the accepted principle that base catalysed sol-gel reactions lead to a denser material'.
Acid catalysis facilitates hydrolysis but retards condensation, whereas the reverse is %
true for base catalysis. Thus it would appear that the lower oxide contents in the acid hydrolysed ethoxides are due to less extensive condensation and, given that base catalysis would not remove all the alkoxide groups by hydrolysis, that condensation in this system is taking place by alcoxolation or olation rather than oxolation. One factor, though, that must be considered is that the acid-hydrolysed samples are dried gels and may, therefore, be subject to contamination with the sodium chloride produced as a by- product in the preparation of the ruthenium alkoxide species. This remains dissolved in the solvent trapped within the gel network and is deposited when the solvent is
evaporated. Thus the proportion of the final product corresponding to RUO2 and Ti02 M
and partially reacted precursors will be lower for gel samples.
Sample Oxide Content /% Unaccounted Mass /%
1 93.6 6.4 2 86.6 13.4 3 84.0 16.0 4 84.1 15.9 5 78.5 21.5 6 102.1 -2.1
Table 4.3: XRF Determined Bulk (5-500 |im) Metal Oxide Contents of the Ru02-Ti02
Samples obtained by sol-gel techniques.
Sample 6, showing anomalous behaviour, derives from an isopropoxide
precursor which is probably a monomer containing bound solvent molecules, due to the steric hindrance of the bulky isopropoxide groups preventing oligomerization. It is, therefore, more reactive than the oligomeric ethoxide precursors as it has been shown' that bound solvent molecules are easier to remove than coordinated alkoxide species.
This allows faster, and possibly more extensive, hydrolysis and may explain the high
oxide content value for this system.
43.4 Powder XRD analysis of the Calcined RuOz-TiOi Mixed Oxide System
The X-ray diffraction patterns for the six samples are shown in Figs 4.6- 4.11. The analysis reveals three major features. Firstly, it is clear that the samples are more likely mixtures of RuOa and Ti02 rather than a solid solution of the two. This becomes apparent from the peak positions. The TiÛ2 has crystallized in the anatase phase as indicated by clear peaks at around 20 = 25.28° and 48.05° and the RUO2 has crystallized in a different, rutile, phase with peaks at values close to 20 = 28° and 35°.
If the products were mixed oxides a single rutile phase should be formed with peaks at 20 values between the pure RuOg and Ti0 2 values, depending upon the ratio of the metals, as has been observed in other work on mixed Ru0 2-Ti0 2 oxide systems^.
Previous work^ has suggested that it is hard to obtain mixed oxides in Ti rich systems.
The peak values for RUO2 in all cases do show a slight displacement from the values
for pure RuOg and it is possible that the precursors do mix to some extent, this being more apparent in the case of the ammonia and water hydrolysed systems (Samples 2 and 3, Figs 4.6-4.8) especially the low volume sample (3, Fig. 4.8), where there is evidence of a shift in the rutile peak at 20 = 28.02° to a value that would correspond to
Chapter 4: Non-Aqueous processing o f RUO2
The second interesting feature of the patterns is the presence of metallic ruthenium in all the samples except the ammonia hydrolysed sample (1). Osaka et al'^ examined the application of sol-gel techniques for the production of iridium dioxide and found that with ethoxide precursors there was a tendency to precipitate metallic iridium upon calcination. This was suggested to be due to the partial charge (as calculated using the method of Livage') on the iridium metal centre being too low and that consequently iridium ethoxides are not readily susceptible to hydrolysis. It would appear reasonable to assume the same of ruthenium in this case as the electronegativity is the same as for Ir (2.2 in Pauling’s scale). The absence of metallic Ru in the
ammonia hydrolysed sample may be due to efficient alcoxolation by hydroxide
groups bound to Ti removing coordinated ethoxide groups bound to Ru rather than these groups being hydrolysed as follows :-
=Ti-OH + RO-Ru- = T i-0-Ru-+ ROH
Acid catalysis retards this condensation reaction as the hydrolysed species involved are often protonated and thus less likely to interact with one another. This
means that titanium species more likely to condense with themselves than with the low
partial charge ruthenium centres leaving unreacted R-OR units which can pyrolyse to the metal. It has also been postulated^ that ruthenium disproportionates to the metal
and the oxide under sol-gel conditions. This may happen in this system; however, the
lack of metal in the ammonia hydrolysed sample (1, Fig 4.6)) suggests that this is not
the case, since if disproportionation were occurring then this sample should also contain some metal.
2—T hota - S c a le JOHNSON MATTHEY TECHNOLOSV CENTRE 2 S -S o p -1 9 9 S 16 ;49
10 28 25 35 40 45
5 0 55 60 65 70