Densities of states in Fe-doped III-V semiconductors: a first-principles study

Atsushi Fujimori; Shoya Sakamoto

REVIEW 3 major objections 5 minor 45 references

Densities of states in Fe-doped III-V semiconductors: a first-principles study

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The onset of ferromagnetism in Fe-doped III-Vs tracks the filling of one minority-spin d level

desk verdict Systematic Fe-doped III-V DFT survey with a clean Fe3+/Fe2+ threshold correlation, but the ferromagnetism mechanism is inferred, not computed. read the letter →

arxiv 1908.02311 v1 pith:ONJOIIII submitted 2019-08-06 cond-mat.str-el cond-mat.mtrl-sci

Shoya Sakamoto , Atsushi Fujimori This is my paper · ORCID

classification cond-mat.str-elcond-mat.mtrl-sci

keywords Fe-dopedIII-Vsemiconductorsdensityofstatesfirst-principlessupercellcalculationferromagneticFe3dimpuritylevelsp-dhybridizationchargetransferlevelcrystalfield

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Using supercell density-functional calculations for nine Fe-doped III-V hosts, the authors try to establish a common electronic-structure picture: replacing Mn with Fe adds one electron, and that electron sits either in majority-spin p-d antibonding states (the Fe3+ configuration) or, in hosts whose valence band lies high enough, in a minority-spin e state (Fe2+). Their central claim is that ferromagnetism appears precisely when the minority-spin $e_{\downarrow}$ state starts to be occupied, which separates ferromagnetic hosts such as GaSb and InSb from paramagnetic hosts such as GaAs. They additionally show that the conduction-band spin splitting from s-d hybridization is far smaller than the valence-band splitting from p-d hybridization, so the familiar s,p-d exchange mechanism cannot explain the high Curie temperatures even in n-type compounds. A reader would care because this gives a simple orbital-occupation criterion for guessing which doped hosts will be ferromagnetic.

What carries the argument

The load-bearing object is the crystal-field level scheme around a substitutional Fe atom, shown in Fig. 1: tetrahedral splitting puts the Fe 3d states into lower e and higher t2 groups, and the t2 levels hybridize strongly with the ligand p orbitals into bonding ($t_b$) and antibonding ($t_a$) levels while the e levels remain essentially nonbonding. This scheme carries the whole argument because it determines the electron configuration: the majority-spin antibonding level $t_{a,\uparrow}$ is full in Fe3+ hosts, and the minority-spin $e_{\downarrow}$ level starts to fill just where the valence band is high enough. The paper uses it to classify all nine hosts, to read off the spin splitting of valence and conduction bands, and to locate the ferromagnetic/paramagnetic boundary at the onset of $e_{\downarrow}$ occupation.

What would settle it

Look for occupancy of the $e_{\downarrow}$ state in a material the paper classifies as Fe3+, for example GaAs or InP, using element-specific resonant photoemission or inverse photoemission; finding a partially occupied $e_{\downarrow}$ in a paramagnetic host, or an empty $e_{\downarrow}$ in a ferromagnetic host, would break the claimed correlation. A cheaper check is to repeat the supercell calculation with a functional that corrects the d-level position and see whether the Fe3+/Fe2+ boundary moves across any of the nine hosts.

Watch

Extended reading notes

Core claim

The central discovery, stated on the paper's own terms, is that Fe in a III-V host does not behave like a simple acceptor: its extra electron relative to Mn lands in one of two places depending on the host's band offsets. In AlP, AlAs, and InP the electron fills the majority-spin antibonding level $t_{a,\uparrow}$, giving the Fe3+ configuration with a total moment of 5 $\mu_B$ per supercell. In GaSb and InSb, and partially in GaAs, InAs, and AlSb, the valence band is high enough that the electron instead occupies the minority-spin level $e_{\downarrow}$, the Fe2+ configuration, reducing the total moment by about 1 $\mu_B$. The paper's key correlation, read from a band-offset diagram anchored at the Fe3+/2+ charge-transfer level of GaAs, is that the hosts with occupied $e_{\downarrow}$ states are exactly the ferromagnetic ones, so ferromagnetism appears when the $e_{\downarrow}$ states start to be occupied. The companion result is that the conduction-band splitting $\Delta E_c$ is an order of magnitude smaller than the valence-band splitting $\Delta E_v$, which the authors take as evidence that s,p-d exchange is not responsible for the high-temperature ferromagnetism even in n-type materials.

Load-bearing premise

The calculation assumes that a generalized-gradient density functional places the Fe 3d levels at the right energies relative to the host valence band in all nine semiconductors, so the Fe3+/Fe2+ boundary, anchored at GaAs's charge-transfer level, classifies every host correctly.

Editorial extensions

If this is right

Hosts with a valence-band maximum above roughly the GaAs Fe3+/2+ charge-transfer line tend to have occupied $e_{\downarrow}$ states and are the natural candidates for ferromagnetic behavior; hosts below the line stay paramagnetic.
The total magnetic moment per Fe is about 5 $\mu_B$ in the Fe3+ limit and drops by about 1 $\mu_B$ when $e_{\downarrow}$ is occupied, so total-moment measurements can fingerprint the charge configuration.
Because $\Delta E_c$ is much smaller than $\Delta E_v$, high Curie temperatures in n-type (In,Fe)As and (In,Fe)Sb cannot be blamed on s-d exchange; the ordering mechanism must be sought elsewhere, for example in double exchange among partly filled $e_{\downarrow}$ orbitals plus short-range superexchange.
The measured conduction-band splitting from tunneling spectroscopy in (In,Fe)As, roughly 32 and 50 meV at 6 and 8 percent Fe doping, agrees with the calculated ~20 meV at 3.7 percent doping, so the s-d exchange strength is consistently weak.
In the Fe2+ limit the electron count would suggest hole doping, but the observed carrier densities are orders of magnitude lower, implying that the carriers are trapped in Fe-rich regions and transport occurs by hopping between them.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the $e_{\downarrow}$-occupation criterion is right, then shifting the Fermi level by doping should move a fixed host across the ferromagnetic boundary; a doping series in GaAs or AlSb would test this without changing the host chemistry.
The same boundary logic suggests a design rule the paper states only implicitly: alloying or strain that raises the host valence band relative to the Fe d levels, or lowers the d levels, should push more hosts into the ferromagnetic regime; the Sb-based hosts already sit there.
Because the calculations use isolated Fe in a 3x3x3 supercell, real Fe-rich nanoclusters could make local $e_{\downarrow}$ occupation happen even in hosts classified as Fe3+, blurring the global boundary and connecting to the observed nanoscale ferromagnetic domains.
A many-body correction to the d-level positions would probably shift the Fe3+/Fe2+ line; if it moved enough to reclassify one of the nine hosts, the correlation with ferromagnetism would need to be rechecked.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Dear colleague,

The short version: this is a useful systematic DFT survey of nine Fe-doped III-V hosts, and the cross-material correlation between Fe3+/Fe2+ configuration and ferromagnetism is worth knowing. The paper is less solid on mechanism: the conclusion that e-down occupation drives ferromagnetism is inferred from a small correlation, not from direct magnetic coupling calculations. I would send it to a referee, but with clear instructions that the causal claim needs softening or more evidence.

What is actually new: the authors compute spin-resolved DOS for a consistent set of nine hosts, classify each as Fe3+, Fe2+, or intermediate, and show a band-offset figure where the Fe3+/2+ charge-transfer level in GaAs works as a threshold: hosts with VBM above that line tend to be ferromagnetic. That is a clean, usable observation. The comparison to (In,Fe)As tunneling data for conduction-band splitting is a nice sanity check and agrees reasonably. The discussion of transport being incompatible with simple Fe2+ hole counting, and pointing to trapped carriers or nanoscale inhomogeneity, is honest and grounded in the cited experiments.

The soft spot is exactly where the stress-test note lands. The e-down-onset criterion is not independently established as causal. The dashed line in Fig. 3 is pinned to the experimental GaAs charge-transfer level, and the host classification is essentially determined by literature VBM offsets. So 'e-down starts to be occupied' and 'VBM above the GaAs threshold' are nearly the same statement. Since the supercell contains only one Fe atom, no exchange constant is computed; the double-exchange idea is plausible but untested. And PBE-GGA is not benchmarked here for d-level placement, so a small systematic error could shift InAs or AlSb across the boundary. The paper mostly hedges with 'might appear' and 'may suggest', but the conclusion leans harder on the mechanism than the evidence supports.

That said, the paper is honest about its limits, cites the relevant experimental literature well, and does not oversell. The DOS data are presumably reproducible from the described methods, though no input files are included. I think this deserves a serious referee: the systematic host comparison is genuinely useful for the dilute ferromagnetic semiconductor community, and the Fe3+/Fe2+ boundary figure will likely be cited. For peer review, I would ask for either direct exchange-constant calculations for representative Fe pairs, or a clear statement that the e-down criterion is a threshold indicator rather than a demonstrated mechanism. The paper is short and readable; a revision along those lines would make it solid.

Cheers.

Referee Report

3 major / 5 minor

Summary. This manuscript reports PBE-GGA supercell calculations of the spin-resolved densities of states of Fe-doped III-V semiconductors (AlP, GaP, InP, AlAs, GaAs, InAs, AlSb, GaSb, InSb) with one substitutional Fe per 3×3×3 cell. The authors show that the extra electron of Fe relative to Mn occupies either majority-spin p-d antibonding states (ta,↑) or minority-spin e states (e↓) depending on the host, and they classify the hosts into Fe3+, Fe2+, and intermediate configurations. The central claim is that ferromagnetism appears when the e↓ states start to be occupied, and the paper further argues that s,p-d exchange is not the origin of high-Tc ferromagnetism because the calculated conduction-band splitting is much smaller than the valence-band splitting. The paper includes a comparison of the computed ΔEc for (In,Fe)As with tunneling spectroscopy and a discussion of alternative mechanisms such as double exchange, superexchange, and nanoscale phase separation.

Significance. The paper provides a systematic and reproducible first-principles dataset for the electronic structure of Fe-doped III-V semiconductors, using a standard all-electron code with consistent parameters. A valuable quantitative result is the agreement between the calculated conduction-band splitting (ΔEc ≈ 20 meV at 3.7% Fe) and the tunneling-spectroscopy value of 31.7–50 meV at 6–8% Fe, expressed through N0α ≈ 0.22 eV. The proposed e↓-occupancy criterion for ferromagnetism is a falsifiable electronic-structure marker that, if confirmed, could guide the search for new ferromagnetic hosts. However, the criterion is currently inferred from a correlation across only nine hosts and is not directly supported by calculations of magnetic exchange interactions, so its significance is contingent on further validation.

major comments (3)

[Section III, Fig. 3] The central claim that ferromagnetism appears when e↓ states start to be occupied is inferred from a correlation between the calculated Fe charge state and experimental Curie temperatures across nine hosts. Because the dashed Fe3+/2+ line is anchored to the experimental charge-transfer level in GaAs and the hosts are ordered by their literature valence-band offsets, the classification into Fe2+/intermediate versus Fe3+ is essentially a re-labeling of the host VBM ordering. Since the VBM position is already known to correlate with ferromagnetism in dilute magnetic semiconductors, the paper should demonstrate that e↓ occupation provides predictive power beyond the VBM alone. The present calculations contain only one Fe atom per supercell and therefore do not compute any Fe–Fe exchange coupling, so the causal role of e↓ occupancy is not established.
[Section III, Table II and Fig. 3] The placement of the Fe3+/Fe2+ boundary relies on the PBE-GGA d-level position relative to each host VBM, but no benchmark is provided against calculations with a Hubbard U or hybrid functional. The authors themselves note that the dashed line is only approximate. Given that a rigid shift of ~0.1–0.2 eV in the Fe 3d level could move InAs or AlSb across the line and destroy the correlation, the paper should quantify the sensitivity of the charge-state assignment to the level alignment, for example by testing one or two hosts with a +U correction or by examining the dependence on the chosen experimental charge-transfer level.
[Section IV] The discussion of double-exchange as the likely ferromagnetic mechanism is speculative: the paper does not compute any exchange constants (J_ij) or total-energy differences for different magnetic configurations. The statement that 'the fact that the TC tends to be higher in Sb-based material ... may suggest that double-exchange interaction is more likely to be responsible' is not supported by the present calculations. Either the authors should perform direct calculations with two Fe atoms to estimate the coupling as a function of e↓ occupancy, or they should explicitly frame this as an open hypothesis and remove the implication that their DOS results resolve the mechanism.

minor comments (5)

[Fig. 2 caption] The caption contains garbled text ('/s32/s33/s34/s35' and similar), apparently a figure-formatting artifact. The caption should be regenerated so that the labels for the total DOS and PDOS curves are readable.
[Abstract and Conclusion] The abstract says the calculations 'suggest' that ferromagnetism appears when e↓ starts to be occupied, while the conclusion says they 'implied' the same. The strength of the claim should be consistent; given the lack of exchange-coupling calculations, the weaker wording is more appropriate.
[Section IV, ref. [34]] The text states that Shinya et al. found the second-nearest-neighbor superexchange to be ferromagnetic but too small to account for high TCs. It would be clearer to give the magnitude of the interaction reported in that work, so the reader can judge the quantitative claim.
[References] Several references are incomplete: [29] lacks an article number or page range, and [38] is an arXiv preprint rather than a published article. The authors should update these entries if published versions exist.
[Eq. for ΔEv] The expression ΔEv ≈ t_pd^2 / (E_p↑ - E_d↑) is introduced without a derivation or reference; a brief explanation of the notation and the origin of the approximation (e.g., perturbation theory for p-d hybridization) would improve accessibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the DOS results and the e-down-occupancy correlation use independent inputs and no parameter is fitted to the ferromagnetism conclusion.

full rationale

The central electronic-structure results come from WIEN2k DFT supercell calculations with PBE-GGA and experimental lattice constants; the Fe3+/Fe2+ classification is read from the calculated spin-resolved partial DOS, not from the magnetization labels. The dashed Fe3+/2+ line in Fig. 3 is anchored to the measured Fe3+/2+ charge-transfer level in GaAs [26] and to literature band offsets [24,25], while the ferromagnetic/paramagnetic classification and Tc values are independent experimental data. No parameter is adjusted to reproduce the magnetization outcome, so the statement 'ferromagnetism might appear when the Fe e_down level starts being occupied' is a post-hoc correlation drawn from independent data, not a fitted input renamed as a prediction. Self-citations [30,31,38] provide XMCD measurements and nanoscale-domain observations; these are external experimental evidence, not load-bearing circular support. The acknowledged PBE-GGA placement uncertainty is a correctness risk, not a circularity, and the paper's cautious 'might/may suggest' wording reflects this. The derivation chain is self-contained and non-circular.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The central claim rests on standard DFT practice and one external empirical input, the GaAs Fe3+/2+ charge transfer level used to draw the boundary in Fig. 3. No new particles, forces, or entities are introduced. The main scientific assumption is that PBE-GGA correctly places the Fe d states relative to the host bands, which is not tested with higher-level functionals.

free parameters (1)

Fe3+/2+ charge transfer level (dashed line in Fig. 3) = positioned to match GaAs, from ref [26]
The dashed line is placed using the known Fe3+/2+ charge transfer level in GaAs and is then used to classify the other hosts as Fe3+ or Fe2+. The central 'ferromagnetism appears when e_down states are occupied' correlation depends on this placement.

assumptions (4)

domain assumption PBE-GGA exchange-correlation functional reliably describes the relative positions of Fe 3d and host p states in these dilute magnetic semiconductors.
The entire Fe3+/Fe2+ assignment in Table II and Fig. 1 depends on the DFT eigenvalue positions; GGA is known to underestimate band gaps and may misplace 3d levels, so this is a load-bearing assumption introduced in the Methods section.
domain assumption A 3x3x3 supercell with one Fe atom (3.7% doping) adequately represents the isolated-impurity limit and the electronic structure at measured higher doping levels.
The paper assumes no significant Fe-Fe interaction at this concentration and compares to experiments at 6-8% doping using N0-alpha scaling, as stated in Section III.
domain assumption The Kohn-Sham density of states can be directly interpreted as the single-particle excitation spectrum relevant to measured band splittings.
The calculated Delta-E_v and Delta-E_c are compared to experimental values in Table II; DFT eigenvalues are not true excitation energies, though this is common practice in this literature.
domain assumption Experimental lattice constants, rather than relaxed ones, are appropriate for these calculations.
The paper uses experimental lattice constants 'for simplicity', which may affect the hybridization strength and hence the Fe3+/Fe2+ boundary, as stated in Section II.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Densities of states in Fe-doped III-V semiconductors: a first-principles study." pith.science (2026). https://pith.science/paper/ONJOIIII

@misc{pith2026190802311,
  author       = {Pith},
  title        = {Pith review of: Densities of states in Fe-doped III-V semiconductors: a first-principles study},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ONJOIIII}},
  note         = {Machine review of arXiv:1908.02311}
}

abstract

The electronic structures of Fe-doped III-V semiconductors were studied by first-principles supercell calculation. It was found that their electronic structures are basically the same as those of Mn-doped ones except that the extra electron of Fe compared to Mn occupies either majority-spin $p$-$d$ hybridized antibonding states ($t_{a,\uparrow}$) or minority-spin $e$ states ($e_{\downarrow}$) and that the center of gravity of the $d$ partial density of states is higher for Fe than for Mn. The present calculations suggest that ferromagnetism appears when the $e_{\downarrow}$ states start to be occupied. The band splitting due to $s$-$d$ hybridization was found to be significantly smaller than the one due to $p$-$d$ hybridization. This indicates that the $s,p$-$d$ exchange interaction is not responsible for the high-temperature ferromagnetism of the Fe-doped ferromagnetic semiconductors even in $n$-type compounds.

Figures

Figures reproduced from arXiv: 1908.02311 by the authors.

**Figure 2.** FIG. 2. Total and partial densities of states of Fe-doped III [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** FIG. 3. Band offsets of III-V semiconductors [24] and Ge [25] wi [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 44 canonical work pages

[1]

Munekata, H

H. Munekata, H. Ohno, S. von Molnar, A. Segm¨ uller, L. L. C hang, and L. Esaki, Phys. Rev. Lett. 63, 1849 (1989)

work page 1989
[2]

H. Ohno, A. Shen, F. Matsukura, A. Oiwa, A. Endo, S. Katsum oto, and Y. Iye, Appl. Phys. Lett. 69, 363 (1996)

work page 1996
[3]

Jungwirth, J

T. Jungwirth, J. Wunderlich, V. Nov´ ak, K. Olejn ´ ık, B. L. Gallagher, R. P. Campion, K. W. Edmonds, A. W. Rushforth, A. J. Ferguson, and P. Nˇ emec, Rev. Mod. Phys. 86, 855 (2014)

work page 2014
[4]

Dietl and H

T. Dietl and H. Ohno, Rev. Mod. Phys. 86, 187 (2014)

2014
[5]

Tanaka, S

M. Tanaka, S. Ohya, and P. Nam Hai, Appl. Phys. Rev. 1, 011102 (2014)

work page 2014
[6]

Dietl, K

T. Dietl, K. Sato, T. Fukushima, A. Bonanni, M. Jamet, A. B arski, S. Kuroda, M. Tanaka, P. N. Hai, and H. Katayama-Yoshida, Rev. Mod. Phys. 87, 1311 (2015)

work page 2015
[7]

Nam Hai, L

P. Nam Hai, L. Duc Anh, S. Mohan, T. Tamegai, M. Kodzuka, T. Ohkubo, K. Hono, and M. Tanaka, Appl. Phys. Lett. 101, 182403 (2012)

work page 2012
[8]

N. T. Tu, P. N. Hai, L. D. Anh, and M. Tanaka, Appl. Phys. Let t. 105, 132402 (2014). 11

work page 2014

Show all 45 references

[9]

L. D. Anh, D. Kaneko, P. N. Hai, and M. Tanaka, Appl. Phys. L ett. 107, 232405 (2015)

2015
[10]

N. T. Tu, P. N. Hai, L. D. Anh, and M. Tanaka, Appl. Phys. Ex press 11, 063005 (2018)

2018
[11]

Y. K. Wakabayashi, Y. Ban, S. Ohya, and M. Tanaka, Phys. R ev. B 90, 205209 (2014)

2014
[12]

N. T. Tu, P. N. Hai, L. D. Anh, and M. Tanaka, Appl. Phys. Le tt. 108, 192401 (2016)

2016
[13]

L. Chen, X. Yang, F. Yang, J. Zhao, J. Misuraca, P. Xiong, and S. von Moln´ ar, Nano Lett. 11, 2584 (2011)

2011
[14]

Y. Ban, Y. Wakabayashi, R. Akiyama, R. Nakane, and M. Tan aka, AIP Adv. 4, 097108 (2014)

2014
[15]

Jungwirth, J

T. Jungwirth, J. Sinova, J. Maˇ sek, J. Kuˇ cera, and A. H. MacDonald, Rev. Mod. Phys. 78, 809 (2006)

2006
[16]

K. Sato, L. Bergqvist, J. Kudrnovsk´ y, P. H. Dederichs, O. Eriksson, I. Turek, B. Sanyal, G. Bouzerar, H. Katayama-Yoshida, V. A. Dinh, T. Fukushima, H. Kizaki, and R. Zeller, Rev. Mod. Phys. 82, 1633 (2010)

2010
[17]

Sato and H

K. Sato and H. Katayama-Yoshida, Jpn. J. Appl. Phys. 39, L555 (2000)

2000
[18]

L. M. Sandratskii and P. Bruno, Phys. Rev. B 66, 134435 (2002)

2002
[19]

A. B. Shick, J. Kudrnovsk´ y, and V. Drchal, Phys. Rev. B 69, 125207 (2004)

2004
[20]

Wierzbowska, D

M. Wierzbowska, D. S´ anchez-Portal, and S. Sanvito, Ph ys. Rev. B 70, 235209 (2004)

2004
[21]

Filippone, G

F. Filippone, G. Mattioli, P. Alippi, and A. A. Bonapast a, Phys. Rev. Lett. 107, 196401 (2011)

2011
[22]

Blaha, K

P. Blaha, K. Schwarz, G. Madsen, D. Kvasnicka, and J. Lui tz, An augmented plane wave+ local orbitals program for calculating crystal properties (2001)

2001
[23]

J. P. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett . 77, 3865 (1996)

1996
[24]

Vurgaftman, J

I. Vurgaftman, J. R. Meyer, and L. R. Ram-Mohan, J. Appl. Phys. 89, 5815 (2001)

2001
[25]

Franciosi and C

A. Franciosi and C. G. V. de Walle, Surf. Sci. Rep. 25, 1 (1996)

1996
[26]

Malguth, A

E. Malguth, A. Hoﬀmann, and M. R. Phillips, Phys. Status S olidi B 245, 455 (2008)

2008
[27]

Haneda, M

S. Haneda, M. Yamaura, Y. Takatani, K. Hara, S. ichi Hari gae, and H. Munekata, Jpn. J. Appl. Phys. 39, L9 (2000)

2000
[28]

Moriya, Y

R. Moriya, Y. Katsumata, Y. Takatani, S. Haneda, T. Kond o, and H. Munekata, Physica E: Low-dimensional Systems and Nanostructures 10, 224 (2001), proceedings of the First International Conference on the Physics and Applicat ions of Spin-Related Phenomena in Semiconductors. 12

2001
[29]

P. N. H. Le Duc Anh and M. Tanaka, Nat. Commun. 7 (2016)

2016
[30]

Sakamoto, L

S. Sakamoto, L. D. Anh, P. N. Hai, G. Shibata, Y. Takeda, M . Kobayashi, Y. Takahashi, T. Koide, M. Tanaka, and A. Fujimori, Phys. Rev. B 93, 035203 (2016)

2016
[31]

Sakamoto, N

S. Sakamoto, N. T. Tu, Y. Takeda, S.-i. Fujimori, P. N. Ha i, L. D. Anh, Y. K. Wakabayashi, G. Shibata, M. Horio, K. Ikeda, Y. Saitoh, H. Yamagami, M. Tan aka, and A. Fujimori, Phys. Rev. B 100, 035204 (2019)

2019
[32]

N. T. Tu, P. N. Hai, L. D. Anh, and M. Tanaka, Phys. Rev. B 92, 144403 (2015)

2015
[33]

Kudrin, V

A. Kudrin, V. Lesnikov, Y. Danilov, M. Dorokhin, O. Vikh rova, D. Pavlov, Y. Usov, I. Antonov, R. Kriukov, S. Zubkov, D. Nikolichev, A. Konakov , Y. Dudin, Y. Kuznetsov, M. Temiryazeva, and N. Sobolev, J. Magn. Magn. Mater. 485, 236 (2019)

2019
[34]

Shinya, T

H. Shinya, T. Fukushima, A. Masago, K. Sato, and H. Katay ama-Yoshida, J, Appl. Phys. 124, 103902 (2018)

2018
[35]

N. D. Vu, T. Fukushima, K. Sato, and H. Katayama-Yoshida , Jpn. J. Appl. Phys. 53, 110307 (2014)

2014
[36]

Gu and S

B. Gu and S. Maekawa, Phys. Rev. B 94, 155202 (2016)

2016
[37]

Nam Hai, L

P. Nam Hai, L. D. Anh, and M. Tanaka, Appl. Phys. Lett. 101, 252410 (2012)

2012
[38]

Sakamoto, L

S. Sakamoto, L. D. Anh, P. N. Hai, Y. Takeda, M. Kobata, R. Nakane, Y. K. Wakabayashi, Y. Nonaka, K. Ikeda, Z. Chi, Y. Wan, M. Suzuki, Y. Saitoh, H. Ya magami, M. Tanaka, and A. Fujimori, arXiv:1902.03742

1902 arXiv
[39]

Y. Yuan, R. H¨ ubner, M. Birowska, C. Xu, M. Wang, S. Prucn al, R. Jakiela, K. Potzger, R. B¨ ottger, S. Facsko, J. A. Majewski, M. Helm, M. Sawicki, S . Zhou, and T. Dietl, Phys. Rev. Mater. 2, 114601 (2018)

2018
[40]

Fukushima, H

T. Fukushima, H. Shinya, A. Masago, K. Sato, and H. Katay ama-Yoshida, Appl. Phys. Express 12, 063006 (2019)

2019
[41]

Kaminski and S

A. Kaminski and S. Das Sarma, Phys. Rev. B 68, 235210 (2003)

2003
[42]

Pinto, L

N. Pinto, L. Morresi, M. Ficcadenti, R. Murri, F. D’Oraz io, F. Lucari, L. Boarino, and G. Amato, Phys. Rev. B 72, 165203 (2005)

2005
[43]

Y. Ban, Y. K. Wakabayashi, R. Nakane, and M. Tanaka, J. Ap pl. Phys. 124, 113902 (2018)

2018
[44]

M. G. Sreenivasan, K. L. Teo, X. Z. Cheng, M. B. A. Jalil, T . Liew, T. C. Chong, A. Y. Du, T. K. Chan, and T. Osipowicz, J. Appl. Phys. 102, 053702 (2007)

2007
[45]

Saito, W

H. Saito, W. Zaets, S. Yamagata, Y. Suzuki, and K. Ando, J . Appl. Phys. 91, 8085 (2002). 13

2002

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Munekata, H

H. Munekata, H. Ohno, S. von Molnar, A. Segm¨ uller, L. L. C hang, and L. Esaki, Phys. Rev. Lett. 63, 1849 (1989)

work page 1989

[2] [2]

H. Ohno, A. Shen, F. Matsukura, A. Oiwa, A. Endo, S. Katsum oto, and Y. Iye, Appl. Phys. Lett. 69, 363 (1996)

work page 1996

[3] [3]

Jungwirth, J

T. Jungwirth, J. Wunderlich, V. Nov´ ak, K. Olejn ´ ık, B. L. Gallagher, R. P. Campion, K. W. Edmonds, A. W. Rushforth, A. J. Ferguson, and P. Nˇ emec, Rev. Mod. Phys. 86, 855 (2014)

work page 2014

[4] [4]

Dietl and H

T. Dietl and H. Ohno, Rev. Mod. Phys. 86, 187 (2014)

2014

[5] [5]

Tanaka, S

M. Tanaka, S. Ohya, and P. Nam Hai, Appl. Phys. Rev. 1, 011102 (2014)

work page 2014

[6] [6]

Dietl, K

T. Dietl, K. Sato, T. Fukushima, A. Bonanni, M. Jamet, A. B arski, S. Kuroda, M. Tanaka, P. N. Hai, and H. Katayama-Yoshida, Rev. Mod. Phys. 87, 1311 (2015)

work page 2015

[7] [7]

Nam Hai, L

P. Nam Hai, L. Duc Anh, S. Mohan, T. Tamegai, M. Kodzuka, T. Ohkubo, K. Hono, and M. Tanaka, Appl. Phys. Lett. 101, 182403 (2012)

work page 2012

[8] [8]

N. T. Tu, P. N. Hai, L. D. Anh, and M. Tanaka, Appl. Phys. Let t. 105, 132402 (2014). 11

work page 2014