FUEL

PROCESSING
TECHNOLOGY
ELSEVIER

Fuel Processing Technology 49 ( 1996) 137- 155

Activity of noble metal-promoted hydroprocessing

catalysts for pyridine HDN and naphthalene
hydrogenation
H.S. Joo *, James A. Guin
Auburn Uniuersity, Chemical Engineering Department, Auburn, AL 36849, USA

Received 10 October 1995; accepted 17 April 1996

Abstract
An important step in the upgrading of coal and coal-waste coprocessing liquids is the removal
of heteroatoms, particularly nitrogen, which tend to poison the downstream cracking operations
required to lower boiling points to the transportation fuels range. This requirement has motivated
research to improve the activity of current hydroprocessing catalysts. In this regard we have
examined the effects of Pt, Ru, and Ir promotion on three commercial Al,O,-supported catalysts
(NiMo, CoMo, and NiW) using the model compounds, pyridine and naphthalene, to investigate
hydrodenitrogenation and ring hydrogenation reactions, respectively.
For pyridine hydrodenitrogenation (HDN), the activity of the sulfided catalysts was better than
that of the oxide forms. For sulfided forms of NiMo, all noble metals tested improved the HDN
activity. The activity of CoMo was improved by Pt and Ir while NiW was not improved by any
noble metals promotion. For ring hydrogenation (HYD) of naphthalene, the P&promoted catalysts
were more active than the original catalysts in the oxide forms, while the activities of both
catalysts were similar after sulfidation. However for ring HYD in the presence of N, the
Pt-promoted catalysts exhibited a distinct advantage.
Keywords: Naphthalene hydrogenation; Noble metal-promoted bydroprocessing catalysts; Pyridine hydrodeni-

trogenation

1. Introduction
Solid fossil fuels such as coal, which are relatively abundant in comparison with
petroleum, will become major fuel sources in the future [Il. A variety of processes for

??

Corresponding author.

0378-3820/96/$15.00
Copyright 0 1996 Elsevier Science B.V. All rights reserved.
PIf SO378-3820(96)01037-5

138

H.S. Joe, J.A. Guin / Fuel Processing Technology 49 (1996) 137-155

producing coal-derived transportation fuels have been developed in past years. There is
also a great deal of current interest in the coprocessing of coal with waste plastics
and/or rubber to produce transportation fuels [2-51. The primary liquid products from
these processes usually contain relatively high quantities of heteroatoms, especially
nitrogen, oxygen, and sulfur. To produce clean transportation fuels, the first stage liquids
must be upgraded by eliminating the heteroatoms through hydrodenitrogenation (HDN),
hydrodesulfurization (HDS), and hydrodeoxygenation (HDO) reactions. Nitrogen compounds are especially detrimental because they poison acid catalysts used to effect
molecular weight reduction via hydrocracking [6]. Among the heteroatom removal
reactions, C-N bond scission in heterocyclic aromatic compounds is more difficult than
C-S and C-O scission [7]. Ring hydrogenation (HYD) is also necessary as a preliminary step in N removal, as well as in hydrocracking polyaromatics to reduce molecular
weight ranges and lower boiling points.
Bimetallic catalysts such as NiMo, CoMo, and NiW supported on Al,O, are
currently used in commercial hydrotreating processes. However, when coal liquids are
upgraded, it is difficult to obtain adequate heteroatom removal with these conventional
catalysts [81. For this reason, more active hydrotreating catalysts need to be developed.
One pathway toward the development of more active catalysts is the consideration of
polymetallic catalysts. For example, Pt-promoted NiMo/Al,O,
catalysts were used to
upgrade a coal-derived middle distillate and found to be better than a commercial
NiMo/Al,O, catalyst in HDN activity and rate of deactivation [9]. Similarly, studies of
the effects of Ru promotion on CoMo/Al,O,
catalyst showed that the Ru-promoted
catalyst exhibited improved quinoline HDN activity [ 10,111, but interestingly had no
effects on conversion of residual oil and HDS activity [12]. Ir was found to have good
hydrotreating (HDN, HDS, and HDO) activities in model compound (quinoline, thiophene, and diphenylether) reactions with Al,O, [13] and carbon 114-171 supported
transition metal catalysts and bulk transition metal sulfides [18]. Our work extends these
various previous investigations by comparing in a single unified study the effects of
noble metal promotion on all three types of commercial hydrotreating catalysts. Specifically, the promotion of three commercial bimetallic catalysts through the addition of
three noble metals, namely, Pt, Ru, and Ir was investigated. Most of the attention was
focused on Pt based on the results of exploratory experiments. In addition, a detailed
study of effects of catalyst sulfidation was made by comparing the performance of the
oxide and sulfide forms of the catalyst. For reasons detailed in the following, naphthalene and pyridine were chosen as model compounds for our study of catalyst HYD and
HDN activity.
1.1. Pyridine HDN

Many N-containing compounds found in petroleum and synthetic liquids are heterocyclic aromatic forms which are resistant to HDN. The literature on the HDN mechanisms of these compounds has been summarized [7,19,20]. For our work, pyridine was
chosen as an HDN model compound due to the fact that pyridine derived compounds are
the simplest N-heterocycles found in petroleum and coal liquids and consequently, the

H.S. Joo. J.A. Guin / Fuel Processing Technology 49 (19%) 137-1.5.5

Pyridine

Pipetidine

Pentylamine

139

Pentane + etc.

Q+/vv
Pentyipiperidine

O+Q
Pentyipiperidine
Fig.

1.Reaction pathways for pyridine HDN.

pyridine HDN reaction network involves relatively few intermediates as compared to

polycyclic N compounds.
Several previous investigations of pyridine HDN have been performed. The kinetics
of pyridine HDN have been studied with both sulfided [20-261 and reduced catalysts
[27]. As a result of these studies, pyridine HDN has been found to occur through three
reactions as depicted in Fig. 1. In the first reaction, pyridine is initially saturated to form
piperidine by first-order kinetics followed by the formation of pentylamine through
piperidine hydrogenolysis [22]. Mcllvried [23] reported that the piperidine hydrogenolysis step is slower than pyridine HYD with NiMo/Al,O,. Subsequent nitrogen removal
from pentylamine results in pentane and ammonia, as well as some minor isomerized
and lower boiling products. This reaction rate is faster than pyridine HYD and piperidine
hydrogenolysis [23]. In the second and third reactions, as shown in Fig. 1, pentylpiperidine and ammonia are formed by disproportionation of piperidine and pentylamine
molecules (alkyl-transfer reaction) and/or two piperidine molecules. In one study, it
was noted that pentylpiperidine did not react further due to its relative stability [28]. In
our experiments, di-n-pentylamine was also found to a minor extent. More than ten
products have been identified from the pyridine reaction with CoMo/Y zeolite hydrocracking catalyst [281. With regard to the effects of S in pyridine HDN, it has been
reported that thiophene enhanced the overall conversion of pyridine and piperidine [24]
and that H,S promoted piperidine hydrogenolysis but not pyridine HYD [25]. We note
here that the primary purpose of our work was to compare the noble metal-promoted and
unpromoted versions of the hydroprocessing catalysts, rather than a determination of the
detailed mechanism of pyridine HDN.

140

H.S. Joo. J.A. Guin / Fuel Processing Technology 49 (19%) 137-155

N@hthalMle

Tetralin

Decalins

Fig. 2. Reaction pathways for naphthalenehydrogenation.

As noted previously, naphthalene was selected as a model compound for hydrogenation activity measurements in our work. Naphthalene HYD is sequential to form tetralin
and cis and trans decalin as shown in Fig. 2.

2. Experimental
2.1. Materials
Four commercial Al,O,-supported catalysts (Criterion 424 NiMo, Cyanamid 1442B
CoMo, Crosfield 550s NiW, and Harshaw Al-3391, a pure alumina), as shown in Table
1, were used in this study. Chemicals used as received include pyridine (Aldrich, 99%)
and naphthalene (Fisher, purified) as model compounds, hexadecane (Fisher, certified)
as a solvent and H,PtCl, (Aldrich, 8%) RuCl, (Strem, anhydrous), Ru,(CO),, (Strem,
99%) and H,IrCl, (Aldrich, 4%) as catalyst precursors.
2.2. Catalyst preparation
Catalysts were crushed and sieved to between 100 and 200 mesh for reactions. The
noble metals were impregnated on the catalysts by an incipient wetness technique.
Catalysts were dried in air overnight and calcined at 450 C for 3 h. An aqueous solution
of 37% HCl and cyclohexane were used to dissolve RuCl, and Ru,(C0)i2, respectively.
These two different Ru precursors are denoted as Ru(A) and Ru(B), respectively.
2.3. Reaction procedure
All reactions were performed in 20 cm3 316 ss tubing bomb microreactors (TBMR)
which were agitated in a fluidized sand bath as shown in Fig. 3 and replicated at least
twice. Catalysts were presulfided at 300 C for 2 h using dimethyl disultide (DMDS) in
a separate TBMR charged with 1000 psig cold hydrogen pressure. The quantity of
sulfur added for each catalyst was based upon its metals content using 50% excess of the
stoichiometric amount needed for bulk sulfidation (MoS,, Ni,S,, Co,!&, and WS,). All
catalysts in this paper are sulfide forms unless specifically noted as oxide forms. In these

141

H.S. Joo, J.A. Guin / Fuel Processing Technology 49 (1996) 137-15s

Table 1
Properties of reference catalysts
Catalyst

Compositions

Surface area cm* g- 1

Pore vohmle
(ml g-1

Criterion 424

MOO,. 19.5%; NiO, 4%;

P,O,, 8%; AI,O,, balance
MOO,, 13.2%; COO, 3.5%;
SO,, 0.5%; Al,O,, balance
WO,, 18-228; NiO, 4-78;
Al,O,, balance
Al,O,, 99.9%

155

0.47

300

0.82

230

0.5

230

0.69

Cyanamid HDS 1442B

Crosiield Sphexicat 550s
Harshaw Al-3991

studies, a reactant solution (6 g) containing 2 wt.% pyridine and (or) 2 wt.% naphthalene
in hexadecane was used with 0.1 g catalyst loading at 1000 psig cold hydrogen pressure.
In reactions with sulfided catalysts, an additional 1.5 wt.% DMDS was added as a sulfur
source to the TBMR to maintain catalyst sulfidation.

Tbcrmocarple 1

1
CompressedAir

Fig. 3. Schematic diagram of tubing bomb microreactor apparatus.

H.S. Joo, J.A. Guin / Fuel Processing Technology 49 (1996) 137-155

142

2.4. Analysis

Analysis was performed on a temperature-programmed Varian 3200 GC with 30

m X 0.32 mm X 1.0 p_rnRestek Stabilwax-DB capillary column for products of pyridine
reactions and 30 m X 0.32 mm X 0.25 pm J&W DB-5 capillary column for products of
naphthalene reactions. GC-MS and standard additions were performed to identify the
products. Percentage HDN and HYD were calculated based on the product distribution
from the GC analysis. Typical average standard deviations for %HDN and %HYD
measurements from replicate experiments were 3.7% and 2.0% on a scale of O-100%,
respectively.

3. Results and discussion

3.1. Naphthalene hydrogenation

To simplify the presentation of hydrogenation activity, an extent of HYD, A, was

defined according to Eq. (1):
A,=

2M,+5M,
(1)

5(M,+MT+Mo)

where M,, M,, M, are moles of naphthalene, tetralin, and decalin present after
reaction, respectively [29]. The sum in the denominator of Eq. (1) was essentially equal
(within about 2%) to the initial moles of naphthalene charged confirming the conserva-

Oxide

? ?Unpromoted

Sulfide

Pt Pmnoted

Fig. 4. Extent of naphthalene hydrogenation at 310 C for 30 min.

H.S. Joo. J.A. Guin/Fuel

Processing Technology 49 (19%) 137-1.55

143

tion of moles of two-ring aromatic species as shown in Fig. 2. According to this

definition,
A, corresponds
to the moles of H, actually reacting divided by the
maximum H, usage possible (all naphthalene converted to decalins). Thus, A, ranges
from 0 to 1 as the products vary from totally aromatic (no HYD) to completely saturated
(100% HYD to decalins).
Fig. 4 shows the effect of Pt promotion and sullidation on ring HYD of naphthalene
in terms of A, defined by Eq. (1). Oxide and sulfide catalyst forms are shown on the
left and right sides of the figure, respectively.
As a point of reference, the A, of
naphthalene in the absence of any catalyst at 350 C for 30 min was measured to be
0.00, i.e. no reaction. As shown on the left side of Fig. 4, in the original oxide forms,
which could be partially Hz-reduced
under reaction conditions,
the two Ni-based
commercial
catalysts evidence much greater HYD activity than the Co-promoted
catalyst, in keeping with the known good HYD activity of Ni metal. Likewise, with the
oxide catalysts, the extent of hydrogenation (A,) for Pt-promoted catalysts was 100%
and much superior to those of unpromoted ones. It is especially notable that Au for
CoMo was increased from 0 to 1 by Pt. This high HYD activity is in keeping with the

Time (min)

04
0.20

Unpromoted

pt Promoted

Time (min)

Fig. 5. Extent of naphthalene hydrogenation in a mixture of pyridine and naphthalene: (a) 350 C, (b) 380 C.

144

H.S. Joe, J.A. Guin / Fuel Processing Technology 49 (1996) 137-155

known excellent HYD activity of Hz-reduced noble metals such as Ft and suggests that
the impregnated Pt has adequate dispersion on the catalysts. As expected, this HYD
activity is, however, significantly poisoned by sulfidation as indicated by the results for
the sulfide forms on the right side of Fig. 4; however, it is worth noting a recent paper
which shows that the noble metals themselves do retain some hydrogenation activity
even in the presence of sulfur [30]. Following sulfidation, both Pt-promoted and original
catalysts showed essentially the same activity with NiMo > CoMo > NiW. For all
catalysts, the hydrogenation activities were decreased by sulfidation except for CoMo,
showing that generally the hydrogenation sites of the catalysts are poisoned by sulfur.
Because of the effect of sulfur on HYD activity it would appear that little benefit would
accrue from Pt promotion regarding HYD of polyaromatics, however a somewhat
expanded perspective can be gained by referring to Fig. 5 which shows naphthalene
HYD in the presence of added N in the form of pyridine at two temperatures. In
comparison with the sulfide catalysts of Fig. 4, the predominant effect in Fig. 5 is the N
inhibition of the HYD reaction, due to competitive adsorption effects, as evidenced by
the much lower overall A, values. However a second point apparent from Fig. 5, is that
in every case, the Pt-promoted catalyst has a higher A, value as compared to the
corresponding unpromoted catalyst. Taken together, Figs. 4 and 5 suggest that, in the
presence of N poisons the Pt-promoted catalysts can provide enhanced HYD of
polyaromatics, as compared to the original catalysts. Such a behavior could provide a
significant benefit to downstream hydrocracking operations in that polyaromatic saturation is required prior to cracking of the ring. Of course, coal liquids contain many
additional N compounds and their behavior may be somewhat different from the simple
model compound system used here.
3.2. Pyridine HDN
With the catalysts used in this study, analysis of the liquid products revealed only
five compounds, namely, pyridine, piperidine, pentylamine, pentylpiperidine, and di-npentylamine as the major N-containing compounds (more than 0.01%). The major
terminal product found was n-per&me, although minor amounts of additional lower
boiling products were also observed at severe conditions. From the liquid-phase product
analysis, two measures of HDN activity were derived. The first is based upon the wt.%
N in the product liquid as calculated from amounts of the five N-containing products
noted above. Thus, if Nr and NP represent the wt.% nitrogen in the feed and product
mixture, respectively:
%--

%HDN = -

Np

N,
In spite of the apparent simplicity of the above procedure as represented by Eq. (2),
there are some difficulties in application of this method. In batch reactions, such as
performed in this study, some N-removal from the liquid phase will occur due to
irreversible preferential adsorption of N compounds on the catalyst surface, rather than
catalytic HDN reaction per se. To examine the extent of N adsorption, a mixture of
pyridine, piperidine, and pentylamine in hexadecane (6 g solution and 2 wt.% each) and

HS. Joo, J.A. Guin / Fuel Processing Technology 49 (1996)

145

137-155

Table 2
Product distribution with Al,O, and Pt-, Ru-, and k-loaded AI,O, at 350 C for 20 min
Catalyst

None
A203

wt.%
Pyridine

Piperidine

Pentylamine

Pentylpiperidine

n-Pentane

1.89
1.69

0.00
0.00
0.33
0.02
0.21

0.00
0.00
0.04
0.02
0.03

0.00
0.00
0.03
0.01
0.02

0.00
0.00
0.00
0.00
0.00

WA1203

1.32

Ru/Al,O,
Ir/Al,O,

1.62
1.42

0.1 g of sultided NiMo catalyst were charged to a TBMR and agitated at room
temperature for 20 min. The removal of pyridine, piperidine, and pentylamine from the
liquid phase by room-temperature adsorption on the catalyst in this experiment were 9,
14, and 22%, respectively. Since the main purpose of this study was to compare the
noble metal-promoted catalysts with the original unpromoted catalysts, it was of interest
to determine the effect of the promotion procedure on the adsorption of N compounds.
In this regard, a pure alumina (Table 1) was impregnated with three noble metals and
used in pyridine HDN experiments. Table 2 shows the liquid product distribution from
these experiments using 0.1 g of catalyst as well as a blank run in which the HDN of
pyridine without catalyst was examined at 350 C for 20 min. For the blank run, no
products were observed although a 6% pyridine loss was found. In the case of pure
A&O,, no products are formed, however, there is a significant adsorption of pyridine on
the Al,O, as the product concentration of 1.69% is significantly less than the feed
concentration of 2%. In the case of the impregnated aluminas, some reaction products
are formed, however in all four cases the total N-containing species sum to 1.69 f 0.04%.
These four experiments thus suggest that the total adsorption of N species is approximately the same, regardless of the noble metal promotion, even though the product
distribution may be different. Nonetheless, because of the possible contribution of
adsorption, it is probably not possible to attach significance to small differences in
product distribution in the pyridine experiments and for this reason attention will only be
focused on the major differences, i.e. the values of pyridine HDN determined in this
work are considered as relative indicators assuming that adsorption is about the same in
comparable (promoted vs. unpromoted) experiments. One additional point to note from
Table 2 is confirmation that the sulfided noble metals themselves do have catalytic
activity, most notably in the HYD of pyridine to piperidine; however, no hydrogenolysis
activity to further products, e.g. pentane, is observed.
The second indicator used as a measure of catalyst activity in the pyridine experiments was pentane production. Experiments using the NiMo catalyst showed that
pentane did not adsorb preferentially on the catalysts surface and that it did not react to
form other products. Based on its appearance as a major terminal HDN product (Fig. l),
the pentane concentration in the liquid product thus provides good indicator of overall
HDN catalytic activity. Generally, reaction product mixtures with a higher %HDN
calculated from Eq. (1) also had higher wt.% pentane concentrations, showing the two
measures of catalyst activity to be consistent.

146

H.S. Joe, J.A. Guin / Fuel Processing Technology 49 (1996) 137-155

? ?NiMo

gjj

COMO

? ?NM

Fig. 6. Effect of Pt loadings on pyridine HDN at 350 C for 30 min.

Fig. 6 shows the effect of Pt loading on HDN activity in the range of O-2 wt% Pt.
For the three original unpromoted catalysts shown as 0% Pt, the ranking of pyridine
HDN activity is NiMo > CoMo > NiW. The PtNiMo and PtCoMo catalysts are more
active for HDN than the original catalysts, while NiW showed no significant benefit due
to Pt promotion. The lack of improvement with the NiW catalyst, contrasted with the
improvement for the other two catalysts, suggests that Pt addition is not simply an
additive effect, but that there is some interaction between the added Pt and the original
catalysts. Had the effect been simply additive, one would expect about the same degree
of improvement with each catalyst. An additional point from Fig. 6 is that any
improvements in HDN activity from Pt promotion are likely to be realizable with small
amounts, ca. < l%, of Pt. Based on the results in Fig. 6, further experiments as
described below were conducted using 0.8 wt.% Pt loading. Ru(A) and Ir promotion
effects were also examined in parallel experiments in the same ranges of loading. The
Ru(A)- and Ir-promoted NiMo catalysts exhibited highest activities around 1.0 wt.%.
Therefore, in subsequent experiments, 1.0 wt.% Ru- and Ir-loaded catalysts were
examined.
Fig. 7 compares the pyridine HDN activities of the oxide and the sulfided original
and promoted catalysts. The HDN activity ranking of the unpromoted sulfided catalysts
is NiMo > CoMo > NiW, in agreement with the data shown in Fig. 6 at 30 min. Except
for the NiW catalyst, all sulfided noble metal-promoted catalysts showed higher HDN
activity than the original unpromoted catalysts. In contrast to the HYD activity shown in
Fig. 4, all catalysts showed higher HDN activity in their sulfide forms except for Ru(A
and B)NiMo. This is in general agreement with known enhancement effects of sulfidation on HDN [6]. Sulfided catalysts are known to have acidic functionalities which

H.S. Joo, J.A. Guin / Fuel Processing Technology 49 (1996) 137-155

? ?oxide
Fig. 7. Pyridine HDN with oxide and sulfided

147

? ?sulfide

original and promotedcatalysts at 350 C for 20min(*, NiMo

catalysts).

enhance hydrogenolysis,
as opposed to hydrogenation reactions. Both Ru-promoted
NiMo catalysts, in the oxide form, also exhibited an unusual activity for hydrocracking,
as well as being very active for pyridine HDN (100% HDN activity); however, this
activity was reduced by sulfur. Ir promotion was very effective in the sulfide form but
not in the oxide form. In general, the HDN activities were improved by the noble metal
promotion.
Figs. 8 and 9 show the effect of Pt promotion on pyridine HDN liquid product
distribution with oxide and sulfide catalysts, respectively. For the oxide catalysts (Fig.
8), HYD of pyridine to piperidine was much greater with the PtCoMo and PtNiW. For
example, comparing the oxide NiW and PtNiW shows that the concentration of pyridine
with PtNiW (0.04) was much lower than that of NiW (1.3 1) as a result of greater HYD
(pyridine-to-piperidine) with the Pt-promoted catalyst. A similar result is obtained with
CoMo. These effects are in agreement with the increased ring HYD activity of the
promoted oxide catalysts as observed earlier in Fig. 4. For the NiMo oxide, the result of
increased pyridine to piperidine HYD is not as apparent because of the increased
subsequent conversion of the piperidine to other products, i.e. the NiMo catalyst is
generally more active for HDN than the CoMo or NiW catalysts. For sullided catalysts
(Fig. 91, a similar trend of higher pyridine HYD and HDN for the Pt-promoted catalysts
compared to the unpromoted forms was observed, except for the NiW catalyst. This
trend is most apparent from the increased pentane concentration for the Pt-promoted
NiMo and CoMo catalysts. There were no major differences in product distributions
between NiW and PtNiW in the sulfide catalysts showing that Pt promotion was not
effective in increasing the activity of the NiW catalyst.

148

H.S. Joe, J.A. Guin/ Fuel Processing Technology 49 (I 9%) 137-155

Fig. 8. Effect of Pt promotion on pyridine HDN product distributions with oxide catalysts at 350 C for 20
min.

The effect of sulfidation with Pt-promoted catalysts is shown in Fig. 10 where S

denotes the sulfided catalysts. The sulfided catalysts were generally less active for
pyridine HYD but more active for pyridine HDN than their oxide counterparts, as

Fig. 9. Effect of Pt promotion on pyridine HDN product distributors with sulfided catalysts at 350 C for 20
min.

H.S. Joo, J.A. Guin / Fuel Processing Technology 49 (19%) 137-155

149

Fig. 10. Effect of sultidation on pyridine HDN product distribution with Pt-promoted catalysts at 350 C for 20
min.

evidenced by higher pyridine concentration and the appearance of substantial quantities

of the terminal product pentane in the product distributions. The reaction products with
high n-pentane concentrations also contained several unidentified low boiling products.
Considering the product distributions of Fig. 10, together with the first reaction pathway
in Fig. 1, it is apparent that the effect of sulfidation is to enhance hydrogenolysis of
piperidine as opposed to the hydrogenation of pyridine. Since this hydrogenolysis step is
the slow (rate determining) step in the overall pyridine HDN process [23], the sulfided
catalysts have significantly greater HDN activity and produce much larger amounts of
terminal product pentane. All three types of catalysts evidence this same trend, although
the NiW catalyst is less pronounced. This trend is more apparent in Fig. 11, where
selectivities for the five major products expressed as their mole percentages are
compared from experiments with oxide and sulfided catalysts, at the same level of
pyridine conversion, obtained by systematically varying the reaction time. These data
show that the concentrations of piperidine and pentylpiperidine were dominant with
oxide catalysts while the sulfided catalysts had much higher n-pentane and much lower
pentylpiperidine than the oxide forms. These results indicate slower C-N scission
through the ring opening of piperidine and more selective disproportionation reactions
(second and third reactions in Fig. 1) with the oxide forms as compared to the sulfide
forms. This observation agrees with a previous study 1251in which pentylpiperidine was
observed to only a minor extent with the sulfided catalyst. Overall, as noted in the
preceding paragraph, the sulfided catalysts are more effective for pyridine HDN, as
indicated by the higher terminal product (pentane) concentration.

150

H.S. Joo. J.A. &in/

MM0

Fuel Processing Technology 49 (1996) 137-lS5

SNiMo

PtNiMo SPlNiMo

PICOMOSPtcoMo

Fig. 11. Effect of sulfidation on pyridine HDN selectivity at the same level of pyridine conversion at 350 C.

Pyridine HDN activity as a function of time for unpromoted and Pt-promoted sulfided
catalysts is shown in Fig. 12. PtNiMo and PtCoMo show higher HDN activities than the
unpromoted versions yielding essentially 100% HDN in 30 min. The HDN activity of
the NiW catalyst was not enhanced significantly by Pt promotion. The order of pyridine

0
0
A
0

NiMo
COMO
NiW
PtNiMo
?? PtCoMo
A PtNiW

R,
./
0
0

,
10

5
20

Time f(min)
_ .
Fig. 12. Pyridine HDN activity vs. time at 350 C.

30

H.S. Joo, J.A. Guin / Fuel Processing Technology 49 (19%) 137-155

151

Time (min)
Fig. 13. n-Pentane

concentration

vs. time in pyridine HDN experiments

at 350 C.

HDN activity was PtNiMo > PtCoMo > NiMo > CoMo > (PtNiW = NiW). Fig.
shows the associated pentane concentrations vs. time from the same experiments.
noted earlier, pentane concentration is another index of pyridine HDN and yields
same catalyst relative activity ranking as that of %HDN in Fig. 12. Assuming that

. PtCoMo

Time (min)
Fig. 14. First-order

plot of pyridine concentration

conversion

at 3.50 C.

13
As
the
the

1.52

H.S. Joo, J.A. Guin / Fuel Processing Technology 49 (19%) 137-155

Table 3
Rate constants for pyridine hydrogenation
Catalyst

Rate constant (min- )

PtNiMo
PtCoMo

0.16
0.12

NiMo
CoMo

0.12
0.091
0.067
0.058

PtNiW
NiW

hydrogen pressure is approximately constant so that the pyridine to piperidine HYD is

an apparent first-order reaction, Eq. (3) is obtained:

(3)

k't

where Cpyr is the pyridine concentration. Based on Eq. (3), the first-order plot in Fig. 14
and the apparent rate constants in Table 3 were obtained. The catalyst rate constant
ranking for pyridine HYD is the same as that of HDN. This trend is in agreement with
that in Fig. 5 in which the Pt-promoted catalyst showed improved HYD activity in the
presence of N. For both the promoted and unpromoted catalyst versions, NiMo was
more active for HDN than CoMo. This observation is in agreement with the results of
other investigations showing NiMo to be more effective for HDN [7,31].
While the most attention was focused on Pt, other noble metals showed significant
potential for promotion of HDN. The HDN activity with the unpromoted and Pt-, Ru-,
and Ir-promoted catalysts is shown in Fig. 15. The most obvious result in Fig. 15 is an
increase in HDN activity of the NiMo and CoMo catalysts due to both Pt and Ir

NiMo

COMO

NiW

Fig. 15. %HDN of pyridine with unpromoted and Ru-, Ir-, and Pt-promoted catalysts at 350 C for 20 min.

H.S. Joo, J.A. Guin / Fuel Processing Technology 49 (1996) 137-155

153

0.601

0.00

COMO

NIW

Fig. 16. Pentane concentration in pyridine HDN reactions with unpromoted and Ru-, Ir-, and Pt-promoted
catalysts at 350 C for 20 min.

promotion. Also the NiW catalyst was not improved significantly by any metal. As
noted earlier, these results suggest that the noble metal promotion is not simply an
additive effect, but that it depends on the original catalyst composition in synergistic
fashion. Fig. 16 shows the pentane concentrations corresponding to the experiments in
Fig. 15. Again, the relative catalyst activity rankings are the same as those based on
%HDN, with the Ir- and Pt-promoted NiMo and CoMo catalysts being most active.

4. Conclusions
4.1. Naphthalene

HYD

For the oxide catalysts, the Pt-promoted catalysts showed much higher activities than
the unpromoted ones, whereas the activities of both promoted and unpromoted sulfide
catalysts were similar. In the oxide forms, the Ni-promoted catalysts were more active
than Co-promoted versions for HYD. The sulfide catalysts were less active than the
oxide ones, except for CoMo. In the presence of N, the Pt-promoted catalysts exhibited
higher ring HYD activities than the unpromoted versions.
4.2. Pyridine HDN

All noble metals used improved the HDN activity in the NiMo catalysts with Pt
promotion being best. With CoMo, Pt and Ir improved the HDN activity to the same

154

H.S. Joo. J.A. Guin/ Fuel Processing Technology 49 (1996) 137-155

extent. With NiW, none of the noble metals increased the activity. Generally, the NiMo
catalyst was more efficient for HDN than the other catalysts tested either with or without
promotion. Among all the catalysts, PtNiMo was most active for pyridine HDN. For
both original and Pt-promoted catalysts, the sulfide forms were more effective for HDN
than the oxide ones. The results of kinetics studies suggest that the noble metals
promotion accelerates HYD of the N ring, which is the first step in N removal. Finally,
we note that experiments currently ongoing indicate an improved HDN activity for the
Pt-promoted catalysts with both coal and coal-plastics
coprocessing liquids [32]. The
completed results of these studies will be reported in a subsequent paper.

Acknowledgements
This work was supported by the US Department of Energy as part of the research
program of the Consortium for Fossil Fuel Liquefaction Science.

References
[l] R.D. Srivastava and H.G. McIlvried, ACS Div. Fuel Chem. Preprints, 40 (3x1995) 513.
[2] L.L. Anderson and W. Tuntawiroom, ACS Div. Fuel Chem Preprints, 38 (4) (1993) 810.
[3] G.P. Huffman, 2. Feng, V. Majajan, P. Sivakumar, H. Jung, J.W. Tiemey and I. Wender, ACS Div. Fuel
Chem. Preprints, 40 (1) (1995) 34.
[4] E.C. Orr, Y. Shi, J. Liang, W. Ding, L.L. Anderson and E.M. Eyring, ACS Div. Fuel Chem. Preprints, 40
(3) (1995) 633.
[5] A.H. Stiller, D.B. Dadybujor, J. Wann, D. Tian and J.W. Zondlo, ACS Div. Fuel Chem. Preprints, 40 (1)
(1995) 77.
[6] 0. Weisser and S. Landa, Sulfide Catalysts-Their
Properties and Applications, Pergamon, Oxford,
1973.
[7] J.R. Katzer and R. Sivasubramanian, Catal. Rev. Sci. Eng., 20 (1979) 155.
[8] S.-J. Liaw, R.A. Keogh, G.A. Thomas and B.H. Davis, Energy Fuels, 8 (1994) 581.
[9] G.H. Singhal, W.E. Winter, K.L. Riley and K.L. Trachte, US Patent 5 152885, 1991.
[lo] A.S. Hirschon, R.B. Wilson and R.M. Laine, Appl. Catal., 34 (1987) 311.
[ll] A.S. Hirschon, R.B. Wilson and R.M. Lame, Advances in Coal Chemistry, Theophrastus Publications,
Athens, Greece, 1988, 351 pp.
[12] D.K. Lee, 1.C. Lee and S.1. Woo, Appl. Catal., 109 (1994) 195.
[13] J.S. Shabtai, N.K. Nag and F.E. Massoth, J. Catal., 104 (1987) 413.
[14] M.J. Ledoux, 0. Michaux and G. Agostini, J. Catal., 102 (1986) 275.
[15] S. Eijsbouts, V.H. De Beer and R. Prim, J. Catal., 109 (1988) 217.
[16] S. Eijsbouts, C. Sudhakar, V.H. De Beer and R. Prins, J. Catal., 127 (1991) 605.
[17] S. Eijsbouts, V.H. De Beer and R. Prim, J. Catal., 127 (1991) 619.
(181 T.A. Pecoraro and R.R. Chianelli, J. Catal., 67 (1981) 430.
[19] T.C. Ho, Catal. Rev. Sci. Eng., 30 (1) (1988) 117.
[20] M.J. Girgis and B.C. Gates, Ind. Eng. Chem. Res., 30 (1991) 2021.
[21] 2. Sarbak, Acta Chim. Hungarica, 127 (3) (1990) 371.
[22] G.C. Hadjiloizou, J.B. Butt and J.S. Dranoff, J. Catal., 131 (1991) 545.
[23] H.G. Mcllvried, Ind. Eng. Chem. Process Des. Dev., 10 (1) (1971) 125.
[24] C.N. Satterfield and J.F. Cocchetto, AICHE J., 21 (6) (1975) 1107.
[25] R.T. Hanlon, Energy Fuels, 1 (1987) 424.

H.S. Joo. J.A. Guin / Fuel Processing

1261
1271
[28]
[29]
[30]
[31]

Technology 49 (1996) 137-155

C.N. Satterfield, M. Model1 and J.A. Wilkens, Ind. Eng. Chem. Process Des. Dev., 19 (1980) 154.
J. Sonnemans, G.H. Van Der Berg and P. Mars, J. Catal., 31 (1973) 220.
G.C. Hsdjiloizou, Ph.D. Thesis, Northwestern University, Evanston, IL, 1989.
J.A. Guin, X. Zhan and R.S. Linhart, Energy Fuels, 8 (1994) 105.
S.D. Lin and C. Song, ACS Div. Fuel Chem. Preprints, 40 (4) (1995) 962.
T.C. Ho, Catal. Rev. Sci. Eng., 30 (1988) 117.

1321 H.S. JOO, X. Zhan and J.A. Guin, No.4ld.,

AIChE Spring Meeting, Houston,

TX, 1995.

155