Preparation and Characterization of Model Catalysts for the HCl Oxidation Reaction - Atomically-Precise Methods for Synthesis of Solid Catalysts - page 212

produce a Ru-DMF complex and HCl gas was expelled. By adding 20 mg
acetic acid before refluxing, it was possible to increase the release of HCl.
Then 300 mg ethanol and 200 mg of 12 wt% polyvinylpyrrolidone in
methanol were added and mixed vigorously. For the solution to be used for
electrospinning, ethanol was used possessing analysis grade. After calcin-
ation at 475 1C for 30 min. (ramp of 5 1Cmin 1 ) hollow nanotubes were
observed, especially if low amounts of precursor were used. Adding 25 mg
H 2 O or using polyvinylbutyral (PVB, Mowital s B60 H) as spinning polymer
resulted in compact nanofibers instead of nanotubes. The electrospinning
parameters were as follows: voltage: 5 kV; distance: 6 cm; feed rate: 0.3 mL h 1 ;
relative humidity: o 25%
The Ru x Ti 1 x O 2 nanofibers could be prepared by adding the appropriate
amount of Ti(OiPr) 4 after refluxing. In case of x o 0.2 it was also possible to
use RuCl 3 H 2 O as precursor without refluxing.
In order to visualize reaction-induced morphological changes of the
catalysts after the HCl oxidation reaction we used RuO 2 -based nanofibers
which were synthesized by electrospinning as described above. The crystal-
lite size of the RuO 2 nanofibers is about 9 nm according to Rietveld analysis
with a BET surface area of 30 m 2 g 1 . Reaction-induced alteration of the fiber
morphology becomes quite easily visible in scanning electron microscopy
(SEM). In Figure 8.6a and b we show how the morphology of pure RuO 2
nanofibers degrades after HCl oxidation reaction in a flow reactor under
oxidizing feed gas composition (p(HCl) ¼ p(O 2 ) ¼ 200 mbar using a buffer gas
p(Ar) ¼ 600 mbar; the total flow rate of the reaction mixture was 50 mL min 1
(STP)) keeping the catalyst bed at 650 K for 2 h on stream. Clearly, pure RuO 2
is structurally not stable under typical Deacon conditions which may lead
also to catalyst loss under industrial conditions. Temperature treatment
alone in Ar þ O 2 flow does not lead to sintering at temperatures up to 750 K.
Consequently, the observed transport of Ru even at temperatures as low as
573 K suggests the participation of a volatile Ru species formed during the
HCl oxidation reaction. We presume that a chlorine-containing Ru species
serves as chemical transporter. The BET surface area remained nearly con-
stant which is in line with the almost constant reactivity of the fibers during
the atmospheric pressure HCl oxidation at 573 K. Obviously, activity ex-
periments alone are not able to identify morphological instabilities of a
catalyst.
This result conflicts apparently with the observed stability of RuO 2 (110)
and RuO 2 (100) model catalysts. Therefore, the morphology changes of RuO 2
fibers suggest that only RuO 2 facets with orientations other than (110) and
(100) are corroded by HCl. Quite in contrast, mixed RuO 2 -TiO 2 fibers are
morphologically stable under the same HCl oxidation reaction conditions
(cf. Figure 8.6c and d).
The Ru-doped TiO 2 nanofibers were prepared with 15 at.% Ru, resulting in
phase-pure rutile mixed RuO 2 -TiO 2 fibers (Rietveld analysis). The normal-
ized catalytic activity of mixed RuO 2 -TiO 2 nanofibers is as high as that of
pure RuO 2 nanofibers. 13 As shown in Figure 8.6b and d, mixed RuO 2 -TiO 2

d n 9 r 4 n g | 8

.