Effect of Zr on catalytic performance of unsupported Ni(Zr)Mo and Ni(Zr)W sulfide catalysts for quinoline hydrodenitrogenation

Article abstract of DOI:10.1016/j.apcata.2020.117744

To increase the dispersion of active species and full utilization of active metals are of great importance for hydrodenitrogenation (HDN) performance of unsupported hydrotreating catalysts. Herein, a series of unsupported Ni(Zr)MoS and Ni(Zr)WS catalysts were prepared from Ni(Zr) layered double hydroxide (LDH) precursors. The Zr species remarkably promote the dispersion and reducibility of NiMo and NiW composite species. Also, the total HDN rate constants are increased from 2.42 h^?1 to 9.18 h^?1 for Ni(Zr)MoS and from 5.68 h^?1 to 23.0 h^?1 for Ni(Zr)WS, and exhibit a maximum at a Zr/Ni atomic ratio of about 0.04. The HDN selectivities indicate that the Zr species increase the number of superficial active sites without affecting their structure. The present work shows that the crystallinity of LDH precursors is crucial to the structure of unsupported sulfide catalysts, and a suitable amount of Zr could be a dispersive promoter to increase the HDN activity.

Full text of DOI:10.1016/j.apcata.2020.117744

Journal Pre-proof
Effect of Zr on catalytic performance of unsupported Ni(Zr)Mo and
Ni(Zr)W sulfide catalysts for quinoline hydrodenitrogenation
Xinyi Liu, Cen Zhang, Tiefeng Liu, Zongxuan Jiang, Can Li
PII:
S0926-860X(20)30337-9
DOI:
https://doi.org/10.1016/j.apcata.2020.117744
APCATA 117744
Reference:
To appear in:
Applied Catalysis A, General
Received Date:
Revised Date:
Accepted Date:
30 March 2020
15 July 2020
17 July 2020
Please cite this article as: Liu X, Zhang C, Liu T, Jiang Z, Li C, Effect of Zr on catalytic
performance of unsupported Ni(Zr)Mo and Ni(Zr)W sulfide catalysts for quinoline
hydrodenitrogenation, Applied Catalysis A, General (2020),
doi: https://doi.org/10.1016/j.apcata.2020.117744
This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
© 2020 Published by Elsevier.

Effect of Zr on catalytic performance of unsupported Ni(Zr)Mo and
Ni(Zr)W sulfide catalysts for quinoline hydrodenitrogenation
Xinyi Liu a,b, Cen Zhang a,b, Tiefeng Liu a, Zongxuan Jiang a,*, Can Li a,*
a State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, 457 Zhongshan Road, Dalian 116023, China
b University of Chinese Academy of Sciences, Beijing 100049, China
*Corresponding author. Tel: 86-411-84379070 (C. Li); Fax: 86-411-84694447;
Email: canli@dicp.ac.cn (C. Li); zxjiang@dicp.ac.cn (Z. Jiang);
URL: http://www.canli.dicp.ac.cn
1

Graphical Abstract
NiZr layered double hydroxide
1) Precipitation
2) Introduction of Zr
Ni2+
1) Deposition of Mo/W
2) Calcination
Sulfidation
Oxidic NiZrM catalyst
M = Mo or W
Sulfided NiZrM catalyst
M = Mo or W
Highlights
 Zr species decrease the crystalline size of layered double hydroxide precursor
via obstructing the aggregation and growth of the crystallites.
 Zr increases the dispersion and reducibility of oxidic NiMo and NiW active
species, which is favorable to the activation of active phases.
 The introduction of Zr promotes the exposure of active NiMoS and NiWS
species on surface of catalyst, thus increases the hydrodenitrogenation activity.
2

Abstract
To increase the dispersion of active species and full utilization of active metals are
of great importance for hydrodenitrogenation (HDN) performance of unsupported
hydrotreating catalysts. Herein, a series of unsupported Ni(Zr)MoS and Ni(Zr)WS
catalysts were prepared from Ni(Zr) layered double hydroxide (LDH) precursors. The
Zr species remarkably promote the dispersion and reducibility of NiMo and NiW
composite species. Also, the total HDN rate constants are increased from 2.42 h-1 to
9.18 h-1 for Ni(Zr)MoS and from 5.68 h-1 to 23.0 h-1 for Ni(Zr)WS, and exhibit a
maximum at a Zr/Ni atomic ratio of about 0.04. The HDN selectivities indicate that the
Zr species increase the number of superficial active sites without affecting their
structure. The present work shows that the crystallinity of LDH precursors is crucial to
the structure of unsupported sulfide catalysts, and a suitable amount of Zr could be a
dispersive promoter to increase the HDN activity.
Keywords: Unsupported catalyst; Hydrodenitrogenation; Hydrotreating
3

1. Introduction
Hydrodenitrogenation (HDN) is necessary for refinery processes because of the
poisoning effect of basic N-containing compounds in fossil fuel on acidic active sites
over zeolite catalysts. Also, N-containing compounds exhibit a strong inhibiting effect
on hydrodesulfurization (HDS) reaction [1,2], which is detrimental to ultra-deep HDS.
Similar to HDS catalysts, NiMoS/Al2O3, NiWS/Al2O3 or NiMoWS/Al2O3 catalysts are
commonly used for HDN in industry [3-5]. Since the hydrogenation steps before
hydrogenolysis are indispensable for aromatic heterocyclic N-containing compounds
and C-N bond is usually much stronger than C-S bond, ultra-deep HDN is more
challenging than ultra-deep HDS. Around 2001, unsupported catalysts based on
NEBULA technology were reported to show more superior HDS and HDN activities
than conventional alumina supported hydrotreating catalysts [6-7]. Moreover,
unsupported catalysts with different metallic composition were studied and the HDS
performances of them were improved by introducing other appropriate metal
components [8-12]. Unsupported catalysts have great potential for industrial
applications due to their excellent hydrotreating performance, however, they are still
limited by the high metal contents and the resulting high costs. Thus, how to increase
the dispersion and utilization of active metals is one of the key issues for improving the
hydrotreating activities of unsupported catalysts.
Layered double hydroxide (LDH) is consisted of layers composed of divalent metal
cations octahedrally coordinated by hydroxyl groups. While a part of original metal
cations is substituted, the layers will be positively charged and exchangeable negatively
2-
-
charged interlayer anions (such as CO3 , NO3 et al) will appear to balance the charges.
In our previous work, using NiAl-LDH [8], NiZn-LDH [9] and NiZnAl-LDH [10] as
structural templates followed by introducing Mo/W via ion-exchange was demonstrated
to be a practical method to prepare unsupported multi-metallic sulfide catalysts, and the
as-prepared catalysts showed much higher catalytic activity for HDS of 4,6-
dimethyldibenzothiophene compared with alumina supported hydrotreating catalysts.
4

For single metallic Ni-LDH, although there is no stoichiometric interlayer anion, the
layered structure still allows a certain amount of active metal species to insert and
disperses them. Thus, unsupported NiMoWS catalysts using Ni-LDH as a precursor
also exhibited outstanding HDS activity [13]. However, the HDN performance of such
catalysts remains unclear. Hence, in the present work, unsupported Ni(Zr)MoS and
Ni(Zr)WS catalysts with different Zr amounts were prepared, characterized and
evaluated for HDN of quinoline, and the influence of Zr on the HDN performance is
discussed in detail.
2. Experimental
2.1 Catalyst preparation
All chemicals were of analytical reagent grade and used without further
purification. Unsupported Ni(Zr)MoS and Ni(Zr)WS catalysts were prepared according
to the procedure reported in the previous work [13]. Firstly, Ni-LDH precursor was
prepared by a precipitation method. Typically, 40 mmol nickel acetate tetrahydrate was
dissolved in deionized water to form a homogeneous solution, and this solution was
added into a three port flask and kept at 100 oC. Afterwards, a solution of Na2CO3 (0.1
M) was added dropwise under vigorous stirring to adjust the pH value to 10 and a green
precipitate of Ni-LDH was formed. For the preparation of NiZr-LDH precursors, the Zr
content was introduced by a further deposition: a certain amount of zirconium acetate
solution was then added to the green suspension, and a small amount of Na2CO3
solution was also added to keep the pH value constant. After keeping at 100 oC for 1h,
the precipitate was filtered and washed with deionized water, and the LDH precursors
were denoted as N-LDH or NZ-LDH-x. For the preparation of oxidic Ni(Zr)Mo
catalysts, ammonium heptamolybdate containing 20 mmol Mo was dissolved in 40 ml
deionized water in a flask, followed by adding Ni(Zr)-LDH precursors under vigorous
stirring to form a suspension. Then, the suspension was kept at 90 oC until water was
evaporated completely, and the oxidic catalysts were obtained after drying overnight
5

and calcination at 380 oC for 4 h. For the preparation of oxidic Ni(Zr)W catalysts, only
a substitution of ammonium heptamolybdate by ammonium metatungstate containing
20 mmol W was needed. Here, the Ni(Zr)Mo catalysts were denoted as NM or NZM-x
and the Ni(Zr)W catalysts were denoted as NW or NZW-x. Besides, all Ni(Zr)-LDH
precursors were calcined to prepare Ni(Zr) oxides denoted as N-O or NZ-O-x.
2.2 Catalyst characterization
X-ray fluorescence (XRF) elemental analysis was performed on a Magix-2424
XRF spectrometer to determine the elemental compositions. X-ray diffraction (XRD)
patterns were collected on a PANalytical Empyrean powder diffractometer with Cu-Kα
radiation (λ = 0.154 nm) operated at 40 kV and 40 mA. Each XRD pattern was collected
o
from 10o to 80o at a speed of 5 /min. N2-physisorption was performed at 77 K on a
Quantachrome instrument. The surface area was calculated using the multi-point
Brunauer-Emmett-Teller (BET) method by assuming the section area of nitrogen
molecule is 0.162 nm2. The total pore volume was calculated by the amount of nitrogen
adsorbed at P/P0 = 0.98. Before the test, each sample was outgassed in a vacuum (about
o
10-4 Torr) at 300 C for 6 h. Transmission electron microscopy (TEM) images were
collected on a Tecnai G2 F30 S-Twin (FEI Company) instrument at an acceleration
voltage of 300 kV. The distributions and averages of the stacking number and slab
length of MoS2/WS2 crystallites were calculated, and the equations for calculating the
average stacking number (Nav) and slab length (Lav) were listed in the supporting
information. Temperature programmed reduction (TPR) profiles were collected on a
TPR instrument equipped with a thermal conductivity detector (TCD). Before the test,
each sample was preheated in Ar at 130 oC for 1 h. Then, the sample was heated to 900
oC (10 oC/min) in 5 vol.% H2/Ar with a flow rate of 50 ml/min.
2.3 Catalyst performance evaluation
The HDN performance evaluation was performed on a high-pressure fixed-bed
6

continuous flow microreactor using a solution of quinoline in decalin with a nitrogen
contents of 500 ppmw as feed. Prior to the HDN evaluation, the shaped oxidic catalyst
was loaded and in-situ sulfided at 400 oC in 10 vol.% H2S/H2 for 2.5 h. Subsequently,
the HDN reaction was carried out at 300 oC, a H2 pressure of 3.5 MPa, a liquid hourly
space velocity (LHSV) of 4.5-36 h-1 and a H2/oil ratio of 600 V/V. After a stabilization
period of 5 h, the product was collected and analyzed. The total nitrogen amount was
determined on an Antek azotometer and the components were analyzed on an Agilent
6890N gas chromatograph with a DB-FFAP capillary column (30 m × 0.320 mm).
3. Results and discussion
3.1. Structure and reducibility of Ni(Zr)-LDHs and Ni(Zr) oxides
Fig. 1a shows the XRD patterns of Ni(Zr)-LDH precursors, and the diffraction
peaks at 2θ = 33.1°, 38.5°, 52.1°, 59.1°, 62.8° are attributed to β-Ni(OH)2 (JPCDS 14-
0017), which has a typical brucite-like layered structure. The two peaks at 2θ = 16.1°
and 20.3° are different from the standard (001) peak at 2θ = 19.2°, implying two
interlayer spacings. Fig. 1b shows the XRD patterns of Ni(Zr) oxides, and the
diffraction peaks at 2θ = 37.2°, 43.3°, 62.9°, 75.4°, 79.4° are attributed to bunsenite
(JPCDS 47-1049). For either Ni(Zr)-LDHs or Ni(Zr) oxides, all diffraction peaks
associated with Ni species are weakened and broadened dramatically with increasing
the Zr content, indicating that the introduction of Zr remarkably decreases the
crystallinity of β-Ni(OH)2 (Ni-LDH) and such effect is retained after calcination. Also,
the two (001) peaks of Ni-LDH merge gradually with increasing the Zr content,
implying the convergence of interlayer spacings. Besides, the absence of crystalline Zr
species suggests that they are amorphous and highly dispersed on Ni-LDH crystallites.
The reducibilities of NiO and NiZr composite oxide are compared as shown in Fig. 2.
Since bulk ZrO2 can hardly be reduced by H2 [14-17], the consumption of H2 is
attributed to the reduction of NiO. The reduction peak of pure NiO is located at 389 oC.
With increasing the Zr content, this peak is weakened and a new reduction peak at 464
7

oC is increased gradually. The TPR results indicate that Zr species have interaction with
NiO [15], probably through substituting the hydroxyl groups of Ni-LDH, and such
interaction hinders the reduction of NiO. Considering that the Zr species are deposited
after the precipitation of Ni-LDH, we suggest that they are predominantly located on
the surface of Ni-LDH particles. The superficial Zr species inhibit the aggregation and
growth of Ni-LDH crystallites, and such effect is retained after calcination (Scheme 1).
3.2. Structure and reducibility of the oxidic precursors of Ni(Zr)MoS and Ni(Zr)WS
catalysts
The elemental compositions of oxidic Ni(Zr)Mo and Ni(Zr)W catalysts are listed
in Table 1. Either Mo/Ni or W/Ni atomic ratio in each oxidic catalyst is essentially
consistent with the raw materials. Fig. 3 shows the XRD patterns of oxidic Ni(Zr)Mo
and Ni(Zr)W catalysts. The oxidic precursors of NiMoS and NiWS catalysts (NM and
NW) exhibit three peaks at 2θ = 37.2°, 43.3° and 62.9° associated with crystalline phase
of NiO derived from residual Ni-LDH. Moreover, the peaks of NiO phase disappear
with the introduction of Zr, indicating that the decrease in the crystallinity of Ni-LDH
is conducive to the dispersion of Ni species in the oxidic catalysts. For oxidic Ni(Zr)Mo
catalysts (Fig. 3a), both α-NiMoO4 (JPCDS 33-0948) and β-NiMoO4 (JPCDS 45-0142)
are observed. In addition, there is a transformation form α phase into β phase with the
introduction of Zr. Renzo et al [18] and Mazzocchia et al [19] reported that the excessive
Ni content helps to stabilize β-NiMoO4 phase by inserting Ni into the NiMoO4 lattice.
Thus, the transformation of NiMoO4 phase could be attributed to the increasing
dispersion of NiO, which promotes the contact between Ni species and NiMoO4. In the
case of oxidic Ni(Zr)W catalysts (Fig. 3b), the peaks at 2θ = 37.2°, 41.0°, 53.7°, 64.2°
are attributed to a crystalline phase of (NH4)aNibWcOx reported by Soled [20]. Further,
the dispersive promoting effect of Zr on the oxidic catalysts is also proved by N2-
physisorption (Table 2): it makes the surface area larger (from 164 m2/g to 180 m2/g for
oxidic Ni(Zr)Mo catalysts and from 143 m2/g to 170 m2/g for oxidic Ni(Zr)W catalysts)
8

and the pore volume smaller (from 0.33 cm3/g to 0.21 cm3/g for oxidic Ni(Zr)Mo
catalysts and from 0.29 cm3/g to 0.19 cm3/g m2/g for oxidic Ni(Zr)W catalysts). Besides,
each oxidic catalyst exhibits a type IV of isotherm with H2-type of hysteresis (Fig. S1),
which is the characteristic of mesoporous material with nonuniform pore structure
according to the IUPAC classification.
Fig. 4 exhibits the TPR profiles of oxidic Ni(Zr)Mo and Ni(Zr)W catalysts. For
oxidic Ni(Zr)Mo catalysts (Fig. 4a), NM shows a major peak at 580 oC, a weak peak at
about 529 ℃ which overlaps with the most intense peak at 580 ℃ and a broad weak
o
o
o
peak from 700 C to 900 C. The peaks below 700 C could be attributed to the
simultaneous reduction of Ni2+ and polymeric octahedral (Oh) Mo6+ species while the
weak peak at high temperature could be corresponding to the reduction of isolated
tetrahedral (Td) Mo6+ species [21-22]. With the introduction of Zr, the peak at above
700 oC is disappeared, the peak at 580oC is weakened and the peak at 529 ℃is increased.
These results demonstrate that the Zr species promote the transformation of Td Mo6+
species to Oh Mo6+ species and increase the reducibility of Oh Mo6+ species. In the case
of oxidic Ni(Zr)W catalysts, NW shows two peaks at 527 oC and 690 oC. The first peak
o
below 600 C could be assigned to the simultaneous reduction of Ni2+ and polymeric
octahedral (Oh) W6+ species while the second peak over 650 oC could be attributed to
the reduction of dimeric tetrahedral (Td) W6+ species [23-24]. With the introduction of
o
o
Zr, the first peak is shifted to 472 C. Moreover, the peak at 690 C is weakened
gradually and a new peak corresponding to the reduction of Oh W6+ species [23] is
observed at 585 oC. These results show that the effect of Zr on W species is quite similar
to that on Mo species, that is, it promotes the transformation of Td W6+ species to Oh
W6+ species and increase the reducibility of Oh W6+ species. Therefore, it could be
concluded that the introduction of Zr promotes the reducibility of NiMo and NiW
species, which is conducive to the activation of the oxidic catalysts.
3.3. Structure of unsupported Ni(Zr)MoS and Ni(Zr)WS catalysts
9

The crystalline structures of unsupported Ni(Zr)MoS and Ni(Zr)WS catalysts are
shown in Fig. 5. Here, MoS2 (JPCDS 37-1492) and WS2 (JPCDS 08-0237) phase with
very similar patterns are observed. Besides, the peaks at 2θ = 21.8°, 31.1°, 37.8°, 38.3°,
44.3°, 49.7°, 50.1° and 55.2° are corresponding to Ni3S2 phase (JPCDS 44-1418). For
both Ni(Zr)MoS and Ni(Zr)WS catalysts, the peaks associated with Ni3S2 are weakened
with the introduction of Zr, implying a decrease in crystalline nickel sulfide phase.
Since nickel sulfide phase also acts as a support, the exposure of composite NiMo or
NiW species on surface of particles is probably promoted by the introduction of Zr. Fig.
6a-d shows the morphologies of Ni(Zr)MoS (NM and NZM-3) and Ni(Zr)WS (NW and
NZW-3) catalysts. Each sulfided catalyst is composed of nickel sulfide particles with a
size of 10-20 nm and MoS2 or WS2 phase at the striate regions. In addition, for either
Ni(Zr)MoS or Ni(Zr)WS catalysts, it is observed that the distributions of stacking
number (Fig. 6e-h) and slab length (Fig. 6i-l) of MoS2 or WS2 are not distinctly
influenced by Zr species: the average stacking number (Nav: from 2.5 to 2.5 for
Ni(Zr)MoS and from 2.5 to 2.5 for Ni(Zr)WS) and slab length (Lav: from 6.9 nm to 7.0
nm for Ni(Zr)MoS and from 6.5 nm to 6.5 nm for Ni(Zr)WS) are almost the same.
These results indicate that the introduction of Zr scarcely influences the morphologies
of MoS2 and WS2 crystallites.
3.4. Catalytic performances and kinetics of quinoline HDN
A whole reaction network of quinoline HDN is shown in Scheme 2 [3-5,25-26].
Firstly, the heterocyclic ring of quinoline (Q) is hydrogenated rapidly to form 1,2,3,4-
tetrahydroquinoline (THQ-1) and an equilibrium between them is established [27-29].
Thus, Q and THQ-1 are lumped together as the original reactant to calculate the
conversion (Conv., including all conversion mentioned without further description):
CpQ  Cp
Conv.  (1-
THQ-1 )100%
CfQ
where CfQ, Cp and CpTHQ-1 are the concentrations of quinoline (Q) in feed, Q in product
Q
10

and 1,2,3,4-tetrahydroquinoline (THQ-1) in product, respectively. Also, the carbocyclic
ring of Q can be hydrogenated to form 5,6,7,8-tetrahydroquinoline (THQ-5). Both
THQ-1 and THQ-5 can be further hydrogenated to form decahydroquinoline (DHQ).
Then, the first hydrogenolysis step involving C-N bond cleavage takes place on THQ-
1 and DHQ to form propylaniline (OPA) and propylcyclohexylamine (PCHA),
respectively. Afterwards, the nitrogen atom of PCHA is removed via a Hofmann
elimination to form propylcyclohexene (PCHE), and propylbenzene (PB) could be
formed via a direct denitrogenation (DDN) of OPA. Finally, propylcyclohexane (PCH)
is formed via hydrogenation of PCHE. Besides, PCHE could be formed via
hydrogenation of PB. The selectivities (Sel.) are calculated by the equation as follows:
Cpi
Cpi
Cpi
Sel.i 
100% 
100% 
100%
Cp
CfQ CpQ CpTHQ-1
CfQ Conv.

i
i
where Sel.i and Cp are the selectivity and concentration of compound i in product,
i
respectively. For quinoline HDN reaction in this work, i could be THQ-5, DHQ, OPA,
PB, PCH or entire hydrocarbons (HC). In addition, the total HDN rate constants (kHDN
)
could also be calculated by assuming a pseudo-first-order reaction:
Np
kHDN   LHSVln
Nf
where LHSV is the liquid hourly space velocity and Np and Nf are the total N content
in product and feed, respectively.
Fig. 7a and Fig. 8a exhibit the HDN activities of Ni(Zr)MoS and Ni(Zr)WS
catalysts. It is observed that each Ni(Zr)WS catalyst shows a higher HDN activity than
Ni(Zr)MoS catalyst with a same Zr/Ni atomic ratio. Moreover, as shown in Table 3, the
HDN activities of Ni(Zr)MoS and Ni(Zr)WS catalysts are increased by the introduction
of Zr (from 2.42 h-1 to 9.18 h-1 for Ni(Zr)MoS and from 5.68 h-1 to 23.0 h-1 for
Ni(Zr)WS), and both of them exhibit a maximum at a Zr/Ni atomic ratio of about 0.04.
These results imply that the promoting effect of Zr on the HDN activity could be
attributed to the influence of Zr on LDH precursors. Fig. 7b-c and Fig. 8b-c exhibit the
11

selectivities of N-containing intermediates and entire hydrocarbons (including PCHE,
PCH and PB) as a function of conversion. For all catalysts, the selectivity of THQ-5 is
quite low because THQ-5 has a slow formation rate and a fast consumption rate [30].
Therefore, DHQ should be predominantly formed via the hydrogenation of THQ-1.
Besides, since the adsorption of aniline is much weaker than that of DHQ [31] and C-
N(sp2) bond is stronger than C-N(sp3) bond, the hydrogenolysis of THQ-1 to form OPA
is limited. Hence, the first hydrogenolysis step predominantly takes place on DHQ to
form PCHA, and the selectivity of OPA keeps at a rather low level. Here, PCHA is not
detected owing to its high reactivity that the N atom is removed rapidly via a Hofmann
elimination to form PCHE. Afterwards, PCHE can be further hydrogenated to form
PCH. Additionally, considering the conjugation between the nitrogen atom and the
benzene ring, the hydrogenolysis of OPA to PB is also a slow step. Based on the above,
the whole reaction network of quinoline HDN could be simplified (Scheme 3):
(Q+THQ-1) → DHQ → [PCHA] → PCHE → PCH. From Fig. 7b-d and Fig. 8b-d,
each selectivity of THQ-5, DHQ, OPA or entire hydrocarbons as a function of
conversion shows the same trend for all Ni(Zr)MoS or Ni(Zr)WS catalysts, whereas the
selectivities of PB and PCH are increased slightly with the introduction of Zr.
3.5. Effect of Zr on Ni(Zr)MoS and Ni(Zr)WS catalysts and catalytic performances for
quinoline HDN
First of all, the introduction of Zr brings Ni-LDH precursor smaller crystalline size,
and such effect could also be retained after calcination. Considering that Zr is
introduced after the precipitation of Ni-LDH, the Zr species are probably located on
surface. Thus, the dispersive promoting effect of Zr should be derived from the
inhibition on growth and aggregation of Ni-LDH crystallites, and facilitates the reaction
of Ni-LDH with molybdate and tungstate. In addition, the Zr species increase the
surface area of the oxidic catalysts, which is beneficial to the dispersion of the NiMo
and NiW composite species. Further, the introduction of Zr results in a transformation
12

from tetrahedral Mo6+ and W6+ species (harder to be reduced) to octahedral Mo6+ and
W6+ species (easier to be reduced), and also decreases the reduction temperature of the
latter. Hence, it could be concluded that the textures of LDH precursors could strongly
influence the properties of the oxidic catalysts, and the introduction of Zr promotes the
dispersion and reducibility of NiMo and NiW composite species.
The HDN activities of Ni(Zr)MoS and Ni(Zr)WS catalysts are increased by the
introduction of Zr. In comparison, for THQ-5, DHQ, OPA and hydrocarbons which are
converted from N-containing intermediates, the selectivities at the same conversion are
scarcely influenced. The hydrotreating active sites are considered to locate at the edge
of stacked MoS2 or WS2 layers incorporated with Ni atoms (NiMoS or NiWS active
species) [32-34]. Considering that the distributions and averages of stacking number
and slab length of MoS2 or WS2 phase do not distinctly change with the introduction of
Zr, we thus suggest that the Zr species do not influence the structure of active sites and
the increase the number of superficial active sites should be the reason for the increasing
HDN activity. On the other hand, the HDN activities exhibit a maximum at a Zr/Ni
atomic ratio of about 0.04 for both Ni(Zr)MoS and Ni(Zr)WS catalysts. Considering
the location of Zr species, they could retard the contact between active sites and
reactants. As a consequence, the excessive Zr content is detrimental to the HDN activity.
Besides, the selectivities of PB and PCH are increased with the introduction of Zr,
indicating the hydrogenolysis of OPA and the hydrogenation of PCHE is promoted.
Also, the increasing effect on the selectivity of PCH is more pronounced than that of
PB, implying that the hydrogenation of PB is promoted as well. Since the adsorption of
DHQ on active sites is much stronger than that of other intermediates [31], the
adsorption of other species should be critical to their conversion. Hence, we suppose
that the introduction of Zr presumably promotes the desorption of DHQ on Ni(Zr)MoS
and Ni(Zr)WS catalysts.
4. Conclusions
13

A series of unsupported Ni(Zr)MoS and Ni(Zr)WS catalysts with different Zr
amounts were prepared from Ni(Zr)-LDH precursors. Since the Zr species facilitate the
dispersion and reducibility of the oxidic NiMo and NiW composite species, the
activation of active phases and the exposure of active NiMoS and NiWS species on
surface of catalyst particles are promoted. The HDN activities of both Ni(Zr)MoS and
Ni(Zr)WS catalysts exhibit a maximum at a Zr/Ni atomic ratio of about 0.04 owing to
an increase in the number of superficial active sites. Besides, presumably due to the
enhancement of the desorption of DHQ on Ni(Zr)MoS and Ni(Zr)WS catalysts, the
introduction of Zr slightly promotes the conversion of OPA, PCHE and PB, and
increases the selectivities of PB and PCH.
Credit Author Statement
Author contributions:
Xinyi Liu: design and implementation of the experiment, writing and revising the
manuscript.
Cen Zhang: discussion and revising the manuscript.
Tiefeng Liu: construction of the experimental device.
Zongxuan Jiang: academic guidance and design of the experiment.
Can Li: academic guidance and revising the manuscript.
Declaration of interests
The authors declare that they have no known competing financial interests or personal
14

relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This work was financially supported by National Natural Science Foundation of
China (NSFC-Liaoning No. U1508205) and Science and Technology Service (STS)
Network Initiative of the Chinese Academy of Sciences (KFJ-STS-QYZX-015).
References
[1] K. Liu, F. T.T. Ng, Catal. Today 149 (2010) 28-34.
[2] H. Farag, M. Kishida, H. Al-Megren, Appl. Catal. A: Gen. 469 (2014) 173-182.
[3] R. Prins, Advances in Catalysis 46 (2001) 399-464.
[4] E. Furimsky, F. E. Massoth, Catal. Rev. 47 (2005) 297-489.
[5] T. C. Ho, Catal. Rev.-Sci. Eng. 30 (1988) 117-160.
[6] F. L. Plantenga, R. Cerfontain, S. Eijsbouts, F. van Houtert, G. H. Anderson, S.
Miseo, S. Soled, K. Riley, K. Fujita, Y. Inoue, Studies Surface Science and Catalysis.
Tokyo: Kodansha Ltd. 145 (2003) 407-410.
[7] S. Eijsbouts, S.W. Mayo, K. Fujita, Appl. Catal. A: Gen. 322 (2007) 58-66.
[8] L. Wang, Y. Zhang, Y. Zhang, P. Liu, H. Han, M. Yang, Z. Jiang, C. Li, Appl. Catal.
A: Gen. 394 (2011) 18-24.
[9] Y. Chen, L. Wang, Y. Zhang, T. Liu, X. Liu, Z. Jiang, C. Li, Appl. Catal. A: Gen.
474 (2014) 69-77.
[10] Y. Chen, L. Wang, X. Liu, T. Liu, B. Huang, P. Li, Z. Jiang, C. Li, Appl. Catal. A:
Gen. 504 (2015) 319-327.
[11] H. Wang, Z. Yang, X. Zhan, Y. Wu, M. Li, Fuel Processing Technology 160 (2017)
178-184.
[12] H. Wang, Z. Yang, F. Meng, W. Yin, Y. Wu, Fuel 228 (2018) 332-341.
[13] L. Wang, Y. Zhang, Y. Zhang, Z. Jiang, C. Li, Chem. Eur. J. 15 (2009) 12571-
12575.
15

[14] N. Li, A. Wang, M. Zheng, X. Wang, R. Cheng, T. Zhang, J. Catal. 225 (2004)
307-315.
[15] S. Li, G. Liu, S. Zhang, K. An, Z. Ma, L. Wang, Y. Liu, J. Energy Chem. 43 (2020)
155-164.
[16] X. Zhang, D. He, Q. Zhang, Q. Ye, B. Xu, Q. Zhu, Appl. Catal. A: Gen. 249 (2003)
107-117.
[17] Q. Tong, A. Zong, W. Gong, L. Yu, Y. Fan, RSC Adv. 6 (2016) 86663-86672.
[18] F. Di Renzo, C. Mazzocchia, G. Thomas, A. M. Vernay, Reactivity of Solids 6
(1988) 145-155.
[19] C. Mazzocchia, F. Di Renzo, Ch. Aboumrad, Solid State Ionics 32-33 (1989) 228-
233.
[20] S. L. Soled, S. Miseo, R. Krikak, H. Vroman, T. C. Ho, K. L. Riley, US Patent
6,156,695, Exxon Research and Engineering Company, 2000.
[21] J. L. Brito, J. Laine, K. C. Pratt, J. Mater. Sci. 24 (1989) 425-431.
[22] T. Klimova, O. Gutiérrez, L. Lizama, J. Amezcua, Micro. Meso. Mater. 133 (2010)
91-99.
[23] Q. Cui, Y. Zhou, Q. Wei, G. Yu, L. Zhu, Fuel Process. Technol. 106 (2013) 439-
446.
[24] M. Ebrahiminejad, R. Karimzadeh, J. Hazard. Mater. 384 (2020) 121485.
[25] W. Han, H. Nie, X. Long, M. Li, Q. Yang, D. Li, J. Catal. 339 (2016) 135-142.
[26] M. Nguyen, M. T. Fayolle, F. Chainet, G. D. Pirngruber, C. Geantet, Appl. Catal.
A: Gen. 530 (2017) 132-144.
[27] C. N. Satterfield, J. F. Cocchetto, Ind. Eng. Chem. Proc. Des. Dev. 20 (1981) 53-
62.
[28] O. Y. Gutiérrez, A. Hrabar, J. Hein, Y. Yu, J. Han, J. A. Lercher, J. Catal. 295 (2012)
155-168.
[29] S. Albersberger, H. Shi, M. Wagenhofer, J. Han, O. Y. Gutiérrez, J. A. Lercher, J.
Catal. 380 (2019) 332-342.
16

[30] M. T. Nguyen, M. Tayakout-Fayolle, G.D. Pirngruber, F. Chainet, C. Geantet, Ind.
Eng. Chem. Res. 54 (2015) 9278-9288.
[31] V. La Vopa, C. N. Satterfield, J. Catal. 110 (1988) 375-387.
[32] R. R. Chianelli, G. Berhault, B. Torres, Catal. Today 147 (2009) 275-286.
[33] D. Zuo, M. Vrinat, H. Nie, F. Maugé, Y. Shi, M. Lacroix, D. Li, Catal. Today 93-
95 (2004) 751-760.
[34] Y. Iwata, Y. Araki, K. Honna, Y. Miki, K. Sato, H. Shimada, Catalysis Today 65
(2001) 335-341.
17

▼: β-Ni(OH)2
a
N-LDH
NZ-LDH-1
NZ-LDH-2
NZ-LDH-3
NZ-LDH-4
NZ-LDH-5
10
20
30
40
50
60
70
2Theta (o)
▼: NiO
b
N-O
NZ-O-1
NZ-O-2
NZ-O-3
NZ-O-4
NZ-O-5
20
30
40
50
60
70
80
2Theta (o)
Figure 1. XRD patterns of Ni(Zr)-LDHs (a) and Ni(Zr) oxides (b).
18

389
464
N-O
NZ-O-1
NZ-O-2
NZ-O-3
NZ-O-4
NZ-O-5
100
200
300
400
500
600
700
800
900
Temperature (oC)
Figure 2. TPR profiles of NiO and NiZr oxide.
19

Calcination
Aging at 100 oC
Precipitation
Ni2+
NiO
Zr species
Ni-LDH
1) Introducing Zr
2) Aging at 100 oC
Calcination
Scheme 1. A model for the effect of Zr on the structure of Ni(Zr)-LDH particles.
20

Table 1. The elemental composition of the oxidic precursors of Ni(Zr)MoS and
Ni(Zr)WS catalysts.
Atomic ratio
NiO
MoO3
wt.%
WO3
wt.%
ZrO2
wt.%
Cat.
wt.%
Mo/Ni W/Ni
Zr/Ni
NM
50.2
49.8
49.7
48.8
48.2
47.3
38.7
38.6
38.3
37.9
37.2
36.8
49.8
-
-
-
0.515
-
-
NZM-1
NZM-2
NZM-3
NZM-4
NZM-5
NW
49.5
0.7
1.6
3.1
4.6
6.1
-
0.516
-
0.009
0.020
0.039
0.058
0.078
-
48.7
-
0.509
-
48.1
-
0.512
-
47.2
-
0.508
-
46.6
-
0.511
-
-
-
-
-
-
-
61.3
60.8
60.5
59.6
59.2
58.4
-
-
-
-
-
-
0.510
0.508
0.509
0.507
0.513
0.511
NZW-1
NZW-2
NZW-3
NZW-4
NZW-5
0.6
1.2
2.5
3.6
4.8
0.009
0.019
0.040
0.059
0.079
21

▼: NiO
: α-NiMoO4
a
□
○: β-NiMoO4
NM
NZM-1
NZM-2
NZM-3
NZM-4
NZM-5
10
20
30
40
50
60
70
80
2Theta (o)
▼: NiO
b
NW
NZW-1
NZW-2
NZW-3
NZW-4
NZW-5
10
20
30
40
50
60
70
80
2Theta (o)
Figure 3. XRD patterns of the oxidic precursors of Ni(Zr)MoS (a) and Ni(Zr)WS (b)
catalysts.
22

Table 2.
Textural properties of the oxidic precursors of Ni(Zr)MoS and Ni(Zr)WS catalysts.
Surface
area
(m2/g)
Pore
Surface
area
(m2/g)
Pore
Catalyst
volume
(cm3/g)
Catalyst
volume
(cm3/g)
NM
164
179
183
184
180
185
0.33
0.23
0.22
0.22
0.21
0.21
NW
143
167
170
168
171
173
0.29
0.21
0.20
0.20
0.20
0.19
NZM-1
NZM-2
NZM-3
NZM-4
NZM-5
NZW-1
NZW-2
NZW-3
NZW-4
NZW-5
23

a
580
529
NM
NZM-1
NZM-2
NZM-3
NZM-4
NZM-5
100
200
300
400
500
600
700
800
900
Temperature (oC)
b
527
690
NW
NZW-1
NZW-2
NZW-3
NZW-4
NZW-5
585
472
100
200
300
400
500
600
700
800
900
Temperature (oC)
Figure 4. TPR profiles of the oxidic precursors of Ni(Zr)MoS (a) and Ni(Zr)WS (b)
catalysts.
24

*: Ni3S2
a
*
■: MoS2
*
*
*
*
*
*
*
NM
NZM-1
NZM-2
NZM-3
NZM-4
NZM-5
10
20
30
40
50
60
70
80
2Theta (o)
*: Ni3S2
●: WS2
b
*
*
*
*
*
*
*
*
NW
NZW-1
NZW-2
NZW-3
NZW-4
NZW-5
10
20
30
40
50
60
70
80
2Theta (o)
Figure 5. XRD patterns of Ni(Zr)MoS (a) and Ni(Zr)WS (b) catalysts.
25

a
b
c
d
40
30
20
10
0
30
20
10
0
e
Nav=2.5
i
Lav=6.9nm
0
1
2
3
4
5
6
0
2
4
6
8
10 12 14 16
Slab length (nm)
20nm
20nm
20nm
20nm
Stacking number
40
30
20
10
0
30
20
10
0
f
Nav=2.5
j
Lav=7.0nm
0
1
2
3
4
5
6
0
0
0
2
4
6
8
10 12 14 16
Slab length (nm)
Stacking number
30
20
10
0
40
30
20
10
0
g
Nav=2.5
Lav=6.5nm
k
0
1
2
3
4
5
6
2
4
6
8
10 12 14 16
Slab length (nm)
Stacking number
30
20
10
0
40
30
20
10
0
h
Nav=2.5
Lav=6.5nm
l
0
1
2
3
4
5
6
2
4
6
8
10 12 14 16
Slab length (nm)
Stacking number
Figure 6. TEM images, distributions and averages of stacking number (Nav) and slab
length (Lav) of MoS2/WS2 crystallites of four sulfided catalysts: NM (a, e, i), NZM-3
(b, f, j), NW (c, g, k) and NZW-3 (d, h, l).
26

Scheme 2. The whole reaction network of quinoline hydrodenitrogenation.
27

100
80
60
40
20
0
80
60
40
20
0
a
b
DHQ
THQ-5
0
10
20
30
40
20 30 40 50 60 70 80 90 100
LHSV (h-1)
Conversion (%)
100
80
60
40
20
0
100
80
60
40
20
0
c
d
HC
PCH
PB
OPA
20 30 40 50 60 70 80 90 100
20 30 40 50 60 70 80 90 100
Conversion (%)
Conversion (%)
Figure 7. HDN activities (a) and selectivities (b, c, d) of Ni(Zr)MoS catalysts: NM
(□), NZM-1 (○), NZM-2 (△), NZM-3 (◇), NZM-4 (+), NZM-5 (×). Reactions
conditions: T = 300 oC, P(H2) = 3.5 MPa and H2/oil ratio = 600 V/V.
28

100
80
60
40
20
0
80
60
40
20
0
a
b
DHQ
THQ-5
0
10
20
30
40
30 40 50 60 70 80 90 100
LHSV (h-1)
Conversion (%)
100
80
60
40
20
0
100
80
60
40
20
0
c
d
HC
PCH
PB
OPA
30 40 50 60 70 80 90 100
30 40 50 60 70 80 90 100
Conversion (%)
Conversion (%)
Figure 8. HDN activities (a) and selectivities (b, c, d) of Ni(Zr)WS catalysts: NW (□),
NZW-1 (○), NZW-2 (△), NZW-3 (◇), NZW-4 (+), NZW-5 (×). Reactions conditions:
T = 300 oC, P(H2) = 3.5 MPa and H2/oil ratio = 600 V/V.
29