Effect of promotion of nickel sulfide catalyst with silver on kinetics of decarbonilation of stearic acid

Article abstract of DOI:10.1007/s11172-018-2359-1

The kinetics of liquid-phase decarbonylation of stearic acid in n-dodecane on γ-Al₂O₃ supported nickel sulfide catalyst promoted with silver was experimentally studied at 350 °C. The parameters of the reaction steps were determined and a structural kinetic model was developed. The model was compared with an earlier developed kinetic model for the unpromoted catalyst. It was suggested that an increased reaction selectivity in the presence of silver promoted catalyst was caused by a change in the composition of the adsorption complexes formed by the active sites of the catalyst. This change in the composition of the complexes is probably associated with an increase in the average size of the surface active particles of the catalyst.

Full text of DOI:10.1007/s11172-018-2359-1

Russian Chemical Bulletin, International Edition, Vol. 67, No. 12, pp. 2224—2229, December, 2018
2224
Effect of promotion of nickel sulfide catalyst with silver on kinetics
of decarbonilation of stearic acid
E. A. Katsman, A. S. Berenblyum, V. Ya. Danyushevsky, V. M. Karpov, P. S. Kuznetsov, S. V. Leont´eva, and V. R. Flid
MIREA — Russian Technological University, M. V. Lomonosov Institute of Fine Chemical Technology,
86 prosp. Vernadskogo, 119571 Moscow, Russian Federation.
E-mail: katsman@aha.ru
The kinetics of liquid-phase decarbonylation of stearic acid in n-dodecane on γ-Al2O3
supported nickel sulfide catalyst promoted with silver was experimentally studied at 350 C. The
parameters of the reaction steps were determined and a structural kinetic model was developed.
The model was compared with an earlier developed kinetic model for the unpromoted catalyst.
It was suggested that an increased reaction selectivity in the presence of silver promoted catalyst
was caused by a change in the composition of the adsorption complexes formed by the active
sites of the catalyst. This change in the composition of the complexes is probably associated
with an increase in the average size of the surface active particles of the catalyst.
Key words: decarbonylation, stearic acid, higher olefins, reaction kinetics, kinetic model,
catalyst promotion, inhibition, heterogeneous catalysis.
The scale of production of higher olefins with an even
number of carbon atoms by oligomerization of ethylene is
very high. The reason is that these olefins are key raw
materials for surfactants,1 which are the bases of synthetic
detergents, synthetic grease oils and additives to them,2
high-temperature heat carriers, as well as for a number of
other petrochemical products. As a rule, higher olefins are
part of narrow, for example С14—С16, or wider fractions.
Unfortunately, due to some features in distribution of
ethylene oligomerization products upon the length of
carbon chain, С16—С22 fraction is formed in insufficient
quantities and cannot fully meet the industrial needs in
these compounds.3 Industrial synthesis of individual higher
olefins with an odd number of carbon atoms is not im-
plemented.
Another source for the production of higher olefins is
processing of renewable resources which have either veg-
etable or animal origin.3 This enables one to obtain valu-
able materials by hydrolysis (glycerin and fatty acids,
mainly stearic (St) and oleic acids). Nickel sulfide sup-
ported catalysts emerged as effective systems for selective
decarbonylation of acids to linear olefins C17 in a hydrogen
medium.3—5 The catalysts show reduced activity in the
side reaction of hydrogenation of olefins and a property
to effectively inhibit oligomerization.6,7
tion on the selectivity of the nickel sulfide catalyst in St
conversion to heptadecenes (Hde).
Experimental
Reagents, solvents and materials. Nickel sulphate (NiSO4•
•7H2O, special purity grade, USSR state standart GOST 4465-
74, 99.5%), silver nitrate (AgNO3, Fluka, 99%), sodium hydr-
oxide (special purity grade, USSR state standart GOST 4328-77,
98.0%), sulfuric acid (reagent grade, 94%), dodecane (high
purity grade, Himmed, Russia, 99%), tridecane (high purity
grade), chloroform (reagent grade), ethyl alcohol (rectificate
96%), strearic acid (analytical grade, USSR state standart GOST
9419-78, 99.2%), hydrogen (A, 99.99%), nitrogen (special pu-
rity grade, 1 grade, 99.999%), and carbon monoxide (high pu-
rity grade, 99.8%) were used. As a support, γ-Al2O3 (Dne-
prodzerginsk, USSR state standart GOST 8136-85, AOA-1, ex-
trudate, specific surface area determined by BET is 254 m2 g–1
was applied. All reactants and solvents were used as received.
)
Catalyst synthesis The method and conditions used to prepare
the nickel sulfide catalyst, to promote it with silver, and to reduce
by hydrogen were described earlier.5 The size of the catalyst
granules was 0.08—0.2 mm, the molar ratio of Ag : Ni was 0.87
at Ni content of 3.08%.
It was previously shown5 that the addition of silver increased
the average size of surface catalytic particles from 4.7 to 11.9 nm.
At the same time, the peaks typical of metallic silver were found
in the XRD patterns, while relatively small amounts of NiS and
Ni3S2 sulfides were identified. Hence, it can be assumed that
nickel and silver do not directly interact with each other, although,
the catalytic data5 showed that they were not spatially separated.
The absence of reflections from metallic nickel in difractograms
can be explained by the small size of nickel crystallites and their
low concentration.
Promoting such catalysts with silver significantly in-
creases the selectivity towards olefins and little increase in
St conversion occurs; however, the reasons of the promot-
ing effect are still unclear.3,5
In the present work, an attempt is made to use struc-
tural kinetic modeling to explain the effect of silver addi-
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 12, pp. 2224—2229, December, 2018.
1066-5285/18/6712-2224 © 2018 Springer Science+Business Media, Inc.

Influence of silver promotion on kinetics
Russ. Chem. Bull., Int. Ed., Vol. 67, No. 12, December, 2018
2225
Table 1. Planned conditions for the kinetic experiments on de-
carbonylation of St in the presence of silver promoted nickel
sulfide catalyst supported on γ-alumina oxide at 350 C
from those recommended by the plan. Based on our earlier stud-
ies6,7 we can conclude that the proposed plan is sufficient to
establish the structure and calculate the parameters of the kinetic
model of the given reaction.
Series P(H2)a C(St)a С(Ni)a,b
t/min
Note
In experiments, in which the reaction time was taken as 0 min,
the autoclave was heated to the operating temperature for 15 min,
then rapidly cooled to ~20 C, and the reaction mixture includ-
ing the gas phase products was analyzed.
It is assumed that the results of the analysis of such a mixture
can be taken as the initial concentrations of the reactants. In this
case, the contribution of the chemical transformation into the
non-isothermal mode by heating the reaction mixture to the
operating temperature is excluded. In other experiments, after
reaching the specified temperature, the reaction was allowed to
proceed for a specified time, after which the reactor was rapidly
cooled to ~20 C and the products were analyzed.
1
2
3
4
5
6
7
8
15
10
17
15
15
15
10
17
0.6
0.6
0.6
0.6
1.2
0.6
0.6
0.6
0.02 0, 30, 60, 120
—
0.02
0.02
0.02
0.02
0.04
0.02
0.02
0, 60
0, 60
0, 60
0, 60
0, 30
0, 60
0, 60
—
—
—
—
—
Р(СО) = 5
C(H2O) = 1.1
a Concentrations (mol L–1) and gas pressures (atm.) are calcu-
lated at room temperature.
Analysis of the reaction products. The techniques used for the
analysis were described in detail earlier.6,7 The concentration of
unreacted St was determined by acid-base titration with a color
indicator in a mixture of chloroform and ethanol. Gas-liquid
chromatography with an internal standard (tridecan) was used
to measure the concentrations of heptadecenes (Hde), hepta-
decane (Hd), and St, with St being methylated in the reaction
mixture. The amount of Н2, СО, СО2, N2, and trace amounts
of methane were determined by adsorption chromatography with
an absolute calibration using pure gases. Chromato-mass spec-
trometry and NMR spectroscopy were used to analyze the
products with minor concentration, for example, heptadecyl
stearate (Es) and diheptadecyl ketone (Kn). The values of the
concentrations of the main components of the reaction mixture,
calculated from the analytical results are shown in Table 2.
Construction and analysis of the kinetic model. Methods and
algorithms for data processing, the choice of an adequate ki-
netic model, as well as the analysis of the single-valuedness and
error in determining its parameters are set out in our earlier
b Calculated concentration of the catalyst based on nickel content.
Methods for kinetic experiments. Methods of conducting
kinetic experiments, as well as the experiments indicating the
kinetic mode of the reaction, are described in our earlier publica-
tion.6 A 50 mL autoclave (Autoclave Engineers Inc.) with a stirrer
and low-inertia heating with Wood's alloy was used as a reactor.
Planning kinetic experiments. One of the main goals of ex-
periment planning is to obtain maximum information about the
influence of input factors on output indicators (for example, the
concentration of the substrate on the selectivity of the reaction)
with the minimum number of experiments. Each series in Table 1
contains the planned values of the initial concentrations or pres-
sures of the reactants, as well as the sampling time for analyzes.
We used four various initial hydrogen pressure levels P(H2), two
carbon monoxide pressure values P(CO), and two initial concen-
trations of St C(St), water C(H2O), and catalyst C(Ni) (for Ni).
The level of factors in this plan is set by a researcher. However,
the plan was not drawn up according to the rules of classical
experiment planning, since mathematical models of the kinetics
of chemical reactions are not linear in parameters. Moreover,
unlike the classical one, this plan is not required to be performed
exactly, i.e., the values of factors, for example, the initial hydro-
gen pressure or the reaction time, may be somewhat different
publications.
6—10 The mathematical model is a system of ordinary
differential equations with the values of starting concentrations
as initial conditions (the Cauchy problem). The search for an
adequate kinetic model was carried out according to the classical
scheme of hypothesis suggestion and testing. The considered
hypotheses included only structural models. Hypothesis forma-
Table 2. Measured consentrations (mol L–1) of the components of St decarbonylation (the results of
calculation by the kinetic model are in brackets)*
Series
t/min
C(St)
C(Hde)
C(CO)
C(Hd)
C(CO2)
1
30
60
90
120
60
60
60
60
30
60
60
0.295 (0.296)
0.229 (0.219)
0.193 (0.170)
0.121 (0.135)
0.330 (0.293)
0.360 (0.327)
0.316 (0.325)
0.551 (0.540)
0.236 (0.202)
0.339 (0.305)
0.451 (0.478)
0.207 (0.217)
0.214 (0.246)
0.244 (0.250)
0.268 (0.244)
0.172 (0.189)
0.122 (0.136)
0.145 (0.108)
0.258 (0.258)
0.232 (0.261)
0.162 (0.155)
0.085 (0.099)
0.301 (0.260)
0.319 (0.329)
0.422 (0.372)
0.471 (0.403)
0.256 (0.275)
0.210 (0.217)
0.179 (0.170)
0.526 (0.431)
0.349 (0.335)
1.300 (1.374)
0.155 (0.146)
0.031 (0.024)
0.010 (0.011)
0.044 (0.044) 0.015 (0.014)
0.068 (0.063)
0.079 (0.079)
0.024 (0.031)
0.020 (0.022)
0.023 (0.020)
0.052 (0.053)
0.032 (0.040)
0.021 (0.020)
0.014 (0.012)
0.016 (0.015)
0.018 (0.017)
0.015 (0.015)
0.016 (0.017)
0.028 (0.026)
0.036 (0.035)
0.006 (0.006)
0.016 (0.018)
0.018 (0.019)
2
3
4
5
6
7
8
* The concentrations are determined concerning the volume liquid phase under the conditions of the
kinetic experiments; Hde are heptadecenes, Hd is heptadecane.

2226 Russ. Chem. Bull., Int. Ed., Vol. 67, No. 12, December, 2018
Katsman et al.
tion was based on the Langmuir mechanism with the following
assumptions:
decarboxylation of An to diheptadecyl ketone (Kn)
An
Kn + CO2,
(6)
(7)
(8)
1) an active site can adsorb more than one molecule,
2) the vapor-gas portion of the system (СО, СО2, Н2О, Н2,
N2) was considered as an ideal gas,
water gas shift reaction
3) the solubility of gases and vapors follows the Henrys law.
׳
CO + H2O
CO2 + H2,
The first assumption, in contrast to the classical theory of
Langmuir adsorption, creates a large number of possible com-
positions of adsorption complexes, and a change in the compo-
sition of at least one adsorption complex generates a new hypo-
thesis. The same applies to other details of the mechanism
scheme, for example, the kinetic order of single steps. The sug-
gested hypotheses were tested by the generally accepted least
squares method. The weighted sum of squares of the deviations
of reactant concentrations, which were calculated by the model
from the values obtained experimentally, was used as an objective
function.10 As g weight, we used the inverse square of the error
in determining the corresponding concentration g = 1/2.
The objective function was minimized by the configuration
method, one of the direct search methods that does not assume
the need to calculate the derivatives of the objective function
from the desired parameters. The initial approximations for the
rate constants and equilibria were formed according to known
recommendations.10 For reliability, the minimization of the
objective function was performed using several initial approxima-
tions. When a hypothesis did not provide an adequate description
of the experimental data by the model it was rejected using the
Fisher adequacy criterion. In total, over 100 hypotheses were
considered. If more than one hypothesis turned out to be ade-
quate, the simplest hypothesis was chosen.
dimerization of Hde
2 Hde
Dime.
The characteristics of the active phase of silver pro-
moted nickel sulfide catalyst, as well as the effect of the
Ag : Ni ratio on the reaction parameters were disclosed
earlier.5 The characteristics refer to the size, shape, struc-
ture, and composition of the supported catalytic particles.
It was found that the main routes of St transformations in
the presence of the promoted catalyst were the same as
those in the presence of the unmodified catalyst, although
the addition of silver increased St conversion by a factor
of 1.4. Even more important parameter for the assessment
of the efficiency of the catalyst for decarbonylation was
a 12% increase in selectivity towards olefin. In addition,
in the presence of silver promoted nickel sulfide catalysts,
the rates for St decarboxylation (3) and water gas shift
reaction (7) were lower, hence little CO2 was formed
and its fraction in the selective transformation of St was
only 2—7%.
The mean square relative error of the description of the ex-
perimentally determined concentrations by the mathematical
model was 5%, which is close to the reproducibility of measure-
ments. The characteristics of this model, i.e., the stoichiometric
and kinetic equations of the stages, the values of the equilibrium
constants of adsorption, and the rate constants of single steps, as
well as interpretation are given in the next section.
In order to understand the nature of the promoting
effect of silver, the kinetic models of the decarbonylation
of stearic acid over unpromoted and silver promoted
catalysts can be compared.
As structural kinetic modeling for silver promoted
nickel sulfide catalyst shows, the adsorption complexes of
the reactants with the active sites Z are formed by the fol-
lowing equations:
Results and Discussion
Z + St + 3 H2 = Z(St)(H2)3,
(I)
In our previous publication6 we described the main
transformations of St over the unpromoted nickel-sulfide
catalyst:
Z + St + 2 H2O = Z(St)(H2O)2,
(II)
decarbonylation of St
Z + Hde + St + H2O + H2 =
= Z(Hde)(St)(H2O)(H2),
St
Hde + CO + H2O,
(1)
(2)
(3)
(4)
(5)
(III)
(IV)
(V)
hydrogenation of Hde
Hde + H2
Hd,
Z + H2O + H2 = Z(H2O)(H2),
Z + 2 CO + H2 = Z(CO)2(H2),
Z + 2 H2 + H2O = Z(H2)2(H2O).
decarboxylation of St
St
Hd + CO2,
(VI)
formation of anhydride of steric acid (An)
2 St
An + H2O,
With roman numbers we denoted both the numbers of
the steps of adsorption and the corresponding adsorption
complexes. There appeared to be six, rather than eight
steps, as assumes the kinetic model for the nickel sulfide
formation of heptadecyl stearate (Es)
St + Hde
Es,

Influence of silver promotion on kinetics
Russ. Chem. Bull., Int. Ed., Vol. 67, No. 12, December, 2018
2227
catalyst, and certain differences were observed in the
composition of these complexes. Thus, according to the
kinetic model for the promoted catalyst, the adsorption
complex I formed in the St decarbonylation step contains
an additional hydrogen molecule. Besides, the adsorption
complex VI, which participates in Hde hydrogenation,
includes a water molecule. The corresponding values of
the adsorption equilibrium constants obtained using
a numerical kinetic model are presented below.
deoxygenation, expressed using concentration of reagents,
are as follows:
decarbonylation
w1 = k1KIC(St)C(H2)3C(Ni)/Den,
hydrogenation
w2 = [k2(1)KIIIC(Hde)C(St)C(H2O)C(H2)C(Ni) +
+ k2(2)C(Hde)KVIC(H2)2C(H2O)C(Ni)]/Den,
oligomerization
Equation
(I)
K
8.24•102
3.01•104
7.93•103
1.24•107
3.73•105
5.03•105
(II)
w8 = k8C(Hde)[C(H2)C(Ni)]–0.5
,
(III)
(IV)
(V)
where kj is rate constant of the step j, Ki is the equilibrium
constant of adsorption i,
(VI)
Den = 1 + KIC(St)C(H2)3 + KIIC(St)C(H2O)2 +
+ KIIIC(St)C(Hde)C(H2O)C(H2) + KIVC(H2O)C(H2) +
+ KVC(CO)2C(H2) + KVIC(H2)2C(H2O).
High values of the constants correspond to the almost
complete coverage of the catalytic active sites with reactant
molecules, so that the number of unoccupied active sites
is less than one per 107 sites. Based on the ratio of the
adsorption equilibrium constants, one can conclude that
the concentration of adsorption complexes with Z(St)(H2)3
composition, which are intermediate in decarbonylation,
is four orders higher than the concentration of unoccupied
active sites. However, the major part of the active sites is
combined in with the Z(H2O)(H2) complex.
The total concentration of active sites is conventionally
taken to be equal to С(Ni). The difference between the
reactions occurring over the unpromoted and promoted
catalysts is mainly governed by the difference in the com-
positions of the adsorption complexes I and VI.
It is known8—10 that the problem of calculating the
values for the parameters of a mathematical model often
has many solutions. Therefore, the same calculated con-
centration values for the reactants can correspond to
a multitude sets of rate constants and equilibrium con-
stants. This ambiguity can be accounted for by numerous
reasons. Among these are the stationary state of the system
or the quasi-equilibrium of adsorption before the decom-
position of the adsorption complex, as described in detail
earlier.8 In this case, not all the parameters of the model,
Table 3 shows the kinetic equations of selected stages,
as well as the values of the rate constants and equilibrium
constants. The equations of the main stages include the
concentration of adsorption complexes.
Hydrogenation of heptadecenes proceeds in parallel
by two mechanisms: 2 (1) is Langmuir mechanism, which
proceeds via decomposition of the adsorption complex III;
2 (2) is Ili—Riedel mechanism which involves the interac-
tion of the olefin with the adsorption complex VI. At the
same time, the kinetic equations of the main stages of St
Table 3. Kinetic equations and the values of the rate constants and equilibrium constants of the
stages*
Step
Kinetic equation
k
K
1
w1 = k1C(I)
w2(1) = k2(1)C(III){1 – C(H2)C(Hd)/[K2C(Hde)]}
w2(2) = k2(2)C(Hde)C(VI){1 – C(H2)C(Hd)/[K2C(Hde)]}
w3 = k3C(II)
1.56•103
3.66•102
5.28•100
4.35•101
1.59•10–3
1.91•10–3
1.93•101
~0
—
3.78•100
3.78•100
—
2 (1)
2 (2)
3
4
5
6
7
8
w4 = k4[C(St)2 – C(An)C(H2O)/K4]
w5 = k5[C(St)C(Hde) – C(Es)/K5]
w6 = k6C(An)
1.72•10–5
2.51•100
—
Contribution of water gas shift reaction is insignificant
1.30
—
w8 = k8C(Hde)[C(Ni)C(H2)]–0.5
1.22•10–4
* Concentrations of liquid and gaseous compounds are represented in mol L–1, time is in min,
catalyst loading (C(Ni)) is present in moles of Ni per 1 L of liquid phase, the values of the rate
constants and equilibrium constants are represented in accordance with the numerical kinetic
model and their dimensions correspond to the kinetic equations.

2228 Russ. Chem. Bull., Int. Ed., Vol. 67, No. 12, December, 2018
Katsman et al.
Table 4. Unambiguosly estimated parameters and parametric
functions (complexes) of kinetic models, their values and errors
of calculations
intermediate, the kinetics of an admixture Kn formation
cannot be adequately described.
As it was found earlier, an unusual behavior of nickel
sulfide catalysts involves an increase in heptadecene se-
lectivity with increasing hydrogen pressure.3—7 This is
associated with the known ability of metal hydrides, in-
cluding nickel and silver, to inhibit the chain polymeriza-
tion/oligomerization of olefins, in particular, the radical-
chain (thermal) polymerization.3—7 This feature of hydride
behavior is consistent with the empirical kinetic equation
of the corresponding step 8 (see Table 3), which contains
negative orders of 0.5 for the catalyst and hydrogen con-
centrations.
In the group of nickel sulfide catalysts for decarbonyl-
ation of fatty acids, an unusual tendency is observed: the
loss of heptadecenes due to a secondary hydrogenation
with heptadecane formation decreases with increasing
hydrogen pressure. This was typical of the undoped cata-
lyst; however, this effect is even more pronounced in the
presence of a silver promoted catalyst. Kinetic modeling
provides a rational explanation for this fact. As follows
from the composition of the adsorption complex I (see the
scheme presented above), the target decarbonylation step
(1) is sharply accelerated and up to the third order of
magnitude with increasing pressure of hydrogen, which is
not consumed at this step. The reason is that with in-
creasing pressure of hydrogen the quasi-equilibrium
concentration of the above mentioned complex involved
in decarbonylation increases. In agreement with the com-
position of the corresponding adsorption complexes III
and VI, an increase in the rate of secondary hydrogenation
steps 2 (1) and 2 (2) is less strong. Thus, an increase in
hydrogen pressure does not only inhibit the side reaction
of oligomerization of target heptadecenes,3—7 but also
increases the rate of formation of the target heptadecenes
relative to their consequent hydrogenation. The total effect
becomes apparent in an increase in olefin selectivity with
increasing hydrogen pressure.
Line
Parametric
function
Numerical
value
Logarithmic calcul-
ation error ( )
a
b
c
d
e
f
g
h
i
KIV/(KVIk2(2)
)
4.67•100
4.84•10–1
1.59•10–3
4.93•10–1
1.22•10–4
3.32•10–4
1.09•100
1.91•10–3
1.40•10–1
3.78•100
2.51•100
0.08
0.08
0.05
0.10
0.11
0.12
0.14
0.24
0.46
0.46
0.68
k1KI/(KVIk2(2)
)
k4
k3KII/(KVIk2(2)
)
k8
k6K4
k2(1)KIII/(KVIk2(2)
)
k5
KV/(KVIk2(2)
)
j
k
K2
K5
but only some packages of parameters, the nonlinear
parametric functions (NPF), can be unambiguously de-
termined (calculated and represented in the form of "mean
error"). If the parameter value is calculated unambigu-
ously, such an estimate is referred to as "unmixed".8—10
Table 4 represents six uniquely estimated NPFs of our
kinetic model, their corresponding values (lines a, b, d, f,
g, i) and "unmixed" estimates of the sought for parameters
of the kinetic model (lines c, e, h, j, k).
The unambiguously estimated nonlinear parametric
functions ("packages", lines a, b, d, f, g, i in Table 4) include
the products of adsorption equilibrium constants and the
rate constants of the decomposition of the corresponding
adsorption complexes, as well as the ratio of adsorption
equilibrium constants. Evidently, the NPFs include several
parameters of the model, between unknown values of
which a functional relationship exists. The inevitability of
the appearance of such NPFs is associated, firstly, with
the construction of a kinetic model based on the Langmuir
mechanism which includes the assumption of a quasi-
equilibrium of the adsorption steps. Secondly, a very small
fraction of unoccupied catalytic sites must be taken into
account. It is in the last case when the contribution of
occupied sites is governed by the ratios of the adsorption
equilibrium constants rather than by the values of these
constants. Therefore, a direct comparison of the values of
model parameters for unpromoted and silver promoted
catalysts is not informative and, as a result, inadvisable.
An unambiguously determined NPF (line f in Table 4)
corresponds to the case of a quasi-equilibrium shifted to
the left (formation of stearic anhydride), which is well-
known in formal chemical kinetics, followed by the de-
composition of its product (decarboxylation of the anhy-
dride to diheptadecyl ketone). An attempt to substitute
steps 4 and 6 by a single common step, made in elaborat-
ing the hypothesis, did not result in an adequate model.
Hence, without the assumption of anhydride of St as an
Now we can try to explain why the contribution of the
side reaction of hydrogenation of olefins over a silver-
promoted catalyst is not as high as found with an unpro-
moted system. As can be seen in the diagram reflecting the
results of kinetic modeling, adsorption complex I contains
an additional hydrogen molecule as compared to the
complex formed in the presence of the unpromoted cata-
lyst. Incorporation of hydrogen in the adsorption com-
plexes formed by the active sites of the catalyst may indicate
that in the absence of hydrogen the activity of these sites
in decarbonylation is low or not evident. It is possible that
hydrogen bonding disorders the structure of the active
sites and thus contributes into an enhancement of the
catalytic activity.
Silver and nickel in silver promoted nickel sulfide
catalyst are found in supported particles of different nature,
the average size of which is 11.9 2.8 nm, and the smallest

Influence of silver promotion on kinetics
Russ. Chem. Bull., Int. Ed., Vol. 67, No. 12, December, 2018
2229
particle size found is 8 nm.5 Since the catalytic activities
of unpromoted and silver promoted nickel sulfide catalysts
are close, it is logical to assume that active sites are located
on nickel-containing particles. In the unpromoted nickel
sulphide catalyst, the average size of these particles is
4.3 1.8 nm, hence, the size of more than 90% of particles
is less than 8 nm.4 Therefore, a plausible explanation of
the difference in the composition of the formed adsorption
complexes is that in the silver promoted nickel sulfide
catalyst the sites are located on particles characterized by
a larger size than in an unpromoted catalyst. Similar find-
ings for other catalysts were described in the previously
published review.11
4. V. Ya. Danyushevsky, P. S. Kuznetsov, E. A. Katsman, A. S.
Kupriyanov, V. R. Flid, A. S. Berenblyum, Russ. Chem. Bull.,
2017, 66, 463.
5. V. Ya. Danyushevsky, V. Yu. Murzin, P. S. Kuznetsov, R. S.
Shamsiev, E. A. Katsman, E. V. Khramov, Y. V. Zubavichus,
A. S. Berenblyum, Russ. J. Phys. Chem. A, 2018, 92, 66.
6. E. A. Katsman, V. Ya. Danyushevsky, P. S. Kuznetsov, R. S.
Shamsiev, A. S. Berenblyum, Kinet. Catal., 2017, 58, 147.
7. E. A. Katsman, V. Ya. Danyushevsky, P. S. Kuznetsov, V. M.
Karpov, H. A. Al-Wadhaf, V. R. Flid, Petrol. Chem.
(Engl. Transl.),2017, 57, 1190.
8. V. G. Gorskii, E. A. Katsman, F. D. Klebanova, A. A.
Grigor´ev, Theor. Exp. Chim. (Engl. Transl.), 1987, 23, 181.
9. E. A. Katsman, A. S. Berenblyum, Paket programm dlya
postroeniya i analiza kineticheskikh modelei i ego primeneniya
[Software for Building and Analysis of Kinetic Models and its
Application], MITKHT, Moscow, 2010 (in Rissian).
10. V. G. Gorskii, Planirovanie kineticheskih experimentov [Ki-
netic Experiment Planning], Nauka, Moscow, 1984, 241 pp.
(in Russian).
This work was financially supported by the Ministry
of Education and Science of the Russian Federation
within the state task of Russian Federation (Project
No. 10.8454.2017/BP).
11. V. I. Bukhtiyarov, M. G. Slin´ko, Russ. Chem. Rev., 2001,
70, 147.
References
1. N. C. Foster, Sulfonation and Sulfation Processes, The
Chemithon Corporation, Seattle, 1997.
2. Chemistry and Technology of Lubricants, Eds R. M. Mortier,
M. Fox, S. T. Orszulik, Springer, New York, 3rd ed., 2010.
3. A. S. Berenblyum, V. Ya. Danyushevsky, P. S. Kuznetsov,
E. A. Katsman, R. S. Shamsiev, Petrol. Chem. (Engl. Transl.),
2016, 56, 663.
Received July 5, 2018;
in revised form October 31, 2018;
accepted November 1, 2018