Possible High-Tc Superconductivity in LiMgN: A MgB2-like Material
O. P. Isikaku-Ironkwe1, 2
1
The Center for Superconductivity Technologies (TCST)
Department of Physics,
Michael Okpara University of Agriculture, Umudike (MOUAU),
Umuahia, Abia State, Nigeria
and
2
RTS Technologies, San Diego, CA 92122
Abstract
The search for superconductivity in materials iso-structural and iso-valent with magnesium
diboride, MgB2, has not yielded any results close enough to the 39K Tc of MgB 2. Lithium
magnesium nitride, LiMgN, resembles MgB2 in that they both have the same averages of
electronegativity, valence electron count and atomic number. Their formula weights are very
close too. Using our recently published chemical symmetry rules for estimating
superconductivity and Tcs of materials, we predict that LiMgN, a semiconductor at room
temperature, with the same material specific characterization dataset (MSCD) as MgB 2,
should be a superconductor, in the ordered α-phase or the orthorhombic phase, when cooled
to 39K.
Introduction
The surprising discovery of high-Tc superconductivity at 39K in MgB2[1, 2] in 2001 excited
expectations that there may be other simple and similar MgB2-like binary or ternary
superconductors with the same electronic, structural and Tc similarity. Predictions and
searches, based on iso-valent, iso-structural and DFT calculations have been undertaken both
theoretically [3, 4, 5, 6] and experimentally [7, 8, 9, 10, 11, 12, 13, 14, 15, 16], without
producing superconductors with Tcs anywhere near 39K. Many chemical-based reasons have
been advanced too on diborides superconductivity [32] but none has produced or predicted
correctly new MgB2-like superconductors with Tc near 39K.
In 2007, we decided to look at this problem from slightly different parameters. We initiated
studies [17, 18, 19], based on empirical material specific correlations of Tc with averages of
electronegativity, valence electrons, formula weight and atomic number. One of the many
materials we identified as a possible MgB2-like superconductor, based on these studies is
1
LiMgN. Further studies revealed the symmetry rules governing superconductors with similar
electronegativity, valence electrons, atomic number and formula weight. Guided by these
symmetry rules [20] we have been able to identify, using the Periodic Table only, and without
recourse to DFT calculations, many potential superconductors. We also showed how to
estimate Tc using the symmetry methods [20] and predicted 21 potential superconductors.
This paper starts by reviewing the structural and electronic properties of LiMgN and methods
for its preparation. Next we present the basis for making our prediction. We next show why
iso-structural and iso-valent features are not sufficient conditions for superconductivity. We
conclude that LiMgN should be a superconductor, based on the simple chemical symmetry
rules.
Preparation of LiMgN
Many routes have been used in the preparation of LiMgN. The preparation of LiMgN from
precursor LiMg annealed in pure nitrogen at 800 degrees centigrade for 8 hours is well
described in reference [21]. The orthorhombic structure type has been prepared from Li3N
and Mg3N2 by rapid cooling from 1000K to room temperature [22]. LiMgN can be also be
prepared by a hydriding thermal reaction between MgH2 and LiNH2 around 80 Kbar H2
pressure and 723K and a subsequent dehydrogenation reaction [23] given by:
LiNH2 + MgH2 = LiMgN + 2H2
Preparation of LiMgN as a commercial safe hydrogen storage compound is described in [24].
Properties of LiMgN
LiMgN is a member of the filled tetrahedral semiconductor family LiMgX(X= N, P, and As) with
a zincblende-like structure [21]. It can form in any of its two cubic phases known as ordered
(α) and disordered (β), with the α- LiMgN more stable, with an experimental energy band gap
of 3.2eV and cubic lattice constant of 4.995 [21]. LiMgN with an orthorhombic structure and
lattice constants, a =7.1586Å, b = 3.5069Å, and c = 5.0142Å, and space group Pnma, has also
been identified [22]. It has a cation-ordered antifluorite-type structure.
The electronic band structure of LiMgN has been extensively studied by ab initio DFT methods
[25, 26, 27, 28] and also experimentally [21, 22, 28]. One study [25] indicates very strong
covalent bonding of Li-N and Mg-N in LiMgN while the others [26, 27, 28] suggest strong ionic
2
bonding of Li-N and Mg-N in disordered β-LiMgN. Raman spectrum studies [28] confirm
antiflorite structure with space group Fm3m and that LiMgN is a filled tetrahedral
semiconductor with band-gap of 3.3eV.
Symmetry Rules: Framework for Predictions
We showed [20] that the maximum Tc of a material may be expressed in material specific
parameters of electronegativity, 𝒳, valence electrons, Ne, and atomic number, Z, given by
Tc = 𝒳
√
Ko
(1)
where Ko is a parameter that determines the value to Tc. Ko = n(Fw/Z) and n is dependent on
the family of superconductors. Fw represents formula weight of the superconductor. For
MgB2, Ko = 22.85 and Fw/Z is 6.26, making n = 3.65. Recently [20] we proposed that similar
superconductors may arise when we compare their averages of electronegativity, 𝒳 , valence
electrons, Ne, and atomic number, Z, and Fw/Z. We distinguished four possible cases, when
at least two of the features are the same, namely: (a) ⟨𝒳, Ne, Z⟩, (b) ⟨Ne, Z⟩, (c) ⟨𝒳, Ne⟩ and
(d) ⟨𝒳, Z⟩. We described such superconductors as similar. We found the symmetry rules apply
within the range 0.75< Ne/√
< 1.02 for most high-Tc superconductors. The symmetry rules,
first proposed in [20] and observable in Table 1, are:
1. Materials with exactly the same average electronegativity, valence electrons and atomic
number (case a) have the same Tc.
2. If two or more materials have the same average valence electrons Ne, and atomic number
Z, (case b) then their Tcs will be proportional to their electronegativities.
3. If two or more materials have the same average electronegativity 𝒳, and valence electrons
Ne, (case c) then their Tcs will be proportional to their average atomic numbers, Z .
4. If two or more materials have the same average electronegativity 𝒳, and average atomic
numbers, Z (case d) then their Tcs will be proportional to their average valence electrons.
Using rule 1 above, we compare the MSCDs of 12 superconductors in Table 1 with the MSCDs
of MgB2 and LiMgN, shown in Table 2. The symmetry of their MSCDs gives us strong ground
to predict that LiMgN will be a superconductor like MgB2 but with Tc of 38.4K
3
Iso-structural and Iso-valent Similarity
In the search for MgB2-like superconductors, iso-structural and iso-electronic similarities were
used for identifying likely candidates. Here we review a few examples. Using MSCD and the
above symmetry rules, it will become clear why their Tcs are far from that of MgB2. MgB2 is
hexagonal with AlB2 type structure. Table 3 gives a list of compounds iso-structural and isoelectronic with MgB2 and their Tcs. Table 4 gives the same compounds and their MSCDs. In
Table 3, we have six sets of compounds out of nine sets that are both iso-structural and isovalent with MgB2. None of them has Tc anywhere near 39K. It is interesting to observe that
none of them has the same electronegativity or atomic number as MgB 2. In Table 4, we show
the MSCDs of the compounds. We find they fall into two classes: those with Ne/√
>1.0.
There are four examples, and none of them is superconducting, as predicted by the symmetry
rules [20]. CaB2 was predicted to be a superconductor [6] with Tc higher than MgB2.
Estimating CaB2 using the symmetry rules [20], it cannot have a higher Tc since its
electronegativity and Ne/√
are less than those of MgB2. Similarly the remaining examples
in Table 4 have much lower electronegativities and Ne/√
than MgB2. We observe a
similarity of Ne and Z in ZrZn2 and SrGa2, suggesting similar properties. ZrZn2 is a known
ferromagnetic superconductor [29]. None of the thirteen compounds studied in Table 3
meets the symmetry criteria for high Tc like MgB2, based on structure and electrovalent
similarity alone.
Discussion
The search for magnesium diboride-like superconductors has revealed that we still do not
know all the parameters that control Tc [20]. Iso-structural and iso-valent similarity [3 - 16]
did not yield Tc even close to 39K (see Tables 3 and 4). This suggests that these parameters
alone were not sufficient in determining Tc. The search for decisive parameters led us to
explore the correlations of electronegativity and atomic number and formula weight with
superconductivity [20]. In that search we discovered the similarity rules with which we have
been able to predict the occurrence of superconductivity from simple periodic table
parameters without recourse to DFT calculations. The prediction of superconductivity in
4
LiMgN (and many other compounds in reference [20] ) stands out as a strong test of the
validity of those rules. Also we may note that nitrides show interesting superconductors and
may have potential high-Tc superconductors [29] like the oxides. Many examples exist of
semiconductors that superconduct [33, 34, 35] at low temperatures. It is known that ionic
bonding also occurs even in the high temperature superconductors [32], with no rules
forbidding it [30]. One of the implications of LiMgN being very similar to MgB 2 is that it may
also be a two-gapped superconductor. The total photon energy distribution of fig. 5 in ref. 27
suggests that. Experimental tests will confirm or refute these predictions.
Conclusion
Even though structurally dissimilar, MgB2 and LiMgN have the same electronegativity,
valence electrons count, atomic number and almost the same formula weight (Table 2).
LiMgN meets all the conditions [20] necessary and sufficient for two materials to have the
same Tc. We conclude that the ordered phase (α) of LiMgN in the cubic or orthorhombic
structure will be found to be superconducting with Tc between 38K and 39K.
Acknowledgements
I acknowledge stimulating discussions with A.O.E. Animalu at University of Nigeria, Nsukka,
in 2008 on LiMgN as a potential superconductor. Again, in 2011 I discussed with M.B. Maple
and J.E. Hirsch at UC San Diego and also with M. J. Schaffer, then at General Atomics, San
Diego and J.R. O’Brien at Quantum Design, San Diego. This research was supported by M. J.
Schaffer.
5
References
1. J. Nagamatsu, N. Nakagawa, T. Muranaka, Y. Zenitani and J. Akimitsu,
“Superconductivity at 39K in Magnesium Diboride,” Nature 410, 63 (2001)
2. A.S. Cooper, E. Corenzwit, L.D. longinotti, B.T. Matthias and W.H. Zachariassen, “
Superconductivity: the transition temperature peak below four electrons per
atom”, Proc. Nat. Acad. Sci , 67, 313-319 (1970)
3. N.I. Medvedeva, A.L. Ivannovskii, J.E. Medvedeva and A.J. Freeman, “Electronic
structure of superconducting MgB2 and related binary and ternary borides”, Phys.
Rev. B 64, 020502 (R) 2001 and references therein
4. H. Rosner, A. Kitaigorodsky and W.E. Pickett, “Predictions of High Tc
Superconductivity in Hole-doped LiBC”, Phys Rev. Lett. 88, 127001 (2002).
5. C. Bersier, A. Floris, A. Sanna, G. Profeta, A. Continenza, E.K.U. Gross and
“Electronic, dynamical and superconducting properties of CaBeSi”, ArXiv:
0803.1044 (2008).
6. Hyoung Jeon Choi, Steven G. Louie and Marvin L. Cohen, “Prediction of
superconducting properties of CaB2 using anisotropic Eliashberg Theory”, Phys.
Rev. B 80, 064503 (2009) and References 1 - 21 in that paper.
7. A. Bharathi, S. Jemima Balaselvi, M. Premila, T. N. Sairam, G. L. N. Reddy, C. S.
Sundar, Y. Hariharan “Synthesis and search for superconductivity in LiBC”
Arxiv:cond-mat/0207448V1 and references therein., Solid State Comm, (2002),
124, 423
8. Renker, H. Schober, P. Adelmann, P. Schweiss, K.-P. Bohnen, R. Heid ,“LiBC - A
prevented superconductor”,Cond-mat/0302036
9. A.M. Fogg, J.B. Calridge, G.R. Darling and M.J. Rossiensky “Synthesis and
characterization of LixBC---hole doping does not induce superconductivity”. Condmat/0304662v1
10. A. Lazicki, C.-S. Yoo, H. Cynn , W. J. Evans, W. E. Pickett , J. Olamit , Kai Liu , and Y.
Ohishi, “Search for superconductivity in LiBC at high pressure: Diamond anvil cell
experiments and first-principles calculations” Phys. Rev. B 75, 054507 (2007)
11. I. Felner “Absence of superconductivity in BeB2”, Physica C 353 (2001) 11 – 13.;
D.P. Young, P.W. Adams, J.Y. Chan and F.R. Franczek, “Structure and
superconducting properties of BeB2” Cond-mat/0104063
12. B. Lorenz, J. Lenzi, J. Cmaidalka, R.L. Meng, Y.Y. Sun, Y.Y. Xue and C.W. Chu,
“Superconductivity in the C32 intermetallic compounds AAl2−xSix, with A=Ca and Sr;
and 0.6<x<1.2” Physica C, 383, 191 (2002)
13. B. Lorenz, R. L. Meng, and C. W. Chu, “High-pressure study on MgB2”, Phys. Rev. B
64, 012507 (2001)
14. R.L. Meng, B. Lorenz, Y.S. Wang, J. Cmaidalka, Y.Y. Xue, J.K. Meen. C.W. Chu“Study
of binary and pseudo-binary intermetallic compounds with AlB2 structure”Physica
C: 382, 113–116(2002).
6
15. R.L. Meng, B. Lorenz, J. Cmaidalka, Y.S. Wang, Y.Y. Sun, J. Lenzi, J.K. Meen, Y.Y. Xue
and C.W. Chu, “Study of intermetallic compounds isostructural to MgB 2, IEEE Trans.
Applied Superconductivity, Vol. 13, 3042- 3046 (2002).
16. Cristina Buzea, Tsutomu Yamashita, “Review of superconducting properties of
MgB2”, Superconductor Science & Technology, Vol. 14, No. 11 (2001) R115-R146
17. O. Paul Isikaku-Ironkwe, “Electronegativity Spectrum Maps: A computational
combinatorial materials synthesis and search tool”
http://meetings.aps.org/link/BAPS.2008.MAR.K1.50http://meetings.aps.org/Meeting
/MAR08/Event/77451
18. O. Paul Isikaku-Ironkwe, “Search for Magnesium Diboride-like Binary
Superconductors” http://meetings.aps.org/link/BAPS.2008.MAR.K1.7
19. O. Paul Isikaku-Ironkwe, “Do ‘magic’ electronegativities exist for
superconductivity?”,http:/meetings.aps.org/link/BAPS.2008.MAR.K1.47
20. O. Paul Isikaku-Ironkwe, “Transition Temperatures of Superconductors estimated
from Periodic Table Properties”, Arxiv: 1204.0233 (2012)
21. K. Kuriyama, K. Nagasawa, K. Kushida, “Growth and band gap of the filled
tetrahedral semiconductor LiMgN”, J. Crystal Growth 237-239 (2002) 2019 – 2022
22. H. Yamane, T. Okabe, O. Ishiyama, Y. Waseda, M. Shimada, “Ternary nitrides
prepared in the Li3N – Mg3N2 system at 900 -1000K”, J. Alloys & Compd. 319
(2001) 124 - 130
23. Jin-Ho Kim, Yong-Mook Kang, Min-Sik Park, Kwang-Taek Hwang, “Synthesis and
hydrogenation properties of lithium magnesium nitride”, Intl. Jour. of Hydrogen
Energy 36 (2011) 9714 – 9718.
24. Y. Nakamori, G. Kitahara, K. Miwa, S. Towata, S. Orimo, “Reversible hydrogenstorage function for mixtures of Li3N and Mg3N2”, Appl. Phys. A 80, 1 -3 (2005)
25. Li Hui-Ping, Huo Zhu-Feng, Huang Mei-Chun, Zhuhong, “Electronic Structures of the
Filled Tetrahedral Semiconductor LiMgN with a Zincblende-Type Structure”, Chin.
Phys. Lett. Vol 20, No. 1 (2003) 114
26. L.H. Yu, K.L. Yao, Z.L. Liu, “Electronic structures of filled tetrahedral
semiconductors LiMgN and LiZnN: conduction band distortion” Physics B 353
(2004) 278 – 286
27. F. Kalarasse, B. Bennecer and A. Mellouki, “Optical properties of the filled
tetrahedral semiconductors LiMgX (X = N, P and As)”, J. Phys. Condens. Matter 18
(2006) 7237 – 7242
28. K. Kuriyama, Y. Yamashita, T. Ishikawa, K. Kushida, “Raman scattering from the
filled tetrahedral semiconductor LiMgN: identification of the disordered
arrangement between Li and Mg”. Phys. Rev. B 75 233204 (2007)
29. C. Pfleiderer, M. Uhlarz, S.M. Hayden, R. Vollmer, H.V. Lohneysen, N.R. Bernhoeft,
G.G. Lonzarich, “Co-existence of superconductivity and ferromagnetism in the dband metal ZrZn2,” Nature 412, 58 – 61 (2001)
30. Shoji Yamanaka “High Tc superconductivity in electron-doped layer structured
nitrides”, Annual Rev. Mater. Sci. (2000) 30: 53 – 82
7
31. Richard P. Messmer, “The chemical bond and superconductivity”, Physica Scripta
37, 595 -604 (1988)
32. L.M. Volkova, S.A. Polyshchuk, F.F. Herbeck “Why Tc of MgB2 is the highest in a
number of diborides” Cond-Mat/0310511 (2003)
33. R. A. Hein, J.W. Gibson, R. Mazelsky, R.C. Miller, J.K. Hulm “Superconductivity in
germanium telluride” Phys. Rev. Lett. 12, 320 -322 (1964)
34. M.L. Cohen, “Superconductivity in many valley semiconductors and semi-metals”
Phys. Rev. 134, A 511 – 521 (1964)
35. M. L. Cohen, “The existence of a superconducting state in semiconductors”, Rev.
Mod. Phys. 36, 240 – 243 (1964).
TABLES
Table 1: Cases of similar superconductors. Cases 1, 3 and 6 show that when the three
parameters of 𝒳, Ne and Z are almost the same for two or more materials, their Tcs are also
almost the same. Adapted from Ref. [20].
Table 2: Material Specific Characterization Datasets (MSCDs) for MgB2 and LiMgN. The exact
match for 𝒳, Ne and Z lends very strong grounds for predicting very close Tc as shown in six
cases in Table 1.
Table 3: Some materials iso-structural and or iso-valent with MgB2. Refrences are indicated.
Table 4: MSCD of materials iso-structural and iso-electronic or iso-valent with MgB2.This table
provides quantitative data for analysis of Tc, based on equation (1) in this paper.
8
𝒳
Ne
Z
Ne/√
Fw
2.30
5.0
24
1.0206
106.913
2.15
5.0
24
1.0206
1.4333
3.0
15.667
2
NbN
MoC
CaAlSi
SrAlSi
1.4333
3.0
ZrN
NbC
2.2
3
4
Nb3Sn
Nb3Ge
Nb3Al
Superconductor
1
Fw/Z
Tc(K)
Ko
4.45
17
7.55
107.951
4.50
14.3
6.79
.7579
95.15
6.07
7.8
7.18
21.667
.6445
142.69
6.59
5.8
6.28
4.5
23.5
.9283
105.231
4.478
10.7
5.24
2.05
4.5
23.5
.9283
104.917
4.465
11.1
5.83
1.65
4.75
43.25
.7181
397.44
9.19
18
15.19
1.65
4.75
38.75
.7631
351.34
9.07
23.2
18.43
1.575
4.5
34
0.7717
305.71
8.991
16-20
13.2—
16.5
11.67
5
1.6
4.5
25
0.9
222.54
8.902
16.8
V3Ga
1.8333 2.6667 4.6667 1.2344
30.63
6.56
0
0
BeB2
6 LiBC
1.8333 2.6667 4.6667 1.2344
29.96
6.42
0
0
Table 1: Cases of similar superconductors. Cases 1, 3 and 6 show that when the three
parameters of 𝒳, Ne and Z are almost the same for two or more materials, their Tcs are also
almost the same. Adapted from Ref.[20]
Material
𝒳
Ne
Z
Fw
Ne/√
Fw/Z
Tc(K)
Ko
1.7333 2.667
7.333 0.9847
45.93
6.263
39
22.85
MgB2
1.7333 2.667
7.333 0.9847
45.26 6.172
38.5
22.85
LiMgN
Table 2: MSCDs for MgB2 and LiMgN. The exact match for 𝒳, Ne and Z and close match of
their Formula weights (Fw), lends very strong grounds for predicting very close Tcs of both
materials, following the chemical symmetry rules of ref. [20].
1
2
9
* = with MgB2
CaBeSi
LiBC
Li0.5BC
BeB2
BeB2.75
CaGa2,
SrGa2,
BaGa2
CaAlSi,
SrAlSi
CaB2
ZrZn2
Iso-structural *
Yes
Yes
No
Yes
No
Yes
Yes
Yes
Yes
Iso-valent*
Yes
Yes
No
Yes
No
Yes
No
Yes
Yes
Same
Electronegativity*
No
No
No
No
No
No
No
No
No
Same Atomic
Number*
No
No
No
No
No
No
No
No
No
Tc (K)
0.4K
0K
0K
0K
0.7K
<1.02K
>39K??
<1.02K
Reference
5
7, 10
9
11
11
14
7.8K,
5.8K
12, 14,
15
6
29
(AlB2, C32,
Hexagonal)
Table 3: Materials iso-structural, iso-electronic and iso-valent with MgB2. Iso-structural and
iso-valent similarity did not influence upward value of Tc to near 39K. Table 4 explains why.
10
Material
𝒳
Ne
Z
Ne/
Fw
Fw/Z
√
Tc
Ko
Ref
Comments
(K)
1.7333
2.6667
7.3333
0.9847
45.93
6.263
39
22.85
1, 16
Ne/√
>1.0
1.8333
2.6667
4.6667
1.2344
29.96
6.42
0
0
7, 10
Ne/√
>1.0
Li0.5BC 2.0
1.8333
BeB2
BeB2.75 1.8667
1.6667
CaB2
3.0
5.0
1.3416
26.29
5.26
0
0
9
Ne/√
>1.0
2.6667
4.6667
1.2344
30.63
6.56
0
0
Ne/√
>1.0
2.7333
4.7333
1.2563
38.74
8.19
<0.7
0
Ne/√
>1.0
2.6667
10.0
0.8433
61.70
6.17
?
?
11
11
6
CaAlSi
1.4333
3.0
15.667
.7579
95.15
6.07
7.8
7.18
8
SrAlSi
1.4333
3.0
21.667
.6445
142.69
6.59
5.8
6.28
9
CaBeSi 1.4333
10
CaGa2
1.4
2.6667
27.333
0.5101
179.52
6.57
<1.02
1.4
14
11
ZrZn2
1.5333
2.6667
33.333
0.4619
222.0
6.66
<1.02
1.4
29
12
SrGa2
1.4
2.6667
33.333
0.4619
227.06
6.81
<1.02
1.4
14
13
BaGa2
1.3667
2.6667
39.333
0.4252
276.77
7.04
<1.02
1.4
14
1
2
MgB2
LiBC
3
4
5
6
7
2.6667
12.667
0.7493
77.18
6.093
0.4
0?
12,
14, 15
12,14,
15
5
Tc likely to be
<39K
Ne/√ <0.8 and
𝒳<1.733
Ne/√
<0.8 and
Ne/√
<0.8 and
𝒳<1.733
𝒳<1.733
Ne/√ <0.8
𝒳<1.733
Ne/√ <0.8
𝒳<1.733
Ne/√ <0.8
𝒳<1.733
Ne/√ <0.8
𝒳<1.733
and
and
and
and
Table 4: MSCD of materials iso-structural and or iso-valent with MgB2 and their Tcs. Note that Ne/√ plays a
key role in determining superconductivity. So too does the value of 𝒳, as shown in equation (1) derived in ref.
[20].
.
11
12