Continuous process for obtaining carbon nanofibers

Article abstract of DOI:10.1023/B:RJAC.0000030347.08283.de

A continuous laboratory installation with a horizontal reactor for obtaining carbon nanofibers by catalytic pyrolysis of methane was developed and tested. The conversion of methane at 600°C was studied and the nanofiber output capacity of the continuous installation under the optimal conditions on a Ni-La₂O₃ catalyst was determined.

Full text of DOI:10.1023/B:RJAC.0000030347.08283.de

Russian Journal of Applied Chemistry, Vol. 77, No. 2, 2004, pp. 187 191. Translated from Zhurnal Prikladnoi Khimii, Vol. 77, No. 2, 2004,
pp. 193 196.
Original Russian Text Copyright
2004 by Rakov, Blinov, Ivanov, Rakova, Digurov.
INORGANIC SYNTHESIS
AND INDUSTRIAL INORGANIC CHEMISTRY
Continuous Process for Obtaining Carbon Nanofibers
E. G. Rakov, S. N. Blinov, I. G. Ivanov, E. V. Rakova, and N. G. Digurov
Mendeleev Russian University of Chemical Engineering, Moscow, Russia
Shubnikov Institute of Crystallography, Russian Academy of Sciences, Moscow, Russia
Received July 1, 2003
Abstract A continuous laboratory installation with a horizontal reactor for obtaining carbon nanofibers by
catalytic pyrolysis of methane was developed and tested. The conversion of methane at 600 C was studied
and the nanofiber output capacity of the continuous installation under the optimal conditions on a Ni La2O3
catalyst was determined.
Carbon nanofibers and nanotubes exhibit a number
of properties which have attracted close attention of
scientists and technologists in recent years. These
materials have a developed surface and possess good
mechanical and emission characteristics, are electrical-
ly conducting, and can be used as catalyst supports,
sorbents, and components of supercapacitors and com-
posites [1, 2]. However, the application of these ma-
terials is hindered by their high cost, which is due
to the expensiveness of some preparation techniques
(e.g., spark procedure [3]) and poor development of
simpler methods. Moreover, the overwhelming ma-
jority of articles, reports, and patents devoted to syn-
thesis of carbon nanofibers and nanotubes describe
batch methods of synthesis, despite that use of con-
tinuous techniques makes it possible to diminish
dramatically the production cost [4].
sity changes. In the process, the bed volume increases
manyfold. The authors of designs mentioned above
partly eliminated these limitations by the nanoagglo-
meration method [6 9]. It remains unclear whether
or not this could ensure a truly continuous operation.
The author of [10] mentioned the difficulties encoun-
tered in developing a stable pseudo-fluidized bed in
the stage of catalyst reduction by hydrogen and car-
ried out the reduction in a fixed filter bed. Chinese
researchers also separated the process into two stages:
catalyst reduction and pyrolysis proper, which made
the procedure batch-like.
A simpler way to perform continuous pyrolytic
synthesis with considerably less pronounced hydro-
dynamic limitations (or without any limitations at
all) consists in use of horizontal tubular reactors with
a continuous motion of a bed of a solid material. Pre-
liminary calculations demonstrated that reactors of
this kind may have a relatively high output capacity
at comparatively small dimensions [13, 14].
The most promising in this regard are pyrolytic
methods, which can use low-cost hydrocarbons as
starting compounds and do not require high tempera-
ture [1, 2, 5]. Fluidized-bed reactors can be used for
performing the process. The fluidized-bed reactors for
catalytic pyrolysis have been created in China [6 9];
laboratory installations have been tested in the Neth-
erlands [10] and in France [11, 12]. The daily output
capacity of the Chinese reactors has been brought to
50 kg of a carbon material containing 70 80% nano-
tubes.
The aim of the present study was to test a continu-
ous installation with a horizontal tubular reactor for
obtaining carbon nanofibers.
EXPERIMENTAL
Carbon fibers were synthesized by deposition of
carbon from the gas phase on a Ni La2O3 catalyst
with Ni content of 23 wt %. The catalyst was pre-
pared from Ni(NO3)2 6H2O (analytically pure),
La(NO3)3 6H2O, and citric acid (both chemically
pure). As starting carbon-containing raw material was
However, the continuous process in a fluidized
bed involves hydrodynamic limitations, since the ini-
tial particles of the catalyst are overgrown with the
product, markedly increase in volume, and their den-
1070-4272/04/7702-0187 2004 MAIK Nauka/Interperiodica

188
RAKOV et al.
with flow meters or rotameters. The temperature in
the reactors was monitored with a Chromel Alumel
thermocouple connected to an MPShchPr-54 milli-
voltmeter with scale-division value of 20 C.
A quartz or ceramic boat with a weighed portion
of a catalyst prepared by the first method was placed
in batch reactors. The catalyst was reduced in a flow
of hydrogen or a nitrogen hydrogen mixture, with
a reactor heated from 20 C to a required temperature
and kept for varied time. Then methane was fed into
the reactor and, after a prescribed time elapsed, the re-
actor was cooled in a flow of methane.
Fig. 1. Schematic of a batch installation for carbon deposi-
tion from the gas phase. (1) Electric furnace, (2) reactor,
(3) boat with catalyst, (4) thermocouple, (5) millivoltmeter,
(6) transformer, and (7) rotameters.
The resulting product was weighed and washed
with 10% nitric acid at 60 C for 4 h, to remove the
catalyst. The purified product was washed with dis-
tilled water to remove nitric acid. The amount of
the product formed was determined by weighing the
samples before and after an experiment.
used town gas containing 98 99% methane and ad-
mixtures of water vapor and mercaptans. Distilled
water was used to prepare solutions and wash the
product to remove the acid.
An X-ray phase analysis was carried out on a
DRON-3M diffractometer (CuK radiation, nickel fil-
ter, accelerating voltage 20 kV). The morphology and
structure of the product were studied with a JEM-
100C transmission electron microscope and Tesla
BS-340 scanning electron microscope. A derivato-
graphic analysis was performed on an MOM deriva-
tograph (Hungary) at a rate of temperature rise in the
The composition of the catalyst and methods of
its synthesis were closely similar to those described
in [15], with the only exception that the Ni : La atom-
ic ratio was taken to be 5 : 6, against 1 : 2 in [15].
The catalyst was prepared by two procedures. In the
first variant, 1.94 g of Ni(NO3)2 6H2O, 3.45 g of
La(NO3)3 6H2O, and 2.00 g of citric acid were
mixed, and the mixture was dissolved in 20 ml of
water. Water was evaporated from the solution, and
the residue was dried at 80 C. The resulting powder
was calcined in air at 700 C for 5 h to give catalyst
no. 1. The second variant of catalyst preparation con-
sisted in burning of the initial mixture. A mixture of
nitrates and citric acid was dissolved in water, with
1 ml of water taken per 1.2 g of the mixture, to give
a transparent solution. The solution was placed in
a furnace heated to 600 C. The product was extracted
15 min after the combustion was complete (catalyst
no. 2). To test the continuous reactor, 100 g of the
catalyst was synthesized by the second procedure,
with the above-mentioned reagent proportions and
modes preserved.
1
range from 1 to 10 deg min . The specific surface
area of the samples was determined by the method of
thermal desorption of nitrogen.
Catalyst samples prepared using different methods
were subjected to an X-ray phase analysis. An anal-
ysis of the diffraction patterns obtained demonstrated
that sample no. 1 is composed of a phase to which
a formula LaNiO3 corresponds, with rhombohedral
crystal lattice [16]. Catalyst no. 2 contains, in addition
to the LaNiO3 phase, trace amounts of other phases.
The specific surface area of catalyst nos. 1 and 2 was
1
36 and 25 m2 g , respectively.
Experiments in the small batch reactor at a temper-
ature of 600 C and conditional linear flow velocity
1
An installation comprising a feeding and gas-flow
controlling unit, a temperature controlling unit, and
a stainless steel reactor was used in the study (Fig. 1).
The experiment was performed by passing successive-
ly from a small batch reactor (inner diameter 28 mm,
length 200 mm) to a larger batch reactor (54 and
400 mm, respectively), and then to a continuous re-
actor (54 and 1000 mm). A Kipp generator or a cy-
linder with a nitrogen hydrogen mixture served as
sources of hydrogen. The gas flow rate was monitored
of methane of 70 cm min (calculated without taking
into account the thermal expansion of the gases) with
sample no. 1 of mass 0.4 g demonstrated that making
longer the process duration leads to a rise in the spe-
cific output of the solid product, with this rise ter-
minating in 20 30 min. The catalyst was preliminar-
ily reduced in a flow of the nitrogen hydrogen mix-
ture, with the reactor heated to 600 C and kept at this
temperature for 15 min. The resulting solid product
was in the form of bent nanofibers 15 60 nm in di-
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 77 No. 2 2004

CONTINUOUS PROCESS FOR OBTAINING CARBON NANOFIBERS
ameter (prevalent diameter 25 35 nm), with 96%
189
purity. Catalyst particles were present at the ends of
nearly all the fibers. The final specific output of the
1
fibers was 4.6 g gcat.
To determine the dependence of the specific out-
put of fibers on the catalyst bed thickness, different
weighed portions of catalyst no. 1 were distributed
uniformly over the same area and placed in the reactor
(five different samples at a time), with the sample
weight increasing or decreasing along the direction
of gas flow in the reactor. The experiment was per-
formed at a temperature of 600 C and a conditional
Fig. 2. Specific output w of the solid product vs. catalyst
2
mass m. Fixed distribution area 220 mm , T = 600 C, linear
1
flow velocity of methane 70 cm min
,
= 30 min. Sample
1
arrangement along the direction of methane flow in the re-
actor: (1) in order of increasing catalyst bed thickness and
(2) in decreasing order.
linear flow velocity of methane equal to 70 cm min
in the course of 30 min, with the catalyst preliminarily
reduced in a flow of the nitrogen hydrogen mixture.
It was found that this dependence shows an extremum
(Fig. 2), with the run of the curve independent of the
order in which the samples studied are arranged along
the gas flow direction and the maximum reached at
a bed thickness of about 1 mm). The rise in the specif-
ic output with increasing bed thickness is presumably
due to the influence of intermediates formed in meth-
ane pyrolysis, and the decrease after reaching the max-
imum value, to hindered delivery of methane to lower
layers of the catalyst bed.
Raising the conditional linear flow velocity of meth-
ane within the range 10 30 cm min at a catalyst
1
Fig. 3. Effect of the conditional flow velocity v of methane
on pyrolysis characteristics: specific output w and methane
bed thickness of about 1 mm, temperature of 600 C,
and preliminary reduction of the catalyst in a flow of
H2 at 600 C for 20 min led to an increase in the spe-
conversion
.
= 30 min, T = 600 C.
1
cific output, which was as high as 8.8 g gcat at a flow
1
velocity of 30 cm min . At the same time, raising
the flow rate entails a decrease in methane conversion
(Fig. 3). The optimal conditional linear flow velocity
1
of methane is 20 cm min , at which the specific out-
1
put is 7.0 8.4 g gcat. In this case, the methane con-
version is close to 65%.
It is noteworthy that experiments without prelimi-
nary reduction of the catalyst yielded the same results.
The method used to synthesize the catalyst also affects
the specific output and the characteristics of the car-
bon product only slightly.
Fig. 4. Schematic of a continuous installation for carbon
deposition from the gas phase. (1) Electric furnace, (2) vi-
bration generator, (3) reactor, and (4) thermocouple.
catalyst bed thickness of about 1 mm, and a duration
of residence of the solid material in the working zone
of the reactor equal to 30 min. This value was 10
times that for the batch installations.
The results obtained for the batch reactors were
used in testing a continuous reactor (Fig. 4), in which
a counterflow motion of gaseous and solid products
was organized. The latter were moved by means of
vibration.
The particles of the product obtained in the con-
tinuous reactor are rounded and have sizes ranging
from 1 to 10 mm. An analysis of their micrographs
demonstrated that the particles are constituted by in-
1
1
1
An output capacity of 5 g h or 3.2 g gcat
h
was achieved at a temperature of 600 C, conditional
1
linear flow velocity of methane equal to 20 cm min ,
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 77 No. 2 2004

190
RAKOV et al.
onto the catalyst in the form of nanofibers 15 to
60 nm in diameter, with conical walls. Catalyst par-
ticles are situated at fiber ends.
(3) It was found that the dependence of the specif-
ic output on the catalyst bed thickness shows an ex-
tremum, with the highest output capacity at a bed
thickness of 1 mm.
(4) It was established that raising the conditional
linear flow velocity of methane at a temperature of
600 C and catalyst bed thickness of about 1 mm leads
to an increase in the specific output of fibers, which re-
1
aches its maximum value at a velocity of 30 cm min .
In this case, the conversion of methane is 40%. The
optimal combination of the output capacity of the
catalyst and of the conversion of methane is achieved
Fig. 5. Micrograph of nanofibers obtained on the continu-
ous installation by pyrolysis of methane on the catalyst
Ni La O .
1
2
3
at methane flow velocity of 20 cm min .
(5) The process of pyrolysis in a continuous re-
actor was tested. The maximum output capacity was
5 g h or 3.2 g gcat h .
terwoven fibers with conical walls (Fig. 5). The fibers
show wide scatter in diameter, from 15 to 60 nm,
and, occasionally, even 100 nm and more. Catalyst
particles not dissolved in washing are contained with-
in some of the fibers.
1
1
1
ACKNOWLEDGMENTS
The specific surface area of the fibers obtained
The study was in part supported financially by the
Russian Foundation for Basic Research (grant no.
01-03-33225) and Ministry of Education of the Rus-
sian Federation ( Chemical Technology Program,
project no. 06.06.017).
1
was 105 m2 g . The X-ray diffraction patterns of
the fibers show a signal corresponding to an angle
of 26.25 , which is characteristic of the interplanar
spacing d002 of multilayer carbon nanotubes and nano-
fibers, and exceeds somewhat that of graphite.
As follows from the derivatograms obtained, active
oxidation of the samples in air begins at a temper-
ature of about 450 C. The content of amorphous car-
bon and sorbed gases, which are removed at temper-
atures that are considerably lower than the combustion
onset temperature of the product, does not exceed 2%
of the original mass of a sample. The incombustible
residue, whose content is about 0.5 1.0%, is com-
posed of the catalyst undissolved in washing (nickel
particles).
REFERENCES
1. Fenelonov, V.B., Poristyi uglerod (Porous Carbon),
Novosibirsk: Inst. Kataliza, 1995.
2. Rakov, E.G., Usp. Khim., 2001, vol. 70, no. 10,
pp. 934 972.
3. Rakov, E.G., Usp. Khim., 2000, vol. 69, no. 1,
pp. 41 59.
4. De Jong, K.P. and Geus, J.W., Catal. Rev. Sci. Eng.,
2000, vol. 42, no. 4, pp. 481 510.
5. Chesnokov, V.V. and Buyanov, R.A., Usp. Khim.,
CONCLUSIONS
2000, vol. 69, no. 7, pp. 675 692.
6. Luo, G., Li, Z., Wei, F., et al., Physica B, 2002,
vol. 323, pp. 314 317.
7. Wang, Y., Wei, F., and Luo, G., Physica B, 2002,
vol. 323, pp. 327 329.
8. Li, Z., Luo, G., Wei, F., and Wang, Y., Chinese
J. Process Eng., 2002, vol. 2, no. 3, pp. 224 229.
(1) Pyrolysis of methane was studied in laboratory
horizontal tubular batch reactors at a temperature of
600 C and a conditional linear flow velocity of methane
1
equal to 70 cm min on a fixed bed of a Ni La2O3-
based catalyst preliminarily reduced in a flow of a ni-
trogen hydrogen mixture. It was found that the spe-
cific output of a solid product reaches its maximum
9. Wang, Y., Wei, F., Luo, G., et al., Chem. Phys. Lett.,
2002, vol. 364, nos. 5 6, pp. 568 572.
1
value of 4.6 g gcat in 30 min.
10. Nijkamp, M., Hydrogen Storage Using Physisorp-
tion: Modified Carbon Nanofibers and Related Ma-
terials, Dissertation, Utrecht, 2002.
(2) It was established, using scanning and trans-
mission electron microscopy, that carbon is deposited
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 77 No. 2 2004

CONTINUOUS PROCESS FOR OBTAINING CARBON NANOFIBERS
191
11. Venegoni, D., Serp, Ph., Feurer, R., et al., Carbon,
dunarodnaya konferentsiya Uglerod: fundamental’
nye problemy nauki, materialovedenie, tekhnologiya
17 19 oktyabrya 2002 g., Moskva (1st Int. Conf. Car-
bon: Fundamental Problems of Science, Materials
Science, and Technology, Moscow, October 17 19,
2002), Moscow, 2002, p. 170.
2002, vol. 40, no. 10, pp. 1799 1807.
12. Morançais, A., Corrias, M., Serp, Ph., et al., 203rd
Meeting Electrochem. Soc., Paris, France, Apr. 27
May 2, 2003, Preprint.
13. Abashev, R. et al., Nanotechnologies in the Area of
Physics, Chemistry and Biotechnology, 5th ISTC SAC
Seminar: Proc., St. Petersburg, Russia, May 27 29,
2002, St. Petersburg: Ioffe Inst., 2002, pp. 171 174.
15. Liang, Q., Gao, L.Z., Li, Q., et al., Carbon, 2001,
vol. 39, no. 6, pp. 897 903.
16. Golub, A.M., Sidorik, L.S., Nedil’ko, S.A., and Fe-
doruk, T.I., Izv. Akad. Nauk SSSR, Neorg. Mater.,
vol. 14, no. 10, pp. 1866 1869.
14. Rakov, E.G. et al., Abstracts of Papers, 1-aya Mezh-
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 77 No. 2 2004