100
Q. Wang et al. / Materials Research Bulletin 74 (2016) 98–102
increased to 673 K, three weak peaks appear at 2
u
= 31.8, 44.2 and
presented. The average Ru
SiO and RuP crystallite size of 10 nm for the RuP/SiO
calculated by the Scherrer equation. It can also be seen from above
results that the synthesis of RuP and Ru P can be controlled by
2
P crystallite size of 13 nm for the Ru
2
P/
ꢀ
46.0 , which are attributed to the RuP phase, indicating the
2
2
were
formation of RuP. As the reaction temperature increases to 823 K,
the intensity of these peaks increases which suggests the growing
2
up of RuP crystalline size. Seven diffraction peaks locating at about
simply adjusting Ru/P molar ratio. The phosphor content in the
precursor before heat treatment was higher than that in the final
ꢀ
2
u
= 29.1, 31.8, 32.5, 35.8, 44.2, 46.0, and 53.6 are observed at the
treatment temperature of 823 K (Fig. 1e), which could be identified
to the (0 0 2), (0 01), (110), (111), (112), (211) and (0 31)
crystallographic planes of RuP, respectively. The XRD patterns for
RuP are similar to that reported previously [26], while the
synthetic temperature (673 K) was much lower in the current
method.
2
product of RuP or Ru P, which can be explained by that the part of
TPP will be evaporated at the temperature of 723 K higher than its
boiling temperature (650 K).
For comparison, the SiO
2
supported RuP with 23 wt.% Ru
loading was also prepared by conventional phosphate as phosphor
source. The XRD patterns of samples reduced at different
temperatures are shown in Fig. S1. It can be seen that the RuP
can be synthesized by conventional method at 873 K for 6 h. This
result was similar to the others reported [8].
3.2. The influence of molar ratio of ruthenium and phosphorus on
ruthenium phosphide
2
For HDS reaction of DBT, the SiO supported RuP with 2.0 wt.%
2
Besides this, the Ru P can also be obtained by changing molar
ruthenium content was synthesized. The XRD patterns of
ratio of Ru/TPP from 1:3 to 1.5:1. The XRD patterns of samples with
3 wt.% Ru loading prepared at 723 K and different Ru/TPP molar
ratios are shown in Fig. 2. When the molar ratio of Ru and P was
phosphide catalysts prepared by TPP and conventional methods
ꢀ
2
are shown in Fig. S2. Only the peak at 22.9 attributed to SiO
2
was
observed due to the low loading or high dispersion of metal which
is beyond the detection limit of XRD.
ꢀ
ꢀ
ꢀ
below 1:1, the main peaks at 31.8 , 44.2 and 46.0 corresponding
to RuP phase were observed. As the Ru/P ratio was changed to 2:1,
ꢀ
ꢀ
the peaks at 38.1 and 47.1 attributed to Ru
2
P phase appeared.
3.3. TEM images of samples
ꢀ
However, the peak at 44.0 belonging to metal Ru existed. And
when the molar ratio was 1.5:1, the peaks at 30.5 , 38.1 , 40.6 , 47.1
and 53.5 all corresponding to Ru
ꢀ
ꢀ
ꢀ
ꢀ
2
TEM investigation of 2.0 wt.% RuP/SiO catalyst prepared by TPP
method reveals that the particle size of RuP was about 2–5 nm and
ꢀ
2
P phase (JCPDS 65-2382) can be
sphere-shaped RuP particles uniformly dispersed on the surface of
SiO (Fig. 3b). Meanwhile, the morphology of RuP/SiO with same
2 2
loading prepared by the conventional phosphate as phosphor
source was also shown in Fig. 3a. It can be seen that the RuP
obviously existed in the form of aggregation with a larger particle
size of 5–15 nm. It suggests the effectiveness of the TPP method in
increasing the dispersion of the active species. As shown in Fig. 3c,
the particle size of Ru
–5 nm and sphere-shaped Ru
of SiO . Besides this, Ru
coexisted. The particle size of RuP prepared by phosphate as
phosphor source was bigger than that prepared by TPP method
because that the higher temperature causes the sintering of the
phosphide. It is also seen from the TEM investigation that
2
P prepared by TPP method was about
P particles dispersed on the surface
P with the bigger particle size (ꢁ 10 nm)
2
2
2
2
2
the particle size of RuP and Ru P prepared by TPP method was
similar.
3
.4. Activity measurements
2 2 2
The activities of RuP/SiO and Ru P/SiO catalysts prepared by
TPP, RuP/SiO
with Ru/SiO
2
prepared by phosphate as phosphor source, together
catalyst were tested in the hydrodesulfurization
2
(
HDS) of DBT. The HDS activities of the four catalysts at different
reaction temperatures were shown in Table 1. The conversions of
DBT on these catalysts increased with increasing reaction
temperature. And the supported noble metal phosphides of RuP
and Ru
Ru/SiO
2
P prepared by TPP method exhibited higher activity than
catalyst. The HDS reaction network of DBT was shown in
2
Scheme S1. The HDS products were analyzed and summarized in
Table S1, and it can be seen that the selectivity of cyclo-
hexylbenzene (hydrogenation product) over phosphides was
higher than that over ruthenium catalyst which indicated the
hydrogenation activity of ruthenium phosphides was higher than
that of ruthenium. The result was consistent with that reported by
Bowker et al. [21]. It is known that hydrogenation of DBT favored
the HDS process. Compared with the RuP/SiO
phosphate as phosphor source, the RuP/SiO
method presented much higher HDS activity in the test. In
particular, the 37.6% initial activity of RuP/SiO prepared by the TPP
method is higher than that of RuP/SiO prepared by conventional
2
catalyst prepared by
2
prepared by TPP
Fig. 2. The XRD patterns of the ruthenium phosphides prepared by TPP methods
from Ru/P molar ratio (a) 1:3, (b) 1:2, (c) 1:1, (d) 1.5:1, and (e) 2:1
2
2