Chemistry Reference
In-Depth Information
be; Ni
2
P W Co
2
P W Fe
2
P, WP, MoP. The most active catalysts produced
predominantly phenol and benzene with small amounts of methoxy-
benzene. Catechol and cresol were observed as the intermediates at
short contact times. 5 wt% Pd/C was tested as reference system and
found to be more active but with less favorable selectivity, giving only
catechol. A commercial CoMoS/Al
2
O
3
deactivated fast compared to the
phosphides.
Li and Wang
200
studied the HDO reaction of anisole over Ni
2
P/SiO
2
,
MoP/SiO
2
and NiMoP/SiO
2
. The activity was observed to decrease in the
order Ni
2
P/SiO
2
W NiMoP/SiO
2
W MoP/SiO
2
, which was explained by a
higher d-electron density of the Ni
2
P system. The nature of the HDO
mechanism was investigated in a fixed-bed reactor and mainly three re-
actions were observed to proceed, being the demethylation of anisole,
hydrogenolysis of phenol and hydrogenation of benzene.
Oyama et al.
149
studied a series of Ni
2
P/SiO
2
(6-22 wt%) catalysts,
produced by temperature programmed reduction, in a three-phase
packed bed reactor for HDS (3000 ppm S as benzothiophene), HDN (2000
ppm N as quinoline) and HDO (500 ppm O as benzofuran). Higher
conversions were obtained than with a NiMoS/Al
2
O
3
reference catalyst. A
loading around 18 wt% was found to give good performance with Ni
2
P
crystals of ca. 7.7 to 9.8 nm. While HDS was found to be structure in-
sensitive, the HDN depended upon local structure. This study was a
follow-up study
201
with its focus on the effect of the Ni/P ratio (max. ac-
tivity with Ni : P of 1 : 2) on hydroprocessing properties.
Some studies have emerged that target particle size derived effects. Bui
et al.
202
prepared SiO
2
supported Ni
2
P, WP, MoP, CoP, FeP by both the
phosphide and phosphate methods, for comparing the phosphate route
with the lower-temperature route. The catalysts were tested for HDO of
2-methyltetrahydrofuran as model feed component. However, similar
TOF values were obtained with the two methods, except possibly for the
WP catalyst. Figure 10 shows a comparison of the catalyst performance
for the total conversion of 2-methyltetrahydrofuran (2-MTHF) and the
HDO products selectivity as function of temperature.
Cecilia et al.
203
prepared Ni
2
P/SiO
2
with small crystal sizes by the
phosphite (low-temperature) route. Four loading levels of Ni (2.5, 5, 7.5
and 10 wt%) with variations in the P/Ni ratio were studied. Particle sizes
in the range 14-31 nm were obtained, with the size found to increase with
loading. TOF values were estimated for HDO of dibenzofuran. The values
were always lowest for the 10 wt% (largest particles). However, the
intermediate loading catalysts were observed to possess superior in-
trinsic kinetics. It was also claimed that these catalysts showed high coke
resistance, and also the H
2
O was not a problem when the catalysts were
prepared with a high P/Ni ratio, likely leading to a P modified support.
Duan et al.
158
have looked into different procedures for passivation of
phosphide catalysts (Ni
2
P/MCM-41), namely by using alternatively H
2
Sor
NH
3
instead of a lean mixture containing O
2
.H
2
S passivation was found
to be beneficial for HDS, and it also required no prereduction. NH
3
passivation led to strongly bound N on the Ni
2
P phase. Thus, the results
imply that S can be a beneficial surface species on phosphide catalysts.
