Phase Transformation and Properties of Fe-Cr-Co Alloyswith Low Cobalt Content

Qiuzhi Gao+, Minglong Gong, Yingling Wang, Fu Qu and Jianeng Huang

School of Resources and Materials, Northeastern University at Qinhuangdao,Qinhuangdao, Hebei 066000, China

The Curie temperature, phase transformation, microstructure and magnetic properties of Fe-10Cr-xCo alloys with low Cobalt content wereanalyzed after casted and solid solution quenched, respectively. The results show that the experimental Curie temperature increases from 1043Kto 1065K with differential Co content addition, and which is about 50K higher (as can be called “the degree of superheat”) compared with thecalculated data. Besides, addition of Co promotes the formation of ¡-ferrite phase, and thus leads to the decrease of Vickers hardness. Accordingto X-ray diffraction results, B2 type ordered structure forms, and Cr1.07Fe18.93 was observed in all the samples, whereas, CoFe15.7 only can befound in the sample with the addition of Co is more than 2mass%. [doi:10.2320/matertrans.M2015077]

(Received February 25, 2015; Accepted June 3, 2015; Published July 10, 2015)

Keywords: iron-chromium-cobalt alloys, Curie temperature, microstructure, properties

1. Introduction

Fe-Cr-Co alloys are considered as ductile permanentmagnets due to their higher Curie temperature, bettercorrosion resistance, and more economical production costin comparison to Al-Ni-Co magnets.1­3) Thus, the alloys havebeen applied variously in the related fields, such as electro-acoustical, electronic industry, telephones, tacho-meters,micro-meters, and so on.4,5) In addition, Fe-Cr-Co alloysalso can be used for the heat-resistant components of powerplant devices that work under temperatures up to 823Kbecause of their excellent high temperature stability.

As a group of materials with a miscibility gap, themagnetic properties in Fe-Cr-Co alloys are formed by thespinodal decomposition of high temperature ¡ phase, intoiron-rich phase (¡1) and chromium-rich phase (¡2).6,7) Asshown in previous investigations,4,8) the addition of Co inFe-Cr-Co alloys can raise the decomposition temperature of¡ phase, and also extend the difference in concentrationsbetween ¡1 and ¡2 phase. The magnetic properties of Fe-Cr-Co alloys are mainly affected by the addition of Co with theusual content being about 20mass%, or above 10mass% atleast.2,9) Recently, it also has been reported that adding 2­3mass% Co in 9Cr ferritic heat-resistant steel drasticallyimproves short-term creep strength, while long-term creepproperties tend to deteriorate around 923K.10,11) The additionof Co may indirectly affect microstructural change in thecomplicated martensitic structure. Hence, research on themechanism of low-Co content in simple ternary Fe-Cr-Coalloy is very important to understand the influence of Co onmicrostructure and related properties in high-temperaturealloys and soft-magnetic materials. The Fe-Cr-Co alloys aspermanent magnets were studied sufficiently.12­14) Unfortu-nately, the mechanism of Co in Fe-Cr-Co alloys with low-Cocontent is unclear.

In the present work, Fe-Cr-Co alloys with various low Cocontent were prepared by vacuum arc melting, and then solid-

solution quenched at 1373K. The influences of Co additionon Curie temperature and microstructure of Fe-Cr-Co alloyswere analyzed, and as a comparison, the theoretical Curietemperature of the employed alloys was also calculated.Differential Thermo Analysis (DTA) and X-ray Diffraction(XRD) were employed in order to measure the phasetransformation and composition. At last, the Vickers hardnessand the magnetic properties were investigated by VickersHardness Tester and Vibrating Sample Magnetometer(VSM), respectively.

2. Experimental Procedure

The Fe-10Cr-xCo (x = 1, 2, 3, and 4 in mass%) alloyswere prepared from 99.9mass% Fe, 99.95mass% Cr and99.8mass% Co by vacuum arc melting in a water-cooledcopper crucible. The alloy ingots about 100 g were repeatedlymelted for eight times, and then solid solution quenched at1373K for 60min in quartz capsules vacuum furnace at1 © 10¹3 Pa. In order to obtain phase transformation data, theDTA experiment was carried out in high purity argonatmosphere to avoid oxidation. The samples used for DTAwere cylindrical specimens, ¯5 © 3mm, and the thermaltreatment cycle are applied as follows: the samples werefirstly heated from room temperature to 1273K with a rateof 15K/min, and holding for 5min, then cooled to roomtemperature with the same rate. Phase compositions wereidentified at room temperature by X-ray diffraction (XRD)analysis using Cu K¡ radiation, with a step size of 0.03°.Vickers hardness tests were carried out with an applied loadof 9.8N (100 g) for 15 s. The magnetic hystersis loop wasmeasured by Vibrating Sample Magnetometer (VSM).Microstructure of the samples was observed by opticalmicroscopy (OM). OM samples were prepared by grindingwith a series of SiC paper from 240 to 1500 grit, followed bydiamond paste (0.05 µm) as the final mechanical polishing insequence. After polished, the samples were etched with asolution of water (100mL), hydrochloric acid (20mL), andiron trichloride (5 g).+Corresponding author, E-mail: neuqgao@163.com

Materials Transactions, Vol. 56, No. 9 (2015) pp. 1491 to 1495©2015 The Japan Institute of Metals and Materials

3. Results and Discussion

3.1 Phase transformations and microstructuresThe results of the heat flow as a function of temperature for

the Fe-10Cr-1Co alloy are shown in Fig. 1. It can be foundthat there are three distinct endothermic peaks during heating.The first one around 393K represents the normal absorptionof samples with heating. The second one around 1043K isoriginated from the transition from a ferromagnetic state to aparamagnetic state, namely the so-called Curie transition orthe magnetic transition. The last one is a new phase formationpeak for the transformation from martensite to austenite.During cooling, only one exothermic peak located between773K and 573K can be recognized, which indicates thetransformation from austenite to martensite.

The obtained Curie temperatures (m-CP) of Fe-10Cr-xCoalloys based on DTA data are plotted in Fig. 2. With theincrease of Co addition, the Curie temperature of Fe-10Cr-xCo alloys increases sensitively, which is also higher than theCurie temperature of pure Fe (1042K15)). It can be found thatthe Curie temperature increases from 1043K to 1065K withthe increase of Co content from 1mass% to 4mass%, whichindicates that the addition of Co would increase the magnetictransition temperature in Fe-10Cr alloy. It was reported16) thatthe Curie temperature is about 1033K in Fe-10 atom%Cralloy, and increases under the small Cr concentrations. On thecontrary, the Curie temperature would decrease when theconcentration of Cr in the alloy is exceeded about 6 atom%Cr. The converted atom percent of Cr in the experimental

materials is about 10.3 atom%, obviously higher than6 atom%, leadings to the decrease of Curie temperature.Besides, pure Co is a strong ferromagnets, and its Curie-pointtemperature is up to about 1373K.17) The combined influenceof Cr and Co results in the increase of Curie temperature. Thecalculated Curie temperature of Fe-Cr-Co alloy can beexpressed as:18)

Tc ¼ xFeTCFe þ xCrT

CCr þ xCoT

CCo

þ xFexCrT0c,Fe,Cr þ xFexCoT

0c,Fe,Co ð1Þ

where T Cm is Curie temperature of element m, the values of

T CFe, T

CCr, T

CCo are 1043K,19) ¹311.5K,20) 1450K,19) respec-

tively. The interaction parameters of Curie temperaturebetween elements i and j are T 0

c;i;j. T0c;Fe;Cr ¼ 850K18) and

T 0c;Fe;Cr ¼ 590K.19) xFe, xCr and xCo are the concentrations of

element Fe, Cr and Co, respectively.Based on eq. (1), the calculated Curie temperatures (c-CP)

in prepared Fe-10Cr-xCo alloys are also displayed in Fig. 2.It can be seen that all the calculated data are lower than theexperimental data, which means that there need to be acertain degree of superheat during the actual Curie trans-formation. The degree of superheat provides the driving forceof the transformation from ferromagnetism to paramagnet-ism, whose value is about 50K under the applied heating ratein present work.

The obtained phase transformation temperatures of Fe-10Cr-xCo alloys during heating and cooling based on DTAexperiment are listed in Table 1. With the increase of Cocontent, it is obvious that both austenitic transformationtemperature (As and Af ) and martensitic transformationtemperature (Ms and Mf ) increases. The transformations ofaustenite and martensite finish in an approximately the samerange about 40K and 120K depending on the content of Co,respectively. The variety of Ms point and the activationenergy of martensite can partly be attributed to the influenceof alloying elements.21­23) Ghosh and Olson24) showed thatCo is a subsitutional solid solution alloying element andstrongly affects phase transition temperature of alloys. Crelement, as a stronger carbides/nitrides forming alloyingelement, has nearly no influence on Ms, which means that itdoes not promote or obstruct martensitic transformation.25,26)

The critical value in (J/mol) of the driving force needed totrigger martensitic transformation is27)

��G£¡0c ¼ 1010þ 4009c0:5C þ 1879c0:5Si þ 1980c0:5Mn

þ 1418c0:5Mo þ 1868c0:5Cr

þ 1618c0:5V þ 1653c0:5Nb þ 3097c0:5N

þ 752c0:5Cu þ 714c0:5W

þ 280c0:5Al þ 172c0:5Ni � 352c0:5Co ð2Þ

Fig. 1 Heat flow as a function of temperature for Fe-10Cr-1Co alloy heatedto 1273K with a rate of 15K/min.

Fig. 2 Curie-point temperature of Fe-10Cr-xCo alloys as a function ofCobalt content.

Table 1 Measured phase transformation temperatures with differentialCobalt contents during heating and cooling cycle.

Samples As (K) Af (K) Ms (K) Mf (K)

Fe-10Cr-1Co 1084 1138 735 607

Fe-10Cr-2Co 1088 1128 752 632

Fe-10Cr-3Co 1090 1125 774 659

Fe-10Cr-4Co 1095 1119 795 679

Q. Gao, M. Gong, Y. Wang, F. Qu and J. Huang1492

where c0.5 is the square root of the alloying elementconcentration in mole fraction. Without consideration ofheat-treatment, it is likely that the decrease of the stability ofaustenite will promote martensitic transformation, andfollowed by this opinion, the alloying elements with lowerstability of austenite (e.g., Al, Ni, Co) should have abeneficial role in promotion of martensitic transformation. Itmeans that Al, Ni, Co will decrease the activation energy ofmartensitic transformation, and promote martensite embyosactivation.28) Hence, the increase of Co addition promotesmartensitic transformation and improves martensite-starttemperature, Ms (as shown in Table 1).

The optical micrographs for the samples solution quenchedat 1373K for 60min are represented in Fig. 3. It is clearthat the samples are consisted of martensite and ¡-ferrite(Fig. 3(b)­(d)), except for Fe-10Cr-1Co alloy with approx-imately single phase of martensite (Fig. 3(a)). The phase withbright contrast can be identified as ¡-ferrite, while the phasewith dark contrast and lath shape can be identified asmartensite. These micrographs reflects that the microstructuretransforms from single phase (martensite) to dual phases(martensite and ¡-ferrite) with the increase of Co content.On the other hand, these micrographs also show that themicrostructure is very sensitive to the Co content in theresearched Fe-10Cr-xCo alloys. As is known to all, as anaustenite-stabilizing alloying element, Co can promoteaustenitic transformation, and improve austenite start-trans-formation temperature (As). Thus, with the increase of Coaddition, As temperature is raised and more ¡-ferrite formsduring continuous cooling, which leads to the co-existence ofmartensite and ¡-ferrite after quenching.

To address the quantitative influence of the Co content onphase formation, the ¡-ferrite fraction were quantitatively

analyzed by measuring the square of light contrast with morethan 20 optical photographs of the samples, the thus obtainedresults are displayed in Fig. 4. The content of ¡-ferriteincreases sharply with the increase of Co content, which isalso confirmed that the microstructure of the researched Fe-10Cr-xCo alloys is significantly affected by Co content. Themicrostructure is single ¡-ferrite phase in magnetic Fe-Cr-Coalloy in which the content of Cr and Co are generally above20mass% and 8mass%, respectively.29) The existence of ¡-ferrite is a typical characteristic in permanent magnets, andthe increase of ¡-ferrite including Fe and Co elements reflectsthat the Fe-10Cr-xCo alloys are gradually transformed topermanent magnets with the increase of Co content.

The X-ray diffraction patterns of the Fe-10Cr-xCo alloyssolution quenched at 1373K for 60min are displayed inFig. 5. Based on magnification of the main diffraction peaklocated at 43°³47° (as shown in Fig. 5(b)), it represents a

(a) (b)

(c) (d)

Fig. 3 Optical micrographs for the samples with differential Cobalt contents solution quenched at 1373K for 60min, (a) 1mass% Co,(b) 2mass% Co, (c) 3mass% Co, (d) 4mass% Co.

Fig. 4 ¡-ferrite fraction via Cobalt content as the samples solutionquenched at 1373K for 60min.

Phase Transformation and Properties of Fe-Cr-Co Alloys with Low Cobalt Content 1493

trend of periodic attenuation, which reflects the formation ofsuperlattice ordered structure. Superlattice crystallization is athermodynamically driven process, with the lattice structuredepending primarily on the size distribution and interparticleinteractions.30) Parkin et al.31) found that metallic superlatticestructures can be formed in Co/Cr, Fe/Cr. Besides, super-lattice also can be formed in semiconductors. The chemicalformula of superlattice structure in the researched allos wasCr1.07Fe18.93, and in addition, only a little CoFe15.7, which cannot be detected by XRD, might exist in the samples with thecontent of Co below 2mass%. In addition, with the increaseof Co content (above 2mass%), the formation of CoFe15.7can be identified in XRD patterns. It means that the B2 typeordered structures of intermetallic compound with theformulas of Cr1.07Fe18.93 and CoFe15.7 form in the researchalloys. According to Fig. 5, it also can be distinguished thatthe diffraction peaks of B2-Cr1.07Fe18.93 and B2-CoFe15.7 arepartly overlapping, except the last peak at around 98°. Theincrease of Co content, from 1mass% to 4mass% in Fe-10Cr-xCo alloys, leads to the broadening of diffraction peaks.Thus, the existence of B2-CoFe15.7 phase can be confirmed.The addition of Co results in the reaction between theelements of Fe and Co, and thus more B2-CoFe15.7 forms inFe-10Cr-xCo alloys.

3.2 Vickers hardness and magnetic propertiesFigure 6 compares the distribution of Vickers hardness for

Fe-10Cr-xCo alloys as a function of Co content. It is obviousthat the value of Vickers hardness is strongly dependent onCo content, and decreases with the increase of Co content.The variety of Vickers hardness reflects the change ofmicrostructure. As well known, the Vickers hardness ofmartensite phase is about 300HV, which can be considered asthe highest value comparing to 120HV of ¡-ferrite.32,33) Theformation of ¡-ferrite can be promoted due to the addition ofCo, and thus the increase of ¡-ferrite reduces the obtainedaverage Vickers hardness. Most of added Co presented assolid solution element in matrix, which means that Co mayindirectly affect the complicated martensitic structure.10)

Because of the increase of Co content, the larger grain sizeof prior austenite causes that the formed martensite lath alsobecomes larger during the following continuous cooling.Hence, the reason, causing the decrease of Vickers hardness

with more Co addition, can be attributed to two points, one isthe increase of ¡-ferrite, and the other is the larger formedmartensite lath.

The typical magnetic hystersis loop of Fe-10Cr-1Co alloysmeasured by VSM is represented in Fig. 7, which is similarwith the other experimental samples. It can be found that theexperimental materials can be obtained magnetic saturationas the magnetic field being more than 400 kA/m, whereas thevalue of Hc is 4³6 kA/m. Based on this, Fe-10Cr-xCo alloyswith low Co content can not absolutely defined as soft-magnetic materials, and should be classified as semihardmagnetic materials. The obtained magnetic properties of Fe-

(a) (b)

Fig. 5 X-ray diffraction patterns of the Fe-10Cr-xCo alloys solution quenched at 1373K for 60min, (a) XRD results of all the samples,(b) magnification of the district from 43° to 47° in XRD result of Fe-10Cr-4Co alloy.

Fig. 6 Distribution of vicker’s hardness for Fe-10Cr-xCo alloys as afunction of Cobalt content.

Fig. 7 Magnetic hystersis loop of Fe-10Cr-1Co alloys.

Q. Gao, M. Gong, Y. Wang, F. Qu and J. Huang1494

10Cr-xCo alloys based on the measured magnetic hystersisloop from VSM are listed in Table 2. It can be found that themagnetic properties depend critically on the content of Co.Both the coercivity and the saturation magnetization increasewith Co content rising from 1mass% to 4mass%. Although,the change of the residual magnetic flux density is notobvious with the content of Co.

4. Conclusions

Fe-10Cr-xCo alloys with low Cobalt content were preparedby vacuum arc melting, followed by solid solution quenched.The Curie temperature, phase compositions, microstructureand magnetic properties were investigated, and the followingconclusions can be drawn:(1) The Curie temperature increases with the increase of Co

content, and there is a degree of superheat affecting theexperimental data compared to the calculated data.

(2) The addition of Co promotes the transformation fromaustenite to martensite. In addition, the formation of¡-ferrite is sensitive to Co content, which increasessharply with the increase of Co content from 1mass%to 4mass%.

(3) Cr1.07Fe18.93 phase is matrix phase in Fe-10Cr-xCoalloys, and coexisted with CoFe15.7 under the conditionthat the content of Co is above 2mass%.

(4) Fe-10Cr-xCo alloys with low Cobalt content aresemihard magnetic materials. Both the coercivity andthe saturation magnetization increase with the increaseof Co content.

Acknowledgments

The grant and financial support by the Natural ScienceFoundation­Steel and Iron Foundation of Hebei Province(Grant No. E2014501056), the Youth Research Foundationof Hebei Education Department (Grant No. QN2014338) andthe Fundamental Research Funds for the Central Universities(Grant No. N142303001) are gratefully acknowledged.

REFERENCES

1) Y. Belli, M. Okada, G. Thomas, M. Homma and H. Kaneko: J. Appl.Phys. 49 (1978) 2049­2051.

2) H. Kaneko, M. Homma, T. Fukunaga and M. Okada: Magn. IEEETrans. 11 (1975) 1440­1442.

3) Z. Ahmad, A. Ali, M. Farooque and A. ul Haq: Mater. Des. 30 (2009)3885­3890.

4) O. A. Ushakova, E. H. Dinislamova, M. V. Gorshenkov and D. G.Zhukov: J. Alloy. Compd. 586 (2014) S291­S293.

5) A. H. Taghvaei, H. Shokrollahi and K. Janghorban: Mater. Des. 31(2010) 142­148.

6) F. Vodopivec and B. Breskvar: J. Magn. Magn. Mater. 152 (1996) 213­218.

7) S. Alleg, B. Bouzabata and J. Greneche: J. Alloy. Compd. 312 (2000)265­272.

8) L. Helis, Y. Toda, T. Hara, H. Miyazaki and F. Abe: Mater. Sci. Eng. A510­511 (2009) 88­94.

9) T. Li, W. Li and X. M. Li: Acta Phys. Sin. 54 (2005) 4384­4388.10) K. Yamada, M. Igarashi, S. Muneki and F. Abe: ISIJ Int. 43 (2003)

1438­1443.11) R. G. Faulkner, J. A. Williams, E. G. Sanchez and A. W. Marshall:

Mater. Sci. Technol. 19 (2003) 347­354.12) O. Ushakova and R. Malinina: Solid State Phenom. 190 (2012) 238­

242.13) A. Korneva: Solid State Phenom. 186 (2012) 116­119.14) A. Korneva, G. Korznikova, K. Berent, A. Korznikov, R. Kashaev,

J. Bogucka and K. Sztwiertnia: J. Alloy. Compd. 615 (2014) S300­S303.

15) M. Pajda, J. Kudrnovský, I. Turek, V. Drchal and P. Bruno: Phys. Rev.B 64 (2001) 174402.

16) M. Y. Lavrentiev, K. Mergia, M. Gjoka, D. Nguyen-Manh, G.Apostolopoulos and S. L. Dudarev: J. Phys. Condens. Matter 24(2012) 326001.

17) M. Pajda, J. Kudrnovský, I. Turek, V. Drchal and P. Bruno: Phys. Rev.B 64 (2001) 174402.

18) L. X. Lv, L. Zhen, C. Y. Xu and X. Y. Sun: J. Magn. Magn. Mater. 322(2010) 987­995.

19) A. F. Guillermet: High Temp. High Press. 19 (1987) 639­652.20) Y. Murakami, D. Shindo, K. Oikawa, R. Kainuma and K. Ishida: Acta

Mater. 50 (2002) 2173­2184.21) K. Ishida: J. Alloy. Compd. 220 (1995) 126­131.22) A. Borgenstam and M. Hillert: Acta Mater. 45 (1997) 651­662.23) B. Ravi Kumar, S. Sharma, P. Munda and R. K. Minz: Mater. Des. 50

(2013) 392­398.24) S. Pasebani and I. Charit: J. Alloy. Compd. 599 (2014) 206­211.25) H. G. Armaki, R. P. Chen, S. Kano, K. Maruyama and M. Igarashi:

J. Phys. Conference Series 240 (2010) 012085.26) K. Sawada, M. Takeda, K. Maruyama, R. Ishii, M. Yamada, Y. Nagae

and R. Komine: Mater. Sci. Eng. A 267 (1999) 19­25.27) C. Capdevila, F. Caballero and C. Garcia de Andres: Mater. Sci.

Technol. 19 (2003) 581­586.28) Q. Gao, C. Wang, F. Qu, Y. Wang and Z. Qiao: J. Alloy. Compd. 610

(2014) 322­330.29) H. Kaneko, M. Homma, K. Nakamura, M. Okada and G. Thomas:

Magn. IEEE Trans. 13 (1977) 1325­1327.30) M. B. Sigman, A. E. Saunders and B. A. Korgel: Langmuir 20 (2004)

978­983.31) S. S. P. Parkin, N. More and K. P. Roche: Phys. Rev. Lett. 64 (1990)

2304­2307.32) W. Horvath, W. Prantl, H. Stroi nigg and E. Werner: Mater. Sci. Eng. A

256 (1998) 227­236.33) R. Grange, C. Hribal and L. Porter: Metall. Mater. Trans. A 8 (1977)

1775­1785.

Table 2 The magnetic properties evolution of Fe-10Cr-xCo alloys.

Samples Hc (kA/m) Br (T) Bs (T)

Fe-10Cr-1Co 4.4 0.03 1.87

Fe-10Cr-2Co 5.8 0.021 1.93

Fe-10Cr-3Co 6.2 0.025 1.95

Fe-10Cr-4Co 6.9 0.036 1.98

Phase Transformation and Properties of Fe-Cr-Co Alloys with Low Cobalt Content 1495