Erratum II: “Twofold and fourfold symmetric anisotropic magnetoresistance effect in a model with crystal field” (Journal of the Physical Society of Japan(2015)84 (094710)Doi:10.7566/JPSJ.84.094710)

We found errors in the value of the spin–orbit coupling constant, λ, in the original paper¹⁾ and the previous erratum.²) We revise four parts, which are related to the errors. The revisions do not affect the main results of the paper. (i) In 2nd line from the bottom in left side on page 12 in Ref. 1, we replace “The quantity λ is set to ¼ 0:013 eV for Fe.^36Þ” by “The quantity λ is roughly evaluated to be ¼ 0:052 eV from ¼ ð10 nÞ₀, with n ¼ 6 and ₀ ¼ 0:013 eV for Fe²⁺ (see Table 1.1 in Ref. 36). Here, n is the 3d electron number and ₀ is the spin–orbit coupling constant in the spin–orbit interaction consisting of the total orbital angular momentum and the total spin.^36Þ” (ii) We replace the lower panel in Fig. 11 in Ref. 1 by the following panel: (iii) In 4th line from the bottom in left side on page 10 in Ref. 1, 18th line from the bottom in left side on page 11 in Ref. 1, captions in Figs. 6–10 in Ref. 1, and 10th line from the bottom in right side in Ref. 2, we replace “ ¼ 0:01 eV” by “ ¼ 0:01 eV”. (iv) We replace the whole of Ref. 37 cited in Ref. 1 by “In Sect. 3, we roughly set =H 0:01 and = 0:1 for strong ferromagnets on the basis of the magnitude relation between H, Δ, and λ. First, we adopt H 1 eV for typical ferromagnets (see Sect. 13.1 in the literature cited in Ref. 36). Second, we use 0:1 eV for fcc-Fe in the ferromagnetic state. Here, fcc-Fe in the ferromagnetic state is regarded as a simple system that is similar to Fe₄N. In addition, Δ is evaluated by fitting the dispersion curves obtained by the tight-binding model to those obtained by the first-principles calculation. This calculation method was described in Ref. 47. Finally, regarding λ, we accurately obtain ¼ 0:052 eV for Fe²⁺ as described in Sect. 4. Meanwhile, in Sect. 3 we aim to observe only the qualitative properties of C₂ and C₄ for strong ferromagnets. In addition, properties of C₂ and C₄ are determined by = and =H, and the above H 1 eV and 0:1 eV are the roughly-evaluated values. In Sect. 3, we therefore choose ¼ 0:01 eV (i.e., =H 0:01 and = 0:1). Note that, in Sect. 4, we use ¼ 0:052 eV to analyze C₂ and C₄ for Fe₄N.”