Correction to “reentrant Spin Glass and Large Coercive Field Observed in a Spin Integer Dimerized Honeycomb Lattice”

Yuan‐Qi Zhai,Yi‐Fei Deng,Zhendong Fu,Erxi Feng,Y. Su,Takuya Shiga,Hiroki Oshio,Z. Klencsár,Junhu Wang,Yan‐Zhen Zheng

DOI: https://doi.org/10.1002/adfm.202111468

IF: 19

2023-01-01

Advanced Functional Materials

Abstract:Adv. Funct. Mater. 2021, 31, 2004744 DOI: 10.1002/adfm.202004744 In the originally published version of the article 57Fe Mössbauer spectra (Figure 7) measured of FEN at temperatures of 25 K, 20 K and 14 K (for which significant magnetic hyperfine interactions lead to an especially complicated spectral shape) were analyzed formally with three Lorentzian doublets. In Table S3 of the Supporting Information the numerical values of the center and distance of these doublets were erroneously listed under the same notations as the isomer shift and quadrupole splitting of the single quadrupole doublet component fitted to the spectra measured at higher temperatures. While the formal fit of the spectra with 3 doublet components can describe the spectral shape with fair accuracy, the center and distance of the peaks in these doublets cannot be interpreted as isomer shift and quadrupole splitting Mössbauer parameters. In this correction we present the results achieved by applying a revised fitting methodology for the analysis of the Mössbauer spectra recorded of FEN in the temperature range of 14⋯60 K, by considering two additional spectra with respect to the case included in the original article. A more detailed account of the results achieved via the revised fitting procedure is available from the authors on request. Apart from revisions related to the Mössbauer analysis results, corrections concerning minor errors found elsewhere in the original text is presented here as well. In the frame of the present correction the authors of the original article are joined by two new coauthors, who participated in the establishment of the revised Mössbauer spectrum fitting procedure and in the preparation and critical reading of the correction (Z.K. and J.W.). The list of references is extended by one additional reference[1] in connection with the software that was applied to fit the Mössbauer spectra. The main conclusions of the original work remain unaffected by the corrections. The corrections are as follows. (1) On page 6 the sentences The straight-line fit gives τ0 = −10−11.4(3) s and zν = 9.2(3). The value of the relaxation time τ0 = −10−11.4(3) s is located in the characteristic range of 10−10–10−13 s usually derived for canonical spin glasses. are replaced with The straight-line fit gives τ0 = 10−11.4(3) s and zν = 9.2(3). The value of the relaxation time τ0 = 10−11.4(3) s is located in the characteristic range of 10−10–10−13 s usually derived for canonical spin glasses. (2) The following paragraph is added to the “4. Experimental Section”. The 57Fe Mössbauer spectra of the samples were recorded in transmission geometry by using a 57Co/Rh (925 MBq) source. The spectrometer (made by Topologic Systems) was operated in constant-acceleration mode. The velocity axis of Mössbauer spectra was calibrated on the basis of the spectrum of α-Fe foil kept at ambient temperature. The isomer shift values are quoted with respect to that of α-Fe at room temperature. An Iwatani HE05/CW404 cryostat was used to measure the sample at low temperatures between 14 K and 60 K. The sample contained 18 mg Fe / cm2 in a finely powdered sample of FEN dispersed in boron nitride. The spectra were fitted by using the MossWinn 4.0i software.[1] (3) Figure 7 is replaced with the following figure. (4) The following text replaces the section “2.4. 57Fe Mössbauer Spectroscopy” 2.4. 57Fe Mössbauer Spectroscopy FEN was characterized by 57Fe Mössbauer spectroscopy [51] in the temperature range of 14⋯60 K (Figure 7). For T ≥ 35 K the main component of the spectra is apparently a quadrupole doublet characterized with isomer shift of δ ≈ 1.2 mm s−1 and quadrupole splitting of ΔEQ ≈ 2.7 mm s−1, originating clearly from high-spin Fe2+ ions and thereby confirming the successful preparation of the sample consistently with the results of magnetic measurements. A smaller quadrupole doublet component, with a relative spectral area fraction being below 3.5%, can be identified as high-spin Fe3+ with parameters of δ(T = 60 K) ≈ 0.52 mm s−1 and ΔEQ(T = 60 K) ≈ 0.62 mm s−1 (Table S3), which was presumably formed via low-level oxidation of the sample during its handling. The Fe2+ doublet component displays an asymmetry concerning the areas of the contributing peaks: the area of the peak towards negative velocities exceeds that of the peak centered at ∼2.54 mm s−1 by ∼8.5(2)% of the latter. As the asymmetry does not change appreciably in the temperature range of 35⋯60 K, we attribute it to a minor level of texture present in the measured sample. Based on Figure 7, components with magnetic splitting appear in the spectra at temperatures 30 K and below. By considering the T1/2 ≈ 10−7 s half-life of the Mössbauer excited state of the 57Fe nucleus along with a Larmor precession time of similar magnitude, this is consistent with Eq. (4) according to which the minimum spin relaxation frequency fmin = 1/(2πτm) = 107 Hz is associated with a spin freezing temperature of Tf ≈ 31.2 K (by using the values of τ0 = 10−11.4 s, Tg = 22.2 K and zv = 9.2). This corroborates that the development of magnetic splitting in the Mössbauer spectra of FEN is due to the slowing down of spin fluctuations as temperature approaches the glass transition temperature from above. With decreasing temperature the doublet component gradually transforms to magnetic components, clearly reflecting the presence of a range of Fe2+ spin fluctuation frequencies especially at the temperatures of 25 K and 30 K (Figure 7). The broadening and profile of the main magnetic Fe2+ component present in the spectra for T ≤ 20 K could be best accounted for by assuming the presence of two distinct Fe2+ magnetic components (Table S3) that differ mainly in the apparent magnitude (Bhf) and local orientation (polar angle β (Bhf), measured with respect to the z axis of the local eigensystem of the electric field gradient tensor, EFG) of the hyperfine magnetic field (Bhf). In both cases a deviation from axial symmetry of the EFG (characterized by an asymmetry parameter of η = 0) is clearly evidenced. The differences in the isomer shift (δ) and the main component of the EFG tensor (Vzz) of these components are mild for T ≤ 20 K, suggesting that at these temperatures there are no substantial differences between the electronic structure and ligand environment of Fe2+ ions contributing to these two components. At the same time, substantial differences are indicated in Bhf and β(Bhf), with the latter taking on ∼54° and ∼67° at 14 K for the two distinct magnetic components. These angles represent preferred local orientations (with respect to the EFG) of Bhf in the system, which are also expected to be connected to the development of the proposed spin canting topology. An asymmetric, broadened doublet component still appears in the spectra for T ≤ 20 K, with a relative spectral area dropping to ∼7% at 14 K from ∼ 14% at 20 K, further corroborating that the spin relaxation of Fe2+ ions slows down gradually with the decrease of temperature. For T ≤ 25 K the minor Fe3+ doublet detected at higher temperatures appears to transform gradually to a magnetic component that needs to be taken into account in order to achieve an acceptable fit of the spectra. On the basis of the decomposition of spectra measured at T ≤ 20 K, the electronic and magnetic state of Fe2+ ions appears to be associated mainly with two different iron microenvironments. To elucidate the possible presence of corresponding different Fe2+ states also at higher temperatures, an attempt was made to resolve the doublet feature in the spectra measured at T ≥ 25 K (Figure 7) into two doublet components characterized with possibly different isomer shift and/or Vzz values (Table S3). It was found that the decomposition of the doublet feature into two doublet components is possible in a consistent manner for all the spectra measured at T ≥ 25 K, and that this decomposition results in a considerable improvement in the goodness of the fit with respect to the decomposition to one doublet only. Apart from their relative area fractions and line widths, the main difference between the two doublet components turned out to be in the value of the Vzz. At 60 K the two components represent similar proportions of Fe2+ ions in the sample, suggesting that the corresponding microenvironments can be associated with neighboring [Fe2] units. (5) Table S3 in the original Supporting Information is replaced with the following table. (6) The original References section is complemented with the following reference.