Effect of Cu-Promotion on the Performance of Molybdenum Sulfide for Hydrotreating of FCC Gasoline

Article abstract of DOI:10.1007/s10562-019-02851-2

Abstract: Copper sulfide (either supported or unsupported) has received poor attention as promoter since this chalcogenide displayed modest hydrotreatment activity. In this sense, was evaluated the potential of sulfide CuMo catalyst in the FCC gasoline hydrotreatment. A MoS₂ bulk and the industrial NiMo/Al₂O₃ catalyst were used as reference. According to the results obtained, the NiMo commercial catalyst had higher HDS, HDN, and HYD activity, but with a significant decrease in the octane number (RON). While the CuMo catalysts showed a lower response towards hydrotreatment, maintaining its octane value. According to the results obtained in the characterizations carried out (XRD, XPS, TPR, physisorption of N₂, among others), we propose that copper species could be generating geometric and/or electronic changes in the CuMoS phases, increasing the number of active sites. Additionally, the incorporation of the Cu in the crystalline structure could be modifying its electronic structure, granting a metallic character to the active phase. The work demonstrates the potential of CuMo catalyst for HDT of FCC gasoline. Graphical Abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s10562-019-02851-2

Catalysis Letters
https://doi.org/10.1007/s10562-019-02851-2
Efect of Cu‑Promotion on the Performance of Molybdenum Sulfde
for Hydrotreating of FCC Gasoline
Paulino Betancourt1 · Susana Pinto‑Castilla2
Received: 20 October 2018 / Accepted: 30 May 2019
© Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
Copper sulfde (either supported or unsupported) has received poor attention as promoter since this chalcogenide displayed
modest hydrotreatment activity. In this sense, was evaluated the potential of sulfde CuMo catalyst in the FCC gasoline
hydrotreatment. A MoS2 bulk and the industrial NiMo/Al2O3 catalyst were used as reference. According to the results
obtained, the NiMo commercial catalyst had higher HDS, HDN, and HYD activity, but with a signifcant decrease in the
octane number (RON). While the CuMo catalysts showed a lower response towards hydrotreatment, maintaining its octane
value. According to the results obtained in the characterizations carried out (XRD, XPS, TPR, physisorption of N2, among
others), we propose that copper species could be generating geometric and/or electronic changes in the CuMoS phases,
increasing the number of active sites. Additionally, the incorporation of the Cu in the crystalline structure could be modify-
ing its electronic structure, granting a metallic character to the active phase. The work demonstrates the potential of CuMo
catalyst for HDT of FCC gasoline.
Graphical Abstract
Keywords Copper-molybdenum catalysts · Hydrotreating · FCC gasoline
1 Introduction
Hydrotreating catalysts are used in numerous refnery pro-
* Susana Pinto-Castilla
cesses to remove mainly sulfur, and nitrogen contaminants
present in petroleum. The development of highly active
hydrotreating catalysts has been one of the major interests
in the petroleum industry in order to comply with increas-
ingly stringent environmental regulations. To decrease these
environmental problems, the nitrogen and sulfur content in
the fuel has been, and will continue to be, reduced to ever
lower levels. This is combined with the rapid increase of
demand for transportation fuels, which requires refneries
to produce clean fuels from available sources. In a typical
cspc99@gmail.com
Paulino Betancourt
paulino.betancourt@gmail.com
1
Facultad de Ciencias, Centro de Catálisis, Petróleo y
Petroquímica, Escuela de Química, Universidad Central
de Venezuela, Los Chaguaramos, Caracas 40679, AP,
Venezuela
2
Laboratorio de Fisicoquímica de Superfcies, Centro
de Química, IVIC, Altos de Pipe, km 11, Carretera
Panamericana, Caracas, Venezuela
Vol.:(0123456789)
1 3

P. Betancourt, S. Pinto-Castilla
refnery, gasoline comes mainly from the fuid catalytic
cracking (FCC) gasoline, which contains high olefn content
and, therefore, a high octane number [1–3]. Moreover, most
of the sulfur in the gasoline pool comes from the FCC unit.
Removal of the sulfur from gasoline is “relatively easy”,
but during the hydrodesulfurization (HDS) process using
conventional hydrotreating catalysts, saturation of the olefns
occurs in signifcant amounts, resulting in a large reduction
in the octane number.
acceptors, which converts them into p-type semiconductors,
whose bandgap (Eg) can be modulated, so that the Eg in the
CuxS increases with the decrease in the value of x. Convert-
ing copper sulfdes into interesting materials in hydrotreat-
ing reactions. An example illustrating this, occurs with the
CoMo/Al2O3–CuY catalysts prepared by physical mixing
showed in the thiophene HDS that the presence of CuY zeo-
lite has a remarkable efect in both conversion and selectiv-
ity [12], probably due to the presences of Cu2S phase in
Y-zeolite.
Therefore, the key technical problem for HDS of FCC
gasoline is to perform a deep sulfur removal and, at same
time, reduces the octane loss that occurs during the hydro-
treating process [4]. A viable method to achieve the objec-
tive is to design a catalyst with favorable activity in HDS
but with a lower octane loss. Consequently, the exploration
of an efcient catalyst to remove sulfur from olefn con-
taining feedstocks with minimum loss of octane number is
essential for the development of an active catalyst toward
refnery scale hydrotreatment for the production of “clean”
gasoline. Since the active site for the hydrogenation is dif-
ferent from that of HDS, it may be possible to control the
activity of both sites separately. Copper-based catalysts have
been scarcely investigated as potential hydrogenation/hydro-
treating catalysts over in recent decades [5–8]. It is known
that copper-containing compounds are catalytically active
for hydrogenation processes, such as methanol synthesis and
selective hydrogenation [9], although few papers deal with
copper in HDS catalysts. Among the frst references on the
use of copper in hydrotreatment is due to Harris and Chi-
anelli [10], it was found that the introduction of Cu in unsup-
ported MoS2 (by the reaction of CuCl2 with (NH4)2MoS4)
acted as a poison because the Cu decreased the electron den-
sity in Mo, which leads to the oxidization of Mo in MoS2.
From the energetic, it can be predicted that the energy of the
sulfur-metal bond is signifcantly lower for CuMoS than for
the CoMoS phase. Which means that the very high endother-
micity for the sulfur adsorption on the CuMoS system is pro-
posed as an explanation for the nonpromoting efect of Cu,
probably indicate that the copper passivates the surface of
the Mo-edge. More recently, Kibsgaard et al. [11] revealed
the formation of mixed-metal CuMoS type structures shaped
as single-layer hexagonally truncated triangular MoS2-like
nanoclusters. The atom-resolved images of the copper-pro-
moted reveal that morphology of the CuMoS phase is identi-
cal to that of CoMoS and NiMoS. However, there is a slight
degree of truncation exists in the CuMoS clusters. From the
above it can be concluded that the promoter atoms not only
change the strength of the metal sulfur bond, but also a deep
change in shape is observed as a function of promoter type.
It should be noted that copper sulfdes have diferent
stoichiometries, such as: covellite (CuS), anilite (Cu1.75S),
digenite (Cu1.8S), djurlite (Cu1.95S) and chalcocite (Cu2S).
The vacancies present in these copper sulfdes act as electron
In the work published by Liu et al. [13], the authors
showed that the Ni9.5Cu0.5Mo10 catalyst exhibited high cata-
lytic activity in the HDS of 4,6-DMDBT and FCC diesel.
The authors suggest that the catalyst (with copper) would be
suitable for the production of ultra-low sulfur diesel. In the
previous cases, a clear catalytic efect of the copper phases
is shown, it should be also considered that Cu+ ions are ef-
cient adsorbents for the removal of sulfur, because they can
form π-complexes with thiophenic aromatic rings, as DBT
and its derivates [14, 15]. Another important aspect is how
the organic sulfur is adsorbed on HDS catalysts. Usually the
adsorption occurs through the sulfur atom, and depending
on the substituents in the organic molecule near to the S
atom, there can be strong steric hindrance. Desulfurization
by HDS is extremely difcult for the least reactive deriva-
tives of DBT with methyl groups in positions adjacent to S,
such as methyldibenzothiophene (MDBT) and 4,6-dimethyl-
dibenzothiophene (DMDBT) [16].
Pointing to high hydrotreating selectivity, correspond-
ing to an important hydrodesulfurization activity and a
low hydrogenation activity for olefns is required. Thus,
hydrotreating catalysts must achieve a deep HDS with mini-
mum olefn saturation. For this purpose, a sulfded CuMo
catalyst was synthesized using solvothermal method. The
activity and selectivity of the CuMo catalyst was evalu-
ated by hydrodesulfurization of FCC gasoline. Likewise,
N2-physisorption, XRD, TG–DSC, XPS, TPR, techniques
were used to characterize the structure and physicochemical
properties.
2 Experimental
2.1 Synthesis
All manipulations were carried out under an inert N2(g)
atmosphere using glovebox techniques.
The CuMo sulfded phase with the projected Cu2MoS4
composition was synthesized using a hydrothermal method.
First of all, 40 mg of powder Cu2O (≥ 99.99%, Sigma-
Aldrich) was dispersed in 25 mL deionized water by soni-
cation for 5 min, then 110 mg (NH4)2MoS4 (99.97%, Sigma-
Aldrich) was dissolved in the solution. After stirring by
1 3

Efect of Cu‑Promotion on the Performance of Molybdenum Sulfde for Hydrotreating of FCC…
15 min, the precursor was transferred into a 100 mL Tefon-
lined stainless steel autoclave and maintained at 160 °C for
8 h. The fnal powder produced was separated from solution,
washed with deionized water and ethanol for several times
to remove any possible ions, and then dried at 80 °C under
vacuum for 12 h. This powder was treated under a 10% H2S/
H2 fow (fow rate=4 L h−1) at 300 °C by 4 h. Physical prop-
erties of the catalyst are shown in Table 1.
(UV–Vis DRS) were recorded on a Perkin-Elmer Lambda 40
spectrometer, equipped with an integration sphere (spectral
range of 280–870 nm). Before analysis, the samples were
crushed and sieved using the fraction smaller than 325 mesh,
and dried in an oven at 150 °C for 1 h. TG–DTA analysis
was carried out on an AT device under inert gas fow at a
heating rate of 5 °C min−1. The gaseous products evolved
on heating of the samples were studied using a mass spec-
trometer (Pfeifer Instruments) equipped with a quadrupole
analyzer. In the TPR experiments, the reduction step was
performed with an Ar/H2 mixture, with a heating rate of
10 °C min−1, up to 700 °C. The change in H2 concentra-
tion was followed by a mass spectrometer Omnistar 300,
Pfeifer vacuum. To calculate the absolute values of the H2
consumed, mass spectrometer was calibrated by reduction of
known amount of CuO (0.16–0.315 mm). In addition, prior
to each measurement the sample was preheated in a stream
of Ar at 500 °C for 30 min to remove adsorbed water. The
amount of NO uptake was determined by nitrogen monox-
ide (NO) isotherms. The sample was sulfded in a 5% H2S/
H2 stream at 350 °C for 2 h. After the sulfdation, the sam-
ple was reduced in a hydrogen stream for 1 h, followed by
purging by helium at the same temperature for 5 min. After
these pretreatments, the NO adsorption isotherms represent-
ing both physisorption and chemisorption were recorded at
25 °C. The catalyst was evacuated for 1 h at 1.2×10−7 MPa
vacuum to remove the physisorbed NO and the second iso-
therm representing only the physisorbed NO was generated
in an identical manner. From the diference of these two lin-
ear and parallel isotherms, the amount of chemisorbed NO
was determined. N2 adsorption and desorption isotherms of
the sulfde solids (after presulfdation) were measured using
a Micromeretics ASAP 2010c system. Prior to the experi-
ments, the samples were degassed (P <10−1 Pa) at 150 °C
for 6 h. Specifc surface areas (SSA) were calculated by the
BET method, the total pore volume (TVP), and average pore
diameter (APD) were determined by nitrogen adsorption at
a relative pressure of 0.98 and pore size distributions from
the adsorption isotherms by the BJH method.
The catalytic response of the Cu2MoS4 was compared
with the extended studied MoS2 phase. The solid was pre-
pared from in situ by thermal decomposition of tetrabutyl-
ammonium thiomolybdate in a Parr reactor by heating the
solid to 400 °C in a H2/H2S 15% gas mixture [17]. The MoS2
presents a mixture of relatively small particles and sintered
particles. Length and stacking vary from three to 49 nm, and
from 3 to 16 layers, respectively.
An industrial NiMo/Al2O3 hydrotreatment catalyst
(extrudes of 1.6 mm in diameter and 5 mm in averaged
length) was used, and its characteristics are given in Table 1.
This was sulfded under the same conditions as the unsup-
ported solids.
2.2 Characterization of the Catalysts
The elemental contents of Cu and Mo were determined by
plasma atomic emission method (ICP). Some standardized
dissolution procedures were necessary to dissolve the sam-
ple. The structure of the samples was examined by X-ray
powder difraction analysis on a Bruker D-8 difractom-
eter using Cu Kα radiation (λ=1.5406 Å) at a scan rate of
2° min−1. The difractograms obtained were compared using
the standard JCPDS fles. The XPS spectra were recorded on
a VG Instrument type ESCALAB 200 XLi system. The sam-
ple excitation was done by Al Kα X-rays (1486.6 eV). The
intensities were estimated by calculating the integral of each
peak after smoothing and subtracting the “S-shaped” back-
ground and ftting the experimental curve to a combination
of Gaussian and Lorentzian distributions. Peak shifts due to
charging of the sample were corrected by taking the C1s line
at 284.6 eV as reference. EDX analysis was obtained with a
JEOL 2010 microscope. UV–vis difuse refectance spectra
2.3 Catalytic Activity Evaluation
In order to examine the performance of the catalyst, compar-
ative tests were carried out for CuMo, MoS2 and a commer-
cial-supported NiMo catalyst. Catalyst activity tests were
conducted in a 20 mL continuous fow fxed-bed reactor with
FCC gasoline (Table 2). The catalysts were sulfded in situ
before the evaluation test under the following conditions:
the sulfdation liquid stream containing 10 wt% CH3S2 in
heptane, the pressure was 0.91 MPa, the hydrogen to oil vol-
ume ratio was 150:1, the volume space velocity was 1.6 h−1,
and the temperature was raised to 300 °C at a heating rate
of 10 °C min−1 and maintained for 2 h at this temperature.
Table 1 Physical and chemical properties of sulphided CuMo and
NiMo catalysts
Property
CuMo
MoS2
NiMo/Al2O3
S/Moa
1.6
29
0.10
7.7
4.1
1.7
33
0.11
3.71
2.4
2.1
SSA (m2 g−1
)
211
0.48
11.2
5.2
TPV (cm3 g−1
APD (nm)
)
NO adsorption (mL g−1
aXPS
)
cat
1 3

P. Betancourt, S. Pinto-Castilla
Table 2 Properties of Venezuelan FCC gasoline
Table 3 Efect of catalysts on hydrotreatment reactions
Feedstock CuMo MoS2 NiMo/Al2O3
Feedstock
Density (g cm−3
)
0.771
158.9
5.3
1.93
102.4
90.4
79.6
Density (g cm−3
Sulfur (ppm)
%HDS (total)
)
0.771
158.9
–
0.771 0.766 0.768
121.4 141.4 65.2
Sulfur (ppm)a
Nitrogen (ppm)b
H/C
Average molecular weight
RON
23.6
11.0 58.9
% conversion S-com-
pounds
BT
DBT
4,6-DMDBT
%HDN
H/C
RON
MON
Hydrocarbon (vol%)
Parafns
Isoparafns
Olefns
Naphtenes
Aromatics
–
–
–
–
1.93
90.4
79.6
42.1
24.3
7.4
18.9 73.0
9.5
8.0
57.0
19
92.4
MON
Hydrocarbon (vol%)c
Parafns
3.0
7.3
4.8
1.88
89.0
82.2
1.86
88.9 85.6
81.6 79.3
1.84
Isoparafns
Olefns
Naphtenes
32.4
19.7
12.4
30.8
Aromatics
4.8
7.3
–
–
–
–
–
9.6
41.8
3.8
14.8
30.1
aASTM D4294-98, bASTM D4629-02, cASTM D5443M
32.4
19.7
12.4
30.8
36.9
12.3
13.4
30.1
When the presulfding process was over, the evaluation test
started. The reaction operating conditions were 3.1 MPa,
350 °C, LHSV of 3.53 h−1, and H2/oil ratio of 150 (v/v).
The reaction products were condensed and collected peri-
odically from a gas–liquid separator, and the conversion
values reported here are those obtained when the steady
state is reached. A carbon balance between the reactor feed
and efuent showed closures of > 95% for all of the cata-
lyst activity data reported herein. The nitrogen and sulfur
were analyzed according to ASTM D4629-02 and ASTM
D4294-98 methods, respectively. An HP 6890-series gas
chromatograph, equipped with an atomic emission detector
was used to analyze individual sulfur species in the gasoline
feed and hydrotreated products. The total aromatic content
was determined by the standard method of fuorescence
indicator adsorption. The octane number was determined by
gasoline engine testing. The hydrocarbon composition and
octane number (GC-RON and GC-MON) in feedstock and
products were analyzed by using a 60 m PIONA (parafn,
isoalkane, olefn, naphthene, and aromatics) silica capillary
column (Agilent 6890 gas chromatograph equipped with an
atomic emission detector).
(19.7 vol%), saturates (37.2 vol%), and aromatics (naphtenes
and aromatics, 43.2 vol%).
Catalyst activity test results under the operating condi-
tions are listed in Table 3, indicating that CuMo, MoS2 and
NiMo/Al2O3 had diferent HDS and HDN conversions. The
blank product characteristics are close to those of the feed. It
is noticeable that was little aromatic saturation, or change in
the ratio of H/C achieved by the catalytic solid. The results
display that the commercial catalyst NiMo has higher HDS,
HDN, and hydrogenation (HYD) activity, but lower octane
number (RON), while CuMo catalyst show that it have lower
hydrotreating activity, but have higher RON compared with
commercial catalyst. Hence, the octane number of gasoline
treated by CuMo remains practically unchanged compared
with the FCC gasoline feedstock. Ruette et al. [21] evaluated
the role of Cu in the hydrodenitrogenation (HDN) of pyri-
dine on molybdenum sulfde surface with a CNDO method
reparameterized for diatomic binding energies. They indi-
cate that the Mo–S bonds are stronger in Mo2CuS than in
Mo2S, which does not facilitate adjacent vacancy forma-
tion around the adsorption site and prevents the pyridine
hydrogenation. Probably, we are in a similar case, where
the incorporation of copper decreases the hydrogenation of
olefns. The catalytic activity of CuMo catalyst is related to
surface defects in the crystal lattice. However, the proper-
ties and stability of these defects are determined by the bulk
atomic and electronic structure. Thus, it is crucial to under-
stand the catalytically stable phases in order to understand
the catalytic activity that it exhibit. The CuMo catalyst was
characterized with respect to ICP, BET surface area, pore
3 Results and Discussion
Previous reports for a similar FCC gasoline [18, 19], has
pointing out that thiophenic, benzothiophenic and diben-
zothiophenic with diferent numbers of substituent chains
are presents in the sample (see Table 3). These compounds
may have diferent origins, according to Lefaive et al. [20]
come from the cracking of long chain alkyl thiophenes
and the addition of H2S to olefns or diolefns followed by
cyclization. As shown in Table 2, gasoline contains olefns
1 3

Efect of Cu‑Promotion on the Performance of Molybdenum Sulfde for Hydrotreating of FCC…
volume and pore size measurement, TPR, TGA, XRD, and
XPS to study the physicochemical properties of the catalyst.
The detailed discussion on these analyses is given below.
The elemental compositions catalysts were measured by
ICP. The result of elemental analysis for the sample syn-
thesized was Cu2.35MoS3.96, which shown that there were
a few defects in the crystal structure. The crystal structure
of CuMo was analyzed by XRD (Fig. 1a). Studies showed
that difraction peaks could indicate a tetragonal phase of
Cu2MoS4. Chen et al. [22], reported the formation of a simi-
lar type of mixed phase formation of Cu2MoS4, which was
also synthesized by hydrothermal treatments. The crystal
structure of the Cu2MoS4 material consists of sheets, sep-
arated by 2.49 Å along the c-axis, formed by interlinked
copper and molybdenum atoms tetrahedrally coordinated to
bridging sulfur atoms. No evidence was found to indicate
the presence of copper oxides/sulfdes or MoS2 segregated
phases, which indicates that the catalyst possess high purity.
Signifcant changes were observed in the XRD pattern
of CuMo after hydrotreating reaction (Fig. 1b). Despite
maintaining the Cu2MoS4 phase, the XRD pattern reveals
the presence of a small amount of MoS2 phase. No copper-
containing phase was observed, but it is not ruled out that
the copper sulfde entities are segregated and dispersed over
the molybdenum sulfde, probably at the edges of the MoS2
fringes. Besides the morphology and crystal structure of the
MoS2, it appears that the MoS2 phase has a peak at 14.39°,
which implies a displacement of the maximum of the MoS2
002 peak from 2θ = 14.48°. The change of the frst XRD
peak to a lower angle indicates an expansion of the distance
between layers in the MoS2.
Chemical composition of this material was investigated
by X-ray photoelectron spectroscopy (XPS) analysis (see
Table 4). The full XPS survey spectrum revealed the pres-
ence of the elements Cu, Mo, and S and trace amounts of
C and O. For Mo3d5/2 and Mo3d3/2 were found at values of
229.8 and 233.0 eV which are characteristic of Mo4+ ions
in compounds of the MoS2 type [23–26]. Additionally, the
peak of Mo3d exhibit a slight shifts compared to pure MoS2,
providing clear evidence of the interaction between copper
and Mo, in CuMo catalyst. The nearby S2s peaks were ft-
ted into two peaks at 225.7 and 224.2 eV, suggesting two
chemical states of S species bonding to the Mo and Cu ions.
The S2p peak at binding energies of 164.3 and 162.3 eV
are attributed to the S2p1/2 and S2p3/2 orbitals of the diva-
lent sulfde ions (S22−), in accordance with the formation
of MoS2-type phase. Elemental or oxidized sulfur species
2−
(SO3 or SO42−) must be discarded since they appear at
substantially higher binding energies. While the Cu2p3/2 and
Cu2p1/2 signals were found at values of 933.2 and 953.4 eV.
No satellite bands indicative of high-binding energy were
observed (typically at 940–943 eV for the Cu2p3/2 peak);
confrming that copper is in its Cu(I) [27]. This clearly
shows that no redox changes occurred during the preparation
of CuMo catalyst. The sulfurization degree of Mo species
has been calculated based on the area of the XPS peak, as
shown in Table 4. There are slight diferences in the sul-
furization degree by copper incorporation, respect to MoS2.
It is worth noting that no transformation was observed in
the XPS analysis after exposing the sample to air, it can be
considered in this case as an air-stable material. In addition,
the quantitative calculations in the XPS spectrum of CuMo
catalyst indicated that the molar ratio of Cu in Cu2MoS4
(30.4%) was close to its stoichiometric value (28.6%). The
energy dispersive X-ray (EDX) elemental mapping results
(not shown) for CuMo suggest that these three elements in
the samples are distributed homogeneously.
Figure 2 shows UV–Vis absorption spectra of sulfded
CuMo prepared by hydrothermal synthesis method. The
frst absorption can be associated with the d–d transitions of
Cu(I), while the latter can be attributed to the charge transfer
from S2− to Cu(I) or from S2− to Mo(IV). In the spectrum,
a strong continuous absorption was observed in the whole
wavelength region studied. This could mean that this solid
has metallic properties instead of semiconducting that are
common in transition metal sulfdes.
Fig. 1 XRD patterns: a sulfded CuMo catalyst (red pattern), and b
the same catalyst after gasoline HDS test (blue pattern)
Table 4 Spectral parameters
Catalyst
Binding energy (eV)
Cu 2p1/2 Cu 2p3/2 Ni 2p3/2 Mo 3d3/2 Mo 3d5/2 S 2p1/2 S 2p3/2
Mosulf/MoTotal
obtained by XPS analysis
CuMo
MoS2
NiMo/Al2O3
953.4
–
933.2
–
–
–
–
233.1
232.3
232.7
229.7
229.1
228.8
162.6
163.2
–
161.6
162.0
–
0.76
0.88
–
853.1
1 3

P. Betancourt, S. Pinto-Castilla
370–440 °C (near the boiling point of sulfur, 443.6 °C) and
the completed decomposition at about 520 °C. This indicates
that sulfur is found in the sample not in one, but in several
chemically bonded states, as indicated by the following TPR
results.
Temperature programmed reduction analysis was carried
out to determine the nature of the diferent sulfdes species
present on the catalyst surface, as well as to study the reduc-
ibility of the catalyst. Due to the limitations of the instru-
ments and the common temperature used for the produc-
tion of FCC gasoline (temperature generally no more than
400 °C), the TPR data were collected before 600 °C.
The CuMo sulfde catalyst has two peaks below 600 °C,
one at 277 °C and other one above 400 °C (Fig. 4a). For the
MoS2 (Fig. 4b) catalyst, in the low-temperature region of
25–250 °C a hydrogen-consumed peak has been assigned
to hydrogenation, of a well-defned and highly reactive
Fig. 2 UV–Vis absorption spectra of the CuMo catalyst
The method of combined thermogravimetry and difer-
ential scanning calorimetry with mass spectrometry of the
volatile products of the samples allowed the detection of the
various features of the catalyst (Fig. 3). Under the conditions
of analysis, thermal oxidation occurs both on the surface and
in the bulk of the sample. The total mass losses were near
70 wt%. The volatile composition of the product shown by
the mass peaks corresponded mainly to the ratio m/z=18
(H2O) and 64 (SO2). The initial mass loss was monotonous
without noticeable steps. CuMo bulk solid loses interbed-
ded water at approximately 218 °C. This may be due to the
structural binding of H2O, in particular within the interlayer
space of the MoS2 nanocrystallites. During decomposition
of the CuMo solid, two smooth endothermic events were
observed, one at 265 °C and the other close to 360 °C. Then
a broad and strongly exothermic event occurred in the range
Fig. 3 Thermogravimetry and diferential scanning calorimetry of the
CuMo catalyst
Fig. 4 H2-TPR profles of the sulfded solids: a CuMo and b MoS2
1 3

Efect of Cu‑Promotion on the Performance of Molybdenum Sulfde for Hydrotreating of FCC…
S species located on the edge sites of MoS2 crystallites
(weakly bonded sulfur). Which are believed to be responsi-
ble for the active sites [28]. In the high-temperature region
of 250–600 °C, a further hydrogen-consumption peak was
regarded as being related to the reduction of stoichiometric
sulfde sulfur.
turnover numbers (number of reaction events per unit time
and per active site) in other areas of heterogeneous catalysis
[39], we chose to compare the catalysts on an active site
basis.
From the above results, it is known that two parameters
are relevant to describe the promotional behavior: (i) a geo-
metric parameter that is related to the relative number of
promoted sites in the CuMoS, and (ii) changes in the elec-
tronic structure or in the binding energy of the sulfur on
metal sites.
The main peak of MoS2 was observed to be centered at
154 °C, whereas the peak was at a much higher tempera-
ture, 290 °C in CuMoS, indicating that copper interacts with
molybdenum and in consequence, the reducibility of CuMoS
was higher than that of MoS2. The hydrogen consumption
peaks were due to the hydrogenation of weakly bonded S
species in copper and molybdenum sulfdes [29, 30], which
was corroborated by H2S amounts detected in the mass spec-
troscopy coupled to TPR. Schefer and Mangnus [31, 32]
have suggested that the hydrogenation of weakly bonded
S species is present in the highly active CoMoS phase.
Removal of the S species could generate coordinatively
unsaturated sites, which are essential for catalytic activity.
At this point, it is also necessary consider the reduction of
Cu and Mo species, would occur simultaneously to sulphur
vacancy creation.
The insertion of Cu strongly afects the electronic struc-
ture in the CuMo catalyst, which improves the interaction
and may be the reason for the exhibited metallic character.
The observed presence of promotion by Cu would be
explained as an intrinsic efect of CuMoS S-edge, which we
tentatively attribute to a weak Cu–S interaction and good H
adsorption strength. It means that copper improves the acti-
vation of hydrogen as a center of adsorption, and that copper
also increases the ability of sulfur atoms to bind protons.
We speculate that variations of the bonding of H on the
CuMoS S-edge could be a dominant factor that determines
both the DDS reaction and the HYD reaction, since too
weak/strong bonding of H would signifcantly hamper the
key reaction steps involving the H transfer from the crystal
edges.
To examine the effect of copper in more detail, NO
uptake in the catalysts was measured. The results are shown
in Table 1, for the CuMo catalyst the NO adsorption was
4.1 mL g−1. On the other hand, the adsorption of MoS2 was
of 2.4 mL g−1, which is notoriously lower than CuMoS cata-
lyst. This revealed that the copper atoms are located at the
edges sites. The notorious lower quantity of MoS2–NO in
comparison to CuMoS–NO revealed that the copper atoms
are located at the edges sites. From Topsøe studies [33], NO
adsorption shows that there is a correlation between NO
on the promoter (Co) sites and the HDS catalytic activity.
Similar to our results using sulfde CuMo catalyst [34–36].
The nitrogen adsorption–desorption isotherm of the
sulfde catalyst showed that the adsorbed volume is sub-
stantially smaller, presenting a microporosity contribution.
Based on the Brunauer–Emmett–Teller (BET) method, the
surface area of CuMo was estimated at 29 m2 g−1. With
respect to MoS2 (33 m2 g−1), the surface area of the CuMo
catalyst was slightly lower. The decrease in surface area in
the sulfde catalyst after the addition of the promoter to the
sulfde catalysts of Mo is also reported by other investiga-
tions [37, 38]. The surface area of unsupported Mo sulfde
catalysts was affected by copper, which influenced the
morphology and/or the degree of aggregation of the CuMo
phase.
It is expected that this paper will provide an approach
for the exploration of a new hydrotreating catalyst with
the desired properties for commercial use. However, up
to now many questions about the precise mechanisms of
hydrodesulfurization catalysts based on CuMo have not yet
been answered. More theoretical and experimental research
should be carried out in the future.
4 Conclusions
The present study provides a new view of the struc-
ture–activity relationship and an approach to the control
of activity using an unsupported CuMo sulfde catalyst by
hydrothermal preparation conditions. Its catalytic response
shows moderate conversions in the hydrotreatment of FCC
gasoline; however, the decrease in the octane number is
lower than the industrial NiMo/Al2O3 catalyst. Conse-
quently, this research provides a novel CuMo catalyst pro-
moted by copper that can be used for the high performance
HDS reaction. We propose that geometric and electronic
modifcations in the crystalline structure could cause an
improvement in the catalytic response. Nevertheless, it must
be further investigations to understand the synergistic efect
of Cu and its reaction mechanism.
An additional critical issue for the practical application of
HDT catalyst is their activity. Because we were not testing
preparations of copper–molybdenum catalyst with optimal
surface area, it was important to try to compare the activi-
ties of these materials on a basis that was independent of
the surface area. Given the successful use of the concept of
Acknowledgements Authors thanks PDVSA for the FCC gasoline
samples used in this paper.
1 3

P. Betancourt, S. Pinto-Castilla
21. Munaretto A, Ruette F, Sánchez M, Rodtfguez-Arias E (2006)
Ciencia 14:83–95
References
22. Chen W, Chen H, Zhu H, Gao Q, Luo J, Wang Y, Zhang S, Zhang
K, Wang C, Xiong Y, Wu Y, Zheng X, Chu W, Song L, Wu Z
(2014) Small 10:4637–4644
1. Liu J, Zhao Sh, Chen X, Shen B (2016) Fuel 166:467–472
2. Nadeina KA, Klimov OV, Pereima VY, Koryakina GI, Danilova
IG, Prosvirin IP, Gerasimov EY, Yegizariyan AM, Noskov AS
(2016) Catal Today 271:4–15
23. He J, Chen L, Wang F, Liu Y, Chen P, Au CT, Yin SF (2016)
ChemSusChem 9:624–630
3. Viswanadham N, Negi BS, Garg MO, Sundaram M, Sairam B,
Agarwal AK (2007) Fuel 86:1290–1297
24. Scofeld J, Hartree-Slater J (1976) J Electron Spectrosc Relat Phe-
nom 8:129–137
4. Brunet S, Mey D, Pérot G, Bouchy C, Diehl F (2005) Appl Catal
A 278:143–172
25. Feduschak T, Akimov A, Morozov M, Uymin M, Zaikovskii V,
Prosvirin I, Vosmerikov A, Zhuravkov S, Vlasov V, Kogan V
(2016) C R Chim 19:1315–1325
5. Busca G (2014) Metal catalysts for hydrogenations and dehydro-
genations. Heterogeneous catalytic materials. solid state chem-
istry, surface chemistry and catalytic behaviour, vol 9. Elsevier,
Amsterdam, pp 297–343
26. Moulder JF, Stickle WF, Sobol PE, Bomben KD (1992) Handbook
of X-ray photoelectron spectroscopy. Perkin-Elmer Corporation
Physical Electronics Division, Eden Prairie
6. Chen B, Dingerdissen U, Krauter JGE, Lansink Rotgerink HGJ,
Möbus K, Ostgard DJ, Panster P, Riermeier TH, Seebald S, Tacke
T, Trauthwein H (2005) Appl Catal A 280:17–46
7. Ishutenko D, Mozhaev A, Salnikov V, Nikulshin P (2016) React
Kinet Mech Catal 119:615–627
27. Yin M, Wu CK, Lou Y, Burda C, Koberstein JT, Zhu Y, O’Brien
S (2005) J Am Chem Soc 127:9506–9511
28. Afanasiev P (2006) Appl Catal A 303:110–115
29. Rodríguez-Castellón E, Jiménez-López A, Eliche-Quesada D
(2008) Fuel 87:1195–1206
8. Nikulshin PA, Tomina NN, Pimerzin AA, Stakheev AY, Mashko-
vsky IS, Kogan VM (2011) Appl Catal A 393:146–152
9. Zhao YF, Yang Y, Mims C, Peden C, Li J, Mei D (2011) J Catal
281:199–211
30. López Cordero R, López Agudo A (2000) Appl Catal A 202:23–35
31. Schefer B, Dekker NJJ, Mangnus PJ, Moulijn JA (1990) J Catal
121:31–46
32. Mangnus PJ, Riezebos A, Vanlangeveld AD, Moulijin JA (1995)
J Catal 151:178–191
10. Harris S, Chianelli RR (1986) J Catal 98:17–31
11. Kibsgaard J, Tuxen A, Knudsen KG, Brorson M, Topsøe H,
Lægsgaard EJ (2010) J Catal 272:195–203
33. Topsøe NY, Tuxen A, Hinnemann B, Lauritsen JV, Knudsen KG,
Besenbacher F, Topsøe H (2011) J Catal 279:337–351
34. Yao HC (1981) J Catal 2:440–444
12. Boukoberinea Y, Hamada B (2016) Arab J Chem 9:S522–S527
13. Liu H, Yin C, Li H, Liu B, Li X, Chai Y, Li Y, Liu C (2014) Fuel
129:138–146
35. Topsoe NY, Topsoe H (1993) J Catal 139:641–651
36. Sobczynski A, Zmierczak W (1991) React Kinet Catal Lett
44:511–516
14. Hernández-Maldonado AJ, Yang RT (2004) J Am Chem Soc
126:992–993
37. Yoosuk B, Hyung Kim J, Song C, Ngamcharussrivichai C, Prasas-
sarakich P (2008) Catal Today 130:14–23
15. Dai W, Zhou YP, Li SN, Li W, Su W, Sun Y, Zhou L (2006) Ind
Eng Chem Res 45:7892–7896
38. Wang W, Zhang K, Li L, Wu K, Liu P, Yang Y (2014) Ind Eng
Chem Res 53:19001–19009
16. Meille V, Schulz E, Lemaire M, Vrinat M (1997) J Catal
170:29–36
39. Boudart M, Djega-Mariadassou G (1984) Kinetics of heterogene-
ous catalytic reactions, vol 4. Princeton University Press, Prince-
ton, pp 118–154
17. Trakarnpruk W, Seentrakoon B (2007) Ind Eng Chem Res
46:1874–1882
18. Hatanaka S, Yamada M, Sadakane O (1997) Ind Eng Chem Res
36:1519–1523
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional afliations.
19. Cheng WC, Kim G, Peters AW, Zhao X, Rajagopalan K, Ziebarth
MS, Pereira CJ (1998) Catal Rev Sci Eng 40:39–79
20. Lefaive P, Lemberton JL, Pérot G, Mirgain C, Carriat JY, Colin
JM (2002) Appl Catal A 227:201–215
1 3