustration of a ledge mechanism. Carbides precipitate on the terraces of the coherent austenite-ferrite interface, and the band spacing is equal to the super ledge height of the interface. The ledge mechanism, which can explain the formation of the straight carbide bands, has been widely supported since carbide precipitation on the terraces of the interface was directly observed in 0.2%C–12%Cr steel by transmission electron microscopy. Recently, considering diffusion of carbide formers along the super ledge, a numeral model in which segregation and solute drag are embedded has also been suggested.42–44) However, the ledge mechanism cannot explain the overwhelming experimental results that the carbide bands are curved.

Schematic illustration of ledge mechanism.41)
Figure 9 shows a schematic diagram of a bowing mechanism. The straight austenite-ferrite interfaces bow out toward the direction of interface migration. The bowed interface is incoherent and has high energy. These high energy curved boundaries migrate again when the immobilized interfaces escape from carbide pinning.45)

Schematic illustration of bowing mechanism.
Figure 10 shows an illustration of a quasi-ledge mechanism46) which can explain the curved bands with a constant band spacing. The carbide band spacing is the minimum carbide nucleation spacing based on the classical nucleation theory.

Schematic illustration of quasi-ledge mechanism.46)
However, even in the same steel sheet, the band morphology changes with the transformation temperature and location in a ferrite grain.47,48) Miyamoto et al.49) analyzed the orientation relationship between austenite and pro-eutectoid ferrite, observing the interphase precipitation morphology in ferrite. The EBSD technique makes it possible to obtain the orientation of austenite that transformed to martensite after quenching to room temperature. The Miyamoto group showed that straight carbide bands were observed at the non-KS interface and proposed that the ledge structure can be extended to the non-KS orientation interfaces, which partially consist of coherent and less mobile interfaces. The small ledges on the surface can explain both the curved and straight band formations. Yen et al.50) also suggested three-dimensional interface structures to explain the band morphologies based on high resolution TEM observation.
Although the conditions that determine whether the ledge mechanism or the quasi-ledge mechanism occurs have not yet been clarified, the morphology of the carbide bands is undoubtedly associated with the energy of the interface between austenite and ferrite.
3.4 Eutectoid mechanismFigure 1151) shows a schematic illustration of a eutectoid mechanism. The direction of carbide growth is parallel to the direction of ferrite growth. The carbides precipitate on the edge of the ferrite allotriomorph, which causes the ferrite not to migrate in the vertical direction of the interface. This model also treats carbide fibers. As illustrated in Fig. 11, fiber carbides precipitate when the angle θ1 between the ferrite-carbide interface and the austenite-ferrite interface is equal to the angle θ2 between the austenite-carbide interface and the austenite-ferrite interface, and carbide bands are generated when θ2 is less than θ1. While the eutectoid mechanism can explain which carbide becomes particle or fiber, it cannot predict the change in the band spacing depending on the transformation temperature and chemical compositions.

Schematic illustration of eutectoid mechanism.51)
In a solute-drag nucleation model, after migration of the austenite-ferrite interface stops due to the concentration of carbon, carbide formation on the interface reduces the solute drag effect, and the increased mobility allows interface migration. In this model, the solute drag effects stops interface movement again when carbon concentrates on the interface. The repetition of this sequence produces carbide bands with a constant spacing, as shown in Fig. 12.52) The solute drag nucleation model can successfully explain the carbide shape changing from the band to the fiber with the lowered mobility of the interface between ferrite and austenite. This model, however, cannot explain the change of the carbide shape depending on the transformation temperature and chemical compositions since it does not treat the interface structure in detail.

Schematic illustration of solute-drag mechanism.52)
Recently, an improved solute-drag nucleation model was proposed, in which the austenite-ferrite interface velocity is associated with solute drag. The carbide band spacing decreases due to the distance from the ferrite nucleation point.49)
3.6 Solute-depletion modelFigure 1353) shows a schematic illustration of a solute-depletion model. While the interface begins to move after carbide precipitation in the same manner as in the solute-drag nucleation model, the interface migrates to the carbide former-depleted area. The solute-depletion model also cannot predict the change in the carbide band spacing caused by the change of the transformation temperature and chemical compositions.

Schematic illustration of solute-depletion mechanism.53)

Change in balance of tensile strength and elongation depending on amounts of Ti and Mn.11)
As described above, many models have attempted to explain interphase precipitation, but no model seems to provide a perfect explanation. Moreover, there may be a possibility of the simultaneous existence of multiple models, depending on the carbon content or transformation temperature.
A hot-rolled high strength ferritic-pearlitic steel with tensile strength of 590 MPa was developed by strengthening the ferrite matrix with interphase precipitated TiC. The composition of the steel was 0.06%C–0.25%Si–0.5%Mn–0.2%Ti. The coiling temperature was controlled to obtain the maximum strength. The developed steel was applied to automotive suspension parts and wheels.54)
In the 1990s, a general-use high strength steel with interface precipitated carbides was developed. For the precipitation of TiC on the Run-Out-Table after hot-rolling, manganese, which changes the austenite-ferrite transformation temperature, was added to control tensile strength and elongation. For example, manganese was added to 0.07%C–0.1%Ti steel to maximize TS×EL,10) as shown in Fig. 14. The ferrite in the ferrite-pearlite steel was strengthened by interphase precipitated TiC. The size of the TiC was about 10 nm in the developed steel with tensile strength of 590 MPa.
(b) Ferritic-martensitic steelIn 1995, interphase precipitated TiC was applied to a ferritic-martensitic steel to obtain high performance automotive steel sheet. The composition of the developed steel was 0.08%C–1.5%Si–1.8%Mn–0.1%Ti.55) The aim of carbide dispersion in the ferrite is to reduce the difference of hardness between ferrite and martensite so as to obtain high fatigue strength and high stretch-flange formability. Titanium carbides with a diameter from 5 nm to 10 nm were precipitated in the ferrite of the developed steel. Figure 1555) shows the stretch-flange formability of the developed steel in comparison with the conventional steels.

Hole expanding ratio of ferritic-martensitic steel strengthened by interface precipitation.55)
In 2008, a high fatigue strength steel was developed by controlling the diameter of TiC in the ferrite of a ferritic-martensitic steel containing 0.08%C–1.5%Si–1.8%Mn–0.09%Ti.56) After hot-rolling, the steel was water-cooled to 953 K followed by air-cooling for a time from 5 s to 10 s to cause TiC interphase precipitation.
Recently, the tensile properties of interphase precipitation-strengthened ferritic-martensitic steel and ferritic-bainitic steel were analyzed.57,58) The results indicate that VC precipitation in ferrite results in a decrease in the extent of strain partitioning in the tensile test specimen.
(c) Ferritic steelA new ferritic steel was developed by applying interface precipitation to obtain high strength.59) The newly-developed steel contained ultra-fine carbides and exhibited excellent stretch-flange formability. While a uniform ferrite microstructure had been known to exhibit high stretch-flange formability, high tensile strength of more than 780 MPa had not been realized due to the difficulty of obtaining ultra-fine carbides. In the developed steel, ultra-fine carbides were successfully precipitated, and tensile strength of more than 780 MPa was achieved by adding carbide-forming elements. In the developed steel, Mo addition generated ultra-fine (Ti,Mo)C instead of TiC, as shown in Fig. 16.13)

Transmission electron micrograph and energy dispersion X-ray spectra of (Ti,Mo)C in ferritic low carbon steel.13)
The stretch-flange formability of the developed steel is higher than that of the conventional low temperature transformed steels. The elongation of the developed steel is also higher than those of the conventional steels, as shown in Fig. 17.13) Since the developed steel contains less silicon, it has a smooth surface, which results in higher fatigue strength compared to the conventional steels.60) Moreover, even when the developed steel is held at a higher temperature than the coiling temperature, coarsening of (Ti,Mo)C is minimal as shown in Fig. 18.17) This means that strength remains high after heat treatment. Since the austenite-ferrite interface migrates through the entire steel, carbide nucleation occurs throughout the material, contributing to uniform dispersion of fine carbides. The alloy design of the developed steel allows production of 980 MPa and 1180 MPa grade precipitation-strengthened ferritic steels.60) The soaking temperature in the hot rolling process limits the amount of carbide dissolution. A much larger amount of the multiple carbide (Ti, Mo)C is solved compared to TiC, and this contributes to realizing high precipitation strength. A detailed analysis of work-hardening in the interphase precipitation-strengthened steels has been started.61,62)

Hole expanding ratio of ferritic steel.13)

Change in hardness of ferritic steels precipitation-strengthened by fine carbides of TiC and (Ti,Mo)C.17)
Recently, the precipitation pathway of (Ti, Mo)C has been investigated by the HR-TEM and 3D-AP techniques.63) HR-TEM observation indicated that embryo clusters that were coherent with ferrite grew into NaCl type nanometer-sized carbides through the GP zone. The interphase precipitation pathway will also be analyzed in detail in the near future. These analyses are expected to lead to the establishment of a new work-hardening theory through the discovery of new phenomena.
4.2 V-bearing high carbon steelDevelopment of V-bearing high carbon steels as non-quenched steels started in the 1970s.64) In the 1980s, ferritic-pearlitic non-quenched steels strengthened by VC or (Nb,V)C were developed for plate products.9) In the 2010s, a new steel was developed by adding 0.3% to 0.5% of vanadium to S45C steel (Japan Industrial Standard) for automobile connecting rods.65,66) The steel rod is reheated at different temperatures in each part in order to obtain different hardnesses in one product. The interphase precipitated VC is used to strengthen ferrite instead of the conventional quenching and tempering process. The high temperature reheated part exhibits higher yield strength of more than 1000 MPa as a result of the existence of a large amount of fine VC after hot-forging and isothermal holding, while the low temperature reheated part exhibits lower yield strength than 600 MPa to secure good machinability. Cu addition was also attempted with the aim of precipitation strengthening by Cu precipitates.67)
Precipitation strengthening is one effective method for producing non-quenched high strength steels. Figure 19 shows the relationship between the carbide volume fraction and ferrite hardness of interphase precipitation-strengthened ferritic steel. Figure 19 indicates that nanometer-sized carbides can provide a high carbide density, which is effective for strengthening steel. Since interphase precipitation occurs in many metals, the development of new functional metal materials by using interphase precipitation is also expected in the future.

Change in hardness with carbide volume fraction.