EFFECTS OF SILICON ADDITION ON GRAIN BOUNDARYSEGREGATION BEHAVIORS AND MECHANICAL

PROPERTIES IN a-IRON

J.C. Kim*, N.H. Heo*, J.G. Na, J.S. Woo*** and G.M. Kim*****Materials and Corrosion Laboratory, Korea Electric Power Research Institute, Taejon 305-380,Korea, **Division of Metals, Korea Institute of Science and Technology, Seoul 136-791, Korea,

***Techical Research Laboratory, Pohang Iron and Steel Co., Pohang, Kyungbuk 790-785, Korea,****Department of Materials Engineering, Chungnam National University, Taejon 305-764, Korea

(Received August 18, 1997)(Accepted in revised form December 19, 1997)

1. Introduction

Great attention has been paid to the study of grain boundary segregation, because the phenomenoninfluences many mechanical and chemical properties of materials [1]. The study of Fe-Si alloys is ofgeneral interest because silicon is a common alloying element for steels and influences materialproperties. Fe-Si alloys with 3–4 wt.%Si are widely used as industrial transformer steels, where grainorientation improves the magnetic properties. Addition of silicon may improve the resistance to stresscorrosion cracking [2], and changes the secondary hardening behavior in steels [3]. In all these cases,interfacial segregation phenomena are suspected to play a significant role. Repulsion or cosegregationat surfaces or grain boundaries between silicon and impurities has been reported in Fe-Si alloys [4–6].

In this study, effects of silicon addition on grain boundary segregation behaviors and mechanicalproperties ina-iron are investigated.

2. Experimental Procedures

Pure iron and 3.37- and 4.26wt.%Si alloys were prepared from vacuum induction melting. The alloyswere homogenized and hot-rolled to 2.5 mm plates. The alloys were annealed at 900°C for 3.6 ks, andsubsequently reannealed (i.e. 776 ks at 450, 500 and 550°C, 72 ks at 600°C, 3.6 ks at 700 and 800°C)in a neutral salt bath. Tensile tests were carried out in liquid nitrogen on samples machined in the rollingdirection and having a dimension (25 mm in gauge, 6.25 mm in width and 1.8 mm in thickness). Inorder to investigate grain boundary segregation behaviors of solutes, AES samples, machined fromtensile specimens and chilled with liquid nitrogen, were fractured in an ultra high vacuum chamber ofabout 13 1027 Pa. An incident beam energy of 5 KeV and about 10 grain boundary facets from eachsample were used for Auger analyses. The sulfur 150 eV, nitrogen 379 eV, carbon 272 eV, phosphorus120 eV, silicon 92 eV and iron 703 peaks were selected for the analyses. All the differential Augerpeaks were normalized by the iron 703 eV peak, and were quantified according to the following method[7]:

Pergamon

Scripta Materialia, Vol. 38, No. 7, pp. 1071–1076, 1998Elsevier Science Ltd

Copyright © 1998 Acta Metallurgica Inc.Printed in the USA. All rights reserved.

1359-6462/98 $19.001 .00PII S1359-6462(98)00009-8

1071

I x 5 I x,0@~ fX - fX,0!$1 - exp~-1/lXcosu !% 1 fX,0#

I Fe 5 I Fe,0$1- OX ÷ Fe

N

fX 1 OX ÷ Fe

N

~ fX 2 fX,0!exp~-1/lFecosu !% (1)

and

I X,0

IFe,05

SX

SFe(2)

wherefx is the monolayer mole fraction of an element X in an iron-based alloy system;lX and lFe

indicate the inelastic mean free paths (IMFP) in monolayers of theX and Fe Auger electrons;u is theemission angle of the Auger electrons to the surface normal;SX and SFe are the relative sensitivityfactors between the pure elements and silver, obtained from the Handbook of Auger Electron Spec-troscopy [8];IX andIFe are the Auger peak amplitudes of the elementX and the bulk Fe, respectively;IX,0 and IFe,0 are the intensities of the relative signals from the pure bulk samples;fX,0 is the molefraction ofX in the bulk. The value of cosu in equation. (1) can be replaced by 0.74 [9,10]. The IMFPof the X Auger electron is evaluated by the equation of Seah and Dench [11]:

lX 5538

EX2 1 0.41~aEX!

1/ 2 ~in monolayers) (3)

where EX is the Auger electron energy in eV of the elementX. a is the monolayer thickness innanometres of the elementX given by

a3 5A

pN(4)

whereA is the atomic weight,N is Avogadro’s number andr is the bulk density [12,13].

3. Results and Discussion

The ultimate tensile strength of the pure iron varied with annealing temperature, as shown in Fig. 1.The pure iron exhibited severe embrittlement at 600 and 700°C, in which the tensile strengths were

Figure 1. Change in ultimate tensile strength of the pure iron with aging temperature.

EFFECTS OF SILICON ADDITION1072 Vol. 38, No. 7

about 270 MPa, and at 500°C showed a maximum, 630 MPa, in ultimate tensile strength. The pure ironshowed nil ductility in the temperature range. Figure 2 shows changes in hardness and ultimate tensilestrength of the 3.37wt.%Si alloy. The solid solution hardening effect of silicon was prominent in thisalloy, as in other research [14], resulting in higher hardness levels which ranged 186 to 220 HV incomparison to those, about 70 HV, of the pure iron. The ultimate tensile strengths, however, werewithin the range 130 and 334 MPa, which were overally lower than those in the pure iron. Mechanicalproperties of the 4.26wt.%Si alloy were not investigated, due to the extreme brittleness. After annealingat 700°C showing the minimum in ultimate tensile strength, the pure iron, as shown in Fig. 3(a), showeda typical intergranular fracture in which fracture paths followed prior austenite grain boundaries. Withincreasing or decreasing the annealing temperature from 700°C, the fracture mode of the pure iron,however, changed from intergranular to cleavage.

In as-rolled condition, the 3.37wt.%Si alloy showed a typical intergranular fracture. As shown in Fig.3(b), the alloy annealed at 700°C was, however, fractured in a transgranular fracture mode with aportion of intergranular.

Figure 2. Changes in hardness and ultimate tensile strength of the 3.37wt.%Si alloy.

Figure 3. Fracture surfaces after annealing at 700°C: (a) the pure iron and (b) the 3.37wt.%Si alloy.

EFFECTS OF SILICON ADDITION 1073Vol. 38, No. 7

In order to investigate the effect of silicon addition on grain boundary segregation behavior ofimpurities, Auger analyses were performed on pure iron and samples containing silicon, annealed at700°C. The sulfur content at grain boundaries, as shown in Fig. 4, lowered with increasing siliconcontent, in comparison to that of the pure iron, and such a silicon addition effect is clearly shown fromthe comparison between Figs. 5 and 6.

Figure 5 shows changes in grain boundary segregation behavior in the pure iron of impurities withannealing temperature. Carbon and nitrogen contents at the grain boundaries increased with decreasingtemperature, but the sulfur content showed a decreasing tendency below 700°C. Intergranular brittle-ness in the pure iron, when referred to Figs. 1 and 5, is, as a result, due to the sulfur segregated at thegrain boundaries. This has also been confirmed in several researches [15–17]. It was suggested that thedecreasing tendency of sulfur below 700°C was due to the solubility limit of sulfur in pure iron andhence the precipitation of iron sulfide followed by the depletion of sulfur in the matrix [17]. However,a recent study of pure iron [15] showed that the precipitate in pure iron was not sulfide but oxide. It is,therefore, suggested that below 700°C the decrease in sulfur content with decreasing temperature isprobably due to the competitive grain boundary segregation between sulfur and carbon or nitrogenrather than the solubility limit of sulfur. This suggestion can partly be supported by researches [18,19],which reported a competitive grain boundrary segregation between the impurities in iron.

Figure 4. Auger spectra obtained from samples annealed at 700°C.

Figure 5. AES analyses of the pure iron.

EFFECTS OF SILICON ADDITION1074 Vol. 38, No. 7

As shown in Fig. 6, the competitive grain boundary segregation behavior between silicon and carbon[20–23] were also observed in the 3.37wt.%Si alloy. As mentioned above, the ultimate tensile strengthof the 3.37wt.%Si alloy ranged over 130 to 334 MPa, and, in the present temperature range, the sulfurlevels at the grain boundaries were lower than those of the pure iron. Nonetheless, the fracture modewas, as shown in Fig. 3(b), mostly transgranular except a trace of intergranular fracture. This impliesthat the fracture behavior in the silicon-bearing alloy may be attributed to lower cleavage fracturestrength than intergranular fracture strength.

3. Conclusions

Effects of silicon addition on grain boundary segregation behaviors and mechanical properties ina-ironhave been investigated. Competitive grain boundary segregation between sulfur and carbon or nitrogen,which caused the decrease of sulfur content at the grain boundaries below 700°C, was observed in pureiron. The intergranular brittleness was mainly caused by the sulfur segregated at the grain boundaries.The sulfur content at the grain boundaries decreased with increasing silicon content. In the 3.37wt.%Sialloy, silicon and carbon or nitrogen segregated competitively to the grain boundaries, and such acompetition was also observed between sulfur and carbon or nitrogen. The transgranular fracture modein the silicon-bearing alloy may be attributed to lower cleavage fracture strength than intergranularfracture strength.

References

1. E. D. Hondros and M. P. Seah, Int. Met. Rev. 22, 262 (1977).2. R. O. Ritchie, Metall. Trans. 9A, 35 (1978).3. M. S. Bhat, W. M. Garrison, and V. F. Zackay, Mater. Sci. Eng. 41, 1 (1979).4. M. Es-Souni and A. Mosser, Surf. Sci. 199, 439 (1988).5. P. Lejcek, M. Koutnik, J. Bradler, V. Paidar, and A. Potmesilova, Appl. Surf. Sci. 44, 75 (1990).6. P. Lejcek and S. Hofmann, Acta Metall. 39, 2469 (1991).7. N. H. Heo, Acta Mater. 44, 3015 (1996).8. L. E. Davis, N. C. MacDonald, P. W. Palmberg, E. Riach, and R. E. Weber, Handbook of Auger Electron Spectroscopy,

2nd edn. Physical Electronics, Edina, MN (1976).9. M. P. Seah, Surf. Sci. 32, 703 (1972).

10. J. C. Shelton, J. Electron Spectrosc. 3, 417 (1974).11. M. P. Seah and W. A. Dench, Surf. Interf. Anal. 1, 2 (1979).

Figure 6. AES analyses for intergranular fracture facet of the 3.37wt.%Si alloy.

EFFECTS OF SILICON ADDITION 1075Vol. 38, No. 7

12. G. W. C. Kaye and T. H. Laby, Tables of Physical and Chemical Constants, 14th edn., Longman, London (1973).13. R. C. Weast, ed., Handbook of Chemistry and Physics, 58th edn., CRC Press, Cleveland, OH (1977–78).14. W. B. Morrison and W. C. Leslie, Metall. Trans. 4A, 379 (1973).15. N. H. Heo, Surf. Interface Anal. submitted.16. K. S. Shin and M. Meshii, Acta Metall. 31, 1559 (1983).17. C. L. Briant, Acta Metall. 33, 1241 (1985).18. G. Tauber and H. J. Grabke, Ber. Bunsenges. Phys. Chem. 82, 298 (1978).19. S. Suzuki, S. Tanii, K. Abiko, and H. Kimura, Metall. Trans. 18A, 1109 (1987).20. H. J. Grabke, Steel Res. 57, 178 (1986).21. F. L. Williams and D. Nason, Surface Sci. 45, 377 (1974).22. V. S. Sundaram and P. Wynblatt, Surface Sci. 52, 569 (1975).23. P. Wynblatt and R. C. Ku, Surface Sci. 65, 511 (1977).

EFFECTS OF SILICON ADDITION1076 Vol. 38, No. 7