FeLi[PO4]: Dissection of a Crystal Structure - Inorganic 3D Structures - page 83

C or Si. This feature should be related to the adamantane structures found in other
(II)-(VI) compounds, such as ZnS, BeO and BeS (blende and wurtzite).
The regularity of the [ -MgS] 6 3 layers might be related to the aromatic

structure of graphite itself, in such a way that in each Na-S contact would be
involved 1.5 electrons. This has been correlated with the rotational disorder of
the SO 4 groups, whose O atoms try to capture a pair of electrons. Because the
electrons are delocalized around the aromatic rings [ 4 ], this would provoke the
simultaneous rotation of the sulphate groups. Looking at Fig. 10a, d, g, j, m , itis
noteworthy that such a hexagonal phase has only been observed in Na 2 [SO 4 ], but
neither in Mg[SO 4 ] nor in FeLi[PO 4 ].

C

4.2 The Olivine-Like Structures and the PbO-Type Blocks

We have analysed so far the graphite-like layers. However, the alternative -MgS

skeletons, derived from a PbO-type array, merit a special comment. Thus, the
skeletons in III-, II- and I-Na 2 [SO 4 ] represented in Fig. 10b, e, h form a continuous
sequence of images, which will be discussed next.
In the III-phase (

C

-[MgS] subarray forms a distorted PbO-
type structure with an S-Na-S angle of 170 , greater than in FeS (118 ) (Fig. 8e ). By
increasing the temperature, the angle becomes closer to linearity, 174 in the
II-phase (

C mcm) (Fig. 10h ) , the C

nma) and becomes linear (180 ) in I-Na 2 [SO 4 ](

6 3 /mmc), where
simultaneously the graphite-like layers become regular hexagons (Fig. 10a, b ) .
If the sequence is completed with Fig. 10n and k , corresponding to FeLi[PO 4 ],
one can obtain a view of all intermediate steps in the complete transition from
PbO to NiAs. Thus, the PbO-type structure is stabilized in the ambient pressure
phase of triphylite FeLi[PO 4 ] (Fig. 10n ) and becomes closer to the NiAs type in
the high temperature, high pressure,

P P

-FeLi[PO 4 ][ 2 ]. It is also worth mentioning
the simultaneous regularity of the PbO-type array in triphylite FeLi[PO 4 ] (Fig. 10n )
with the strong distortion of its NiAs-type structure represented in Fig. 10o , and how
the regularity and distortion are inverted in

b

b

-FeLi[PO 4 ], as shown in Fig. 10k, l .
A question that arises is why the PbO-type subarray has not been observed in
Na 2 [SO 4 ]. A possible reason for this is that Na 2 [SO 4 ] tries to reproduce the structure
of pseudo-compounds derived from this composition. Thus, the only possibility
should be C -Mg[SO 4 ], as observed in the mentioned Cmcm- and Pnma-phases
(III- and II-Na 2 [SO 4 ]), showing arrangements like that of Fig. 10p . This is equiva-
lent to say that Fe(LiPO 4 ) could also be formulated as Fe 2+ (LiPO 4 ) 2 , equivalent to
a Fe-stuffed

-Mg[SO 4 ] structure. Surprisingly, the structure shown in Fig. 10l is
that of Mg[SO 4 ] itself (Fig. 10p )[ 20 ] .
In principle, a PbO-type array should not be discarded because, as far as we
know, no systematic HT study has been carried out on Mg[SO 4 ]. If stable, such a
phase should be obtained at very high temperatures.
Another interesting question arising from the above discussion is why triphylite
FeLi[PO 4 ] transforms into the

C

b -phase ( C mcm) without reaching the P 6 3 /mmc,