Synergism between Ni and W in the Ni-W/C sulfide catalyst in hydrodenitrogenation of pyridine and hydrodesulfurization of thiophene

Article abstract of DOI:10.1135/cccc19990735

Parallel hydrodenitrogenation (HDN) of pyridine and hydrodesulfurization (HDS) of thiophene was studied over active-carbon-supported Ni, W, and Ni-W sulfide catalysts at 2 MPa and at 280 and 320 °C. Synergism between Ni and W was observed both in HDN and HDS reactions: the activity of the Ni-W catalyst was higher than the sum of the activities of the Ni and W catalysts. However, the synergistic increase in activity was much higher in HDS than in HDN. This led to a characteristic shift in the HDN/HDS selectivity, which was strongly shifted to the HDS side over the Ni-W catalysts as compared with the Ni and W catalysts. HDS was faster than HDN over the Ni-W catalyst, the rate of both reactions being about the same over the Ni catalyst and HDN being faster than HDS over the W catalyst. The selectivity of all the catalysts was shifted to the HDN side with decreasing temperature. The data are a new example for generalisation of the rule that the synergism in activity of bimetallic sulfide Co-Mo, Ni-Mo, and Ni-W catalysts is higher in HDS than in hydrogenation and HDN.

Full text of DOI:10.1135/cccc19990735

Ni-W/C Sulfide Catalyst
735
SYNERGISM BETWEEN Ni AND W IN THE Ni-W/ C SULFIDE
CATALYST IN HYDRODENITROGENATION OF PYRIDINE AND
HYDRODESULFURIZATION OF THIOPHENE
1,
Daniela GULKOVÁ and Miroslav ZDRAŽIL *
Institute of Chemical Process Fundamentals, Academy of Sciences of the Czech Republic,
1
Rozvojová 135, 165 02 Prague 6, Czech Republic; e-mail: zdrazil@icpf.cas.cz
Received Novem ber 16, 1998
Accepted Jan uary 16, 1999
Parallel h ydroden itrogen ation (HDN) of pyridin e an d h ydrodesulfurization (HDS) of
th ioph en e was studied over active-carbon -supported Ni, W, an d Ni-W sulfide catalysts at 2 MPa
an d at 280 an d 320 °C. Syn ergism between Ni an d W was observed both in HDN an d HDS
reaction s: th e activity of th e Ni-W catalyst was h igh er th an th e sum of th e activities of th e
Ni an d W catalysts. However, th e syn ergistic in crease in activity was m uch h igh er in HDS
th an in HDN. Th is led to a ch aracteristic sh ift in th e HDN/HDS selectivity, wh ich was
stron gly sh ifted to th e HDS side over th e Ni-W catalysts as com pared with th e Ni an d W cata-
lysts. HDS was faster th an HDN over th e Ni-W catalyst, th e rate of both reaction s bein g
about th e sam e over th e Ni catalyst an d HDN bein g faster th an HDS over th e W catalyst.
Th e selectivity of all th e catalysts was sh ifted to th e HDN side with decreasin g tem perature.
Th e data are a n ew exam ple for gen eralisation of th e rule th at th e syn ergism in activity of
bim etallic sulfide Co-Mo, Ni-Mo, an d Ni-W catalysts is h igh er in HDS th an in h ydrogen a-
tion an d HDN.
Key w ord s: Hydroden itrogen ation ; Hydrodesulfurization ; Ni-W sulfide catalyst; Heteroge-
n eous catalysis; Nickel; Tun gsten ; Hydrogen ation .
Various h ydrotreatm en t processes usin g bim etallic Co-Mo, Ni-Mo, or Ni-W
sulfide catalysts are widely used in th e production of liquid fuels, lubricat-
in g oils an d feeds for petroch em istry. A ph en om en on of key im portan ce in
th e h ydrotreatm en t catalysis is th e syn ergistic effect between th e two m etal
sulfides in bim etallic sulfide catalysts: th e activity of Co-Mo, Ni-Mo, an d
Ni-W catalysts is m uch h igh er th an th e sum of activities of th eir Co, Ni,
Mo, an d W com pon en ts. Data an d th eories of th is syn ergism h ave been
sum m arized in m an y reviews, e.g. in refs1–5
.
Most of th e relevan t literature was devoted to iden tification of surface
sites or ph ases respon sible for syn ergism an d h en ce th e structure of th e syn -
ergistic catalysts is relatively well un derstood n ow. Th e discussion of th is
Collect. Czech. Chem. Commun. (Vol. 64) (1999)

736
Gulková, Zdražil:
structure is outside th e con text of th e presen t con tribution . On th e oth er
h an d, th e m ech an ism of th e syn ergistic effect is m uch less un derstood.
An in terestin g ph en om en on related to th e m ech an ism of syn ergism is
th at th e m agn itude of syn ergism (th e ratio of th e activity of a bim etallic
sulfide system to th e sum of activities of th e correspon din g m on om etallic
sulfide catalysts) is n ot th e sam e for all reaction s in volved in h ydro-
treatm en t.
For in stan ce, th e m agn itude of syn ergism in h ydrogen olysis of C–S bon ds
is always h igh er th an in h ydrogen ation . As a result, th e rate ratio
h ydrogen olysis of C–S bon d/h ydrogen ation is always m uch h igh er over bi-
m etallic syn ergistic system s Co(Ni)-Mo(W) th an over th e correspon din g
m on om etallic Co, Ni, Mo, an d W catalysts. Th is gen eral rule h olds for all
com bin ation s Co(Ni)-Mo(W) as in dicated by n um erous data on th e form a-
tion of various in term ediates in h ydrodesulfurization (HDS) an d on parallel
h ydrogen ation an d HDS (refs6–8).
A secon d exam ple of un equal m agn itude of syn ergism in various reac-
tion s is provided by th e fact th at th e m agn itude of syn ergism seem s to be
system atically h igh er in HDS th an in HDN. Th e result is th at th e rate ratio
HDN/HDS (i.e. th e HDN/HDS selectivity) is m uch lower over bim etallic syn -
ergistic system s th an over th e correspon din g m on om etallic sulfide cata-
lysts. However, th e available literature data are rath er lim ited an d all
con cern Mo-based catalysts on ly. Th e HDN/HDS selectivity in a parallel re-
action s of pyridin e an d th ioph en e with a Ni-Mo/C catalyst was m uch lower
th an th e selectivities of th e correspon din g Ni/C an d Mo/C sam ples9. Th e
HDN/HDS selectivities of Co-Mo/Al2O3, Ni-Mo/Al2O3, Co-Mo/C, an d
Ni-Mo/C catalysts were m uch lower th an th e selectivities of th e correspon d-
in g Mo/Al2O3 an d Mo/C sam ples. However, data for th e correspon din g
Co/Al2O3, Ni/Al2O3, Co/C an d Ni/C catalysts were n ot given 10
.
Th e purpose of th e presen t work is to fin d out wh eth er th e h igh er m agn i-
tude of syn ergism in HDS th an in HDN observed for Mo-based system s is
also exh ibited by th e W-based catalysts. No relevan t literature data on th e
subject (i.e., HDN an d HDS data for all th ree catalysts Ni, W, an d Ni-W un -
der th e sam e con dition s) are available.
Th e above m en tion ed rules on th e relative m agn itude of syn ergism in
various h ydrotreatm en t reaction s h old in depen den tly of th e type of cata-
lyst support6–10. From th is poin t of view, an y support is in prin ciple suitable
for th e purpose of th e presen t work. However, th e use of alum in a, a con -
ven tion al support, is disadvan tageous. Th e Ni/Al2O3 catalyst is kn own to
possess a very low activity as com pared with activities of W/Al2O3 an d
Collect. Czech. Chem. Commun. (Vol. 64) (1999)

Ni-W/C Sulfide Catalyst
737
Ni-W/Al2O3 an d th is m akes a com parison of all th ree catalysts at th e sam e
tem perature rath er difficult. A carbon support was selected in th e presen t
work because th e activity of th e Ni/C catalyst is n ot so differen t from th e
activities of th e W/C an d Ni-W/C sam ples. An an oth er argum en t for usin g
carbon support is th at tun gsten is very difficult to sulfide wh en supported
on alum in a (i.e. ref.11), but n ot on carbon . Th e m odel reaction s were HDS
of th ioph en e an d HDN of pyridin e.
EXPERIMENTAL
Catalysts
Th e active carbon used was a com m ercial product (GA-05, Sloven ské Lučobné závody,
Hnúš/a, Slovakia) with a BET surface area of 1 100 m 2 g–1, a pore volum e 0.75 cm 3 g–1, an d a
m icropore volum e of 0.35 cm 3 g–1. It was crush ed an d sieved to th e particle size fraction
0.16–0.32 m m , purified successively by refluxin g several tim es with water, eth an ol, an d
butan ol, h eatin g with dim eth yl sulfoxide at 100 °C, wash in g several tim es with cold water,
refluxin g several tim es with water an d dried in a vacuum rotary evaporator before im pregn a-
tion .
Th e W/C, Ni/C, an d Ni-W/C catalysts con tain ed 21.5 wt.% WO3, 7 wt.% NiO, an d 21.5
wt.% WO3 an d 3 wt.% NiO (based on th e weigh t of th e catalyst), respectively. Th eir m olar
com position was 0.93 m m ol WO3, 0.93 m m ol NiO, an d 0.93 m m ol WO3 an d 0.4 m m ol NiO
per gram of catalyst, respectively. Th e con ten t of NiO in th e m on om etallic Ni/C was h igh er
th an in th e bim etallic Ni-W/C catalyst for activity reason s: th e con version s ach ievable with
th e 3 wt.% NiO/C catalyst were lim ited by th e catalyst bed volum e available in th e reactor
used. It was establish ed in th e low con version region th at th e activity of th e Ni/C catalyst
was proportion al to th e NiO loadin g. Th e m easurem en t with 7 wt.% NiO/C catalyst allowed
to obtain data in a broader ran ge of con version s. It was assum ed th at th e HDN/HDS selectiv-
ity of th e Ni/C catalyst is essen tially in depen den t of loadin g. (Th is assum ption is supported
by our un publish ed results th at selectivity in th e h ydrotreatm en t over m on om etallic sulfide
catalysts is in sen sitive to loadin g.)
Th e W/C, Ni/C, an d Ni-W/C catalysts were prepared by con ven tion al im pregn ation with
am m on ium paratun gstate, n ickel n itrate, an d successive im pregn ation with am m on ium
paratun gstate an d n ickel n itrate, respectively. Th e catalysts were dried in a rotary vacuum
evaporator at 100 °C an d were n ot calcin ed. Th ey were presulfided in situ as described be-
low.
Apparatus
Data were obtain ed in a tubular flow reactor (stain less steel, i.d. 2 or 4 m m , depen din g on
catalyst ch arge) with a fixed catalyst bed an d with all reactan ts in th e gas ph ase. Catalyst
ch arge, W, was varied in th e ran ge 0.01–1 g; th e feed rates, Fi, of h ydrogen , th ioph en e (TH),
an d pyridin e (PY) were 400, 0.09, an d 0.08 m m ol h –1, respectively. Th e feed was gen erated
in a sim ple evaporator (n ot bubble saturator) wh ere stream of h ydrogen passed over a liquid
m ixture of TH an d PY. Com position at th e in let an d outlet of th e reactor was an alysed usin g
an on lin e gas ch rom atograph (122 °C, glass colum n , 2.5 m , 3 m m i.d., packed with 60/80
Collect. Czech. Chem. Commun. (Vol. 64) (1999)

738
Gulková, Zdražil:
m esh Carbopack B/4% Carbowax 20 M/0.8% KOH, Supelco). Th e products of th e TH reac-
tion were C4 h ydrocarbon s (C4); oth er carbon -con tain in g products were n egligible. Th e
products of th e PY reaction were C5 h ydrocarbon s (C5), piperidin e (PI), an d pen tylam in e
(PA). Th e am oun t of PA was always m uch lower th an th at of PI. Th ese in term ediate satu-
rated am in es (PI an d PA) are togeth er n am ed SA. Pen tylpiperidin e, wh ich was foun d as an
in term ediate durin g th e reaction of pyridin e in th e absen ce of sulfur in th e feed12 was n ot
foun d in our case. Products H2S an d NH3 were n ot an alyzed.
Procedure
Catalysts were presulfided in situ at atm osph eric pressure with a H2S/H2 m ixture (1 : 10) at
400 °C for 2 h . Feed rates Fi were th e sam e in all experim en ts an d space tim e W/F was
ch an ged by variation of th e catalyst ch arge W. Th e run at each space tim e value was per-
form ed with a fresh catalyst ch arge.
Th e com position of th e reaction m ixture was expressed in m ole fraction s, ai: aTH = nTH
n 0TH , aPY= nPY/nP0Y , aSA= nSA/nP0Y , an d aC5 = nC5/nP0Y , wh ere ni0 an d ni are th e in itial an d fin al
n um ber of m oles, respectively. Th e m ole fraction s are related to th e con version s: xTH = 1 –
/
a
TH, xPY = 1 – aPY, xSA = aSA, an d xC5 = aC5, wh ere xTH is th e con version of TH, xPY is th e
overall con version of PY, xSA is th e con version of PY to SA, an d xC5 is th e con version of PY
to C5.
RESULTS AND DISCUSSION
Kinetics
Th e result of th e m easurem en t was th e depen den ce of th e reaction m ixture
com position ai on space tim e W/F (aTH on W/FTH an d aPY, aSA, an d aC5 on
W/FPY) at two tem peratures for each catalyst. Th e data obtain ed at 320 °C
are sh own in Fig. 1. Th e features of th e data obtain ed at 280 °C were sim ilar
to th ose at 320 °C but th e activity of th e catalysts was clearly lower an d it
was n ecessary to use h igh er space tim es W/F to cover th e wh ole ran ge of
m ole fraction s (con version s).
Th e kin etic curves ai = f(W/F) were correlated by sim ple form al kin etics
with th e aim to evaluate con version -in depen den t activity an d selectivity
param eters (th e ch an ge in th e h ydrogen partial pressure in th e course of
th e reaction was n eglected because h ydrogen was fed in large excess). It was
n ot th e purpose of th e presen t work to study th e detailed kin etics of th e re-
action s.
Th e reaction of th ioph en e followed a sim ple pseudo first-order equation :
aTH = exp (–kTH*W/FTH),
(1)
Collect. Czech. Chem. Commun. (Vol. 64) (1999)

Ni-W/C Sulfide Catalyst
739
−1
cat
wh ere kTH is th e rate con stan t (m m ol h –1
g
). Th e reaction of pyridin e was
correlated by a sim plified reaction sch em e of two con secutive pseudo
first-order rate equation s:
PY + n H2 → SA
(2)
(3)
SA + m H2 → C5 + NH3
(for codes PY, SA, an d C5, see Experim en tal). In tegration of th e correspon d-
in g rate equation s gives th e depen den ce of m ole fraction s on space tim e:
aPY = exp (–kPY*W/FPY),
(4)
1.0
1.0
a
a
i
i
0.8
0.8
Ni-W
W
0.6
0.6
0.4
0.2
0.0
0.4
0.2
0.0
0.0
0.2
0.4
0.6
W/F, g
0.8
h mmol
1.0
0
1
2
3
4
5
6
–1
cat
1.0
a
i
0.8
Ni
0.6
FIG.1
0.4
0.2
0.0
Com position of th e reaction m ixture versus
space tim e in th e parallel h ydrode-
n itrogen ation of pyridin e an d h ydro-
desulfurization of th ioph en e over Ni, W ,
an d Ni-W sulfide catalysts at 320 °C: ꢀ, TH
(th ioph en e); ∆, PY (pyridin e; ꢀ, SA (satu-
rated am in es); ꢀ, C5 (C5 h ydrocarbon s)
0
1
2
3
4
5
6
–1
W/F, g
h mmol
cat
Collect. Czech. Chem. Commun. (Vol. 64) (1999)

740
Gulková, Zdražil:
aSA = (kPY/(kC5 – kPY))*((exp (–kPY*W/FPY)) – (exp (–kC5*W/FPY))) ,
(5)
aC5 = 1 – ((kC5/(kC5 – kPY))*exp (–kPY*W/FPY)) +
+ ((kPY/(kC5 – kPY))*exp (–kC5*W/FPY)) ,
(6)
−1
cat
wh ere kPY an d kC5 are th e rate con stan ts (m m ol h –1
steps (2) an d (3), respectively.
g
), correspon din g to
Th e con stan ts kTH, kPY, an d kC5 were obtain ed by n on -lin ear regression of
th e experim en tal data by Eqs (1) an d (4)–(6); th ey are sum m arized in Table I.
Th e curves calculated usin g th ese equation s an d rate con stan ts are sh own
in Fig. 1 for 320 °C. Th e quality of th e fit of data at 280 °C was sim ilar.
Magnitude of Synergism
Th e purpose of th e presen t work was to com pare th e m agn itude of syn er-
gism in HDS an d HDN. A possibility of doin g th at is to use kin etic con -
stan ts from Table I. Th e m agn itude of syn ergism is defin ed as th e ratio of
TABLE I
−1
cat
Kin etic con stan ts (m m ol h –1
g
)
in parallel h ydroden itrogen ation of pyridin e an d
h ydrodesulfurization of th ioph en e
HDS
HDN
Catalyst
kTH
kPY
kC5
AHDN
280 °C
Ni/C
0.1
0.1
3.5
0.4
0.5
1.4
0.3
0.4
2.3
0.2
0.3
1.0
W/C
Ni-W/C
320 °C
Ni/C
0.5
0.5
0.6
0.8
3.0
1.1
1.4
0.5
0.6
3.4
W/C
Ni-W/C
19.1
17.9
Collect. Czech. Chem. Commun. (Vol. 64) (1999)

Ni-W/C Sulfide Catalyst
741
activity of th e bim etallic catalyst to th e sum of activities of th e correspon d-
in g m on om etallic catalysts. Th is ratio equals un ity if n o syn ergism occurs
bein g h igh er th an un ity in th e case of positive syn ergism .
Th e HDS activity is well ch aracterized by th e rate con stan t kTH. Th e m ag-
n itude of syn ergism in HDS, MSHDS, is th us defin ed by th e ratio
MSHDS = kTH,Ni-W/((3/7)*kTH,Ni + kTH,W) ,
(7)
wh ere kTH,j is th e HDS activity for th e catalyst j. Th e con ten t of NiO in th e
Ni-W/C catalyst was lower th an in th e Ni/C catalyst an d th is is wh y th e
value of kTH,Ni was m ultiplied by 3/7 in Eq. (7) (see Experim en tal).
Th e defin ition of th e HDN activity is n ot so straigh tforward. Th e overall
reaction of pyridin e to C5 h ydrocarbon s an d am m on ia con sists of two steps
(Eqs (2) an d (3)) an d n eith er kPY n or kC5 alon e defin es th e activity. Th e
overall con version of PY to HDN products depen ds on both th ese con stan ts
in a rath er com plex m an n er described by Eq. (6). However, th e HDN activ-
ity of a catalyst, AHDN, can be defin ed as th e reciprocal value of W/FPY at
wh ich aC5 = 0.5. Th e dim en sion of AHDN is th e sam e as th at of th e rate con -
−1
cat
stan ts, i.e. m m ol h –1
g
. Th e values of AHDN were obtain ed from th e curves
aC5 = f(W/FPY) calculated usin g th e con stan ts kPY an d kC5, an d are presen ted
in Table I. Th e m agn itude of syn ergism in HDN, MSHDN, is defin ed by th e
ratio
MSHDN = AHDN,Ni-W/((3/7)*AHDN,Ni + AHDN,W) ,
(8)
wh ere AHDN,j is th e HDN activity for th e catalyst j; AHDN,Ni was corrected for
th e con ten t of NiO by m ultiplyin g by 3/7.
Th e m agn itudes of syn ergism were evaluated for HDS an d HDN at tem -
peratures 280 an d 320 °C usin g th e con stan ts in Table I; th e results are sum -
m arized in Table II. It can be seen th at th e m agn itude of syn ergism is m uch
h igh er in HDS th an in HDN.
Synergistic Shift in HDN/HDS Selectivity
An oth er possibility of com parin g th e m agn itude of syn ergism in HDS an d
HDN is a com parison of th e HDN/HDS selectivities of th e catalysts. Th e
selectivities at 280 an d 320 °C are evaluated graph ically in th e con version
Collect. Czech. Chem. Commun. (Vol. 64) (1999)

742
Gulková, Zdražil:
versus con version graph s in Figs 2 an d 3, respectively. Th e curves sh own
were calculated usin g th e con stan ts given in Table I. In th e case of a sim ilar
m agn itude of syn ergism in HDN an d HDS, th e selectivity of th e bim etallic
sulfide Ni-W catalyst sh ould be in term ediate between th e selectivities of th e
Ni an d W sulfides. However, th e data in Figs 2 an d 3 sh ow th at th e selectiv-
ity of th e Ni-W catalyst is stron gly sh ifted to th e HDS side com pared with
th e Ni an d W catalysts. Th is in dicates th at th e m agn itude of syn ergism is
m uch h igh er in HDS th an in HDN.
Th e data of th e presen t work were obtain ed un der m odel con dition s
wh ich differ from th e con dition s of th e in dustrial h ydrotreatm en t m ain ly
in th ree poin ts: th e sim ple low m olecular weigh t m odel com poun ds were
used an d th e feed did n ot con tain an y h ydrocarbon s an d h ydrogen sulfide.
TABLE II
Magn itude of syn ergism a between Ni an d W in th e Ni-W/C sulfide catalyst
Tem perature, °C
HDS
HDN
280
320
28
28
3
4
a
For defin ition s see Eqs (7) an d (8).
100
100
a
b
W
Ni
Ni
W
50
50
Ni-W
Ni-W
0
0
0
50
100
0
50
100
x
, %
x
, %
TH
TH
FIG. 2
Selectivity in th e parallel reaction s of pyridin e an d th ioph en e over Ni, W, an d Ni-W sulfide
catalyst at 280 °C: a overall con version of pyridin e versus h ydrodesulfurization con version ; b
con version of pyridin e to C5 h ydrocarbon s versus h ydrodesulfurization con version
Collect. Czech. Chem. Commun. (Vol. 64) (1999)

Ni-W/C Sulfide Catalyst
743
Th e m ost im portan t of th ese poin ts is probably th e very low partial pres-
sure of h ydrogen sulfide in our experim en ts (it was produced on ly by HDS
of TH). It is well kn own th at th e effect of h ydrogen sulfide on various
h ydrotreatm en t reaction s is differen t. Hydrodesulfurization is m ostly
stron gly in h ibited by h ydrogen sulfide. Hydroden itrogen ation is in fluen ced
less, th e effect of h ydrogen sulfide bein g m ildly positive or n egative, de-
pen din g on th e partial pressure of h ydrogen sulfide. It is expected th at th e
addition of h ydrogen sulfide to th e feed sh ould sh ift th e selectivity
HDN/HDS to th e HDN side un der our reaction con dition s. However, it is
believed th at th is effect sh ould be rath er sim ilar for all th e catalysts studied
an d th at th e differen t m agn itude of syn ergism in HDN an d HDS (Table II)
an d th e syn ergistic sh ift in selectivity (Figs 2 an d 3) are qualitatively in de-
pen den t of th e partial pressure of h ydrogen sulfide.
Importance of Results for Theory of Hydrotreatment
Th e results of th e presen t work are a n ew con tribution to th e gen eralization
of th e ph en om en on of a differen t m agn itude of syn ergism in various
h ydrotreatm en t reaction s (com parison of syn ergism in HDN an d HDS over
Ni, W, an d Ni-W catalysts un der th e sam e con dition s h as n ot been reported
in literature). However, in terpretation of th is ph en om en on rem ain s diffi-
cult. Th e differen ce of HDS from h ydrogen ation an d HDN is th at th e reac-
100
100
b
a
W
W
Ni
50
50
Ni
Ni-W
Ni-W
0
0
0
50
100
0
50
100
x
, %
x
, %
TH
TH
FIG. 3
Selectivity in th e parallel reaction s of pyridin e an d th ioph en e over Ni, W, an d Ni-W sulfide
catalyst at 320 °C: a overall con version of pyridin e versus h ydrodesulfurization con version ; b
con version of pyridin e to C5 h ydrocarbon s versus h ydrodesulfurization con version
Collect. Czech. Chem. Commun. (Vol. 64) (1999)

744
Gulková, Zdražil:
tion product, H2S, in teracts deeply with th e catalyst. Sulfur is con tain ed n ot
on ly in th e HDS reactan ts an d products but also in th e sulfide catalyst. Th e
outstan din g position of th e C–S bon d h ydrogen olysis am on g oth er
h ydrotreatm en t reaction s as regards th e m agn itude of syn ergism seem s to
in dicate th at th e m ech an ism of syn ergism m igh t be associated with th e
way in wh ich sulfur species are tran sferred th rough th e catalytic cycle. A
m ore detailed discussion of th is possibility is beyon d th e scope of experi-
m en tal data obtain ed in th e presen t paper.
However, th e ch aracteristic sh ift in h ydrogen ation /HDS or HDN/HDS
selectivities to th e HDS side accom pan yin g syn ergism in activity m igh t be
useful for th e th eory of syn ergism also from an oth er poin t of view. For th e
catalyst of a given com position , th e selectivity is an easily m easurable, use-
ful in dication of th e exten t of con tact or degree of co-operation of th e two
tran sition m etal sulfides in th e syn ergistic pair. Th e selectivity of th e cata-
lyst with th e best con tact or co-operation of both com pon en ts is in an ex-
trem e position at th e HDS side. An y sligh t dam age of th is optim um con tact
or co-operation (at th e sam e overall com position ) is in dicated by a sh ift of
h ydrogen ation /HDS or HDN/HDS selectivities to th e h ydrogen ation or th e
HDN side, respectively. Catalytic activity of in dustrial Co-Mo, Ni-Mo, an d
Ni-W sulfide catalysts decreases durin g tim e on stream an d durin g oxida-
tive regen eration (for a recen t discussion an d referen ces, see e.g. refs4,5).
Am on g various causes of th e deactivation , selectivity is expected to be
m uch m ore sen sitive to segregation of active com pon en ts (decreasin g th e
degree of utilization of th e com pon en ts for syn ergism ) or to decom position
of active syn ergistic ph ases th an to surface cokin g an d support texture
sin terin g. In an con ven tion al approach , segregation of active com pon en ts
or decom position of active syn ergistic ph ases is studied by various spectro-
scopic tech n iques4,5. It is suggested th at m easurem en ts of HDS/h ydrogen a-
tion or HDN/HDS selectivities, such as sh own in Figs 2 an d 3, m igh t be a
con ven ien t altern ative way of gettin g addition al in form ation .
Remarks on the HDN/HDS Selectivity
It is worth m en tion in g two poin ts con cern in g th e HDN/HDS selectivity n ot
directly associated with th e m ain purpose of th e presen t work.
a) Over con ven tion al in dustrial Co-Mo, Ni-Mo, an d Ni-W catalysts, HDN
is always m ore difficult th an HDS (refs13–15). It is som etim es con cluded th at
th is relative reactivity is gen erally valid for all sulfide catalysts. However, it
h as recen tly been sh own th at HDN is faster th an HDS over som e un con ven -
tion al sulfide catalysts: HDN of pyridin e an d HDS of th ioph en e over Pt/C
Collect. Czech. Chem. Commun. (Vol. 64) (1999)

Ni-W/C Sulfide Catalyst
745
an d Ir/C sulfide catalysts16 an d Ni/C, an d Mo/C sulfide catalysts9,
h ydrotreatm en t of ligh t catalytic cycle oil over un supported Fe-W an d
Fe-Mo sulfides17, an d in h ydrotreatm en t of coal-derived n aph th a over
Ir/Al2O3 an d Pt/Al2O3 sulfide catalysts18. It can be seen in Figs 2 an d 3 th at
also W/C catalysts ran k am on g catalysts wh ere HDN can be faster th an
HDS.
b) It seem s th at th e activation en ergy of HDN is always lower th an th at of
HDS. Th is leads to a decrease in th e HDN/HDS selectivity with in creasin g
tem perature. Th is was observed for Ni/C (ref.9), Mo/C an d Ni-Mo/Al2O3
(ref.19), Ni-MoAl2O3 an d un supported Fe-Mo (ref.17), an d Co-Mo/Al2O3-SiO2
(ref.20) sulfide catalysts an d over a Mo2N catalyst21. A com parison of
selectivities sh own in Figs 2 an d 3 for 280 an d 320 °C, respectively, in di-
cates th at th is rule also h olds for W/C an d Ni-W/C sulfide catalysts (th e
data for th e Ni/C catalyst in Figs 2 an d 3 con firm our previous results with
an oth er Ni/C catalyst9).
CONCLUSIONS
Th e m agn itude of th e syn ergistic effect resultin g from a com bin ation of Ni
an d W in a Ni-W sulfide catalyst is h igh er in th e HDS of th ioph en e th an in
th e HDN of pyridin e. As a result, th e HDN/HDS selectivity of th e Ni-catalyst
is stron gly sh ifted to th e HDS side com pared with th e Ni an d W catalysts.
HDN of pyridin e is m ore difficult th an HDS of th ioph en e over th e Ni-W
catalysts but it is easier th an HDS over th e W catalyst. Th e HDN/HDS selec-
tivity is sh ifted to th e HDN side by decreasin g tem perature. It is suggested
th at th e HDN/HDS selectivity m easurem en ts can be useful to in dicate if a
decom position of active syn ergistic structures h as occurred durin g th e cata-
lyst life an d regen eration ; decom position of th e syn ergistic structure sh ould
be accom pan ied by a sh ift of th e selectivity to th e HDN side.
Financial support by the Grant Agency of the Czech Republic (grant No. 104/96/0573) is gratefully
acknowledged. The authors thank the Chemické Lučobné Závody Co., Slovak Republic, for activated
carbon.
REFERENCES
1. Prins R., de Beer V. H. J., Somorjai G. A.: Catal. Rev.-Sci. Eng. 1989, 31, 1.
2. Delmon B.: Catal. Lett. 1993, 22, 1.
3. Vasudevan P. T., Fierro J. L. G.: Catal. Rev.-Sci. Eng. 1996, 38, 161.
4. Topsøe H., Clausen B. S., Massoth F. E.: Hydrotreating Catalysis, Science and Technology.
Springer, Berlin 1996.
Collect. Czech. Chem. Commun. (Vol. 64) (1999)

746
Gulková, Zdražil:
5. Eijsbouts S.: Appl. Catal., A 1997, 158, 53.
6. Zdražil M., Kraus M.: Stud. Surf. Sci. Catal. 1986, 27, 257.
7. Zdražil M.: Bull. Soc. Chim. Belg. 1991, 100, 769.
8. Zdražil M. in: Transition Metal Sulphides, Chemistry and Catalysis (T. Weber, R. Prins and
R. A. van Santen, Eds), p. 273. Kluwer, Dordrecht 1998.
9. Drahorádová A., Zdražil M.: Collect. Czech. Chem. Commun. 1992, 57, 2515.
10. Vít Z.: Fuel 1993, 72, 105.
11. Mangnus P. B., Bos A., Moulijn J. A.: J. Catal. 1994, 146, 437.
12. Sonnemans J., Neyens W. J., Mars P.: J. Catal. 1974, 34, 230.
13. Katzer J. R., Sivasubramanian R.: Catal. Rev.-Sci. Eng. 1979, 20, 155.
14. Ho T. C.: Catal. Rev.-Sci. Eng. 1988, 30, 155.
15. Lewandowski M., Sarbak Z.: Appl. Catal., A 1997, 156, 181.
16. Vít Z., Zdražil M.: J. Catal. 1989, 119, 1.
17. Ho T. C., Jacobson A. J., Chianelli R. R., Lund C. R. F.: J. Catal. 1992, 138, 351.
18. Liaw S. J., Raje A., Lin R., Davis B. H. in: Hydrotreating Technology for Pollution Control.
Catalysts, Catalysis and Processes (M. L. Occelli and R. Chianelli, Eds), p. 197. Dekker,
New York 1996.
19. Vít Z.: Catal. Lett. 1992, 13, 131.
20. Ho T. C.: Ind. Eng. Chem. Res. 1993, 32, 1568.
21. Liaw S. J., Raje A., Bi X. X., Eklund P. C., Graham U. M., Davis B. H.: Energy Fuels 1995,
9, 921.
Collect. Czech. Chem. Commun. (Vol. 64) (1999)