Heat effect during the low-temperature oxidation of nickel nanoparticles in a porous matrix

Article abstract of DOI:10.1134/S0036024411110045

A model of the heat-generating reaction of the low-temperature oxidation of nickel nanoparticles in nickel-supported catalysts is discussed. We use the continuous measurement of magnetization to estimate the heat effect of the reaction, allowing us to measure the heating of nickel nanoparticles in the support matrix, due to the temperature dependence of magnetization.

Full text of DOI:10.1134/S0036024411110045

ISSN 0036ꢀ0244, Russian Journal of Physical Chemistry A, 2011, Vol. 85, No. 11, pp. 2050–2053. © Pleiades Publishing, Ltd., 2011.
Original Russian Text © P.A. Chernavskii, M.I. Ivantsov, G.V. Pankina, V.V. Lunin, 2011, published in Zhurnal Fizicheskoi Khimii, 2011, Vol. 85, No. 11, pp. 2194–2197.
BRIEF
COMMUNICATIONS
Heat Effect during the LowꢀTemperature Oxidation
of Nickel Nanoparticles in a Porous Matrix
P. A. Chernavskii, M. I. Ivantsov, G. V. Pankina, and V. V. Lunin
Faculty of Chemistry, Moscow State University, Moscow, 119991 Russia
eꢀmail: chern@kge.msu.ru
Received January 20, 2011
Abstract—A model of the heatꢀgenerating reaction of the lowꢀtemperature oxidation of nickel nanoparticles
in nickelꢀsupported catalysts is discussed. We use the continuous measurement of magnetization to estimate
the heat effect of the reaction, allowing us to measure the heating of nickel nanoparticles in the support
matrix, due to the temperature dependence of magnetization.
Keywords: heat effect, lowꢀtemperature oxidation, nickel nanoparticles, porous media.
DOI: 10.1134/S0036024411110045
INTRODUCTION
mina with pore sizes of 8, 10, and 20 nm (Engelhard),
and commercial 25% Ni/Cr O3 nickel catalyst
2
Many catalytic reactions, particularly Fischer–
Tropsch synthesis, are accompanied by a considerable
(
Engelhard). A 20ꢀmg weighted portion of the catalyst
was placed in the measuring cell of a vibratingꢀsample
magnetometer, which simultaneously served a flowꢀ
through microreactor [2]. The catalysts were reduced
exothermic effect ( Н = –160 kJ/mol), which preꢀ
Δ
sents certain difficulties associated with heat removal
and reactor control. In addition, a local increase in
temperature leads to a decrease in the process selectivꢀ
ity with respect to heavy hydrocarbons. The thermoꢀ
couples used to control temperature measure the temꢀ
perature of the gas phase, which is only indirectly
related to the temperature on the surface of the nanoꢀ
particles of a catalytically active metal located in the
pores of an inert support. Nevertheless, the true temꢀ
perature in the catalyst granules can differ appreciably
from the temperature of the reaction gas flow around
the granules.
Some of the effects observed during the lowꢀtemꢀ
perature oxidation of cobalt nanoparticles in a porous
matrix of an inert support can be attributed to the
nonisothermal behavior of the process [1]. In this
work, we propose a method that makes it possible to
estimate the heat effect within a catalyst granule using
the example of a model reaction of the lowꢀtemperꢀ
ature oxidation of metallic nickel nanoparticles
located in the pores of supports of different chemical
nature. The concentration of nickel was varied from
in a flow of hydrogen at 300°С for 1 h and then cooled
to room temperature. At room temperature, the flow
of hydrogen was replaced with a flow of special purity
grade Ar and then with a flow of air. The resulting
changes in magnetization were recorded with a freꢀ
quency of 1 Hz.
The average size of Ni particles in the freshly
reduced catalyst was determined from the field depenꢀ
dence of magnetization according to the Langevin
equation [3]. Figure 1 shows the field dependences of
magnetization for Ni particles with average sizes of 5
and 9 nm. There is no hysteresis in the dependences of
the magnetization on the field (H) for any of the studꢀ
ied catalysts, which is indicative of the superparamagꢀ
netic properties of Ni particles at room temperature.
The concentration of metallic nickel was found by
extrapolating the magnetization to an infinite field in
the magnetization—1/Н coordinates. The temperature
dependence of magnetization at a field intensity of
6
kOe was derived for each of the catalysts under study.
5
5
to 30 wt %, and the nickel particle size ranged from
to 9 nm.
RESULTS AND DISCUSSION
Figure 2 depicts the time dependence of magnetiꢀ
zation after replacing the flow of Ar with air. The drop
EXPERIMENTAL
The catalysts were prepared by impregnating a supꢀ in magnetization is due to the oxidation of Ni nanoꢀ
port with nickel nitrate Ni(NO3)2 6H O and subseꢀ particles. Oxidation at room temperature is convenꢀ
quent drying at 80 . The supports were silica gel with tionally called lowꢀtemperature oxidation, which is
·
2
°С
2
a specific surface area of 200 to 300 m /g and a pore well described in terms of the Cabrera–Mott model
size of 50 to 6 nm (Fujisilysia Chemical LTD), aluꢀ [4]. Intense oxidation ceases after the formation of a
2050

HEAT EFFECT DURING THE LOWꢀTEMPERATURE OXIDATION
2051
3
σ × 10 , А/(m g)
σ
,
1.2
arb. units
(a)
2
1
0
1
2
1.0
0.8
0.6
0.4
0.2
0
−
−
Δσ
(b)
1156
1176
1196
1216
τ
, s
1
0
1
Fig. 2. Time dependence of magnetization after replacing
the flow of Ar with air; the arrow shows the instant of the
transition to air.
Assume that the distribution of distances between
individual Ni particles in the sample under study is
subject to the Poisson law. Then the average distance
between the centers of the Ni particles is defined as
−
–
1/3
L
= 0.55(
λ
)
, where is the density of Ni particles
λ
in the sample. However, this estimate of the average
distances between the particles does not take into
account the finite sizes of the Ni particles. We thereꢀ
fore used the results of [5], which gives the exact
expressions for estimating the average distances
between the particles of a given size, subject to a
Poisson distribution of the distances between the
centers of the particles. The calculated average disꢀ
−
8
−
4
0
4
8
−
3
H × 10 , Oe
Fig. 1. Field dependences of magnetization for Ni particles
with an average size of (a) 5 and (b) 9 nm.
relatively thin oxide layer. The end of the oxidation tances between the Ni particles are appreciably
process is accompanied by an increase in magnetizaꢀ higher than those calculated from a simple Poisson
tion, as is evident from Fig. 2. In this case, the distribution.
increase in magnetization is produced by cooling the
The density of Ni particles was found as the ratio of
sample after the exothermic stage of formation of a the number of particles in the sample under study to
thin oxide layer. It is characteristic that as this takes the sample volume. The number of particles can in
turn be determined if we know the average particle size
and weight of Ni in the sample. The average particle
size and weight of Ni were determined through magꢀ
netic measurements. The table lists the average disꢀ
tances between the particles, the particle sizes, and the
respective observed heat effects, which are expressed
in the form of the difference between the temperature
on the surface of Ni particles and the temperature
measured by the thermocouple for a number of the
studied catalysts (room temperature was about 293 K).
Figure 4 depicts the dependence of heat effect on the
average distance between nickel particles. It is evident
from the table and Fig. 4 that the temperature differꢀ
ence between the thermocouple and nickel particles
place, the external thermocouple located in the bulk
of the catalyst sample detects no changes in the temꢀ
perature. Magnetization remains unchanged with
the further passage of air. Our experiment showed
that the difference in magnetization Δσ in Fig. 2
depends on the concentration of Ni particles per
unit volume of the catalyst. In fact, Δσ signifies the
drop in magnetization caused by heating Ni particles
during oxidation. The temperature dependence of
magnetization (thermomagnetic curve) was derived
for each of the Niꢀcontaining catalysts under study.
Figure 3 shows the thermomagnetic curve for the
reduced Ni/Cr O catalyst. The heat effect of the
2
3
reaction can be determined from the temperature can be as great as 100 K or above; however, it shrinks
dependence of magnetization, as is evident from relatively rapidly with increasing average distance
Fig. 3.
between the Ni particles. The observed effect is similar
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 85
No. 11 2011

2052
CHERNAVSKII et al.
σ
, arb. units
Δσ
0
.8
.4
0
0
Δ
T
3
00
400
500
600
700
, K
T
Fig. 3. Temperature dependence of magnetization for Ni/Cr O3 catalyst.
2
Δ
T
, K
Δ
T
120
, K
120
8
0
0
0
8
0
0
4
4
10
30
50
70
L
, nm
0
20
40
60
80
100
Ni/CrO , %
3
Fig. 4. Dependence of heat effect on the average distance
between particles.
Fig. 5. Dependence of heat effect of the concentration of
Ni/Cr O3 in a mixture with alumina.
2
to the processes of combustion of heterogeneous mixꢀ scarce inclusions of fuel can be described in terms of
tures, where a percolation phase transition takes place percolation on random nodes.
[
6]. It was shown in [6] that the concentration limit of
The absence of a threshold value characteristic of a
percolation phase transition for the distance between
the particles is apparently due to the Ni particle size disꢀ
the combustion of heterogeneous condensed mixtures
is directly related to the occurrence of a geometric
(
percolation) phase transition in disordered systems, tribution, since average particle sizes were used to calꢀ
and that the combustion of condensed mixtures with culate the average distances between the particles.
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 85
No. 11 2011

HEAT EFFECT DURING THE LOWꢀTEMPERATURE OXIDATION
2053
Table
diluent. Figure 5 shows the dependence of heat effect
on the concentration of Ni/Cr O3 in a mixture with
4
7
6
2
m
×
10 , g
d
×
10 , cm
L
×
10 , cm
Δ
T, K
alumina. In this case, the generation of heat gradually
decreases with decreasing concentration of the cataꢀ
lyst. These data suggest that there is no thermal interꢀ
action between the individual particles of the catalyst.
The observed heat effect also occurs in other exoꢀ
thermic processes in porous media prone to exceeding
the percolation limit of the concentration of active
particles in a porous matrix. The Fischer–Tropsch
synthesis on Coꢀsupported catalysts can serve an
example. It is difficult to apply the described method
to these systems because of the relatively weak temꢀ
perature dependence of magnetization in the range
5
4.0
7.4
7.0
8.0
6.0
5.0
4.0
5.0
8.0
1.6
5.0
3.4
3.4
4.0
3.0
2.0
4.0
7.0
6.0
6.0
6.0
110
38
25
38
0
2.0
5.0
2.0
2.5
2.0
0
1
5.6
.4
0.0
85
0
9
1
15.0
6.0
7.0
8.0
0
of 180–220°С, which is optimum for the Fischer–
Tropsch synthesis on cobalt catalysts [7].
1.7
3.0
3.7
0
19
0
ACKNOWLEDGMENTS
Note: m is the weight of the Ni, d is the average Ni particle size, L is
the average distance between Ni particles, and ΔT is the temꢀ
perature difference.
This work was supported by the Russian Foundaꢀ
tion for Basic Research, project no. 11ꢀ03ꢀ00501ꢀa.
REFERENCES
CONCLUSIONS
1
. A. V. Mugtasimov, N. V. Peskov, G. V. Pankina, et al.,
Zh. Fiz. Khim. 85, 1 (2011) [Russ. J. Phys. Chem. A
85, 217 (2011)].
For an average distance between Ni particles that
exceeds 30–40 nm, the heating of the particles is thus
not observed, at least within the accuracy of the magꢀ
netization measurement. To all appearances, the level
of heat losses is fairly high for distances greater than
2
3
4
5
. P. A. Chernavskii, A. Y. Khodakov, G. V. Pankina, et al.,
Appl. Catal. A 306, 108 (2006).
. A. K’akay, M. W. Gutowski, L. Takacs, et al., J. Phys. A:
3
0–40 nm, and the particles do not affect one another
Math. Gen. 37, 6027 (2004).
during oxidation. Calculations show that the volume
concentration of Ni particles in this case is less than
. N. Cabrera and N. F. Mott, Rep. Prog. Phys. 12, 163
(1948).
1
1% (the percolation limit for this system).
. S. Torquato, B. Lu, and J. Rubinstein, Phys. Rev. A 41
,
A crucially different picture is observed in studying
2059 (1990).
the oxidation of mixtures of a catalyst with an inert
6. P. S. Grinchuk and O. S. Rabinovich, Fiz. Goreniya
diluent. The catalyst was Ni/Cr O3, which exhibits the
Vzryva 40, 41 (2004).
2
maximum heat effect (the average distance between
7. R. L. Espinoza, A. P. Steynberg, B. Jager, and
the Ni particles is 16 nm);
γ
ꢀAl O was selected as the
A. C. Vosloo, Appl. Catal. A 186, 13 (1999).
2
3
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 85
No. 11 2011