Effect of pretreatment on textural characteristics of an alumina-supported nickel-tungsten catalyst
Descripción
~ut~~als Chemistry and Physics, 24 (1990) 443-456
EFFECT AN
OF
PRETREATMENT
ON
443
TEXTURAL
CHARACTERISTICS
OF
ALUMINA-SUPPORTED NICKEL-TUNGSTEN CATALYST
Z.FOLTYNOWICZ and Z.SARBAK Faculty of Chemistry , Adam Mickiewicz University 60-780 Poznah (Poland) Received April 27,1989 : accepted June 1.1989
ABSTRACT The results of the investigation of the porous structure of oxide
,
reduced and sulfided
forms of alumina supported . Microdistribution
tungsten catalyst are presented was
determined ,
adsorption
on
the
intrusion porosimetry different
basis
of
the
while macrodistribution
forms
of
low
nickel
of the pore5
temperature
was determined
nitrogen
by mercury
. The change in pore size distribution the
NiWfAl,O,
catalyst
relation to the condition5 of its activation
is .
discussed
of in
It was revealed
that the process of reduction by hydrogen as well as simultaneous reduction-sulfiding
by a mixture of hydrogen and hydrogen sulfide
result in new surface species formation which affect the porous pattern of catalyst5 .
Catalysts for hydroprocesses cations: Cr
,
MO or W
cations as promotors often
used
catalysts
are obtained by a deposition of
as active
components and Fe
, onto a gamma alumina support are
prepared
components and promotors: CO-MO
,
from
a
mixture
Ni-Mo and Ni-W
, Co or Ni . The
of
most
active
, with nickel
considered the most preferable for economic reasons .
444 Classical support
preparation
impregnation
,
of the promotors
of
with
the
above
a nitrate
,
during
catalysts
solution
before
is
based
containing
or after
on
cations
impregnation with . Preparations
ammonium paramolybdate and ammonium metatungstate
obtained in this way are dried and calcined . As a result , oxide forms are produced . Oxide catalysts prior to use are activated by reduction with hydrogen
and
temperatures
sulfidation . As a
with ,
result
hydrogen
catalysts
sulfide active
at
higher
in hydrocarbon
hydrogenation and hydrogenolysis of such bonds as C-N , C-S , C-O or C-Metal are obtained . Depending
on
the
applied
conditions
in
the
preparation
procedure , catalysts with a specific structure of the surface are yielded
. The structure
type
catalytically
of
in
,
turn
active
determines
,
the
amount
and
. From the point of view
centres
of catalysis
it is the very importnnt to define these centres
Henceforth,
this is the mnin
.
workers
,
However
they
field of study of many research
often
tend
to
neglect
the
problems
concerning the porous structure As is well know , the type of pores and their sizes as well as the type and concentration
of active centres hnve
influence on
the activity of the catalysts . The question of the porosity is of considerable importance in the processes of molecule transportation to and from the catalyst active centres , particulary in the case when heavy fractions of petroleum or coal liquids are Large
organic
move freely active
centres
respective well
as
the
on which
reagent
,
taking with
catalyst
used
the
products
.
,
place
in
hydrogen
NiW/A150, in
reuson
access
would
be
and hydrogen is of
one
of
active
petrochemical
very
leave
centres
. to
system sulfide
as the
and as a result
important
the
formed
freely
deactivated
porous
to the
and mny undergo
blocked
to the
should
access
products
should
set
it is very
, which
processing
chemisorbed
molecules
the
to gain
conversion
would
have
the catalyst
mentioned
activation
the
, pores not
in order
may be
unreacted
consequence changes
they
the nbove products
UP
system
. Also
would
.
building
pore
. Otherwise
molecule5
above
the
conversion
catalysts other
molecules
inside
used
Due
to
the
examine
the
during of
oxide
active and
. In
therma 1 forms
of
catalysts
carbochemical
445
EXPERIMENTAL Catalyst preparation In the experiments described below a nickel-tungsten catalyst was used . It the
was made of 5.1 wt% of NiO and 22.0 wt% of WO, by
impregnation
technique
labelled as 0-NiW/Al,O,
. The
catalyst
in oxide
form was
.
The oxide
form was reduced with hydrogen at temperature of -1 for 2 hours. The reduced 400°C at the flow rate of 100 cm3 min .
form was labelled as H-NiW/Al,O, ,
On the other hand
the sulfided
form of the catalyst was
marked as S-NiW/Al,O, and was obtained under the same conditions as the previous one . However
,
instead of hydrogen
, a mixture . Details
of 10 ~01% of hydrogen sulfide in hydrogen was applied
of the preparation conditions were desccribed in paper [ll.
Determination of the pore size distribution in the pore radius range 5-150 A ,
The microdistribution temperature
(77 K) nitrogen adsorption measurments
out on a Sartorius microbalance the measurements
low
were carried
(Gravimat , type 4133) . Prior to
the samples were outgassed at 523 K in a vacuum
of 10-s Torr . The pore size microdistribution
in the pore radius range from
5-150 %,was calculated from the desorption branch of the isotherm by the method of Cranston and Inkley performed onan In
results
I21 .A11 calculations were
IBM computer using a specially elaborated routine.
,the data
on
cumulative
pore
volume
,
Vcum
the
mean pore radius r cumulative pore surface area S as well cum ’ av as plots of the pore volume and pore surface area distributions as functions of the pore radius were obtained . The
pore
size
75-75000
%, was
performed
on
pressure
macrodistribution obtained
a Carlo
range
macrodistributions
Erba
0-1000 ,
by AG
in
mercury 65a
atm.
the
pore
porosimeter The
radius
range
intrusion measurements in the
parameters
i.e. AV/Alg(rl,AS/Alg(rl
computer based on the Ritter and Drake method
of
applied pore
were calculated by I31 .
The t-plots representing the relationships between the amount of
nitrogen
adsorbed
V
layer t were calculatedaby
and
the thickness of the adsorption the Lippens-de Boer method [41 based
on the result of nitrogen adsorption measurements.
446
RESULTS AND DISCUSSION The pretreatment
of the oxide form of NiW/Al,O,
catalyst
in
hydrogen or hydrogen and hydrogen sulfide mixture results in new species formation on surface of the catalyst such
a
and
species
its
existance
in
[ll . Appearance of
different
geometrical
configurations should affect the surface area as well as Porous structure of different forms of NiW/Al,O, catalyst . Nitrogen adsorption as well as mercury instrusion measurments confirmed our assumptions . The total amount of adsorbed nitrogen in case of 0-NiWfAlsO, and S-NiW/Al,O, is comparable , whereas in the
case
of
H-NiW/Al,O,
nitrogen capacity catalysts
and
it
is
distiguished
by
much
larger
. Adsorption isotherms for nitrogen on reduced
their
precursor
are
given
in
Figs. 1-3.
Va (cm3 g'-*)
P/P, Fig. 1. Nitrogen adsorption - desorption isotherms 0-NiW/AlZ03 catalyst : 'A'-adsorption; 'o'-desorption.
for
an
447 Va
(cm3 g's)
I
0.0
0.1
0.2
0.3
0.3
0.F
0.5
0.7
0.8
0.Q
1.C
P/P, Fig.
2. Nitrogen
H-NiW/A1,03
Va
adsorption
:
catalyst
- desorption
'A'-adsorption;
isotherm
for
an
for
an
'o'-desorption.
(cm3 g-2)
1
53o/m
P/P, Fig.
3. Nitrogen
S-NiW/A1203
adsorption
catalyst
- desorption
: 'A '-adsorption;
isotherm
'IX'-desorption.
All isotherms are of type exibit
a
sharp
0-NiW/Al,O,
knee
IV in BDDT .
(point B)
classification
However
(Fig.1) approches the line P/P_=
whereas those on H-NiW/Al,O, a final upward turn
,
the
. They
isotherm
on
1 asymptotically
,
(Fig.2) and S-NiW/Al,O, (Fig.3) show
. The isotherm on 0-NiW/Al,O,
IFig.1) shows
hysteresis with a shoulder point at p/pm= 0.35 , indicating small pores
.
The
hydrogen-reduced
form
shows
a
hysteresis
loop
extending up to p/p*- 0.58 (Fig.2) . This indicates the origin of a
great
number
of
larger
pores
.
Simultaneous
sulfiding results in smaller pore formation on
the
S-NiW/Al,O,
isotherm
(Fig.3) occurs
reduction
and
. The shouder point at
lower relative
pressure (p/pa= 0.52) . The hysteresis loops are of a mixed shape and resemble those of E or B'type with admixture of A , according to de Boer
[51 . This
is equivalent
to the presence of tubular
capillaries with wide and narrower parts as well as so-called ink bottle capillaries , which is reflected on the plots of pore size distribution given in Fig. 4-6 .
As/Ar (ms g-l A-+)
AV/Ar 1BJ(crn3 g i h-%)
Pore radius (A) as functions of pore radius Fig. 4. pore size microdistributions for an 0-NiW/A1203 catalyst : a) pore volume - 'A'; b) pore surface area - '0'.
449
AV/Ar
103(cm3
Pore Fig.
5. Pore
As/Ar
g-l h-1)
radius
(&I
size microdistributions
for an H-NiW/A1,03
catalyst
10J(cm3
6. Pore
surface
g-+ A-1)
Pore Fig.
as functions
: a) pore volume b) pore
AV/Ar
catalyst
radius
area
- '0'.
(rn2 g-1 A-1)
(A)
size microdistributions
for an S-NiW/A1,03
of pore
- 'A';
As/Ar
radius
(ml g-i %I-l)
as functions
: a) pore volume b) pore
surface
of pore
- 'A'; area
- 'a'.
radius
450
The pore size distribution for 0- and H-NiW/Al,O, forms in the radius range 15-150 fi (so called micropores
I611 is practically amplitudes. In the
monodispersed with wide maxima of different precursor 0-NiW/Al,O, pores
less
than
.
(Fig.4a)
The
significant surface As
a
60
porous
change
of the pore
most A
wide
with
structure
during
volume
is contained
dominating of
hydrogen
radius
0-NiW/Al,O, ,
treatment
at
in
35
A
undergoes
a
which
causes
a
layer reduction as well as catalyst component migration
result
of
surface
layer restructurization
upon
hydrogen
treatment narrower pores are formed with dominating radius at 25 A (Fig.Sa) ; became
local maxima
levelled but
about 100 A for the 0-NiW/Al,O,
some knee
came
into view
sample .
at 60 A
This
family of pores developed completely during reduction -sulfiding in HZ-H,S
mixture
,
which
can
be
seen
in
Fig.6a
as
a
new
intensive peak exactly at 60 A . As a consequnce of sulfiding new pores are formed , which
is reflected on the plot of pore size
distribution as a maximum at 85 A . From Figs. 4b-6b we can see clearly that the main contribution to cumulative pore surface area comes exactly from the same pores as in the case of cumulative pore volume.
Table 1. Characterization of the pore
structure on the
basis of
nitrogen adsorption-desorption measurements. Sample
Pore volume Vc,,,,, Mean pore cm3g-l
Surface area
radius F av A
m2g-1 Cranston Inkley method S ClJUl
0-NiW/l,O, H-NiW/l,O, S-NiW/l,O,
0.37 0.48 0.34
388 342 353
182 231 165
BET method SET
238 305 210
The analysis of nitrogen adsorption data is shown in Table I . we can see , both processes of pretreatment effect the parameters of pore structures in opposite manner . Hydrogen
As
treatment
leads to the significant
increase of Vcum
, Scum as
well as SBST , while pretreatment with the mixture of hydrogen and hydrogen sulfide cauaea a decline of above mentioned
451
parameters H,S
. The diminishing of pore structure parameters due to
presence
,
formation sulfide
ions
0-NiW/Al,O, hydrogen phase and
can
pore
.
[7,81 surface
atmosphere
, while
to
(macropore)
on
On
the
NiO
phase
,
hand easily
[l]
. The
of
of
new
exchanged
be
explain
enlargement
basis
are
other
can
layer would
species
for
some
parts
converted presence
in of
the 28% increase
Vcum
larger
is a result
of the
such
a
of Sewn of
large
.
formation surface
the
oxygens
species
30%
The cumulative
area of pores
Scum,determined
by the CI
[21 , is much smaller than the specific surface area of the
method
samples from
explained
terminal
in the surface
SBET
tions
be
where
estimated
the nature
of the sample
of both methods.
indicate
the
relationships
existence
,
.
by the BET method surface
The values of
presented
These
as well
of mean
macropores in Fig.7
result
as from the assump-
statistical
. The
,
differences
pore
linearity
confirms
the
of
radius Va-
presence
t of
macropores.
‘a
(cm= s-1)
mo,
150 -
Fig.
7. Relationship
between
the amount
and
the
of the
adsorption
thickness
a) 0-NiW/A1203
- 'A'
b) 0-NiW/A1,03
of
nitrogen
layer
- '0'
c)
t for
adsorbed catalysts
0-NiW/A1203
V
a
:
- '0'
452
In order to describe the porous structure more completly , the macrodistribution
in
(the pores falling
the
radius
range
75 to
75000 A) has also been studied on the basis of the results of mercury
intrusion
measurements
. The
analysis
of
the
results
obtained is shown in Table II. Table II. Characterization of the pore structure on the basis
of
mercury porosimetry measurements. Pore volume Vcum Mean pore
Sample
0-NiW/Al,O, H-NiW/Al,O, S-NiW/Al,O,
is
m2g-l
6.24 7.66 5.52
837 1015 1416
0.40 0.36 0.19
contribution
to
insignificant
.
precursor
sample
reduced
sulfiding
caused
50% decline
at
least
the
two
processes
of
formation
the
results
of
.
assumption
. However
is
‘av
new
upon
treatment
The
connected
of
of
V
of
Vc_
walls
of
fiav
takes
of
pores
value place
shows
that
The
change
.
of
value
in
surf ace
:
formation data
.
we claim
macropores
-
, while Vvum
is
give
their layer
a large
the for
our
increase
of
S-NiW/Al,O,
smaller
,
than
for
species
on
. The enlargement
restructurization
macropore
the
sulfide
new bulky
volume
layer
On
confirms
of H-NiW/Al,O,
surface
reducing
.
hydrogen-hydrogen
or
reduced
the
that
enlargement
which
the
and i?av we may conclude
increases significantly , but ‘av 0-NlW/Al,O, . Reduction and sulfiding the
of of
into
adsorption
, in the case
with
Scum
Taking
species
Rav
area
, while
parameter
their
surface
hydrogen
change
surface
pretreatment
slightly this
CUllI
nitrogen
total
Hydrogen
affected
and
pore
the
Scum only
behavior
restructurization basis
area
A
only
consideration
Eav Surface
cm3g-l
The macropore sample
radius
structure
is
also
shown
in
detail in Figs. 8-10, which represent differential curves of the pore volume (Figs. 8a-10a) as well as pore surface distribution (Figs. 8b-lob) as a function of the pore radius. Generally area
comes
distinguishes pore
surface
some small is
, the
only contribution
from the by its
pores
own porosity
distribution local
less are
maxima , while
open to the mesopores
.
to the total pore surface
than
200
pattern. practically the
plot
fi
.
Each
Differential
sample
is
curves
of
monodispersive for
H-NiW/Al,O,
with (Fig.9b)
453
AV/Alsr
(cm3 s-2 &-+I
As/Algr
1
(cm2 g-% W-it
i
lo2
103
lo4
1Q5
Pore radius (%i) Fig.6. Differential curves of the pore volume(A) and the pore surface area (a)macrodistributions the pore
AV/Algr
Ae/Algr
(cma g-1 &-A)
I
(cm2 g-i A-1)
G‘B
~1,~ ,
120 I10 loo
as functions of
radius for an 0-NiW/Al,O, catalyst.
‘i
0.7
GIlI
Il.8
830
D.6
75 Eo
: xl ZI IO 0
0.6
i
i
:
0.4
‘n.3
cl.2 n.i
1
i
10
103
104
0.0
10=
Pore radius (h) Fig.9. Differential curves of the pore volume (A)and the Pore surface area(o) microdistributions radius for an H-NiW/Al.O, catalyst.
as functions of the pore
454
AV/Algr
(cm3 CT1 A-i)
As/Algr
I&or
(cm2 g-% A+)
, I.0
I
t
102
103
1%5
104
Pore radius (8) Fi!3.10. Differential curves of the pore volume (A)and the pore surface (n) macrodistributions
as functions of the
pore radius for an S-NiW/Al,O, catalyst.
On the contrary , pore volume distribution is polydispersive In the case of 0-NiW/Al,O, pore structure components smoothes
(Fig.SbI the sample polydispersity of
is a result
on
the
.
A1202
out polydispersity,
of
imposition
matrice specially
.
of Ni
Hydrogen
and W
oxides
pretreatment
in the region
of pores
with radius below 40% fi , with the formation of sharp maxim% at 180 %I (Fig.9b) . These pores became nearly completely reduced in sulfide atmosphere , but eimultaneous restructurization of the surface layer results in formation of a
hydrogen-hydrogen
new family of pores in the radius range 15% to 25% % (Fig.l%b) . Hydrogen
treatment
macropores
led to the development
, already present
of three branches
in the 0-NiW/Al,O, precursor
of
,i.e.
range 1000 - 75%00 A . This results in a Vcum increase for H-NiW/Al,O, sample , After hydrogen-hydrogen sulfide treatment the first (2000 - 4000 A) and third (more than 1000% 5) maxima decline about 40% , which is connected with a Vcu, in the radius
decrease . The changes described above in poroua structure of the oxide form
of
NiWfAl,O,
catalyst
*
occurins
after
the
thermal
are due to chemical
and hydrogen sulfide
activation in hydrogen
processes taking place on the pore surface . The species occuring on the surface of pores were X-Ray
,
Diffraction
,
,
UV-Vis , Raman
Thermal
, Magnetic Susceptibility
Spectroscopy
:
Infrared Spectroscopy
Differential
Diffuse Reflectance Spectroscopy measurments
identified
. The following should be pointed out
using numerous methods
Analysis
(Faraday method1
and EPR
[ll .
Upon thermal activation in the atmosphere of oxygen different species are formed on the surface of pores which determine
the
porous structure
and
polymeric
of the sample
species
containing
. The existence
octahedrally
octahedrals
is more
or
similar
to NiWO, , WO,
and
ions
A part of these
. Morever
less deformed
W '+
coordinated
coupled by oxygen bridges W 6+_O_W6+ was found structures
of dimeric
,
formation
of
heteropolytungstates
was indicated . Reduction
with
hydrogen
the
of
above
leads
sample
to
a
transformation of some of the species found on the pore surface into others . Namely , the presence of NiO ant+binuclear spegfes single
with
or
respectively
double
, was confirmed
bridges
.
partial reduction
. On the other hand
and
the
sulfiding
changes complete
of
of
porous
oxide
structure
substitution
of
form ,
1
1
and W, 0, W, 0, [ [ Heteropolytungstates undergo a
oxygen
,
simultaneous
leads to
which
terminal
even
is due
oxide
to
ions
reduction
more a
partial
(in the
existing on the surface of pores) for volumetrically
profound or
species
big sulfide
ions .
CONCLUSION The
qualitative
and
quantitative
changes
induced
in
the
formed species by applying thermal activation in the mixture of hydrogen and hydrogen sulfide affect the textural the examined catalyst are
produced
by
the
properties
of
. In our opinion the most profound changes change
of
terminal
oxide
ions
for
considerably biger sulfide ions . The described changes in the catalyst texture are confirmed by the catalytic
properties
.
It was
concluded
that the sulfide
form of the NiW/Al,O, catalyst is characterized by high activity in the
hydrodenitrogenation
derivatives
[1,9] .
of
carbazole
and
its hydrogenated
456
This
indicates
thermal
that
the
pore
in
the
mixture
activation
sulfide allows ,
carbazole
for a diffusion
tetrahydrocarbazole
system of
produced
in result
hydrogen
and
of relatively and
big molecules
hexahydrocarbazole
catalytically active centres and for a diffusion direction
of
hydrogen to
of the
in the opposite
.
REFERENCES Z.Sarbak
,
Catalysts
Structure (in Polish)
and
Activity
of
,Adam Mickiewicz
Hydrodenitrogenation Press University
Poznari , 1985. R.W.Cranston and F.A.Inkley ,Adv.Catal. ,9 (19571 143. H.L.Ritter and L.E.Drake ,Ind.Eng.Chem.,Anal.Ed.,l7
(1945) 782.
B.C.Lippens and J.H.de boer ,J.Catal. ,4 (1965) 319. J.H.de Boer ,The Structure and Properties of Porous Materials Butterworths ,London ,1985 ,p.68. ,Pure Appl.Chem. ,45 (1976) 71.
6
R.L.Burwell
7
Z.Sarbak ,Surface Technol. ,26 (1985) 331
8
Z.Foltynowicz
and
Z.Sarbak
,Surface
Coatings
Technol.
(1988) 29. 9
Z.Sarbak *React. Kinet.Catal.Lett.
,32 (1986) 435.
.35
Lihat lebih banyak...
Comentarios