1590
S. Strekopytov, C. Exley / Polyhedron 24 (2005) 1585–1592
into the solid phase and more particularly into HAS. To
ꢀ
trend towards the formation of HAS such that struc-
B
understand how F influenced HAS formation it is
tural (NMR) and compositional (microprobe) analyses
of the solid phase showed that it was approximately
identical to that precipitated from the previous parent
solution (1.0Si/2.0Al/1.0F).
informative to compare solid phases formed for each
parent solution stoichiometry in the absence and pres-
ence of fluoride.
ꢀ
(
i) 0.5Si/2.0Al ± 0.5F – In the absence of F the solid
Thus in spite of quite obvious differences in the solid
ꢀ
phase was predicted to be a mixture of HASA and
Al(OH)3(s) and this was borne out by a Si:Al ratio of
phases formed in the absence of added F the structure
and composition of HAS formed in its presence were
0
ence of HAS was that all Al was present in octahedral
.27 in the solid phase. Further confirmation of the pres-
approximately identical, probably an HAS -like struc-
A
ture which included F in a single chemical environment
19
A
2
7
geometry ( Al d
2.5 ppm) and Si was present in
9
( F NMR d
and Al:F ratio of ca. 5–6.
(iv) 2.0Si/1.0Al ± 1.0F – In the absence of F the so-
lid phase was predicted to be predominantly HAS and
ca. ꢀ131 ppm) at a Si:F ratio of ca. 2–3
max
max
2
equal mixture of Q (3Al) ( Si d
ꢀ79.1 ppm) and
3
max
2
9
ꢀ
ꢀ
Q (2Al) ( Si d
ꢀ81.1 ppm). In the presence of F
3
max
the Si:Al ratio of the solid phase was 0.25 which again
suggested a mixed phase. Interestingly, the ratios of
Si:F and Al:F in the solid phase were ca. 2.0 and 8.5,
respectively, and these may prove to be helpful in iden-
tifying the location of F in the solid phases. Again all Al
B
some HAS . The Si:Al ratio of the solid phase, ca. 0.94,
A
confirmed the presence of the former and this was sup-
2
7
29
ported further by solid state Al and Si NMR which
showed that Al was present in both octahedral (dmax
2.5 ppm) and tetrahedral (dmax 55.0 ppm) geometries
and that Si was coordinated primarily through
2
7
was in octahedral geometry ( Al d
presence of HASA was confirmed by Si NMR, though
2.7 ppm) and the
max
2
this time the proportion of Q (3Al) ( Si
9
d
78.2 ppm) was very much higher than Q (2Al) ( Si
Q (1-2Al) linkages (d
max
ꢀ90.3 ppm). Microprobe and
3
max
9
3
2
ꢀ
NMR characterisation of the solid phase which formed
ꢀ
3
1
9
d
ꢀ82.1 ppm). F NMR gave dmax at ꢀ151.1 ppm,
in the presence of F showed that in respect of both Al
and Si it was identical to that formed in the absence of
max
1
9
which, based upon the chemical shift ( F d
max
ꢀ
ꢀ150.1 ppm) we obtained for a pure precipitate of alu-
minium hydroxyfluoride, was probably indicative of F
F . This was unusual in that for previous parent solu-
ꢀ
tions the additional presence of F had always resulted
associated with Al(OH)3(s), and ꢀ131.6 ppm which we
in structural changes manifested as increases in the pro-
portion of Q (3Al) linkages relative to Q (1-2Al). No
such changes occurred when F was incorporated into
believe showed F associated with HAS .
A
3
3
ꢀ
(ii) 1.0Si/2.0Al ± 1.0F – In the absence of F the solid
phase was predicted to be primarily HAS and this was
the predominantly HAS -like phase precipitated from
B
A
1
this parent solution. However, the F NMR chemical
9
supported by a Si:Al ratio of 0.43 and characteristic
2
7
chemical shifts for both Al NMR (dmax 2.5 ppm) and
shift at dmax ꢀ132.8 ppm was similar to those obtained
2
9
ꢀ
Si NMR (d
ꢀ79.6 and ꢀ82.3 ppm). In the presence
for substitution of F into HAS -like structures
A
max
ꢀ
of F the Si:Al ratio of the solid phase was 0.41 which
was indicative of HASA and the Si:F and Al:F ratios
were ca 2.0 and 5.5, respectively. The lower ratio of Al
(ꢀ131.6, ꢀ131.1 and ꢀ131.9 ppm) and this may indicate
that F was present in a similar chemical environment in
both HAS structures. The solid phase was different from
to F in this phase was probably indicative of a higher
ꢀ
the F-containing HAS -like solid phases in that the ra-
A
[
F ] to [Al] in the parent solution whereas this effect
tioꢁs of Si:F (ca. 4) and Al:F (ca. 4) showed an increase
in Si and a decrease in Al relative to F when compared
was limited by a significant reduction in the proportion
of aluminium hydroxyfluoride in the mixed phase as was
suggested by the almost complete disappearance of the
with HAS -F. These apparently contradictory changes
A
in the content of F in the solid phase relative to both
Si and Al may actually be informative in respect of the
mechanism of formation of HAS via HAS For exam-
1
9
F NMR chemical shift at ca. ꢀ150 ppm. Solid state
2
7
29 19
Al, Si and F NMR showed that this precipitate
B
A
was almost identical to that isolated from the previous
parent solution (0.5Si/2.0Al/0.5F) except that the major-
ple, it was assumed that the solid phase which formed in
ꢀ
the absence of F was a mixture of HAS and HAS
B
A
1
9
ity of F was now associated with HAS ( F NMR d
and this was supported by NMR. However, the Si:Al ra-
tio of this ꢀmixedꢁ phase was very close to 1.0 and this
may indicate that the predominant forms of HAS were
A
max
ꢀ
131.1 ppm).
ꢀ
(iii) 1.0Si/1.0Al ± 1.0F – In the absence of F the so-
lid phase was predicted to be HAS with perhaps some
actually HAS (ca. 60% of the total) and an intermedi-
B
A
HAS and the latter was borne out by a Si:Al ratio of
B
ate form HASAB (ca. 40% of the total) in which the fur-
ther incorporation of Si(OH) has not yet triggered the
dehydroxylation reaction which would result in 50% of
its constituent Al switching from an octahedral to a tet-
rahedral geometry (Fig. 5). It is our understanding to
date that the rate at which dehydroxylation takes place
0
supported by both Al NMR, which indicated a small
.63. Again the presence of a proportion of HAS was
2
B
4
7
2
7
amount of Al in tetrahedral geometry ( Al NMR d
max
2
3.8), and Si NMR which showed that Si was now
9
5
coordinated through both Q (3Al) and Q (1-2Al) link-
3
3
ꢀ
ages [6]. The presence of F appeared to reverse the
is highly dependent upon the [Si(OH) ] which is free to
4