Hydrocracking vegetable oil on borate-containing catalysts: Effect of nature of support

Article abstract of DOI:10.1134/S0023158417050032

The structure, texture, and acid properties of platinum catalysts on oxide (Al₂O₃, ZrO₂, ZrO₂–Al₂O₃) and borate-containing supports (B₂O₃–Al₂O₃, B₂O₃–ZrO₂) are studied. The catalysts are tested in the process of hydrocracking sunflower-seed oil at 380°C, 4.0 MPa, and a weight stock feed rate of 1.0 h^–1. It has been found that aluminum oxide (A) contains the γ-Al₂O₃ phase, zirconium dioxide (Z) includes 85 and 15 rel. % of the monoclinic (M) and tetragonal (T) phases, respectively, while zirconium dioxide with the addition of 2.5 wt % Al₂O₃ (ZA) comprises 14 and 86 rel. % of the M–ZrO₂ and T–ZrO₂ phases, respectively. The B₂O₃–Al₂O₃ (BA) and B₂O₃–ZrO₂ (BZ) systems modified with boron oxide (20 wt %) are X-ray amorphous. A Pt/BA catalyst differs from a Pt/A catalyst, while a Pt/BZ catalyst has a larger specific surface area and acidity than Pt/Z and Pt/ZA catalysts and contains Bronsted acidic centers (BACs) along with Lewis acidic centers (LACs). Only LACs are present on the surface of Pt/A, Pt/Z, and Pt/ZA catalysts. The LAC/BAC ratio in Pt/BA and Pt/BZ catalysts is 0.3 and 1.0, respectively. All the catalysts provide complete oil conversion to give C₅₊ hydrocarbons with a yield of 81.7–87.3 wt %. Pt/A catalyzes mainly decarboxylation and hydrogenation–dehydration reactions, while Pt/Z and Pt/ZA provide decarboxylation. The yield of diesel fraction reaches 71.8–73.9 wt % with an n-alkane content of 94.0–95.9 wt %. One-stage oil hydrocracking with the prevalence of hydrodecarbonylation and hydrogenation–dehydration reactions occurs on Pt/BA and Pt/BZ catalysts for 20 h to give the yield of the diesel fraction of at least 81.4 and 74.4 wt % and the total content of iso-alkanes and cycloalkanes of at least 28.3 and 60.7 wt %, respectively.

Full text of DOI:10.1134/S0023158417050032

ISSN 0023-1584, Kinetics and Catalysis, 2017, Vol. 58, No. 5, pp. 563–576. © Pleiades Publishing, Ltd., 2017.
Original Russian Text © Yu.A. Chumachenko, M.V. Trenikhin, V.P. Talzi, T.I. Gulyaeva, E.A. Paukshtis, 2017, published in Kinetika i Kataliz, 2017, Vol. 58, No. 5, pp. 574–588.
Hydrocracking Vegetable Oil on Borate-Containing Catalysts:
Effect of Nature of Support
Yu. A. Chumachenkoa, *, M. V. Trenikhina, b, V. P. Talzia, T. I. Gulyaevaa, and E. A. Paukshtisc
aInstitute of Hydrocarbon Processing, Siberian Branch, Russian Academy of Sciences, Omsk, 644040 Russia
bOmsk Scientific Center, Siberian Branch, Russian Academy of Sciences, Omsk, 644024 Russia
cBoreskov Institute of Catalysis, Siberian Branch, Russian Academy of Sciences, Novosibirsk, 630090 Russia
* e-mail: juliana@ihcp.ru
Received December 29, 2016
Abstract—The structure, texture, and acid properties of platinum catalysts on oxide (Al2O3, ZrO2, ZrO2–
Al2O3) and borate-containing supports (B2O3–Al2O3, B2O3–ZrO2) are studied. The catalysts are tested in the
process of hydrocracking sunflower-seed oil at 380°C, 4.0 MPa, and a weight stock feed rate of 1.0 h–1. It has
been found that aluminum oxide (A) contains the γ-Al2O3 phase, zirconium dioxide (Z) includes 85 and
15 rel. % of the monoclinic (M) and tetragonal (T) phases, respectively, while zirconium dioxide with the
addition of 2.5 wt % Al2O3 (ZA) comprises 14 and 86 rel. % of the M–ZrO2 and T–ZrO2 phases, respectively.
The B2O3–Al2O3 (BA) and B2O3–ZrO2 (BZ) systems modified with boron oxide (20 wt %) are X-ray amor-
phous. A Pt/BA catalyst differs from a Pt/A catalyst, while a Pt/BZ catalyst has a larger specific surface area
and acidity than Pt/Z and Pt/ZA catalysts and contains Bronsted acidic centers (BACs) along with Lewis
acidic centers (LACs). Only LACs are present on the surface of Pt/A, Pt/Z, and Pt/ZA catalysts. The
LAC/BAC ratio in Pt/BA and Pt/BZ catalysts is 0.3 and 1.0, respectively. All the catalysts provide complete
oil conversion to give C5+ hydrocarbons with a yield of 81.7–87.3 wt %. Pt/A catalyzes mainly decarboxyl-
ation and hydrogenation–dehydration reactions, while Pt/Z and Pt/ZA provide decarboxylation. The yield
of diesel fraction reaches 71.8–73.9 wt % with an n-alkane content of 94.0–95.9 wt %. One-stage oil hydro-
cracking with the prevalence of hydrodecarbonylation and hydrogenation–dehydration reactions occurs on
Pt/BA and Pt/BZ catalysts for 20 h to give the yield of the diesel fraction of at least 81.4 and 74.4 wt % and
the total content of iso-alkanes and cycloalkanes of at least 28.3 and 60.7 wt %, respectively.
Keywords: one-stage hydrocracking, vegetable oil, supported platinum catalysts, borate-containing alumi-
num oxide, borate-containing zirconium dioxide, renewable diesel fuel
DOI: 10.1134/S0023158417050032
The stricter requirements on the performance of ing components of this group of catalysts are fixed on
diesel fuel and in conditions of limited global oil the surface of (a) inert (SiO2 [15] or Sibunit [16]), or
reserves favors the involvement of renewable raw
materials based on vegetable oil [1], deep fat, and ani-
mal fat [2, 3] in the hydrogenation processes for man-
ufacturing diesel fuel. The green diesel obtained by
lipid (fatty acid triglycerides) hydrocracking is charac-
terized by the absence of sulfur and aromatic com-
pounds, resistance to oxidation, a high cetane number
(up to 80 points), and lowered cloud point [4].
(b) acidic (SAPO-11 [17], SAPO-31 [18], Al-SBA-15
[19], HY, ZSM-5, USY, β [20], SiO2–Al2O3 [21])
materials.
In the presence of catalysts related to groups 1 and 2a,
the process of triglyceride hydrocracking includes the
stages of primary hydrocracking (HC) of triglycerides
(TGs) into propane and carboxylic acids (abbreviated
form HC-TGs, designated below as reaction (I)),
hydrodeoxygenation (HDO) of carboxylic acids by
decarboxylation (DCO2)
The catalysts used for triglyceride hydrocracking
can be divided into the following groups:
(1) Massive systems containing only the hydrogenating
component (NiMoS, CoMoS [5], Raney Ni [6]);
(2) Supported systems containing a hydrogenating
component in the form of the Co(Ni)–Mo(W)–S
phase [7, 8] or disperse particles (Group VII metals Pt,
Pd, Ni [9–11], phosphide Ni2P [12], carbide Mo2C
[13], and nitride Mo2N [14]). The listed hydrogenat-
Rх–COOH → Rх–H + CO2,
hydrodecarbonylation (HDCO)
Rх–COOH + H2 → CO + H2O + Rх–H,
hydrogenation–dehydration (H–DH2O)
Rх–COOH + 3H2 → 2H2O + Rх–CH3,
(II)
(III)
(IV)
563

564
CHUMACHENKO et al.
and hydrogenolysis (HH) of C–C bonds in hydrocar-
bons Rх–H and Rх–CH3 (reaction (V)) [22].
The calcination of pseudobemit at 550°C for 16 h
resulted in Al2O3. A zirconium oxynitrate solution, as
well as an aluminum nitrate and zirconium oxynitrate
solution, heated to 90°C was treated with an ammonia
solution to precipitate zirconium dioxide hydrate or
mixed hydroxide, respectively, until the pH of the sus-
pension reached 10.5. The resultant precipitates were
aged in the mother liquor for 1 h, separated by filtra-
tion, washed with distilled water until the pH of the fil-
trates became neutral, and dried at 120°C. The calci-
nation of zirconium dioxide hydrate and mixed
hydroxide at 600°C for 16 h resulted in ZrO2 and
ZrO2–Al2O3 supports, respectively. The calculated
content of aluminum oxide in the latter was 2.5 wt %.
As the linear ester fragments Rх-COO–, contain-
ing 16 and 18 carbon atoms (х = 15 and 17), are preva-
lent in a triglyceride structure, hydrocracking over cat-
alysts of groups 1 and 2a leads to mixtures of normal
hydrocarbons with an enhanced fraction of C15–C18
components. Obviously, these mixtures could not be
used as diesel fuel because of their poor low-tempera-
ture properties. The first way to solve this problem
consists in using the platinum catalysts of group 2b for
the hydroisomerization of the hydrocarbon product
obtained on the group 1 or 2a catalysts. A second,
more promising way is to conduct one-stage hydroc-
racking of triglycerides over the group 2b catalysts. In
this case, the transformations of hydrocarbons Rх–H
A suspension of pseudobemit was impregnated
with heat to 90°C in an aqueous solution of orthoboric
and Rх–CH3 via cracking (VI) and isomerization acid with a concentration of 110 g/L and evaporated at
(VII) proceed along with reactions (I)–(V).
100°C. The dry residue was calcined at 550°C for 16 h
to prepare the B2O3–Al2O3 support. A suspension of
zirconium dioxide hydrate containing ~23 g of ZrO2
heated to 90°C was mixed with 10.2 g of solid orthob-
oric acid and evaporated at 100°C. The calcination of
the solid residue at 600°C for 16 h led to the prepara-
tion of the B2O3–ZrO2 support. The calculated con-
tent of boron oxide in B2O3–Al2O3 and B2O3–ZrO2
corresponded to 20 wt %.
In the design of new efficient catalysts for manu-
facturing green diesel, their accordance with environ-
mental safety requirements, economy of synthesis,
and concurrent accomplishment of reactions (I)–(IV)
and (VII) has fundamental significance. In contrast to
the group 2b catalysts, a bifunctional system contain-
ing 0.5 wt % platinum and borate-containing alumi-
num oxide (B2O3–Al2O3) meets all these requirements
[23]. Another bifunctional system based on borate-
containing zirconium dioxide (B2O3–ZrO2) with the
same platinum content also ensures the isomerization
of n-alkanes [24]. The B2O3–ZrO2 and B2O3–Al2O3
systems are known to exhibit a higher activity than the
corresponding ZrO2 and Al2O3 oxides in acid-con-
trolled reactions of the dehydration of alcohol [24, 25],
caprolactam synthesis [26, 27], etc.
The Al2O3, ZrO2, ZrO2–Al2O3, B2O3–Al2O3, and
B2O3–ZrO2 supports are designated hereinafter as A,
Z, ZA, BA, and BZ, respectively.
Preparation of Catalysts
Catalysts were prepared by impregnation until the
saturation of the support fractions with a particle size
of 0.2–0.5 mm with a chloroplatinic acid solution;
then, they were dried at 120°C and calcined in an air
flow at 500°C for 4 h. The use of support fractions of
this particle size allowed us to reduce the possible
effect of the external and internal mass transfer on the
results of the testing of the catalysts [7, 28, 29]. The cal-
culated platinum content in the catalysts was 0.5 wt %.
Based on these notions, the aims of this work are
the following:
—to study the physicochemical characteristics of
platinum catalysts on oxide (Al2O3, ZrO2, ZrO2–
Al2O3) and borate-containing supports (B2O3–Al2O3,
B2O3–ZrO2), including their acidity;
—to study the effect of the nature of these supports
on the main routes of the transformation of vegetable
oil under the conditions of hydrocracking and on the
yield and composition of the resulting diesel fraction.
Determination of Catalyst Characteristics
The X-ray powder diffraction analysis of the cata-
lysts was performed on a Bruker D8 Advance diffrac-
tometer (Germany) using monochromatized CuKα
radiation. The samples were scanned in the angle
range 2θ = 5°–80° with a step of 0.05°. The diffraction
patterns were decoded using a Bruker EVA software
package with the aid of the ICDD PDF-2 powder dif-
fraction database. The number of phases and the size
of the coherent scattering region (CSR) were deter-
mined using the TOPAS 3.0 software.
Transmission electron microscopy (TEM) was car-
ried out on a JEOL JEM-2100 electron microscope
(Japan) with accelerating voltage of 200 kV and lattice
plane resolution of 0.145 nm. The computer treatment
EXPERIMENTAL
Preparation of Supports
In the synthesis of the supports, we used pseudobemit
AlOOH containing ~23 wt % Al2O3 (analytical grade),
orthoboric acid H3BO3 (reagent grade), aqueous solu-
tions of zirconium oxynitrate ZrO(NO3)2 ⋅ 2H2O
(analytical grade) at a concentration of ~24
/L,
g
ZrO2
aluminum nitrate Al(NO3)3 ⋅ 9H2O (analytical grade) at a
concentration of ~0.6 /L, and ammonia NH4OH
g
Al2O3
(high purity grade) at a concentration of ~95
/L.
g
NH3
KINETICS AND CATALYSIS
Vol. 58
No. 5
2017

HYDROCRACKING VEGETABLE OIL ON BORATE-CONTAINING CATALYSTS
565
of the TEM images and determination of particle size mined from the νNH line profile in the region 3400–
1800 cm–1 [30].
were conducted using the Digital Micrograph Gatan
software package.
27
11
The Al and B NMR spectra were recorded by
MAS NMR on a Bruker Avance-400 spectrometer
using the magic angle spinning of the samples (10 kHz).
Catalyst Testing
The catalysts were tested in the process of oil
hydrocracking on a flow installation in a reactor with an
unmovable layer at 380°C, pressure of 4.0 MPa, stock feed
rate of 1.0 h–1, and a hydrogen : stock = 1400 N m3/m3 vol-
ume ratio for 1 h. The internal diameter of the reactor
was 0.65 cm, the length was 52 cm, the catalyst weight
was 7.5 g, and the ratio of the catalyst layer’s length to
the reactor’s diameter was 74. Resource testing for
Pt/BA and Pt/BZ samples was carried out under the
same conditions for 20 h. The catalysts were prelimi-
narily reduced in a hydrogen flow at 500°C for 1 h.
The raw material used consisted of refined sunflower-
seed oil containing 10.9 wt % oxygen, 231 ppm water,
acid number 0.10 mg KOH/g.
The products of oil hydrocracking in the reactor’s
output were separated in a cooled separator into gas-
eous and C5+ hydrocarbon components. The volume
of the gaseous products was measured with a wet-gas
meter. The weight of the gaseous products was calcu-
lated from the volume and density measured in an
27
The Larmor frequency was 104.23 MHz for Al and
11
128.33 MHz for the B nuclei. The duration of the
90-degree pulses was ~2.5 μs and the relaxation delay
for both nuclei was 1 s. A 1 M AlCl3 solution and
BF3[O(C2H5)2] were used as the external references
for 27Al and 11B NMR, respectively.
The texture characteristics of the catalysts were
determined by analyzing the nitrogen adsorption–
desorption isotherms at –195.6°C obtained on a
Micromeritics ASAP-2020 volume vacuum static
installation (United States). The specific surface area was
calculated by the BET method using an adsorption iso-
therm in the relative pressure range of 0.05–0.25. The
specific pore volume was determined from the nitrogen
adsorption data at a relative pressure of 0.990.
The temperature-programmed ammonia desorp-
tion (TPD-NH3) for the studied catalysts was conducted
on a Micromeritics AutoChem II 2920 chemisorption ideal gas approximation based on the results of the
analyzer equipped with a thermal conductivity detec- chromatographic analysis. The gaseous products were
tor in the temperature range 150–500°C [23].
analyzed in the online mode on a two-channel Khro-
mos GKh-1000 gas chromatograph (Russia). The
content of H2, CO, and CO2 was determined on a
packed column 3 m long and 4 mm in diameter filled
with the Porapak R phase using a thermal conductivity
detector. Helium was used as the carrier gas. Hydro-
carbons C1–C4 were separated on a DB-1 capillary
column 60 m × 0.25 mm × 0.50 μm (Agilent Technol-
ogies, United States) using a flame ionization detec-
tor. Argon was used as the carrier gas. The analysis was
performed in the temperature-programming mode:
40°C, pause 5 min; heating at a rate of 4 K/min to
150°C, pause 15 min.
The IR spectra of the adsorbed pyridine (Py) were
registered on a Shimadzu FTIR-8300 Fourier-trans-
form IR spectrometer (Japan) at ambient temperature
with a resolution of 4 cm–1 and the number of scans
(300). The samples of Pt/BA and Pt/BZ were mixed
with barium fluoride BaF2 in a weight ratio of 1 : 4.
Each of these samples and each of the Pt/A, Pt/Z, and
Pt/ZA catalysts were pressed to form pellets with a
thickness of up to 0.02 g/cm2. The pellets were
degassed in the IR cell at 480°C for 1 h and then
treated with pyridine at 150°C for 15 min. The weakly
bound pyridine was removed by the evacuation of the
catalysts at the same temperature for 30 min until a
residual pressure of 0.13 Pa was reached. The concen-
tration (N) of Lewis acidic centers (LACs) and Bron-
sted acidic centers (BACs) was calculated as μmolPy/g
The hydrocarbon hydrocracking products C5+ were
weighed and the water content was determined
according to the state standard 2477–65.
The qualitative composition of C5+ products was
determined on an Agilent 6890/5973N gas chromato-
graph–mass spectrometer (Agilent Technologies)
equipped with an HP-5ms capillary column (30 m ×
0.25 mm × 0.25 μm) using helium as a carrier gas.
Analysis was performed in the following temperature-
programming mode: 40°C, pause 2 min; heating at a
rate of 10 K/min to 250°C, pause 15 min. The com-
pounds were identified using the NIST 0.5 mass spec-
tral library.
by the formula
where A (in cm–1) is the
N = A ρA0 ,
integral absorption of the IR bands in the regions 1440–
1465 and 1520–1535 cm–1, respectively, ρ (in g/cm2) is
the pellet thickness, and A0 (cm/μmol) is the molar
coefficient of the integral absorption presented in the
monograph [30] and equal to 3.5 and 3.0, respectively.
The strength of the BACs was assessed by the value of
the proton affinity (PA, kJ/mol) calculated by the
equation
РА = [log(3400 − νNH) − 0.115] 0.0023,
The fractional composition of the C5+ products
was assessed by simulated distillation on a Shimadzu
GC-2010 gas chromatograph with an Rtx-2887 capil-
where 3400 is the wave number (in cm–1) of the undis-
turbed N–H bond of the pyridinium ion, νNH (in cm–1)
is the wave number of the center of gravity for a lary column (10 m × 0.53 mm × 2.65 μm) (Restek,
stretching vibration band of the pyridinium ion deter- United States) and a flame ionization detector using
KINETICS AND CATALYSIS Vol. 58 No. 5 2017

566
Intensity
CHUMACHENKO et al.
(M) М–ZrO2 phases (curve 3) with a CSR size of 11.3
and 14.3 nm, respectively. The diffraction pattern of
the Pt/ZA catalyst (curve 4) displays more intense
reflections of the T–ZrO2 phase, while reflections at
2θ angles of 24° and 31.5° responsible for the М–ZrO2
phase are less intense. There are no other reflections
typical for this phase. Therefore, the content of the
T‒ZrO2 phase in the Pt/ZA catalyst increases to
86 relative %, while the content of М–ZrO2 decreases
to 14 relative %. The size of the CSR for the T–ZrO2
phase increases to 11.7 nm, while that of the М–ZrO2
phase decreases to 9.6 nm. Nonetheless, the size of the
CSR for the T–ZrO2 phase in the Pt/ZA catalyst does
not reach the critical value of 20–24 nm; at larger val-
ues, only the equilibrium М–ZrO2 phase would be
stable in the ZA support [32]. Thus, the size factor
ensures the stabilization of the T–ZrO2 phase in the
ZA support due to the formation of anion-deficient
solid solutions of the Zr1–хAlхO2–0.5х type [32].
According to the published data [33], the complete
stabilization of the T–ZrO2 phase in the ZA support
prepared by a similar method is achieved at a content
of about 7 wt % Al2O3. The diffraction pattern of the
Pt/BZ catalyst (curve 5) displays a broadened halo in
the range 2θ = 25°–35°, along with weak reflections
from the T–ZrO2 and М–ZrO2 phases. This indicates
the X-ray amorphous state of the BZ support due to
the retarded crystallization of T–ZrO2.
T
M
M
T
T
T
5
T
T
4
3
2
1
M
M
M
M
M
M
M
M
M
M
γ
γ
γ γ
100
20
30
40
50
2θ, deg
60
70
80
Fig. 1. X-ray diffraction patterns for Pt/A (1), Pt/BA (2),
Pt/Z (3), Pt/ZA (4), and Pt/BZ (5) catalysts. Designa-
tions: γ, γ-Al O phase; T, tetragonal phase of ZrO ; M,
2
3
2
monoclinic phase of ZrO .
2
helium as the carrier gas. The temperature program
was as follows: 30°C, heating at a rate of 17 K/min to
280°C, pause for 10 min. The percentage of fractions
was calculated according to ASTM D 2887 in the fol-
lowing boiling temperature ranges: boiling beginning–
150°C for the gasoline fraction, 150–350°C for the
diesel fraction, and 350°C–end of boiling for heavy
gasoil.
Morphology of Platinum Catalyst Supports
The hydrocarbon composition of the diesel fraction
was determined on a Khromos GKh-1000 gas chromato-
graph with a DB–1 capillary column (100 m × 0.25 mm ×
0.50 μm) (Agilent Technologies) equipped with a
flame ionization detector using argon as the carrier gas
in the following temperature mode: 40°C, pause for
5 min, heating at a rate of 4 K/min to 250°C, pause for
360 min.
The yield of hydrocracking products (gases, C5+
hydrocarbons, water) and separate fractions was cal-
culated based on the results of the chromatographic
analyses and the material balance data.
According to the TEM images and the correspond-
ing electron diffraction patterns (not given), the A
support contains the γ-Al2O3 phase identified by inter-
plane distances 0.456 nm (111), 0.280 nm (220), and
0.198 nm (400). The presence of this phase in the A
support is confirmed by analysis using X-ray diffrac-
tion (Table 1). The particles of the γ-Al2O3 phase are
mainly lamellar with their size ranging from 3 to 10 nm
or needle-like with parameters 3 × 30 nm.
The structure of the BA support includes the nee-
dle-like and arbitrary-form aggregates formed by the
lamellar crystallites 3–5 nm in diameter (Fig. 2a). The
lattice parameters for these crystallites calculated from
the electron diffraction patterns are 0.197 and 0.240,
respectively, and correspond to interplane distances
with indices (400) and (311) typical for γ-Al2O3. Fur-
thermore, the BA support includes aggregates of irreg-
ular shape that have an amorphous structure (Fig. 2b),
which appears in the electron diffraction pattern as a
diffuse ring. These observations agree closely with the
previously established phase composition of the
RESULTS AND DISCUSSION
Phase Composition of the Catalysts
The results of X-ray powder diffraction for the cat-
alysts are given in Table 1 and Fig. 1. The Pt/A sample
contains the γ-Al2O3 phase with a CSR size of 4.6 nm
(Fig. 1, curve 1). The Pt/BA catalyst differs from it by
the lack of reflections (curve 2) typical for γ-Al2O3,
and the emergence of a scattered halo in the region of Pt/BA catalyst (Table 1). Since the X-ray diffraction
2θ (20°–30°), which indicates the formation of X-ray method allows detecting the crystalline phase with a
amorphous highly dispersed aluminoborates [31] in particle size of larger than 3 nm in such systems, we
the BA support. The Pt/Z sample contains 15 and can affirm that the content of the γ-Al2O3 phase in the
85 rel. % of the tetragonal (T) T-ZrO2 and monoclinic
BA support is small. Thus, it consists mainly of the
KINETICS AND CATALYSIS
Vol. 58
No. 5
2017

HYDROCRACKING VEGETABLE OIL ON BORATE-CONTAINING CATALYSTS
Table 1. Physicochemical properties of catalysts
567
Catalyst
Pt/Z
Characteristic
Pt/A
Pt/BA
Pt/ZA
86/14*
Pt/BZ
Phase composition
γ
а
15/85*
a, М**, Т**
–
Crystallite size, nm
4.6
–
11.3 (Т)
14.3 (М)
11.7 (Т)
9.6 (М)
Specific surface area, m2/g
179
203
59
67
150
Specific pore volume, cm3/g
0.48
0.50
0.14
0.15
0.16
Amount of AC, μmolNH3/g
weak
average
strong
Total
Amount of LAC, μmolPy/g
Amount of BAC, μmolPy/g
Proton affinity, kJ/mol
9
164
153
326
222
–
18
293
187
498
54
6
121
85
212
46
–
5
105
77
187
59
24
258
71
353
42
200
1191
–
44
–
–
–
1226
Designations: γ, γ-Al O phase; a, amorphous phase; T, tetragonal phase of ZrO ; M, monoclinic phase of ZrO ; * T/M ratio; ** trace
2
3
2
2
amounts.
amorphous phase represented by highly dispersed alu- arbitrary shape ranging from 5 to 20 nm in size and
minoborates.
particles with parameters 5 × 30 nm whose projections
have a rectangular shape with smoothed angles (Fig. 2f).
At the same time, it contains conglomerates with an
amorphous structure, along with T–ZrO2 particles,
which is confirmed by the presence of a diffuse ring on
the electron diffraction pattern. It is difficult to deter-
mine the size of the T–ZrO2 crystallite by analyzing
the X-ray amorphous system (as in the case of the
Pt/BA catalyst) using X-ray diffraction. We can only
assume that this oxide is present in the BZ support in
small amounts. This observation provides an addi-
tional idea on the structure of the BZ support that was
previously formed based on the data on the phase
composition of the Pt/BZ catalyst (Table 1) and the
data available in the literature [37].
To describe the morphology of the support parti-
cles in all the samples mentioned above, we used
micrographs where the platinum particles are invisi-
ble. The study of the morphology and dispersity of
platinum particles in the described catalysts will be the
subject of our subsequent publications.
The TEM micrograph of the Pt/Z catalyst (Fig. 2a)
displays the planes of the crystal lattice with a period-
icity of 0.316 and 0.370 nm. These interplane distances
correspond to indices (–111) and (110) typical for
М‒ZrO2. Since the М–ZrO2 phase is prevalent in the
Pt/Z (Table 1), no crystallographic planes related to
T–ZrO2 are visible in Fig. 2c. Therefore, we can
believe that the Z support contains agglomerates of
zirconium dioxide particles of the T and M modifica-
tions, which have an isometric shape and a size rang-
ing from 10 to 20 nm (Fig. 2d). The range for the size
of these crystallites agrees closely with the X-ray diffrac-
tion data (Table 1), while their shape coincides with the
shape described in the literature [34].
The TEM images of the Pt/ZA catalyst (not given
here) containing mainly the T–ZrO2 phase provide a
resolution that is typical for this oxide of the planes of
a crystal lattice with a periodicity 0.360 nm and indices
(100). At the same time, we failed to detect crystallo-
graphic planes related to М–ZrO2, which can be
explained by its significantly lower content in the ZA
support compared with T–ZrO2 (Table 1). The shape
of the support particles and their size range coincides
with the corresponding characteristics of the Z support
and is confirmed by the published data [35, 36] and the
results of the X-ray diffraction analysis (Table 1).
Local Environment of Al and B Atoms
in Platinum Catalyst Supports
The 27Al NMR spectrum for the Pt/A catalyst
(Fig. 3, spectrum 1) is similar to that of γ-Al2O3 and
contains the signals of the aluminum atom located in
the octahedral (10 ppm) and tetrahedral (70 ppm)
voids of the oxide. In the spectrum of Pt/BA (spec-
trum 2), the relative intensity of the signals from the
The TEM image of the Pt/BZ catalyst (Fig. 2e)
exhibits the planes of a crystal lattice with periodicity
0.360 nm and indices (100), which refer to the T–ZrO2
phase. The BZ support contains T–ZrO2 particles of AlO4 tetrahedra is lower compared with the AlO6 octa-
KINETICS AND CATALYSIS Vol. 58 No. 5 2017

568
CHUMACHENKO et al.
(a)
(b)
(400)
50 nm
10 nm
(c)
(c)
(−111)
(d)
(d)
0.316 nm
0.316 nm
0.370 nm
0.370 nm
(110)
(100)
10 nm
((ee))
50 nm
((ff))
0.360 nm
50 nm
10 nm
Fig. 2. TEM micrographs and corresponding electron diffraction patterns for catalysts Pt/BA (a and b), Pt/Z (c and d), Pt/BZ
(e and f).
hedra than in the Pt/A spectrum. This is due to the the AlO5 pentahedra or the AlO4 tetrahedra distorted
substitution of the AlO4 tetrahedra in the structure of
γ-Al2O3 by the BO4 tetrahedra [38]. The spectrum of
Pt/BA also displays a weak signal with a chemical shift
of ~35 ppm related to the aluminum atoms in the AlO5
pentahedra [39]. The spectrum of the Pt/ZA catalyst
(spectrum 3) consists of two signals that belong to the
aluminum atom in the AlO6 octahedra (5 ppm) and
because of the formation of the Zr1–хAlхO2–0.5х (40 ppm)
solid solution [40].
The 11B NMR spectra of the Pt/BA and Pt/BZ cat-
alysts (Fig. 4, spectra 1, 2) contain two overlapping
signals with chemical shifts of ~5 and ~16 ppm related
to tetrahedral (BO4) and trigonal (BO3) boron,
respectively [37, 38]. There is a relationship between
KINETICS AND CATALYSIS
Vol. 58
No. 5
2017

HYDROCRACKING VEGETABLE OIL ON BORATE-CONTAINING CATALYSTS
Intensity
569
Intensity
3
2
2
0.5 × 108
1
1
0.5 × 108
50
40
30
20
10
0
–10 –20
120 100 80
60
40
20
0
–20 –40
Chemical shift, ppm
Chemical shift, ppm
11
27
Fig. 4. B MAS NMR spectra of catalyst: (1) Pt/BA,
(2) Pt/BZ.
Fig. 3. Al MAS NMR spectra of catalyst: (1) Pt/A,
(2) Pt/BA, (3) Pt/ZA.
ports with the B2O3 content of at least 5 wt %. Further
its growth up to 20 wt %, apparently, should not result
in the growth of the number of BACs. Thus, we can
assume that at the same calculated B2O3 content in the
Pt/BA and Pt/BZ catalysts, the number of BACs in
the first sample should be larger than in the second.
the coordination of the boron atom in the BA and BZ
support structures and their acidity, which is consid-
ered by the Tanabe rule [41]. The rule consists of the
fact that acid centers (ACs) form in mixed oxides when
their model structures contain an excess positive or
negative charge. The magnitude and sign of the charge
are dependent on the coordination number of the cat-
ion whose content is lower, i.e., on B3+, which is situ-
ated in the BA and BZ supports in the tetrahedral oxy-
gen environment. Since the coordination number of
the O2– ions in the BO4 structural units of these sup-
ports is 2, the calculation by the Tanabe rule indicates
the presence of an excessive charge of –0.25e per B–O
bond. It should be noted that the B3+ cations in the
trigonal oxygen environment, according to the Tanabe
rule, do not create an excess charge in borate-containing
systems. Therefore, irrespective of the coordination of
the aluminum and zirconium atoms, it is the presence of
the BO4 structural units that should provide the distinct
acidic properties of the BA and BZ supports.
The presence of an excess negative charge of the
B–O bonds in the model structures of the BA and BZ
supports indicates the formation of BACs [41]. These
centers are present in borate-containing systems as
proton-donating OH groups incorporated into the
environment of tetrahedral coordinated boron [42,
43]. The structural BO4 units in the BA support con-
taining from 3 to 20 wt % B2O3 was found to both in
the volume and on the surface. The number of BACs
linearly increases with the B2O3 content in this range
[42]. In the BZ support with the B2O3 content up to
5 wt %, zirconium dioxide particles are coated with a
monolayer composed of BO4 structural units [44],
whereas, with the B2O3 content ranging from 5 to 20 wt %,
the BO3 structural units are arranged as clusters or new
layers over the layer of the BO4 monolayer [45, 46].
Textural Characteristics of the Catalysts
Table 1 shows the textural parameters of the cata-
lysts. To facilitate the analysis and subsequent discus-
sion, we combined the Pt/A, Pt/Z, and Pt/ZA sam-
ples into a group of oxide catalysts, while the Pt/BA
and Pt/BZ were combined into borate-containing cat-
alysts.
The specific surface area of the Pt/BA catalyst is
slightly larger than the Ssp of the Pt/A sample, whereas
the specific surface area of the Pt/BZ sample is by a
factor of >2.5 higher than the Ssp of the Pt/Z sample.
Thus, the presence of the amorphous phase in borate-
containing catalysts favors the enhancement of their
specific surface area. The specific pore volume in the
borate-containing Pt/BA and Pt/BZ catalysts is
retained at the level of their oxide analogs (0.5 and
0.16 cm3/g, respectively). The specific surface area
and specific pore volume for the Pt/Z and Pt/ZA cat-
alysts have values close to each other.
Acidic Properties of the Catalysts and Supports
The results of the study on the acidity of catalysts
and supports by TPD-NH3 are given in Tables 1 and 2,
respectively. In terms of the thermal ranges of ammo-
nia desorption, the surface ACs of BA [27] and BZ
[44] supports are divided into weak (150–200°C),
average (200–350°C), and strong (350–500°C). A
Therefore, BACs form on the surface of the BZ sup- similar approach was used for the consideration of the
KINETICS AND CATALYSIS Vol. 58 No. 5 2017

570
CHUMACHENKO et al.
higher than on the Pt/BZ catalyst (86 μmolPy/g),
Table 2. Acidic properties of supports
which indicates the higher acidity of the Pt/BA cata-
lyst compared with the Pt/BZ catalyst and confirms
the results obtained in the study of borate-containing
catalysts by the TPD-NH3 method (Table 1). At the same
time, the BACs in the Pt/BA catalyst are stronger and
their concentration is higher by a factor of 4.5 than in
the Pt/BZ catalyst (Table 1). Therefore, the previous
presumption on the higher concentration of these cen-
ters in the Pt/BA catalyst than in thePt/BZ catalyst is
considered to be reasonable.
Support
Acidic centers,
μmolNH
g
A
BA
Z
ZA
BZ
3
Weak
11
165
133
309
24
285
173
482
4
89
77
6
92
79
23
327
69
Average
Strong
Total
170
177
419
AC in the other supports and all platinum catalysts.
Tables 1 and 2 show that the centers of the average
strength provide the largest contribution to the acidity
of all the catalysts and supports.
As compared with the Pt/A oxide system, the total
acidity of the Pt/BA catalyst is larger by a factor of 1.5.
This is mainly due to the increase in the number of the
average AC (by a factor of 1.8) on its surface and, to a
lesser extent, the weak and strong AC. The acidity of
the Pt/Z and Pt/ZA samples is at the same level. The
general acidity of the Pt/BZ catalyst is by a factor of
1.7 higher than the acidity of the Pt/Z oxide system,
which is explained by the content of the average AC on
its surface being twice as large.
Thus, borate-containing catalysts differ from oxide
samples by both larger number of their average AC and
their increased overall acidity. Among borate-contain-
ing catalysts, the Pt/BA sample exhibits the most pro-
nounced acidic properties, while the Pt/A catalyst is
the most acidic among the oxide samples.
A comparison of the data presented in Tables 1 and
2 displays that the features of the variations in the acid-
ity characteristics (the total number of centers and
content of the average acidity centers) for platinum
catalysts and the corresponding supports are close to
each other.
To identify the LACs and BACs, we used the IR
spectroscopy of the probe molecules (pyridine). The
IR spectra of pyridine adsorbed on oxide catalysts
(Fig. 5a, spectra 1–3) show absorption bands (a.bs.) in
the region 1450–1440 cm–1 related to pyridine com-
plexes with LACs, i.e., Py–L complexes. The spectra
of pyridine adsorbed on borate-containing catalysts
(Fig. 5b, spectra 4, 5) exhibit absorption bands in the
regions 1535–1520 and 1465–1460 cm–1, which
belong to pyridinium ions (PyH+) and Py–L, respec-
tively. The formation of PyH+ ions indicates the pres-
ence of BACs on the surface of the Pt/BA and Pt/BZ
catalysts. These ions are supposedly formed upon the
proton transfer from hydroxyl groups in a tetrahedral
boron environment to pyridine molecules.
Borate-containing catalysts are known to acceler-
ate the acid-controlled dehydration reactions of alco-
hols to alkenes [47–49] and the skeletal isomerization
[50] and oligomerization [51] of alkenes. Their activity
in these reactions is caused by the presence of BACs
and the enhanced content of the average AC on BA
and BZ supports [49, 50, 52]. Since both alcohol and
alkenes can be formed as intermediate products on the
transformation of oxygen-containing compounds into
diesel fraction hydrocarbons, we can expect Pt/BA
and Pt/BZ catalysts to be the most efficient in the
one-stage process of oil hydrocracking.
The Role of Support in Hydrocracking Sunflower-Seed
Oil on Bifunctional Platinum Catalysts
The results of testing oxide (Pt/A, Pt/Z, Pt/ZA)
and borate-containing (Pt/BA, Pt/BZ) catalysts show
that the oil undergoes hydrotransformation on all
samples preferably into organic C5+ products (with a
yield of 81.7–87.3 wt %), as well as gaseous products
and water (Table 3).
The direction of oil transformation typical for each
group of platinum catalysts can be determined by the
CO/CO2 ratio and weight concentration of gaseous
C1 + C2 and C3 + C4 products. The numerator and
denominator of the CO/CO2 ratio include gaseous
products of the hydrodeoxygenation (HDO) reactions
resulting in hydrocarbons with an odd number of car-
bon atoms, i.e., hydrodecarbonylation (HDCO) and
decarboxylation (DCO2) products, respectively. The
sum C1 + C2 (methane + ethane) is represented by the
hydrogenolysis products of C5+ hydrocarbons on the
metal centers (MCs) of catalysts, i.e., the products of
the HG–MC reactions. The sum C3 + C4 (propane +
butane + butenes) includes the products of hydroc-
racking of triglycerides (HC–TG) and C5+ hydrocar-
bons on AC (HC–AC), i.e., the products of the HC–
TG and HC–AC reactions.
During the oil hydrotransformation on oxide and
borate-containing catalysts, we detected ratios of
CO/CO2 < 1 and CO/CO2 > 1, respectively (Table 4).
The LAC concentration on the Pt/A catalyst is
larger than on the Pt/Z and Pt/ZA catalysts (Table 1), This indicates that the contribution of the DCO2 reac-
which agrees closely with the difference in the general
acidity of these samples. The total concentration of the
LACs and BACs on the Pt/BA catalyst (254 μmolPy/g) is
tion in the presence of Pt/Z, Pt/ZA, and Pt/A cata-
lysts is greater than in the HDCO reaction; however,
in contrast, the contribution of HDCO is higher than
KINETICS AND CATALYSIS
Vol. 58
No. 5
2017

HYDROCRACKING VEGETABLE OIL ON BORATE-CONTAINING CATALYSTS
571
the DCO2 contribution in the presence of the Pt/BA
(а)
Absorbance
and Pt/BZ catalysts. The different behavior of oxide
and borate-containing catalysts in these reactions is
caused by the presence of BACs in the Pt/BA and
Pt/BZ catalysts, which accelerate the HDCO reaction
because of the addition of proton to one of the oxygen
atoms of the carboxyl group activated on an MC with
its subsequent elimination [53–55].
Py–L
3
2
Py–L
Py–L
The content of the C1+C2 alkanes in the gaseous
products of oil hydrotransformation increases in the
series of Pt/BZ < Pt/BA < Pt/A < Pt/ZA < Pt/Z cat-
alysts (Table 4). Thus, the HG–MC reactions proceed
on oxide catalysts faster than on borate-containing
catalysts. This is probably explained by the higher
hydrogenating activity of the platinum centers in oxide
catalysts than in borate-containing ones. The study of
the dispersity and the state of the charge of the hydro-
genating component in these samples, which will
check the validity of this assumption, will be the sub-
ject of our subsequent publications.
0.1
1
1400 1450 1500 1550 1600 1650
Wavenumber, cm–1
(b)
Absorbance
The content of C3 + C4 hydrocarbons in gaseous
products of oil hydrotransformations increases in the
series Pt/ZA < Pt/Z < Pt/A < Pt/BA < Pt/BZ (Table 4).
The study on the acidity of these samples (Table 1)
allows us to assume that the contribution of the HC–
TG and HC–AC reactions for the borate-containing
catalysts is larger than for oxide catalysts because of
the increased content of the average AC. However, the
Pt/BA sample showing the maximal acidity according
to TPD-NH3 (Table 1), leads to a lower content of the
C3 + C4 hydrocarbons in gaseous products than the
Pt/BZ catalyst. This is explained by the fact that
Pt/BA, where the ratio NLAC/NBAC = 0.3, produces
fewer C4 components (8.6 wt %) than the Pt/BZ cata-
lyst (42.4 wt %), where NLAC/NBAC = 1.0. Thus, the
HC–AC side reactions proceed on the Pt/BA sample
slower than on the Pt/BZ sample. The contribution of
these reactions over borate-containing catalysts is
determined only by the value of the NLAC/NBAC ratio.
Py–L
PyH+
×4
5
PyH+
Py–L
0.02
4
1400 1450 1500
1550 1600 1650
1700
Wavenumber, cm–1
Fig. 5. IR spectra of pyridine adsorbed on catalysts: a, Pt/A
(1), Pt/Z (2), Pt/ZA (3); b, Pt/BA (4), Pt/BZ (5).
amounts of iso-alkanes, cycloalkanes, and arenes.
These data indicate that the oxygen-containing prod-
ucts of HC–TG reactions are completely transformed
over all catalysts into hydrocarbons.
The data on the qualitative composition of C5+
products of oil hydrocracking on Pt/BA, Pt/BZ,
Pt/A, Pt/Z, and Pt/ZA samples enables us to assess
the contribution of the HDO reactions on borate-con-
taining and oxide catalysts and to determine the routes
for the deoxygenated compounds’ transformation into
other hydrocarbons. Moreover, the state of aggrega-
tion of the C5+ products and the ratio of the main fuel
fractions is directly affected by their composition.
Over the Pt/BA and Pt/BZ catalysts, the products
We also revealed other behavior features of plati-
num catalysts in oil hydrocracking process. First, the
C5+ products obtained over oxide catalysts may
include alkenes in spite of the supposed enhanced
hydrogenating activity of the catalysts. In contrast,
there are no alkenes at the start of the process over
contain n-alkanes (up to C24), dimethyl-substituted borate-containing catalysts, which supposedly display
reduced hydrogenating activity. Second, it is the
borate-containing catalysts that produce the C5+
(up to C8), and monomethyl-substituted (C9–C20 and
C9–C18, respectively) iso-alkanes, alky-substituted
cyclohexanes (up to C12), and a few arenes (C7–C9). products containing the largest number of iso-alkanes.
These observations convey the suggestion that the differ-
ence in the properties of the catalysts of both groups is
likely to be due to the state of the platinum centers and
the nature of the AC (the absence or presence of BACs).
At the same time, the products formed over Pt/A,
Pt/Z, and Pt/ZA catalysts contain preferably
n-alkanes (up to C30) and alkenes (C17 over Pt/A and
C15–C23 over Pt/Z and Pt/ZA), as well as trace
KINETICS AND CATALYSIS Vol. 58 No. 5 2017

572
CHUMACHENKO et al.
Table 3. Yields of products and fractions upon oil hydroc-
Considering the routes of alkene formation in the
oil hydrocracking reaction, let us note that alkenes can
result from reactions of two types: HDO of unsatu-
rated oxygen-containing compounds (carboxylic
acids, aldehydes, and alcohols) and dehydrogenation
(DH) of n-alkanes. Since each of these reactions pro-
ceeds with the participation of the MC catalyst, the
DH reactions under hydrocracking conditions should
compete for these centers with the HDO reactions.
Therefore, it is most likely that DH will proceed much
slower than HDO. Thus, the main directions of alkene
formation lie in the reactions related to the elimination
of oxygen (DCO2, HDCO, and H–DH2O) from the
unsaturated products of triglyceride hydrocracking.
The considerations noted above, in turn, lead to the
conclusion that C=C bonds in triglyceride molecules
in the conditions of oil hydrotransformation do not
undergo complete hydrogenation.
racking over different catalysts*
Catalyst
Yield, wt %
Pt/A Pt/BA Pt/Z Pt/ZA Pt/BZ
Products
gas
14.5
9.9 11.2 12.0 19.1
hydrocarbons C5+
water
Fractions
gasoline
diesel
heavy gasoil
81.9 86.0 87.2 87.3 81.7
5.5
0.2
5.5
3.5
3.3
0.0
2.3
2.3
0.0 23.8
73.9 75.1 72.3 71.8 55.6
7.8 7.4 14.9 15.5 2.3
* Hydrocracking conditions: temperature 380°C, pressure 4.0 MPa,
–1
weight stock feed rate 1.0 h , catalyst testing duration 1 h.
The C5+ products of hydrocracking containing iso-
alkanes seem to emerge over the Pt/BA and Pt/BZ
catalysts through the bifunctional carbocation mecha-
nism of alkane hydroisomerization [56]. The presence of
alkenes in the reaction medium of the process compen-
sates the first stage of the mechanism—the n-alkanes’
DH reaction over the MC. The presence of BACs pro-
vides the second stage of the mechanism, the alkene
protonation reaction. The resultant carbenium ions
undergo a rearrangementof the carbon chain to form
alkylcarbenium ions. Further, the transfer of hydride
ions from n-alkane molecules to alkylcarbenium ions
takes place.
Table 4. Composition of gaseous products and diesel frac-
tions upon oil hydrocracking over different catalysts*
Catalyst
Composition,
wt %
Pt/A Pt/BA Pt/Z Pt/ZA Pt/BZ
Gases
CO
CO2
C1+C2
C3+C4
Total
21.2
28.7
35.3
14.8
31.9
21.8
24.3
22.0
9.1
27.6
55.6
7.7
14.6
36.0
42.6
6.8
22.6
7.9
10.6
58.9
100.0 100.0 100.0 100.0 100.0
During the moderate hydrogenating activity of the
platinum centers, borate-containing catalysts are most
likely to provide such an alkene concentration under
hydrocracking conditions that carbenium ions can
appear in the BACs. It is reasonable to suppose that
the increased hydrogenating activity of the platinum
centers and the presence of LACs only in the oxide
catalysts favors the alkene transformation (as the key
intermediates in the process of hydrogenation enno-
bling of triglycerides) mainly into n-alkanes.
According to the current notions on the routes of
the sunflower-seed oil transformation in the course of
its hydrocracking over borate-containing catalysts, a
marked contribution to the formation of a fractional
composition of C5+ products is provided by the skeletal
isomerization reactions of unsaturated hydrocarbons
and HC–AC. Since the HC–AC reactions over Pt/BZ
proceed more intensely than over Pt/BA, the liquid
hydrocracking products formed over the first sample
show a higher content of the gasoline fraction (23.8 wt %)
and a lower content of the diesel fraction (55.6 wt %)
and heavy gasoil (2.3 wt %) compared with the prod-
ucts obtained over in the second sample (Table 3). The
fractional composition of the C5+ products of oil
hydrocracking over the oxide catalysts is affected by
the HDO and HG–MC reactions. Therefore these
products are in a solid state at the ambient temperature
CO/CO2
0.7
1.5
0.3
0.4
2.9
n-Alkanes
ΣC10-C14
C15
0.8
5.9
1.6
1.8
3.2
2.7
0.7
7.3
0.6
0.7
7.0
0.8
8.7
0.3
0.2
1.3
0.5
0.1
C16
C17
C18
70.4
15.9
1.2
20.8
7.5
78.6
4.1
78.4
4.5
C19+C20
Total
OH
0.5
2.7
2.6
95.9
36.5 94.0 94.0
11.1
ΣC10–C14
C15
C16
0.5
0.1
0.2
0.9
0.9
1.5
4.1
7.4
3.2
9.2
0.4
0.2
0.3
1.4
1.3
2.5
6.0
0.3
0.2
0.3
1.7
1.2
2.3
6.0
82.6
2.1
2.2
1.2
0.4
0.5
88.9
C17
C18
27.0
12.1
4.6
C19+C20
Total
(ΣC15+ΣC17)/
(ΣC16+ΣC18)
ΣOH/Σn-alkanes 0.04 1.74
63.5
4.2
1.7
14.1
13.1
1.5
0.06 0.06 8.02
* Hydrocracking conditions are same as noted in Table 3.
KINETICS AND CATALYSIS
Vol. 58
No. 5
2017

HYDROCRACKING VEGETABLE OIL ON BORATE-CONTAINING CATALYSTS
573
and the yields of the diesel fraction over them are close
Yield, wt %
90
to each other (71.8–73.9 wt %). On the whole, the data
presented in Table 3 indicate that the Pt/BA sample is
the most promising among all the platinum catalysts
in terms of the highest yield of the diesel fraction
(75.1 wt %) during oil hydrocracking.
1
2
3
85
80
75
70
65
60
55
50
4
The value of the weight ratio (ΣC15 + ΣC17)/(ΣC16 +
ΣC18) in diesel fractions provides the possibility to
determine the main HDO reactions that occur over
the Pt/BA, Pt/BZ, Pt/A, Pt/Z, and Pt/ZA catalysts.
The HC-TG reactions during the hydrocracking of
sunflower-seed oil produce enhanced amounts of fatty
acids with 16 and 18 carbon atoms. Consequently, the
numerator and denominator of this ratio include the
hydrocarbons resulting from the DCO2 + HDCO and
H–DH2O reactions of these acids, respectively, or the
products of the HDO reactions that lead to hydrocar-
bons with an odd and even number of carbon atoms,
respectively.
1
2
3
4
0
2
4
6
8
10 12 14 16 18 20
Time, h
Fig. 6. Dependences of yield of C products (1, 2) and
diesel fraction (3, 4) on hydrocracking duration in catalysts
Pt/BA (1, 3) and Pt/BZ (2, 4).
5+
The ratio (ΣC15 + ΣC17)/(ΣC16 + ΣC18) > 1 for all
diesel fractions (Table 4). This means that the pre-
dominance of both the DCO2 and HDCO reactions
over all platinum catalysts is due to the nature of the
hydrogenating component. However, Table 4 also
indicates that the value of this ratio depends on the
nature of the support of the platinum catalyst. The
reduction of fatty acids into alkanes of a diesel series
via reaction (IV) includes consecutive stages of acid →
aldehyde → alcohol → alkene → alkane. On the one
hand, the first three stages of this process are acceler-
ated if the BACs of the bifunctional catalyst partici-
pate in the addition of a proton to the oxygen atoms of
the C=O and C–OH bonds activated on the MC [57].
On the other hand, the BA and BZ supports, due to
their distinct Bronsted acidity and enhanced concen-
tration of the average AC [58], are more active in the
reaction of alcohol dehydration to alkenes compared
with the A [47] and Z and Pt/BZ samples, and the
(ΣC15 + ΣC17)/(ΣC16 + ΣC18) ratios are lower than in
the fractions obtained over Pt/A, Pt/ZA, and Pt/Z
(Table 4). Hence it appears that, due to the arguments
listed above, the H–DH2O reactions proceed faster in
borate-containing catalysts than in oxide samples.
Based on the all the ideas and data presented in
Table 4 above, we have drawn the additional conclu-
sion that the increased concentration of the average
AC and the presence of BACs in a platinum catalyst
favor a more intense course of the H–DH2O reac-
tions.
A comparison of the CO/CO2 and (ΣC15 + ΣC17)/
(ΣC16 + ΣC18) ratios allows us to assess the role of the
DCO2, HDCO, and H–DH2O reactions in the for-
mation of gaseous and deoxygenated C5+ products of
oil hydrocracking over platinum catalysts with differ-
ent supports. According to the data of Table 4, the
preferable directions of HDO are DCO2 in the Pt/Z
and Pt/ZA samples; DCO2 and H–DH2O, in the
Pt/A samples; and HDCO and H–DH2O, in the
Pt/BA and Pt/BZ samples.
To select the catalysts for one-stage oil hydrocrack-
ing, it is reasonable to use the value of the weight ratio
of ΣOH/Σn-alkanes in the diesel fractions obtained in
them. In this ratio, the sum of the other hydrocarbons
(OHs) is the total content in the diesel fraction of iso-
alkanes, cycloalkanes, and alkenes equal to 100 – Σn-
alkanes (in percentage). It was shown previously that
unsaturated hydrocarbons as the main intermediates
in the HDO reactions of triglyceride hydrocracking
products undergo isomerization and cyclization only
in borate-containing catalysts. Therefore, the diesel
fractions obtained in the Pt/BA and Pt/BZ samples
exhibit an ΣOH/Σn-alkane ratio that is larger than 1
(Table 4). In contrast, the HDO of triglyceride hydro-
cracking products and the hydrogenation of unsatu-
rated hydrocarbons in oxide catalysts result in diesel
fractions where this ratio is below 1 (Table 4). Since
cycloalkanes and iso-alkanes are among the compo-
Let us note that the dehydration of alcohol also
occurs over catalysts showing only Lewis acidity [58].
Since the LAC concentration and the number of aver-
age AC in Pt/A is markedly larger than in Pt/ZA and
Pt/Z, this reaction in the Pt/A catalyst should proceed
more intensely than on the other two. In turn, the ratio
(ΣC15 + ΣC17)/(ΣC16 + ΣC18) in the diesel fraction
formed in the Pt/A catalyst is much smaller than in the
diesel fractions obtained in Pt/ZA and Pt/Z (Table 4).
Thus, we can suppose that the Pt/A sample is the most
active in the H–DH2O reaction among all the oxide
catalysts due to the higher probability of realizing in it nents of the diesel fraction that provide the high per-
the stage of alcohol dehydration into alkenes.
formance characteristics at a low temperature, it is the
KINETICS AND CATALYSIS Vol. 58 No. 5 2017

574
CHUMACHENKO et al.
products of oil hydrogenation over these catalysts.
ΣOH/Σn-alkanes
Thus, the difference in the properties of the Pt/BZ and
Pt/BA catalysts suggests that the methods of generat-
ing alkylcarbenium ions and their subsequent trans-
formation are determined by the BAC strength and the
NLAC/NBAC ratio in borate-containing catalysts. The
alkylcarbenium ions obtained from alkenes with the
participation of the strong BACs of the Pt/BA sample
(Table 1) undergo moderate cracking because in this
case NLAC/NBAC = 0.3. Due to the presence of weak
BACs in the Pt/BZ sample and as a consequence of its
NLAC/NBAC ratio = 1.0 (Table 1), the formation of
alkylcarbenium ions is supplementarily provided by
the LACs. Since the abstraction of the hydride ion
from the tertiary carbon atom in LACs occurs rather
readily, we can suppose that the alkylcarbenium ions
formed from iso-alkanes over the Pt/BZ sample
quickly undergo hydrocracking due to the alternation
of β-elimination and isomerization [59, 60].
9
8
7
6
5
4
3
2
1
1
2
2
1
0
2
4
6
8
10 12 14 16 18 20
Time, h
Fig. 7. Dependences of ΣOH/Σn-alkanes ratio in diesel
fraction on hydrocracking duration in Pt/BA (1) and
Pt/BZ (2) catalysts.
Finally, among the two borate-containing samples,
the Pt/BZ catalyst should be considered as the most
efficient oil hydrocracking catalyst in terms of the
highest content of cycloalkanes and iso-alkanes in its
diesel fraction (88.9 wt %). However, the yield of the
diesel fraction in this catalyst is considerably lower than
in Pt/BA (Table 4). Therefore the possibility of the stable
production of the diesel fraction with the maximal yield
and enhanced OH content will play a key role in the
choice of the optimal borate-containing catalyst.
ΣOHС10–С14, wt %
ΣOHС15–С20, wt %
60
100
80
60
40
20
1
2
3
4
4
50
40
30
20
10
0
3
Assessment of the Stability of Borate-Containing
Catalysts in the Hydrocracking Process
of Sunflower-Seed Oil
2
1
It was found in the course of the resource testing of
the Pt/BA and Pt/BZ samples that oil hydrocracking in
these catalysts for 20 h leads to the formation of C5+ prod-
ucts free of oxygen-containing compounds. The yield
of these compounds in the Pt/BA catalyst is constant
and constitutes 86.1 wt % (Fig. 6, dependence 1),
whereas the yield in the Pt/BZ sample gradually
increases to 85.3 wt % (dependence 2). The yield of
the diesel fraction increases up to 81.4 wt % in Pt/BA
(dependence 3) and up to 74.4 wt % in Pt/BZ (depen-
dence 4). The total OH content in the diesel fractions
obtained in Pt/BA and Pt/BZ decreases to 28.3 and
60.7 wt %, respectively, during hydrocracking, while
the ratio of ΣOH/Σn-alkanes drops to 0.39 and 1.54,
respectively (Fig. 7). The latter fact indicates the deac-
tivation of borate-containing catalysts. The deactiva-
tion can be caused by two reasons: degradation of the
metallic function of the catalyst due to the agglomera-
tion of the platinum particles under hydrothermal
conditions and the decrease of its acidity due to the AC
being blocked by the unsaturated products of the
alkene oligomerization. In the case of the deactivation
of Pt/BA, both reasons probably come into play
because the C15–C18 alkenes appear over time and the
0
2
4
6
8
10 12 14 16 18 20
Time, h
Fig. 8. Dependence of ΣOH
(1, 2) and
C10–C14
ΣOH
(3, 4) in diesel fraction on hydrocracking
C15‒C20
duration over Pt/BA (1, 3) and Pt/BZ (2, 4) catalysts.
borate-containing catalysts that have attracted interest
for further investigation.
It should be noted that the diesel fraction obtained
over Pt/BZ has an ΣOH/Σn-alkane ratio larger than
that in Pt/BA (Table 4). This is explained by two fea-
tures. In the first place, iso-alkanes + cycloalkane
C10–C14 and iso-alkanes C17–C18 provide the largest
contribution to the sum of the OHs formed in the
Pt/BZ and Pt/BA catalysts, respectively (Table 4). In
the second place, ΣOH C10–C14 (82.6 wt %) > ΣOH
C17–C18 (39.1 wt %). This fact, in turn, indicated that
OHs with a carbon atom number ≤18 undergo HC–
AC side reactions over the Pt/BZ catalyst more
actively than over the Pt/BA sample. Note that a sim-
ilar conclusion was drawn previously based on the data total content of the C10–C14 (Fig. 8, curve 1) and C15–C20
on the content of C4 hydrocarbons in the gaseous
OH (curve 3) decreases in the resultant diesel frac-
KINETICS AND CATALYSIS
Vol. 58
No. 5
2017

HYDROCRACKING VEGETABLE OIL ON BORATE-CONTAINING CATALYSTS
575
tions. Consequently, the activity of the acid centers in
the Pt/BA catalysts responsible for the hydrocracking
(HC–AC) and isomerization reactions decreases. The
deactivation of the Pt/BZ leads to a decrease of the
total content of OH C10–C14 in the resultant diesel
fraction (Fig. 8, curve 2) and, in contrast, the total
content of OH C15–C20 increases (curve 4). Thus, a
decrease of activity occurs only for centers that provide
HC–AC reactions.
In the course of oil hydrocracking in the Pt/BZ and
Pt/BA catalysts over 20 h, the yield of the diesel frac-
tion increases to 74.4 and 81.4 wt %, while the total
content of the iso-alkanes and cycloalkanes decreases
to 60.7 and 28.3 wt %, respectively. The Pt/BZ catalyst
is the most efficient.
ABBREVIATIONS
The supposed reasons for the deactivation of
borate-containing oil hydrocracking catalysts can be
refined by studying samples that have operated for
more than 100 h. Phase composition, the dispersity of
the hydrogenating component, and the nature and
number of surface organic deposits, as well as the tex-
tural and acid properties of these samples will be con-
sidered in our subsequent publications.
As a whole, in spite of the fact that the yield of the
diesel fraction in the Pt/BZ sample after 20 h of testing
is 7% lower than in the Pt/BA sample; the content of
OH in this fraction in the first sample is much higher
(by 32.4 wt %) than in the second sample. Conse-
quently, we can draw the conclusion that Pt/BZ
should be considered as the best catalyst for oil hydro-
cracking at 380°C. At the same time, we showed in our
previous work [23] that processing for 20 h at 400°C in
the Pt/BA sample resulted in a diesel fraction with a
yield of 82.0 wt % and OH content of 76.1 wt %. Con-
sequently, both samples, Pt/BZ and Pt/BA, can
behave as efficient catalysts for the one-stage hydroc-
racking sunflower-seed oil.
BAC
H–DH2O
HH–MC
Bronsted acidic centers;
Hydrogenation–dehydration reactions;
Hydrogenolysis of C5+ hydrocarbons on
metallic centers;
Hydrodeoxygenation;
Hydrodecarbonylation;
Hydrocracking;
Hydrocracking of C5+ hydrocarbons on
acidic centers;
Hydrocracking of triglycerides;
Dehydrogenation;
HDO
HDCO
HC
HC–AC
HC–TG
DH
DCO2
OHs
Decarboxylation;
Other hydrocarbons;
Acidic centers.
Lewis acidic centers;
Metallic centers;
Triglycerides.
ACs
LACs
MCs
TGs
ACKNOWLEDGMENTS
CONCLUSIONS
Yu. A. Chumachenko thanks Dr. A.V. Lavrenov for
the statement of the problem and tasks of this work.
Platinum catalysts in oxide (Z, ZA, A) and borate-
containing (BZ, BA) supports provide the hydroc-
racking of sunflower-seed oil at 380°C, 4.0 MPa, and The authors thank E.N. Kudrya for the determination of
stock feed rate of 1.0 h–1 to give deoxygenated C5+
products with a yield of 81.7–87.3 wt %.
the qualitative and fractional composition of the C5+
products of oil hydrocracking, Dr. N.N. Leont’eva for
the X-ray powder diffraction analysis of the catalysts,
and E.A. Buluchevskii for his help in testing the cata-
lysts. This work was performed with the use of equip-
ment of the Omsk Shared Facility Center, Siberian
Branch, Russian Academy of Sciences.
The Pt/Z, Pt/ZA, and Pt/A catalysts contain a
lower average AC (121, 105, and 164
μmolNH g,
3
respectively), and include only LACs. Decarboxyl-
ation reactions prevail in oil hydrocracking over the
Pt/Z and Pt/ZA catalysts, while the hydrogenation–
dehydration and decarboxylation reactions are pre-
dominant over the Pt/A catalyst. The yield of the die-
sel fraction in these catalysts reaches 71.8–73.9 wt %.
However, the enhanced content of n-alkanes (94.0–
95.9 wt %) confines the possibility of its direct appli-
cation as a fuel.
REFERENCES
1. García-Dávila, J., Ocaranza-Sánchez, E., Rojas-
López, M., Muñoz-Arroyo, J.A., Ramírez, J., and
Martínez-Ayala, A.L., Fuel, 2014, vol. 135, p. 380.
2. Bezergianni, S., Dimitriadis, A., and Chrysikou, L.P.,
One-stage oil hydrocracking with the predomi-
nance of the hydrodecarbonylation and hydrogena-
tion–dehydration reactions proceed in the Pt/BZ and
Pt/BA samples, which contain a larger average AC
Fuel, 2014, vol. 118, p. 300.
3. Baladincz, P. and Hancsók, J., Chem. Eng. J., 2015,
vol. 282, p. 152.
4. Bezergianni, S. and Dimitriadis, A., Renewable Sus-
(258 and 293
respectively) and include
μmolNH g,
tainable Energy Rev., 2013, vol. 21, p. 110.
3
BACs along with LACs. The yield of the diesel fraction
over them is 55.6 and 75.1 wt %, respectively, while the
total content of iso-alkanes and cycloalkanes in this
fraction reaches 88.9 and 63.5 wt %, respectively.
5. Zhang, H., Lin, H., and Zheng, Y., Appl. Catal., B,
2014, vols. 160–161, p. 415.
6. Studentschnig, A.F.H., Schober, S., and Mittelbach,
M., Energy Fuels, 2013, vol. 27, p. 7480.
KINETICS AND CATALYSIS Vol. 58 No. 5 2017

576
CHUMACHENKO et al.
7. Kubička, D. and Kaluža, L., Appl. Catal., A, 2010, 33. Turakulova, A.O., Golubina, E.V., Lokteva, E.S.,
vol. 372, p. 199.
Korotkov, A.V., and Lunin, V.V., Russ. J. Phys. Chem.,
2011, vol. 85, no. 3, p. 402.
8. Kubička, D., Bejblová, M., and Vlk, J., Top. Catal.,
2010, vol. 53, p. 168.
34. Cerrato, G., Bordiga, S., Barbera, S., and Morterra, C.,
Appl. Surf. Sci., 1997, vol. 115, p. 53.
9. Herskowitz, M., Landau, M.V., Reizner, Y., and
Berger, D., Fuel, 2013, vol. 111, p. 157.
35. Ravichandran, A.T., Catherine Siriya Pushpa, K., Rav-
ichandran, K., Karthika, K., Nagabhushana, B.M.,
Mantha Srinivas, and Swaminathan, K., Superlattices
Microstruct., 2014, vol. 75, p. 533.
36. Morterra, C., Cerrato, G., Ferroni, L., Negro, A., and
Montarano, L., Appl. Surf. Sci., 1993, vol. 65–66,
p. 257.
37. Sinhamahapatra, A., Pal, P., Tarafdar, A., Bajaj, H.C.,
and Panda, A.B., ChemCatChem, 2013, vol. 5, p. 331.
38. Hansen, M.R., Jakobsen, H.J., and Skibsted, J., J.
Phys. Chem. C, 2008, vol. 112, p. 7210.
39. Dumeignil, F., Guelton, M., Rigole, M., Amoureux, J.-P.,
and Fernandez, C., Colloids Surf., A, 1999, vol. 158, p. 75.
40. Moreau, S., Gervais, M., and Douy, A., Solid State
Ionics, 1997, vols. 101–103, p. 625.
41. Tanabe, K., Misono, M., Ono, Y., and Hattori, H.,
New Solid Acids and Bases: Their Catalytic Properties,
Amsterdam: Elsevier, 1989.
42. Sato, S., Kuroki, M., Sodesawa, T., Nozaki, F., and
Maciel, G.E., J. Mol. Catal. A: Chem., 1995, vol. 104,
p. 171.
10. Kikhtyanin, O.V., Rubanov, A.E., Ayupov, A.B., and
Echevsky, G.V., Fuel, 2010, vol. 89, p. 3085.
11. Chen, N., Gong, S., Shirai, H., Watanabe, T., and
Qian, E.W., Appl. Catal., A, 2013, vol. 466, p. 105.
12. Zhao, S., Li, M., Chu, Y., and Chen, J., Energy Fuels,
2014, vol. 28, p. 7122.
13. Qin, Y., He, L., Duan, J., Chen, P., Lou, H., Zheng, X.,
and Hong, H., ChemCatChem, 2014, vol. 6, p. 2698.
14. Monnier, J., Sulimma, H., Dalai, A., and Caravaggio, G.,
Appl. Catal., A, 2010, vol. 382, p. 176.
15. Han, F., Guan, Q., and Li, W., RSC Adv., 2015, vol. 5,
p. 107533.
16. Sugami, Y., Minami, E., and Saka, S., Fuel, 2016,
vol. 166, p. 376.
17. Wang, C., Liu, Q., Songa, J., Li, W., Li, P., Xu, R., Ma, H.,
and Tian, Z., Catal. Today, 2014, vol. 234, p. 153.
18. Smirnova, M.Yu., Kikhtyanin, O.V., Smirnov, M.Yu.,
Kalinkin, A.V., Titkov, A.I., Ayupov, A.B., and Erma-
kov, D.Yu., Appl. Catal., A, 2015, vol. 505, p. 524.
19. Duan, J., Han, J., Sun, H., Chen, P., Lou, H., and
Zheng, X., Catal. Commun., 2012, vol. 17, p. 76.
20. Ishihara, A., Fukui, N., Nasu, H., and Hashimoto, T.,
Fuel, 2014, vol. 134, p. 611.
21. Liu, Y., Sotelo-Boyás, R., Murata, K., Minowa, T.,
and Sakanishi, K., Chem. Lett., 2009, vol. 38, no. 6,
p. 552.
22. Huber, G.W., O’Connor, P., and Corma, A., Appl.
Catal., A, 2007, vol. 329, p. 120.
23. Chumachenko, Yu.A., Lavrenov, A.V., Buluchevskii,
E.A., Arbuzov, A.B., Gulyaeva, T.I., and Drozdov,
V.A., Catal. Ind., 2016, vol. 8, p. 56.
24. Cortez-Lajas, L.A., Hernández-Enríquez, J.M., Cas-
tillo-Mares, A., Rivera-Armenta, J.L., Sandoval-
Robles, G., García-Serranoy, L.A., and García-Ala-
milla, R., Rev. Mex. Ing. Quim., 2006, vol. 5, p. 321.
25. De Farias, A.M.D., Esteves, A.M.L., Ziarelli, F., Cal-
darelli, S., Fraga, M.A., and Appel, L.G., Appl. Surf.
Sci., 2004, vol. 227, p. 132.
26. Xu, B.-Q., Cheng, S.-B., Zhang, X., and Zhu, Q.-M.,
Catal. Today, 2000, vol. 63, p. 275.
27. Curtin, T., McMonagle, J.B., and Hodnett, B.K., Appl.
Catal., A, 1992, vol. 93, p. 91.
28. Priecel, P., Čapek, L., Kubička, D., Homola, F.,
Ryšánek, P., and Pouzar, M., Catal. Today, 2011,
vol. 176, p. 409.
29. Anand, M., Farooqui, S.A., Kumar, R., Joshi, R.,
Kumar, R., Sibi, M.G., Singh, H., and Sinha, A.K.,
Appl. Catal., A, 2016, vol. 516, p. 144.
43. Urbano, F.J., Aramendia, M.A., Marinas, A., and
Marinas, J.M., J. Catal., 2009, vol. 268, p. 79.
44. Xu, B.-Q., Cheng, S.-B., Jiang, S., and Zhu, Q.-M.,
Appl. Catal., A, 1999, vol. 188, p. 361.
45. Xu, B.-Q. and Cheng, S.-B., Acta Phys.-Chim. Sin.,
2001, vol. 17, p. 443.
46. Liang, Z., MS Thesis, Singapore: National Univ. of
Singapore, 2009.
47. El-Hakam, S.A. and El-Sharkawy, E.A., Mater. Lett.,
1998, vol. 36, p. 167.
48. Peil, K.P., Galya, L.G., and Marcelin, G., J. Catal.,
1989, vol. 115, p. 441.
49. Reta, J.L., García, R., Sandoval, G., and Rivera, J.L.,
Rev. Mex. Fis. S, 2009, vol. 55, p. 98.
50. Bautista, F.M., Campelo, J.M., Garcia, A., Luna, D.,
Marinas, J.M., Moreno, M.C., and Romero, A.A.,
Appl. Catal., A, 1998, vol. 170, p. 159.
51. Lavrenov, A.V. and Duplyakin, V.K., Kinet. Catal.,
2009, vol. 50, no. 2, p. 235.
52. Li, C. and Chen, Y.-W., Catal. Lett., 1993, vol. 19, p. 99.
53. Senol, O.I., Ryymin, E.M., Viljava, T.R., and
Krause, A.O.I., J. Mol. Catal. A: Chem., 2007, vol. 277,
p. 107.
54. He, Z. and Wang, X., Catal. Sustainable Energy, 2013,
vol. 1, p. 28.
55. Yang, Y., Chen, J., and Shi, H., Energy Fuels, 2013,
vol. 27, p. 3400.
56. Akhmedov, V.M. and Al-Khowaiter, S.H., Catal. Rev.,
2007, vol. 49, p. 33.
30. Paukshtis, E.A., Infrakrasnaya spektroskopiya v getero-
gennom kislotno-osnovnom katalize (Infrared Spectros-
copy Applied to Heterogeneous Acid–Base Catalysis),
Novosibirsk: Nauka, 1992.
31. Korontsevich, A.Yu., Molodozhenyuk, T.B.,
Vorob’ev, V.N., and Abidova, M.F., Uzb. Khim. Zh.,
1991, no. 4, p. 17.
32. Murav'eva, G.P., Fionov, A.V., Lunina, E.V., Turaku-
lova, A.O., Oleinikov, N.N., and Lunin, V.V., Dokl.
Chem., 2000, vol. 371, nos. 1–3, p. 39.
57. Chen, J., Shi, H., Li, L., and Li, K., Appl. Catal., B,
2014, vol. 144, p. 870.
58. El-Sharkawy, E.A., Al-Shihry Shar, S., and Ahmed, A.I.,
Adsorpt. Sci. Technol., 2003, vol. 21, p. 863.
59. Weitkamp, J., Appl. Catal., 1983, vol. 8, p. 123.
60. Souverijns, W., Martens, J.A., Froment, G.F., and
Jacobs, P.A., J. Catal., 1998, vol. 174, p. 177.
Translated by I. Kudryavtsev
KINETICS AND CATALYSIS
Vol. 58
No. 5
2017