L Journal of Alloys and Compounds 334 (2002) 173–181 www.elsevier.com / locate / jallcom Thermodynamic optimization of the B–Fe system * T. Van Rompaey , K.C. Hari Kumar, P. Wollants KU Leuven, Departement Metaalkunde en Toegepaste Materiaalkunde, Kasteelpark Arenberg 44, 3001 Heverlee, Belgium Received 2 July 2001; accepted 18 July 2001 Abstract The B–Fe binary system has been optimized using the CALPHAD method, utilizing published experimental thermochemical and phase diagram data. Two models for the solid solubility of B in b.c.c. Fe and in f.c.c. Fe are presented, B as a interstitial, and B as a substitutional constituent. Thermodynamic calculations based on the interstitial and the substitutional model were in good agreement with experimental data.  2002 Elsevier Science B.V. All rights reserved. Keywords: Transition metal alloys; Phase diagram; Thermodynamic modelling 1. Introduction available experimental data. As compared to the most recent reassessment [7], five additional sources of ex- The knowledge of the thermodynamic properties of the perimental data were considered [12–16]. B–Fe system can be of interest in several fields of materials engineering: e.g. B–Fe is a subsystem of the B–Fe–Nd alloys with superior magnetic properties [1,2], 2. Experimental data and the high modulus TiB -reinforced steel composite is 2 based on the B–Fe–Ti ternary [3]. Furthermore, boron is In the B–Fe system, six equilibrium phases are present, used as an alloying element to enhance the hardenability of viz. the liquid (L), three terminal solid solutions, b.c.c. Fe steels and to form amorphous alloys. ( aFe and dFe), f.c.c. Fe ( gFe) and rhombohedral B ( bB), Several critical evaluations of the B–Fe system have and two intermediate phases Fe B and FeB. The Fe B 2 3 been reported in literature [4–7]. Thermodynamic calcula- phase reported by Takahashi et al. [13] and by Khan et al. tions of the system were also presented [4,7–10]. Chart [4] [17] is generally considered to be metastable [18]. The proposed a six-parameter model for the liquid and an occurrence of the FeB phase, reported by Voroshin et al. 2 estimation for the dilute solutions of B in b.c.c. and f.c.c. [19], was not reproduced by other authors. Fe. Kaufman et al. [8] used a regular solution for the liquid For the B–Fe system, only limited phase diagram data phase, and a subregular model for the b.c.c. and f.c.c. are available. (Table 5 presented at the end of the phases. Ohtani et al. [9] considered the interstitial solid Discussion includes literature data on the invariant equilib- solution of B in b.c.c. and f.c.c. Fe to calculate the ria, as reported by several investigators [19–22]). Kneller Fe–B–C ternary. On the other hand, Hallemans et al. [7] and Khan [23] suggest that the Fe B compound melts 2 suggested substitutional solubility of B in both Fe phases. congruently. Due to the discrepancy with all other data, The latter description and the calculation by Pan [10] used this is not used in the optimization. the SGTE recommended expressions for pure Fe and B, Experimental phase boundary data are available for the given by Dinsdale [11]. liquidus [20], and for the solid solubility of B in Fe In this work a thermodynamic description of the stable [22,24–30]. In Table 1, these data and the corresponding phases in the B–Fe system is presented, based on the experimental techniques are presented. In this optimization only the most recent data by Cameron and Morral [22] on the solid solubility of B in Fe are considered. They used *Corresponding author. the autoradiography technique which allows to discrimi- E-mail address: Tim.VanRompaey@mtm.kuleuven.ac.be (T.Van Rom- paey). nate between B in solution, B contained in precipitates, or 0925-8388 / 02 / $ – see front matter  2002 Elsevier Science B.V. All rights reserved. PII: S0925-8388(01)01777-7