Rocksalts Designed as Super-electrides , Drude Metals , and as Possible High Temperature Superconductors


This web page will update the two previous essays on super-electrides and Drude metals , Electrons and Perovskites . The design and synthesis of super-electrides having the rocksalt crystal structure will be explored in this brief study .

Rocksalts are very simple solid state compounds having an MX empirical formula , and in which both the M cations and the X anions have a six-fold octahedral coordination to each other . The objective is to design MX compounds with stoichiometric numbers of atom-sized cation vacancies octahedrally surrounded by anions . These large voids in the lattice will be filled either by singlet electrons or by pairs of electrons having opposite spins (in effect , Cooper pairs) . Because these void electrons are coordinated by the negative charges on the surrounding anions , they will be suspended in the void centers . More importantly , though , the coordinating anions will completely isolate and free them from any bonding to their parent atomic kernels . Such completely free , unassociated electrons within matter are referred to as Drude electron gas , and the resulting electronic solid is a Drude electron material .

Drude electron materials chemically resemble solid state electrides , and it should be possible to synthesize them in the same way as electrides . Conventional electrides are prepared by the solution (or insertion) of a highly reactive metal dopant usually an Alkali or Alkaline Earth element into a suitable inert host substrate . The electride forms when the valence electron or electron pairs on the inserted atoms are separated from their parent kernel . The electride will be stable at STP if both the cation kernels and their separated electrons are energetically stabilized within the host lattice .

For example , sodium metal can be dissolved in pure , anhydrous liquid ammonia to prepare inky , blue-black solutions of the sodiumammonia electride . Formation of the electride is energetically favorable because sodium is a powerful reducing agent (E0ox = 2.71 V) and by separating its 3s1 valence electrons from its Na1+ kernel (which has the very stable neon electronic configuration , 2s2 2p6) it can substantially reduce its system energy . Both the Na1+ parent kernels and their separated 3s1 electrons are electronically stabilized in the electride . The electrophilic Na1+ cations form weak Na–NH3 coordinate covalent bonds , having an approximate octahedral coordination , with the nucleophilic lone pairs on the ammonia molecules surrounding them . The electrons , which are located in the void spaces in between the ammonia molecules , are probably bonded to the positive dipoles on the NH3 molecules by van der Waals dipolar bonds . Numerous references concerning the electrides were cited in the Electrons web page .

In the two previous web pages chemical techniques to prepare new solid state electrides were discussed . Unlike the conventional electrides known up to now – including even the zeolite ones the new electrides were designed to have stoichiometric numbers of atom-sized void spaces in the lattice in which the “popped” electrons reside . Because of the relatively high concentrations of the separated electrons in the compounds they were referred to as super-electrides . The free electrons will have much more room within the voluminous cation vacancies than they would in the interatomic voids . Their parent cation kernels will have been incorporated into the crystal lattice as an integral part of the normal crystal structure of the compound , which hopefully will be stable and isolable at STP .

If the completely free , unassociated electrons in the cation vacancies are able to migrate through the lattice under a potential difference (voltage) applied across the crystal's physical dimensions , the super-electride will be a Drude electron material , not a conventional Fermi-Dirac metal . In a Fermi-Dirac metal the free , mobile electrons are in orbitals – the crystal orbital / metallic bond /conduction band – and their energy distribution is described by Fermi-Dirac statistics . As a Drude electron material the super-electride should be superconducting , since its free , non-orbital electrons are no longer bound to their parent kernels and there should be no resistance to their migration through the lattice . If there are channels of atom-sized void spaces in the lattice containing the Drude electron gas it will be able to flow through the crystal unopposed at all temperatures – there will be no upper temperature limit to the superconduction – as long as the voids and the channels remain clear and unblocked by stray or impurity atoms . The Drude electron materials could prove to be high temperature superconductors , and as such they could become materials of great interest to solid state researchers .


The Hypothetical Rocksalt Drude Electron Materials , A0[ ]M2Cl4


One challenge in designing these new materials is in adapting the common crystal structures as suitable containers for the void spaces and their Drude electron gas . Another challenge , once an acceptable crystal container has been chosen , is to find appropriate cations and anions with which to construct its lattice and to produce the electride in it . In the Electrons web page the zinc blende structure , having cation vacancies with tetrahedrally coordinating anions , was suggested as the container . In the Perovskites web page perovskites having cation vacancies with twelve coordinating anions in the walls of the voluminous ReO3 cavities were examined as possible new super-electride candidates . The rocksalt structure , with octahedrally coordinating anions for the cation vacancies , will now be studied as the crystal container for the super-electrides .

The rocksalt empirical formula , MX , is quadrupled to obtain M4X4 . The first two M cations are replaced with A0and [ ] , where A0 is a very reactive Alkali or Alkaline Earth zerovalent atom whose underlying cationic kernel prefers an octahedral coordination in its ionic salts , and [ ] is a cation vacancy in the lattice ; ie. it should normally be filled with a cation , but is now empty . The revised rocksalt formula is now A0[ ]M2X4 . If X is a chloride anion , we can (try to experimentally) insert one gram atom equivalent of an Alkali or Alkaline Earth elementary metal into two gram formula equivalents of a suitable , unreactive MCl2 ionic salt . The M2+cations in the MCl2 host should also prefer an octahedral coordination by chloride anions in their ionic salts :

A0 + 2 MCl2 ------ [heat , inert atm.] -------> A0[ ]M2Cl4 = A1+[*]M2Cl4 or A2+[**]M2Cl4 ;

A0 = Li , Na , K , Rb , Mg , Ca and Cd ; M = Mg2+ and Ca2+ .

Note that cation vacancies are used to contain the popped electrons from the reducing A0 atoms . The electrons will then be surrounded on all sides by anions (in this case , chlorides) . The anions will suspend the electrons in the void centers , and will shield them from their parent cation kernels . This will ensure that there is no association between the electrons and their original atoms , and that they can't be recaptured by them . The popped electrons will therefore be free Drude electron gas , and the resulting super-electrides might be high temperature superconductors .

The F-centers (discussed in the Electrons web page) are produced by the resonance of free electrons in empty anion vacancies in the lattices of Alkali Metal halides such as NaCl . These free electrons must be orbital in nature , because they are resonating in the empty 3s frontier orbitals of the Na1+ cations octahedrally surrounding the anion void spaces . Since they are associated electrons , they must be classified as Fermi-Dirac electron gas and not as Drude electron gas . Because only “ordinary” metallic solids (and low temperature superconductors) might result from the presence of Fermi-Dirac electron gas in a crystal lattice , the aim in these recent web pages has been to design super-electride compounds having popped electrons in empty cation vacancies , so as to obtain Drude electron gas in them , and potential high temperature superconductors .

Placing an electron or pair of electrons in a cation vacancy surrounded by anions is undoubtedly thermodynamically unfavorable , and could result in an increase in the system's chemical energy , rather than a decrease in energy as is required for a chemical reaction to occur . In order to make the overall electride formation thermodynamically favorable the chemical bonding of the cation from the guest atom in the host lattice must be especially strong , so that a net energy decrease in the system of reactants is realized . In this present case the guest cations will be forming strong ionic bonds with the host anions in the rocksalt lattice . Hopefully this coulombic bonding – which is considerable in typical ionic salts such as NaCl – will offset the unfavorable energetics of locating electrons in cation vacancies , surrounded by anions .

There are really only two suitable MCl2 host lattices suitable for receiving the reducing A0 dopant atoms : MgCl2 and CaCl2 . Anhydrous magnesium chloride has the layered cadmium chloride crystal structure , which is similar to that of the cadmium iodide structure , differing from it in the stacking of the layers (CdCl2 has a cubic stacking , while CdI2 has a hexagonal packing) :

The magnesium cations in MgCl2 are octahedrally coordinated by the chloride anions , which have a trigonal pyramid coordination to the magnesiums (as in TiS2 , in the above sketch) . Empty interlayer spaces separate the ClMgCl layers [Wikipedia sketch “Cadmium-chloride-3D-polyhedra.png”, PNG image , 98 KB] .

In the super-electride synthesis the reducing A0 atoms would first be inserted in between the MgCl2 layers . The A0 valence electron or electrons would be popped into adjacent interlayer spaces , then the A0MgCl2[ ]MgCl2 intermediate would recrystallize as the cubic rocksalt super-electride A1+[*]M2Cl4 or A2+[**]M2Cl4 :

Li0 (m.p. 180.5 C) + 2 MgCl2 (anhydrous , m.p. 714 C) -------- [heat together in an inert atmosphere such as pure , dry nitrogen or argon] -------> Li0[ ]Mg2Cl4 = Li1+[*]Mg2Cl4 .

Alternately , using magnesium metal powder – much more convenient than the highly reactive lithium metal , which can be difficult to handle – as the source of the Drude electron gas ,

LiCl (m.p. 610 C) + Mg0 (m.p. 650 C) + 1 MgCl2 (m.p. 714 C)

-------- [grind together , pelletize , heat in an inert atmosphere] -------> Li1+[*]Mg2Cl4 .

Caution : Some chemical companies offer extremely finely divided grades of metal reagents . These metal powders with very fine particle sizes are highly reactive , but they are often dangerously pyrophoric (spontaneously combustible in air) , and must be handled in an inert atmosphere , eg. in a nitrogen-filled glove bag . I see in my Aldrich Chemical Co. Catalog Handbook of Fine Chemicals (1992) a listing for “catalog no. 25,398-7 , Magnesium , powder , – 50 mesh , 99+ % , $44.00/Kg”. This is an excellent and quite satisfactory purity level , and – 50 mesh is probably the optimum particle size , providing rapid reactivity with safe handling , ie. it probably isn't pyrophoric , but please read the label on the reagent container for Aldrich's special handling instructions , if any !

Li would be a good size match with Mg in Li1+[*]Mg2Cl4 . The crystal ionic radius of Li1+, per Shannon and Prewitt , CN = 6 (octahedral) is r = 0.76 ; that of Mg2+ is r = 0.72 . LiCl has the cubic rocksalt crystal structure , so Li1+ will accept an octahedral coordination by chloride anions (it's usually tetrahedrally coordinated by oxides and sulfides) . Li1+[*]Mg2Cl4 should have an undistorted cubic symmetry as a rocksalt compound .

CaCl2 (m.p. 775 C) + Mg0 (m.p. 650 C) + MgCl2 (m.p. 714 C)

-------- [grind together , pelletize , heat in an inert atmosphere] -------> Ca2+[**]Mg2Cl4 .

Calcium metal is highly reactive and can be difficult to manipulate , so the cheaper and easier to use magnesium metal powder could provide the Drude electron gas in this preparation .

The last example with calcium reminds us that magnesium metal itself could be the dopant reducer in the MgCl2 host lattice :

Mg0 (m.p. 650 C) + 2 MgCl2 (m.p. 714 C)

------ [grind together , pelletize , heat in an inert atmosphere] ------> Mg2+[**]Mg2Cl4 , ie. Mg3Cl4 .

At first glance Mg3Cl4 looks like a Robin-Day Class II mixed-valent compound . However , its magnesium atoms are entirely Mg2+. It would (or should) be an even more remarkable chemical compound : a Drude electron material !

Both anhydrous magnesium chloride and magnesium metal are produced in vast quantities worldwide as industrial chemical and metal commodities . At one time both MgCl2 and Mg were produced in the United States by the Dow Chemical Company , which isolated the magnesium from seawater (P.D.V. Manning , “Magnesium , Metal of the Future”, Engineering and Science Monthly , pp. 14-18 , June , 1944 , PDF , 860 KB) . However , the production of magnesium metal via the electrolysis of anhydrous magnesium chloride has been overtaken and supplanted in recent years by the dominant Chinese-operated Pidgeon Process (reduction of MgO by ferrosilicon , Wikipedia article) .

The use of bone-dry , anhydrous reagents in the synthesis of the super-electrides should be obvious even to novice chemists . In the last example , traces of water in the commercial anhydrous MgCl2 would be reduced by the Mg0 in the mixture : Mg0 + H2O -----> MgO + H2 (g) . The hydrogen molecules would diffuse out of the reaction mixture , taking the electride electrons with them , which is obviously highly undesireable . While traces of MgO in the super-electride product probably wouldn't affect its physical and electrical properties very much , loss of the reducer's valence electrons would undoubtedly degrade the material's electrical performance .

The commonest form of magnesium chloride is its hexahydrate salt . Full dehydration of this chemical to its anhydrous form can be a major problem , as merely heating it tends to hydrolyze the salt to a mixture of MgCl2 and Mg(OH)2 . It can be dehydrated to anhydrous MgCl2 only by heating it in a stream of hydrogen chloride gas , which prevents the hydrolysis : H.F. Walton , Inorganic Preparations , A Laboratory Manual , Prentice-Hall , New York , 1948 ; pp. 74-76 . This useful inorganic synthesis textbook can be downloaded for free from the library resources web page [DJVU , 1644 KB ; a suitable DjVu reader for your computer can be downloaded for free from . The WinDjView Reader v. 1.0.3 for older FAT32 Windows can be downloaded for free from FileHorse] .

Another technique for dehydrating chloride salts is to heat them with thionyl chloride , which rapidly combines with water to evolve only gaseous products : SOCl2 + H2O ------> SO2 (g) + 2 HCl (g) . Many hydrated chlorides , including MgCl2.6H2O , have been fully dehydrated with thionyl chloride (GIF image , 73 KB) . The dehydration of CrCl3.6H2O by thionyl chloride has been described by A.R. Pray , “Anhydrous Metal Chlorides”, Inorg. Synth. 5 , T. Moeller et al. (eds.) , McGraw-Hill , New York , 1957 ; pp. 153-156 [PDF , 7324 KB for the entire vol. 5 ; also in vol. 28 , pp. 321-323 , PDF , 18,645 KB for the entire vol. 28] . This might be a useful method for removing the last traces of water from commercial “anhydrous” MgCl2 .

The super-electrides must also be protected from degradation by atmospheric oxygen , carbon dioxide , and of course humidity . Their preparation must be carried out in a furnace under a protective inert atmosphere of pure nitrogen or argon . Failure to protect them during their synthesis would undoubtedly lead only to the production of an oxidized insulator . However , the following simple technique might be helpful to the solid state chemist attempting their synthesis . An appropriate quantity of thoroughly ground reaction mixture would be compressed into a cylindrical pellet in a common laboratory press (eg. Carver press , used for making KBr discs for infrared spectroscopy) . The pellet would be placed into a suitably sized glassy carbon or graphite crucible and covered with a blanket of graphite powder , gently tamped down :

This picture (slightly modified) was scanned from the Alfa-Aesar catalog , p. 1289 (1997) . My thanks to the copyright owner of this image . Alfa-Aesar offers a line of glassy carbon and graphite labware , including the above crucibles . Graphite would be chemically inert toward the reaction mixtures considered here . The ceramic-like precursors and products might bond to the ordinary porcelain , fireclay , or metal crucibles (sodium chloride is used as a salt glaze in some types of pottery) , especially if they melt into a liquid or a viscous , glassy , magma . The sintered pellets very likely wouldn't adhere to the graphite or glassy carbon crucibles , and could be easily removed from them with no loss of product .

The crucible and its contents would be placed in a furnace , which as an added protection could be purged continuously by a stream of nitrogen or argon . The commercial grades of N2 and Ar can be fully deoxygenated by bubbling them through a wash-bottle filled with Fieser's solution (DOI) :

The fully deoxygenated , but now damp nitrogen or argon would be dried by flowing the gas through a bed of drying agent such as Drierite , P2O5 , CaCl2 , etc . in wash-bottles further downstream .


Calcium Chloride as the Host Lattice for the Super-electrides


Like magnesium chloride , calcium chloride is another relatively inert halide salt that could serve as the host for the insertion of the reducing zerovalent metal atoms . The calcium cations are octahedrally coordinated by the chloride anions in anhydrous CaCl2 , while the chlorides have a trigonal planar coordination to the calciums (they have a trigonal pyramid coordination to the Mg2+ in MgCl2 , resulting in its much different crystal structure) :

The above sketch was copied from the Wikipedia web page , Calcium chloride . My thanks to the author of this graphic , and Wikipedia , for implied permission to reproduce it here .

While these coordinations are similar to those of the cations and anions in the rutile crystal structure (GIF image , 57 KB ; GIF image , 62 KB) the atomic packing in calcium chloride is nevertheless quite different from that in rutiles :

The above sketch is a modified version of the image “Hydrophilite” (GIF image , 112 KB) , from the Web . My thanks to the author and/or copyright owner of the original graphic .

The large interchain volumes in the CaCl2 lattice could readily accommodate the zerovalent guest atoms of the reducing metals . As usual a 1 : 2 molar equivalent ratio of A0 metal atoms to CaCl2 host should be observed , to provide empty cation void spaces for the A0 ns1 or ns2 valence electrons when the ions recrystallize into the rocksalt structure . Na1+cations (r = 1.02 , CN = 6) should fit well into a CaCl2 host rocksalt compound with the Ca2+cations (r = 1.00 , CN = 6) :

Na0 (m.p. 98 C) + 2 CaCl2 (anhydrous , m.p. 775 C) -------- [heat together in an inert atmosphere such as pure , dry nitrogen or argon] -------> Na0[ ]Ca2Cl4 = Na1+[*]Ca2Cl4 .

Calcium chloride is an abundant , cheap , industrial chemical , familiar to everyone in northern countries as the “calcium salt” spread on icy , snowy roads and sidewalks in Winter to help remove slippery ice surfaces on them . This commercial product is undoubtedly the hexahydrate , and would obviously be unsuitable for use in the preparation of the super-electrides . As with the magnesium chloride discussed above (and generally) , extremely dry , anhydrous reagents must be used in their synthesis . Anhydrous calcium chloride in various grades is readily available , although the purer and dryer they are , the more expensive they are . Again as with MgCl2 the researcher could treat a reasonably pure sample of CaCl2 with a chemical drying agent such as thionyl chloride to remove the last traces of water in it before the super-electride synthesis .

Dissolving calcium metal in CaCl2 might form the super-electride Ca0[ ]Ca2Cl4 = Ca2+[*]Ca2Cl4 = Ca3Cl4 . Magnesium metal powder , which is cheaper and easier to handle than the highly reactive calcium , could also be investigated in this series of compounds :

Mg0 (m.p. 650 C) + 2 CaCl2 (m.p. 775 C)

-------- [grind together , pelletize , heat in an inert atmosphere] -------> Mg2+[**]Ca2Cl4 .

Cadmium is a mild reducing agent (E0ox = 0.403 V) , and is a soft , low-melting metal . The crystal ionic radius of Cd2+ , r = 0.95 (CN = 6) , is close to that of Ca2+ (r = 1.00 ) . Cd2+ seems to prefer an octahedral coordination by anions (as in the rocksalt CdO and in the layered compound CdCl2) . Cadmium could thus conceivably form a super-electride when dissolved in CaCl2 :

Cd0 (m.p. 321 C) + 2 CaCl2 (m.p. 775 C) ----- [heat in an inert atm.] ------> Cd2+[**]Ca2Cl4 .

Cadmium chloride would be unsuitable as a host for a super-electride . The reactive A0 metal atoms would reduce its Cd(II) to the Cd(I) “cadmous” cation , (Cd2)2+, which is similar to the mercurous cation , (Hg2)2+ (as in calomel , Hg2Cl2 . See : J.D. Corbett , “The Cadmium(I) Ion Cd22+. Raman Spectrum and Relationship to Hg22+ ”, Inorg. Chem. 1 (3) , pp. 700-704 (1962) [DOI] . The Zn(II) in zinc chloride would similarly be reduced to Zn(I) , which is (Zn2)2+ : D.H. Kerridge and S.A. Tariq , “The Solution of Zinc in Fused Zinc Chloride”, J. Chem. Soc. A 1967 , pp. 1122-1125 [DOI]) .

It might be possible to synthesize the super-electrides of manganese , iron , cobalt , and nickel in a magnesium chloride host lattice . They are all mild reducing agents (ie. are base metals) , and the crystal ionic radii of their divalent cations (which are all low energy redox-wise , neither oxidizing nor reducing) are comparable to that of Mg2+ :

manganese : E0ox = 1.185 V , Mn2+ r = 0.83 ; cf. Mg2+ r = 0.72 ;

iron : E0ox = 0.447 V , Fe2+ r = 0.61 ;

cobalt : E0ox = 0.28 V , Co2+ r = 0.65 ;

nickel : E0ox = 0.257 V , Ni2+ r = 0.69 .

All crystal ionic radii cited are for an octahedral coordination with CN = 6 . Crystal ionic radii were from the CRC Handbook of Chemistry and Physics , “Ionic Radii in Crystals”, pp. 12-11 to 12-12 . Unfortunately the tabulation doesn't distinguish between high spin and low spin conditions of the Transition metal cations . I suspect the Mn2+ r = 0.83 value , which is somewhat large , is probably for high spin Mn(II) , while the Fe2+ r = 0.61 value is for low spin Fe(II) . Chloride anions are weak CF ligands , so Transition metal M(II) cations associated with them would undoubtedly be in a high spin state .

The standard redox potentials cited are versus the SHE at STP , and were from : D.R. Lide (ed.) , CRC Handbook of Chemistry and Physics , 87th edition , CRC Press / Taylor & Francis , Boca Raton (FL) , 2006 ; P. Vansek (ed.) , “Electrochemical Series”, pp. 8-20 to 8-25 . Another useful reference with redox potentials is : A.J. Bard (ed.) , Encyclopedia of Electrochemistry of the Elements , various volumes , Marcel Dekker , New York , ca. 1973-1986 .The Wikipedia web page “Table of Standard Electrode Potentials” is a helpful online reference in this regard .

Synthesis of the Mn , Fe , Co , and Ni super-electrides might be attempted in the usual manner :

Mn0 (m.p. 1246 C) + 2 MgCl2 (m.p. 714 C)

-------- [grind together , pelletize , heat in an inert atmosphere] -------> Mn2+[**]Mg2Cl4 ;

Fe0 (m.p. 1538 C) + 2 MgCl2 -------- [heat] -------> Fe2+[**]Mg2Cl4 ;

Co0 (m.p. 1495 C) + 2 MgCl2 -------- [heat] -------> Co2+[**]Mg2Cl4 ; and ,

Ni0 (m.p. 1455 C) + 2 MgCl2 -------- [heat] -------> Ni2+[**]Mg2Cl4 .

Given the refractory nature of manganese , iron , cobalt , and nickel , their zerovalent atoms might be more efficiently obtained from their zerovalent coordination compounds such as manganese decacarbonyl , Mn2(CO)10 , m.p. 154 C ; iron pentacarbonyl , Fe(CO)5 , b.p. 103 C ; and cobalt octacarbonyl , Co2(CO)8 , m.p. 51 C (dec.) [Strem] . Nickel tetracarbonyl , Ni(CO)4 , b.p. 42 C , is a very toxic , hazardous reagent , and is not recommended for this application . Instead , the zerovalent nickel reagent tetrakis(triethylphosphite)nickel(0) , [(C2H5O)3P]4Ni – a white , air-sensitive powder , m.p. 108 C – might be a safer , more easily handled source of the nickel atoms to be inserted into MgCl2 . The preparation of this nickel(0) compound has been described by : M. Meier et al. , “Tetrakis(Triethyl Phosphite) Nickel(0) , Palladium(0) , and Platinum(0) Complexes”, Inorg. Synth. 13 , pp. 112-117 ; F.A. Cotton (ed.) , McGraw-Hill , New York , 1972 [PDF , 13,911 KB for the entire vol. 13] .The original (and I think , better) prodecure was by : R.S. Vinal and L.T. Reynolds , “The Reduction of Nickel(II) Halides by Trialkyl Phosphites”, Inorg. Chem. 3 (7) , pp. 1062-1063 (1964) [DOI] . One equivalent of Transition metal element in the coordination compounds would be combined with two equivalents of MgCl2 under an inert atmosphere (N2 or Ar in a glove bag) , the mixture ground together , then heated to expel the volatile carbon monoxide or triethyl phosphite ligands (fume hood !) . This resulting precursor material would then be pelletized and annealed in a furnace under an inert atmosphere , or protected by a graphite blanket , to complete the super-electride synthesis .

As an interesting variation of this study the four Transition metal elements might form rocksalt super-electrides with their own halides (fluorides and chlorides) ; for example ,

Fe0 (m.p. 1538 C) + 2 FeCl2 (m.p. 677 C) -------- [heat] -------> Fe2+[**]Fe2Cl4 = Fe3Cl4 ; or ,

Fe(CO)5 (b.p. 103 C) + 2 FeCl2 -------- [mix , heat] -------> Fe0Fe2Cl4 “premix” + 5 CO (g)

-------- [heat] -------> Fe2+[**]Fe2Cl4 = Fe3Cl4 .

The M2+ cations in these M3X4 rocksalts would all undoubtedly be in a high spin electronic state , making the materials strongly paramagnetic . However , as the Drude electron gas in the cation voids is shielded from their parent M2+ kernels by the X anions , this Curie paramagnetism should have no significant effect on its electronic properties in the lattice .

Anhydrous MF2 and MCl2 salts (M = Mn , Fe , Co , and Ni) are readily available at a moderate cost (Alfa-Aesar , American Elements) . They can also be synthesized by the researcher using fairly simple methods . As a general precaution they should be handled under anaerobic conditions (Fe2+ in particular is somewhat sensitive to atmospheric oxygen , and is readily oxidized by it to Fe3+) . The anhydrous MF2 salts have the rutile lattice , while the MCl2 salts have the CdCl2 crystal structure (see for example this PNG image of anhydrous FeCl2 , 249 KB) .


Zinc and Magnesium Fluorides Designed as Rocksalt Super-electrides


Zinc is an inexpensive base metal and is a moderately strong reducing agent (E0ox = 0.7618 V) . It would thus be interesting to use it in rocksalt super-electrides if possible . Zn(II) shows a strong preference for a tetrahedral coordination by chloride anions in zinc chloride (and in many other solid state compounds such as ZnO and ZnS , for example , in which there is substantial coordinate covalent ZnX bonding) . However , in ionic crystals with fluoride , such as in ZnF2 (rutile crystal structure , PNG image , 299 KB) and in the perovskite AgZnF3 , zinc is octahedrally coordinated by the fluoride anions . It might therefore be possible to formulate A0[ ]M2F4 rocksalt super-electrides , in which A and/or M is zinc :

Zn0 (m.p. 420 C) + 2 ZnF2 (m.p. 872 C)

-------- [grind together , heat in an inert atmosphere] -------> Zn2+[**]Zn2F4 , ie. Zn3F4 .

As noted above , Zn(I) can form as the reactive , unstable species (Zn2)2+ when zinc metal is dissolved in molten zinc chloride . The question then becomes : which is more energetically stable , the hypothetical Zn(I)–Zn(II) composite (Zn2)2+Zn2+F4 , or the rocksalt super-electride Zn2+[**]Zn2F4 , especially when the synthesis is carried out at an elevated temperature , perhaps at around 600 C or so ? Mercury(I) – “mercurous” – halides , Hg2X2 , readily disproportionate into Hg0 and HgX2 under various physical and chemical conditions . Zn(I) is even more unstable than Hg(I) , so formation of (Zn2)2+Zn2+F4 would probably be much less energetically favorable than that of the desired super-electride Zn2+[**]Zn2F4 .

Magnesium might similarly form Mg3F4 :

Mg0 (m.p. 650 C) + 2 MgF2 (m.p. 1263 C) -------- [heat] -------> Mg2+[**]Mg2F4 , ie. Mg3F4 .

Mg2+ and Zn2+ have nearly similar crystal ionic radii ( r = 0.72 and r = 0.74 , respectively , CN = 6) , so they might form mixed cation rocksalts :

Mg0 + 2 ZnF2 -------- [heat] -------> Mg2+[**]Zn2F4 , ie. MgZn2F4 ; and ,

Zn0 + 2 MgF2 -------- [heat] -------> Zn2+[**]Mg2F4 , ie. ZnMg2F4 .

Writing the formula for the mixed-cation compounds as Mg2+[**]Zn2F4 , for example , is only a formality to show the origin of the popped electrons in the octahedral cation voids . The actual formula could just as well be written as Mg2+Zn2+Zn2+[**]F4 , with a random distribution of the cations and void spaces (which would alternate with the fluoride anions , of course) .

Inorganic fluorides , like organofluorine compounds , seem to have a lower affinity for water and hydration than do the corresponding chlorides . They are thus easier to desssicate in a drying oven to bone-dryness than the hydrated chlorides . However , fluorides are more refractory than the chlorides , and they are more expensive than them , which makes them less attractive economically as host structures for the super-electrides .


Aluminum Rocksalt Super-electrides


The fluorocorundum compound Al4F9 was proposed as an aluminum super-electride in the Perovskites web page . It might be synthesized by the solution of one molar equivalent of aluminum metal powder in three molar equivalents of aluminum fluoride to obtain Al2F3Al[**]F3Al[*]F3 . Suppose an analogous reaction was attempted , but using alumina , Al2O3 , as the host :

Al0 (m.p. 660 C) + 3 Al2O3 (m.p. 2054 C)

------- [grind together , pelletize , heat in an inert atmosphere] -------> Al3O3Al2[**]O3Al2[*]O3 .

The resulting product , Al7O9 , should have the rocksalt crystal structure , although it might not be perfectly cubic in its symmetry . Just as AlF3 (an ionic compound) has a sort of ReO3 structure , Al7O9 (an ionic compound) might similarly have a sort of rocksalt structure . That is , ions with the Al4F9 and Al7O9 stoichiometries would “sort themselves out” into approximations of the corundum and rocksalt lattices , respectively , when the compounds form .

Six decades ago (in 1954) Hoch and Johnson tried to prepare the sub-valent aluminum oxides Al2O and AlO by reacting aluminum metal with alumina . They discovered that Al2O and AlO were stable only at very high temperatures . When the reaction mixtures were cooled down to room temperature both compounds disproportionated back into the Al0 and Al2O3 starting materials :

M. Hoch and H.L. Johnston , “Formation , Stability and Crystal Structure of the Solid Aluminum Suboxides : Al2O and AlO”, J. Amer. Chem. Soc. 76 (9) , pp. 2560-2561 (1954) [DOI] ; C.N. Cochran , “Aluminum Suboxide Formed in Reaction of Aluminum with Alumina”, J. Amer. Chem. Soc. 77 (8) , pp. 2190-2191 (1955) [DOI] . See also T. Forland et al. , “Measurements of Phase Equilibria in the Aluminum – Aluminum Sulfide System”, Acta. Chem. Scand. , Series A28 (2) , pp. 226-228 (1974) [PDF , 375 KB ; DJVU , 109 KB] . The compound AlS has a narrow window of stability between 1010 C and its m.p. of 1060 C .

X-ray diffraction measurements showed that both Al2O and AlO had a cubic symmetry . Suppose AlO was a rocksalt (Al2O might have been a cubic antifluorite [GIF image , 15 KB] compound like Li2O) . Its Al2+ would have been 3s1 electronically . The only physical location for the voluminous 3s orbital in a rocksalt structure would have been in the atomic interstices , which is a highly restrictive space . It was probably too small a volume for the 3s1 valence electrons , resulting in an energetically unfavorable condition in the lattice . The starting materials , Al0 and Al2O3 , were more thermodynamically stable than the AlO , so it collapsed back to them when the thermal energy sustaining it was removed , ie. it was cooled to room temperature .

I believe this is precisely why all of the conventional superconductors known to date are so thermally unstable , and “decompose” to normal metallic solids at higher temperatures . Their Cooper pairs , while being non-orbital Drude electron gas , are physically located in the atomic interstices , which are very small , restrictive spaces . As the atomic kernels are warmed up , they vibrate more and more and fill up even more of the interatomic void space . The Cooper pairs are squeezed to the point where they break up into singlet electrons which are recaptured by their parent cationic kernels and are rebonded in the cations' frontier orbitals , thus restoring the normal metallic state . The super-electrides will hopefully solve this problem by providing the enormous (relatively speaking) atom-sized cation vacancies for the popped valence electrons from the inserted guest atoms .

The new aluminum super-electride proposed here , Al7O9 , is similar to AlO [ie. Al9O9] , except that it is missing deliberately two formula Al3+ cations , so that their cation vacancies can be filled by the popped electrons : (Al3+O2-)7([**]O2-)([*]O2-) . By locating the popped electrons in these atom-sized voids , which are vastly larger than the atomic interstices , the resulting electride should hopefully be thermodynamically stable at all temperatures . The electrons in those cation vacancies are isolated from their parent Al3+ kernels and are free Drude electron gas . If the free electrons can flow through the Al7O9 crystal under an applied p.d. it will be superconducting , maybe even at high temperatures .

Aluminum metal might also be dissolved in anhydrous , bone-dry MgCl2 , in the now-familiar 1 : 3 molar ratio , to obtain another rocksalt super-electride , AlMg3F6 :

Al0 (m.p. 660 C) + 3 MgF2 (m.p. 1263 C)

-------- [grind together , heat in an inert atmosphere] -------> Al3+MgF2[**]MgF2[*]MgF2 .

Note that aluminum is a weaker reducer (E0ox = 1.662 V , 2.069 V in a fluoride environment) than magnesium (E0ox = 2.372 V) , so the proposed compound AlMg3F6 really would be an aluminum super-electride in the inert MgF2 host lattice . An analogous reaction could be attempted using zinc fluoride as the host matrix for the aluminum guest atoms :

Al0 + 3 ZnF2 (m.p. 872 C) -------- [heat] -------> Al3+ZnF2[**]ZnF2[*]ZnF2 = AlZn3F6 .

In this latter reaction the aluminum is a stronger reducer than zinc (E0ox = 0.7618 V) and so could theoretically dump its three valence electrons onto the Zn2+ cations , reducing them cleanly to Zn0. However , the stoichiometry has been arranged so that if the hypothetical compound AlZn3F6 forms a rocksalt structure its lattice will have two cation vacancies per formula unit to hold all three of the aluminum valence electrons . The end result should be the formation of the stable compound AlZn3F6 . Whether it's an aluminum super-electride in ZnF2 or a zinc super-electride in a mixed cation Al–Zn–F matrix is unimportant as long as it's a Drude electron material and is superconducting .

The design of practical , functional superconductors based on aluminum , zinc , and magnesium , which are three of the cheapest and most abundant base metals (except iron , mentioned above) , has been a long-standing objective of mine , ever since I began studying superconductivity twenty-five years ago in the the autumn of 1989 . I think that a successful synthesis their super-electrides , as discussed above , may finally be the best (and probably only) way to accomplish this long-cherished ambition .


Digital Data Storage in Super-electrides


The Drude electron gas free electrons undoubtedly have spins , as do the electrons associated with atomic orbitals . The spins of the singlet free electrons in the cation vacancies of the super-electrides might be oriented by an external magnetic field applied by a magnetic induction probe over the surface of the material . Of course, this is how the magnetic hard drives in computers function . Such magnetic induction applied to super-electrides at their surfaces might orient the singlet electrons in either an up or down position , leading to the possibility of developing singlet electron super-electrides as digital data storage media . This concept is illustrated in the following simple sketch of such a magnetic induction process :

Simply running a current through the material would randomize the spins once again , thus instantly erasing all the data stored in it .


Related web pages in this series about Drude electron materials :

A New Picture of Superconductivity : Lightning Bolt Electrons in a Crystal” ;

Perovskites Designed as Drude Metals and Ambient Superconductors” ;

Chromium as the Guest Atom in Super-electride Drude Metals” ;

Lead , Tin , and Bismuth as the Guest Atoms in Super-electride Drude Metals” ;

* Betaines and Electrides : From Sugar Beets and Baby Shampoo to Superconductors ; and ,

* Drude Electron Materials Having Rutile and Layered Structures” .


[ Index Page ] [ Contact ]