Abstract
We found errors in the higherorder (HO) terms with ð=Þ^{2} in ^{ð} _{0} ^{2} _{;} ^{Þ} of Eq. (44), ^{ð} _{2} ^{2} _{;} ^{Þ} of Eq. (46), ^{ð} _{4} ^{2} _{;} ^{Þ} of Eq. (47), C_{4} of Eq. (62), and c^{0}_{4} of Eq. (63) in the original paper.^{1)} The HO terms did not have a crucial role in the paper. In particular, the HO term in C_{4} (i.e., c^{0}_{4}) was finally ignored because of its small value. Therefore, this correction does not affect the main results of the paper. (Equation Presented) For C_{4} of Eq. (62) consisting of ^{ð} _{4} ^{2} _{;} ^{Þ} , “þc^{0}_{4}” in the righthand side of Eq. (62) must be deleted. In addition, c^{0}_{4} of Eq. (63) should be deleted. On the basis of the correction of C_{4}, we make the following corrections: (I) Sentences on lines 9–14 in Sect. 3.1(ii) starting from “Also, c^{0}_{4} of Eq. (63) …” should be deleted. (II) Another paragraph in Sect. 3.1 starting from “With regard to C_{4},…” must be deleted. Note here that Eq. (71) is regarded as Eq. (62) excluding c^{0}_{4}. (III) The dotdashed curves in Fig. 6 should be deleted. (IV) A paragraph in Sect. 3.1 starting from “We next determine …” must be replaced by “We next determine the effective value of the undefined parameter = by comparing C_{4} obtained by PT with that obtained by the EDM. Note that the value of = is necessary for calculations of C_{2} and C_{4} by the EDM [see Eqs. (58) and (59)], although = is not included in C_{2} and C_{4} obtained by the PT [see Eqs. (61) and (62)]. We here put r_{"}1 = r_{"}ð1 þ Þ; ð4Þ r_{"}2 = r_{"}; ð5Þ where η represents the difference between r_{"1}=r_{"} and r_{"2}=r_{"}. Figure 4 shows the jj= dependence of C_{4} of Eqs. (62) and (59) for the systems with H = 1 eV, Δ = 0:1 eV, λ = 0:01 eV,^{37Þ} _{"}= ¼ 1, r ¼ 0, r ¼ 0:01,^{43Þ} r_{"}=r ¼ 1, and η ¼ 0, 1, and 2. Here, r ¼ 0 and r ¼ 0:01 are set on the basis of those for Fe_{4}N.^{19Þ} The range of jj= is roughly assumed to be 1=2 ≤ λ/δγ ≤ 3/2 considering the aforementioned parameters and δγ/Δ ≪ 1. We find that C_{4} obtained by the EDM is nearly constant and is close to that obtained by the PT [i.e., Eq. (71)] at each η. Therefore, jj= is considered to be valid for 1/2 ≤ λ/δγ ≤ 3/2. We choose λ/δγ = 1/2 in this study.”
Original language  English 

Article number  108001 
Journal  journal of the physical society of japan 
Volume  86 
Issue number  10 
DOIs 

Publication status  Published  2017 Oct 15 
ASJC Scopus subject areas
 Physics and Astronomy(all)