Evidence of Large Magnetostructural Effects in Austenitic Stainless Steels L. Vitos, 1,2 P.A. Korzhavyi, 1 and B. Johansson 1,2,3 1 Applied Materials Physics, Department of Materials Science and Engineering, Royal Institute of Technology, SE-10044 Stockholm, Sweden 2 Condensed Matter Theory Group, Physics Department, Uppsala University, S-75121 Uppsala, Box 530, Sweden 3 AB Sandvik Materials Technology, SE-811 81 Sandviken, Sweden (Received 20 September 2005; published 24 March 2006) The surprisingly low magnetic transition temperatures in austenitic stainless steels indicate that in these Fe-based alloys magnetic disorder might be present at room temperature. Using a first-principles approach, we have obtained a theoretical description of the stacking fault energy in Fe 100cn Cr c Ni n alloys as a function of composition and temperature. Comparison of our results with experimental databases provides a strong evidence for large magnetic fluctuations in these materials. We demonstrate that the effects of alloying additions on the structural properties of steels contain a dominant magnetic contribution, which stabilizes the most common austenitic steels at normal service conditions. DOI: 10.1103/PhysRevLett.96.117210 PACS numbers: 75.50.Bb, 71.15.Nc, 81.05.Zx Fully austenitic stainless steels are composed mainly of Fe, Cr, and Ni, and have the face centered cubic (fcc) crystallographic structure of -Fe. At low temperatures, these alloys exhibit a rich variety of magnetic structures as a function of chemical composition, ranging from ferro- magnetic phase to spin-glass and antiferromagnetic align- ments [1,2]. At ambient conditions the austenitic steels have very low magnetic permeability and are generally regarded as nonmagnetic. Besides the applications where excellent mechanical properties and high corrosion resist- ance are required, these steels represent the primary choice also for nonmagnetic engineering materials. It is well known that the local magnetic moments in Fe, Cr, and Ni survive in their high-temperature paramagnetic states [3–10]. The persisting moments give rise to sizable contributions to the specific heat and entropy. These effects were used for tracing the high-temperature spin fluctua- tions in magnetic transition metals [4,5,11,12]. The mag- netic transition temperatures in Fe-rich Fe-Cr-Ni solid solutions are unexpectedly low ( &100 K) [1]. This sug- gests that in these alloys disordered local magnetic mo- ments might be present at ambient conditions. The picture of nonvanishing magnetic moments in paramagnetic aus- tenitic steels is in line with the high-field magnetization measurements on Fe 100cn Cr c Ni n (14 <n< 21 and c  20) alloys [2]. These experiments reported Curie- Weiss type susceptibility in the pure paramagnetic regime above 130 K. Having in mind the strong crystal field dependence of Fe magnetic moments [6,13,14], we can anticipate a large impact of the magnetic disorder on the cohesive properties of steels. Based on first-principles alloy theories, here we present a direct verification of the spin fluctuations in Fe-Cr-Ni alloys, and reveal the impor- tance of magnetism on the phase stability and mechanical properties of these important materials. The outstanding mechanical performance of austenitic steels emerges from the intrinsic properties of austenite. The stacking fault energy (SFE) is a key microscopic parameter of this phase. It is used for modeling a vast number of phenomena, e.g., plastic deformation, phase transformations, shape memory effects. Using various techniques, the SFE of austenitic steels has been measured as a function of composition [15– 20] and temperature [21– 23], and today commonly accepted databases exist. Numerous empirical models focused on the understanding of the principal factors governing the SFE [15,24,25]. According to the pioneering work by Ishida [24], the SFE, to a good approximation, is proportional to the Gibbs energy difference between the hexagonal close- packed (hcp) and fcc phases. The magnetic free energy vanishes in the hcp Fe ("-Fe) [6] indicating that the local moments disappear in this phase. Therefore, the SFE ap- pears to be a perfect candidate for detecting the footprint of room-temperature spin fluctuations on the cohesive prop- erties of austenitic steels. We used the exact muffin-tin orbitals (EMTO) method [26] to compute the SFE of Fe 100cn Cr c Ni n alloys. This method, in combination with the coherent potential ap- proximation (CPA) [27], is suitable for describing the simultaneous presence of the chemical and magnetic dis- order in Fe-based random alloys [28–30]. In the present study the paramagnetic phase of the Fe 100cn Cr c Ni n alloy was modeled by the quasiternary Fe " 1=2 Fe # 1=2  100cn  Cr " 1=2 Cr # 1=2  c Ni " 1=2 Ni # 1=2  n alloy with randomly distributed magnetic moments oriented up ( " ) and down ( # ). This approximation accurately describes the effect of loss of the net magnetic moment above the transition temperature [7–9]. The most common stacking fault in an fcc crystal, the so-called intrinsic staking fault, may be viewed as a miss- ing 111 layer from an otherwise perfect lattice. The excess free energy F per unit interface area defines the fault energy . Within the axial interaction model [31,32], PRL 96, 117210 (2006) PHYSICAL REVIEW LETTERS week ending 24 MARCH 2006 0031-9007= 06=96(11)=117210(4)$23.00 117210-1  2006 The American Physical Society