A study on the kinetics of hydrogen reduction of molybdenum disulphide powders

Article abstract of DOI:10.1016/j.tca.2008.04.009

In order to achieve direct reduction of molybdenite in presence of a sulphur scavenger such as CaO such that SO₂ emission is completely avoided, it is important to maximise the rate of the partial reaction involving molybdenite and hydrogen (without lime) given the low thermodynamic driving force for this reaction. Accordingly, reaction of molybdenum disulphide powders with hydrogen was investigated by thermogravimetric method. Effect of temperature and concentration on the reaction rate was studied under such conditions that resistance to mass transfer arising from external film, between particles and intra-grain was negligible. The operating temperature ranged between 973 and 1173 K while the hydrogen concentration was varied between 30 and 100%. The experimental data obtained under the above conditions were analyzed by applying the shrinking unreacted core model . The reduction reaction was found to be first order with respect to the gaseous reactant. Pre-exponential factor and activation energy have been determined to be 3.91 × 10³ cm min^-1 and 139.0 kJ mol^-1, respectively. Activation energy obtained from a fitted model, agreed well with the values determined from the model-free methods using isothermal measurements.

Full text of DOI:10.1016/j.tca.2008.04.009

Thermochimica Acta 473 (2008) 61–67
Contents lists available at ScienceDirect
Thermochimica Acta
journal homepage: www.elsevier.com/locate/tca
A study on the kinetics of hydrogen reduction of molybdenum
disulphide powders
a,b,∗
a
b
a
M. Mehdi Afsahi
, Morteza Sohrabi , R. Vasant Kumar , Habib Ale Ebrahim
a
Amirkabir University of Technology, Department of Chemical Engineering, Tehran 15914, Iran
University of Cambridge, Department of Materials Sciences and Metallurgy, Cambridge CB2 3QZ, UK
b
a r t i c l e i n f o
a b s t r a c t
Article history:
In order to achieve direct reduction of molybdenite in presence of a sulphur scavenger such as CaO such
that SO2 emission is completely avoided, it is important to maximise the rate of the partial reaction involv-
ing molybdenite and hydrogen (without lime) given the low thermodynamic driving force for this reaction.
Accordingly, reaction of molybdenum disulphide powders with hydrogen was investigated by thermo-
gravimetric method. Effect of temperature and concentration on the reaction rate was studied under such
conditions that resistance to mass transfer arising from external film, between particles and intra-grain
was negligible. The operating temperature ranged between 973 and 1173 K while the hydrogen concen-
tration was varied between 30 and 100%. The experimental data obtained under the above conditions
were analyzed by applying “the shrinking unreacted core model”. The reduction reaction was found to be
first order with respect to the gaseous reactant. Pre-exponential factor and activation energy have been
Received 20 November 2007
Received in revised form 5 April 2008
Accepted 8 April 2008
Available online 23 April 2008
Keywords:
Molybdenum disulphide
Hydrogen
Kinetics
Reduction
3
−1
−1
Gas–solid reaction
determined to be 3.91 × 10 cm min and 139.0 kJ mol , respectively. Activation energy obtained from
a fitted model, agreed well with the values determined from the model-free methods using isothermal
measurements.
©
2008 Elsevier B.V. All rights reserved.
1. Introduction
Conventional method for producing molybdenum metal is
based upon roasting of molybdenite concentrate, followed by
Commercial source of molybdenum metal and molybdenum
purification of the resultant oxide (MoO ), and a final step of
3
compounds is MoS . Molybdenum disulphide, in addition to its
reduction of purified molybdenum oxide with hydrogen [18]. Over-
all process suffers from disadvantages such as, high number of
processing steps, environmental pollution by SO2 emission from
2
natural occurrence, can be prepared synthetically by several routes
including direct union of elements in pure nitrogen [1], thermal
decomposition of ammonium tetrathiomolybdate or molybde-
num trisulphide [2–7], and by reaction between MoO3 and H2S
or H S/H mixtures [8]. Such preparative techniques result in
roasting of MoS , and low yield due to loss of materials during
2
purification. Industrial importance of molybdenum metal on one
hand and processing difficulties on the other have motivated the
scientist to search for more efficient methods for producing molyb-
denum metal.
2
2
hexagonal crystalline MoS , by far the most common form, the
2
rhombohedral form is found in nature, and may be synthesized
[
6]. Natural and synthetic MoS2 of both crystalline types possess
One such method is based on directly reducing molybdenite in
presence of a sulphur scavenger agent such as CaO [19–24], and the
reaction can be described as
lubricant properties, but the natural hexagonal material is preferred
when cost and overall performance are considered. The excellent
lubricant properties of MoS are attributed to the large spacing and
weak van der Waals bonding between S–Mo–S sandwich layers.
Frictional behaviour and other factors influencing the performance
2
MoS (s) + 2H (g) + 2CaO(s) → Mo(s) + 2CaS(s) + 2H O(g) (1)
2
2
2
It is possible to consider that reaction (1) is made up the follow-
ing two partial reactions (2) and (3):
of MoS as a lubricant have been discussed in review papers [9–11]
2
and a number of research papers [12–17].
MoS (s) + 2H (g) → Mo(s) + 2H S(g)
(2)
(3)
2
2
2
H S(g) + CaO(s) → CaS(s) + H O(g)
2
2
∗ Corresponding author at: University of Cambridge, Department of Materials Sci-
ences and Metallurgy, Cambridge CB2 3QZ, UK. Tel.: +44 7790066854;
fax: +44 1223 334567.
The standard-free energy changes and the corresponding equilib-
rium constants for the above three reactions are given in Table 1,
in the temperature range of 973–1273 K [25]. Reaction (2) is
E-mail addresses: mma50@cam.ac.uk, mmafsahi@gmail.com (M. Mehdi Afsahi).
0
040-6031/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.tca.2008.04.009

6
2
M. Mehdi Afsahi et al. / Thermochimica Acta 473 (2008) 61–67
Table 1
The standard-free energy changes and the equilibrium constants for reactions (1)–(3)
◦
−1
)
T (K)
ꢀG (kJ mol
K (equilibrium constant)
Reaction (1)
Reaction (2)
Reaction (3)
Reaction (1)
Reaction (2)
Reaction (3)
9
1
1
1
73
073
173
273
13.8
5.4
136.1
127.9
119.8
112.0
−61.1
−61.2
−61.2
−61.4
0.2
0.5
1.3
2.8
4.9E−8
5.9E−7
4.6E−6
2.5E−5
3.7E6
9.1E5
2.9E5
1.1E5
−2.8
−10.8
clearly the least favoured reaction thermodynamically, the driv-
2.2. Apparatus and procedure
ing force can only be maintained by keeping the PH2S/PH2 at
−
5
very low values, e.g. the equilibrium value of ratio is 2.5 × 10
Experiments were performed, using a thermogravimetric appa-
ratus (Shimadzu TGA-50, Japan) with the accuracy of 1 ␮g. A
schematic diagram of the experimental setup is shown in Fig. 1.
A ceramic tube with inner diameter of 18 mm placed in an electric
furnace was applied as a reactor. The MoS2 powders were placed
in a quartz cell 13 mm inner diameter and 4 mm height. In order to
make a shallow powder bed, the majority of experiments were car-
ried out with 5 mg of powders. The cell containing the sample was
suspended from an arm of the balance by a platinum wire. The tem-
perature of the furnace was measured by a thermocouple (S-type)
placed under the cell. In order to control the gas flow rate further,
in addition to the rotary flow meters situated before the furnace, a
soap bubble flow meter was also placed after the furnace.
at 1273 K. The main driving force for the desired reaction (1) is
thus derived from reaction (3) due to the excellent scavenging
properties of CaO. For the overall reaction, it can be noted that
the driving force increases marginally with increasing tempera-
ture. The equilibrium value of PH2S/PH2 at 1273 K, for the reaction
1) is 2.8, suggesting that MoS2 can be reduced to Mo in pres-
ence of CaO, more readily than without it. Fixing S with CaO,
also has the potential to eliminate the formation of H S, which is
toxic and polluting replacing it with benign H O as the gaseous
(
2
2
product.
This low driving force for direct reduction with hydrogen is not
specific to MoS but other sulphide compounds also suffer from the
2
same problem. Presence of a sulphur adsorbent with large equilib-
Prior to the experimental runs, the balance was calibrated using
standard weights and the thermocouple was checked by dropping
a tiny piece of platinum metal hanging from a silver wire (mp:
rium constant eliminates the H S gas from the reactions media and
2
shifts the reaction to the right.
It is envisaged that in future, especially when H2 is available
through use of non-C energy, there will be incentive for most or
◦
962 C). The quartz cell containing the weighed sample was sus-
pended from the balance. The furnace was then brought to the
proper position and the sample was heated under a continuous
many metallurgical reductions to be carried out with H rather than
2
−
1
with C-based fuels.
flow of 70 ml min of helium gas with a temperature increasing at
−
1
Given the future interest in hydrogen economy and in the antic-
ipated massive reduction in C emissions, it is surprising that apart
from a few investigations [26–28], no significant work has been
performed during the recent years in this topic.
To model the reaction (1) intrinsic kinetic parameters of reac-
tions (2) and (3) can be helpful in designing suitable process and
equipment. The aim of the present study is to obtain the intrin-
sic kinetic parameters of reaction (2) towards elucidating detailed
fundamental mechanisms.
a rate of 30 K min until the set point was reached. Five minutes
later helium gas flow was switched to that of hydrogen and the
reduction in weight of the sample was recorded every 2 s. At the
end of each run, the hydrogen flow was turned off and the sample
was allowed to cool down under helium atmosphere. In all exper-
iments the gas was entered from the top of the furnace and was
allowed to leave via the bottom of the apparatus.
3. Results and discussion
There is no published paper in the literature on the kinetic
parameters for the hydrogen reduction of MoS2 except for mod-
elling research work by the present authors that has been recently
submitted (Mehdi Afsahi et al., A model for the intrinsic kinetic
of the direct reduction of MoS2 with hydrogen). For other metals
such as Co, Ni and Cu, a few investigations have been reported for
reducing sulphide with hydrogen [29–32].
3.1. Introduction
It is known [33] that molybdenum disulphide will not be sol-
ubilised by moisture, in fact it dissolves only in strongly oxidizing
acids (e.g. aqua regia) to give molybdenum(VI) compounds. In inert
atmosphere as in He, MoS2 is stable and can decompose to Mo2S3
and metallic molybdenum only under vacuo at elevated temper-
ature >1473 K. Heating in air at 773 K produces MoO . Reduction
with hydrogen at 1373 K has been reported to yield Mo2S3
2. Experimental
3
[
33].
2.1. Materials
MoS2 powder was supplied by Alpha Aesar Company (melting
3.2. Characterizations of the molybdenum sulphide
◦
point 1185 C, hexagonal structure, 99% purity and −325 mesh size),
helium gas with purity of 99.999% was used as protective gas during
the heating of the sample before starting the reaction and during
cooling of the furnace at the end of the reaction. Hydrogen used
was of extra pure gas grade (with less than 10 ppm impurity). Since
MoS2 reacts easily with oxygen at the elevated temperatures, and
Particle size of as received molybdenum sulphide was −325
mesh. In order to reduce intra-grain diffusion resistance, the pow-
der was ground for 1 h in a ball mill. To ensure that during grinding
operation, oxidation of MoS2 has not occurred, the powdered sam-
ples were examined by XRD. No changes in the characteristic peaks
before and after grinding were observed. Particle size distribution
of the ground powders was measured by a laser particle size ana-
lyzer (Cilas, Model 1064 Dry). Ground molybdenum sulphide had a
narrow particle size distribution with an average size of 1.77 ␮m.
the amounts of MoS in the majority of experiments were quite low
2
(
ca. 5 mg) the complete elimination of oxygen and water vapour
from the gas stream was crucial. Two traps were, therefore, used in
the gas line prior to the furnace.

M. Mehdi Afsahi et al. / Thermochimica Acta 473 (2008) 61–67
63
Fig. 1. Schematic diagram of the experimental system.
The scanning electron micrograph (JEOL 6340F FEG) was used
3.3. Kinetic measurements
to provide useful information concerning topography and mor-
phology of the particles. Fig. 2 reveals that MoS2 particles had
irregular shapes and highly impervious structure. Fig. 3 indicates
that the molybdenum disulphide consisted of flat shape parti-
cles.
Intrinsic kinetic parameters should be determined under con-
ditions where no mass transfer effects intrude the rate of the
reaction. The effect of hydrogen flow rate on conversion at 1123 K
is shown in Fig. 4 for a constant amount of sample (30 ± 1 mg).
Progress of the reaction has been monitored by considering con-
version (X). The conversion at any time X(t) is defined as the
weight loss of the sample divided by the theoretical maximum
weight loss, i.e., when all the sulphur atoms are assumed to
be removed from the sample. The error in the fraction con-
verted at a given time is estimated as 0.03, arising from errors in
The specific surface area of the ground powders was mea-
sured by BET method (TriStar 3000 V6.04 A) and was found to be
2
−1
5
.09 m g . Thus MoS2 powders are solid particles with each par-
ticle possessing a nonporous structure, hence, as reported there
is a one-to-one relation between surface area and particle size
[
10].
Fig. 2. Scanning electron micrograph showing the irregular and impervious surface
of the MoS2 particles before reaction.
Fig. 3. Scanning electron micrograph showing that MoS2 consists of flat plate par-
ticles with a width of 75 nm.

6
4
M. Mehdi Afsahi et al. / Thermochimica Acta 473 (2008) 61–67
Fig. 4. Effect of hydrogen flow rate on conversion (PH2 = 0.82 bar).
temperature, flow rate and weight measurements. Thus there
is clear trend of increasing conversion with time as the hydro-
−
1
gen flow rate is increased from 100–300–500–700 ml min
especially after 50 min of reduction time. At flow rates higher
,
−1
than 700 ml min , the conversion is unaffected by increas-
ing the flow rate within the experimental error. External mass
transfer resistance can be reduced by increasing the rate of
the gas flow. Critical flow rate, beyond which, no further
effect on the reaction rate could be observed was determined
−
1
to be 700 ml min
. This is considered sufficient to elimi-
Fig. 6. Effect of hydrogen concentration on the rate of reaction (hydrogen flow rate:
nate the external mass transfer resistance. All subsequent runs
were, therefore, carried out at a slightly higher flow rate of
00 ml min
Intra-particle diffusion effects were reduced by evenly spread-
ing the fine particles in the quartz cell. The effect of powder quantity
on the reaction rate is shown in Fig. 5. There is an insignificant dif-
ference between the curves for the 5 mg and the 8 mg samples.
Subsequent runs were, therefore, carried out using 5 mg of pow-
ders.
800 cm3/min; weight of sample: 5 mg; temperature: 1073 and 1173 K).
−1
8
.
ysis have been presented directly without any smoothness in
Fig. 5.
The effect of hydrogen concentration on the reaction rate within
the range of 30–100% was investigated at two different tempera-
tures, 1073 and 1173 K. The data are shown in Fig. 6a and b. The
concentration of hydrogen was regulated by varying the flow rates
of both hydrogen and helium while maintaining the total flow rate
In order to demonstrate the reliable performance of the
equipment, the data obtained from the thermogravimetric anal-
−1
of gases at 800 ml min
.
The reaction temperature was varied within the range of
9
73–1173 K. The effect of temperature on the reaction of MoS pow-
2
ders with pure hydrogen gas at 0.82 bar is shown in Fig. 7.
Fig. 5. Effect of sample weight on the conversion of molybdenum disulphide with
hydrogen.
Fig. 7. Effect of temperature on the rate of reaction (5 mg sample of 1.77 ␮m average
size of particles (PH2 = 0.82 bar).

M. Mehdi Afsahi et al. / Thermochimica Acta 473 (2008) 61–67
65
4. Interpretation of rate data
Some researchers [20,29–32] have reported the growth of fila-
mentary products (or whiskers) in the reduction of metal sulphides
such as nickel, cobalt and copper. However, in the present work, as
is shown in Fig. 8, the product after reduction of MoS2 powders is
a porous material without any filament or whisker.
Since the starting particles are nonporous, experimental data
may be analyzed by using the “shrinking unreacted core model”
[34]. The relations describing such a model could be given as fol-
lows:
A(g) + bC(s) → cC(g) + dD(s)
ꢂ
ꢃꢂ
ꢃ
ꢀ
ꢁ
m
0
C
bk
ꢁ
Ap
(4)
n
A
C
1/Fp
C
−
t = 1 − (1 − X)
= g(X)
0
Ke
FpVp
In the above equation, CA
0
and CC
0
are the bulk concentrations;
n and m are the orders of reaction with respect to gaseous A and
C, respectively. In this equation, b is the number of moles of solid
reacted by 1 mol of gaseous reactant (b = 1/2 for reaction (1)); k
and Ke are the reaction and equilibrium constants, respectively;
Fig. 9. Typical results of the reduction of MoS2 for different values of shape factor.
Eq. (4) has been obtained based on the following assumptions:
(a) particles are nonporous;
ꢁ
is the true density of MoS2 particles; g(X) is integrated form of
the reaction model; Ap and Vp represent surface and volume of
particles, respectively; Fp is the shape factor of particles and equal
to 1, 2 and 3 for flat, cylindrical and spherical grains, respectively; t
is the time and the term (FpVp/Ap) is the characteristic dimension of
the particles. This is equal to half of the thickness for a flat plate and
(
(
b) particles maintain their initial shape and size;
c) the reaction is isothermal;
(
d) impact of structural change of the particles on the progress of
reaction is negligible.
−
6
radius for cylindrical and spherical particles ((0.075/2) × 10 m in
By comparing the SEM images for product and reactant
this work).
(Figs. 2, 3 and 8), it may be concluded that assumptions (a) and (b) is
The relation between conversion X and time may be, therefore,
shown by the following equation:
−1
justified. This reaction is highly endothermic (+213 kJ mol [25]),
but since size of the particles are very small and also since small
amount of particles is used in the experiments therefore isothermal
reaction is an acceptable assumption.
W − W(t)
0
X(t) =
= 2.4962(1 − W(t))/W0
)
(5)
W (M /M
MoS2
0
S
2
Since as sulphur is removed from MoS powders during the reac-
2
In the above equation W0 and W(t) are the initial weight of the
sample and the weight at time t (as measured by the thermogravi-
tion, structural properties of particles such as density and porosity
will also change, and this could affect the results.
metric method). MMoS2 and MS2 are molecular weights of MoS and
2
The shape factor of a particle, provided that the mass transfer
resistances are eliminated, can be determined from the slope of a
straight line obtained by plotting the experimental values of g(X)
against the reaction time t. In Fig. 9, such plots, drawn at 1137 K
have been given.
As it may be observed from this figure, the best correlation is
obtained with Fp = 1, corresponding to the particles having a flat
shape. Similar conclusions can be made from the results at other
temperatures and subsequently, confirmed by SEM images (Fig. 3).
The effect of hydrogen concentration on the reaction rate at two dif-
ferent temperatures (1073 and 1173 K) was investigated. By plotting
g(X) versus time, according to Eq. (6), with Fp = 1, a straight line is
obtained (Fig. 10a and b) with a slope corresponding to
the liberated sulphur, respectively.
When mass transfer resistances are eliminated, the concentra-
tion of gaseous product (Cc) on the solid surface may be assumed
to be negligible and Eq. (4) could be simplified to
ꢂ
ꢃ
ꢀ
ꢁ
bk
ꢁ
Ap
n
A
1/Fp
C
t = 1 − (1 − X)
= g(X)
(6)
0
FpVp
ꢂ
ꢃ
ꢀ
ꢁ
bk
ꢁ
Ap
n
A
S =
C
(7)
0
FpVp
A plot of ln(S) versus ln(CA
0
) yields a straight line, the slope of which,
determines the order of reaction (n). Values of n at 1073 and 1173 K
are both close to unity. This may indicate that the reaction rate is
first order with respect to hydrogen concentration (Fig. 11).
According to Eq. (6), at any selected temperature, a plot of g(X)
against t should yield a straight line with slope S from which the
rate constant (k) at that particular temperature may be deter-
mined. Such curves are drawn in Fig. 12. Applying Arrhenius
−
1
relation, from a plot of ln(k) versus T (Fig. 13), the pre-exponential
term and activation energy of the reaction were determined as
3
−1
−1
3
.9 × 10 cm min and 139.0 kJ mol , respectively.
To verify the kinetic triplet derived in the present work (reaction
Fig. 8. Scanning electron micrograph showing the porous structure of the reaction
product at 1123 K.
model, pre-exponential factor and activation energy) a comparison

6
6
M. Mehdi Afsahi et al. / Thermochimica Acta 473 (2008) 61–67
Fig. 12. Plot of conversion against time at different temperature (data from Fig. 7).
Fig. 10. Plot of integral conversion function against time at different concentrations
of hydrogen (data from Fig. 6a and b).
has been made between the latter and those determined from the
model-free method. A model-free equation for isothermal condi-
tion [35–37] may be considered as follows:
Fig. 13. Arrhenius plot for the reaction between MoS2 powders and hydrogen (data
from Fig. 12).
ꢂ
ꢃ
k0j
EX
represent pre-exponential factor and integrated form of the reac-
tion model. Appling Eq. (8) to isothermal kinetic data within the
range of 0.1 < X < 0.9, values for E were obtained as a function of
conversion with an average of 134.1 kJ mol (Fig. 14).
This value is close to that determined previously, using shrinking
unreacted core model.
−ln tX,i = ln
−
(8)
gj(X)
RT
In the above equation, subscripts i and j refer to isothermal con-
dition and the type of the model, respectively. tX,i is the time
required to reach a certain conversion isothermally, k0 and g (X)
−1
j
Fig. 11. Plot of ln(S) against ln(C) at two different temperatures (used for determin-
Fig. 14. Variation of the activation energy with conversion derived from model-free
ing the order of the reaction).
method applying isothermal kinetic data.

M. Mehdi Afsahi et al. / Thermochimica Acta 473 (2008) 61–67
[2] S.K. Srivastava, B.N. Avasthi, J. Mater. Sci. 28 (18) (1993) 5032–5035.
67
Similarity between the activation energy obtained from two
[
[
3] J. Chen, S.L. Li, Z.L. Tao, J. Alloys Compd. 356–357 (2003) 413–417.
4] H.W. Wang, P. Skeldon, G.E. Thompson, G.C. Wood, J. Mater. Sci. 32 (2) (1997)
97–502.
[5] K. Du, W. Fu, R. Wei, H. Yang, S. Liu, S. Yu, G. Zou, Mater. Lett. 61 (27) (2007)
887–4889.
different methods supports the kinetic triplet obtained from the
shrinking unreacted core model.
It may be observed from Fig. 14 that dependency of activation
energy on conversion is rather weak, decreasing at the early stages
but increasing towards the end of the reaction. The changes in E
with conversion may be described as follows.
4
4
[
[
6] J.C. Wildervanck, F. Jellinek, Z. Anorg, Allgem. Chem. 328 (1964) 309–318.
7] H.W. Wang, P. Skeldon, G.E. Thompson, G.C. Wood, J. Mater. Sci. Lett. 15 (6)
(1996) 494–496.
Nucleation, nuclei growth and diffusion of the gaseous product
provide the effective parameters that will determine the activation
energy. Interaction of these phenomena is such that at the start of
reaction, the activation energy is decreased. However, during a wide
range of conversion, the latter remains almost unchanged. Finally,
at the later stages of the reduction the activation energy is increased
rapidly with conversion. This observation may be explained by not-
ing that at the beginning of the reaction, the product layer imposes
a negligible resistance to the overall rate. However, as the result of
nuclei growth during the reaction, activation energy is decreased.
On the other hand, at the final stages of the reaction, formation of
a thick layer hinders the movement of gases towards the reaction
zone, leading to higher activation energy.
[8] Y. Feldman, E. Wasserman, D.J. Srolovitz, R. Tenne, Science 267 (5195) (1995)
22–225.
9] H.F. Barry, Lubricat. Eng. 33 (9) (1977) 475–480.
[10] J.P.G. Farr, Wear 35 (1975) 1–22.
[11] W.O. Winer, Wear 10 (1967) 422–452.
12] J.M. Martin, in: D. Dowson (Ed.), Molybdenum Disulphide Lubrication, Tribol-
ogy Series, vol. 35, 1999.
13] W. Hai-Dou, X. Bin-Shi, L. Jia-Jun, Z. Da-Ming, Sci. Technol. Adv. Mater. 6 (5)
(2005) 535–539.
2
[
[
[
[
[
[
14] S. Watanabe, J. Noshiro, S. Miyake, Surf. Coat. Technol. 183 (2–3) (2004) 347–
351.
15] J. Xu, M.H. Zhu, Z.R. Zhou, Ph. Kapsa, L. Vincent, Wear 255 (1–6) (2003) 253–
258.
16] N.M. Renevier, J. Hamphire, V.C. Fox, J. Witts, T. Allen, D.G. Teer, Surf. Coat.
Technol. 142–144 (2001) 67–77.
17] L. Cizaire, B. Vacher, T. Le Mogne, J.M. Martin, L. Rapoport, A. Margolin, R. Tenne,
Surf. Coat. Technol. 160 (2–3) (2002) 282–287.
18] K. Othmer, Encyclopedia of Chemical Technology, fourth ed., John Wiley & Sons,
New York, 1998.
19] H. Kay, in: W.A. Krivsky (Ed.), High-Temperature Refractory Metals, Gordon and
Breach, New York, NY, 1968, p. 33.
[
[
In a previous study (Mehdi Afsahi et al., submitted for publica-
tion, A model for the intrinsic kinetic of the direct reduction of MoS2
with hydrogen), the intrinsic kinetic parameters for reduction of flat
pellets were determined, applying nucleation and growth models.
[
[20] F. Habashi, R. Dugdale, Metall. Trans. B 4 (1973) 1865–1871.
[21] P.M. Prasad, T.R. Mankhand, P. Suryaprakash Rao, Miner. Eng. 6 (8–10) (1993)
−
1
The activation energy was found to be 135.4 kJ mol which is close
to the value determined in the present work.
8
57–871.
[22] P. Suryaprakash Rao, P.M. Prasad, Mater. Trans. JIM 34 (12) (1993) 1229–
233.
1
[
[
[
23] J.D. Ford, M.A. Fahim, Metall. Trans. B 6 (1975) 461–464.
24] F. Habashi, B.J. Yostos, J. Metals 29 (7) (1977) 11–16.
25] A. Roine, HSC Chemistry 6.0, Chemical Reaction and Equilibrium Software with
Extensive Thermochemical Database and Flowsheet Simulation, Outokumpu
Research Oy Information Center, August 2006.
5
. Conclusions
In the present work variables affecting the reaction between
MoS2 powders and hydrogen gas were investigated. The data were
analyzed using shrinking unreacted core model and compared with
the free model method. The reaction rate constant was correlated
by the following equation:
[26] P.K. Tripathy, R.H. Rakhasia, Miner. Process. Extract. Metall. 115 (1) (2006) 8–
14.
[27] P.M. Prasad, P. Surya Prakash Rao, S.N. Singh, A.J.K. Prasad, T.R. Mankhand,
Metall. Mater. Trans. B: Process Metall. Mater. Process. Sci. 33 (3) (2002)
ꢀ
ꢁ
345–354.
4
3
−1.67×10
−1
k = 3.91 × 10 exp
(cm min )
[28] R. Padilla, M.C. Ruiz, H.Y. Sohn, Metall. Mater. Trans. B: Process Metall. Mater.
Process. Sci. 28 (2) (1997) 265–274.
[29] R.E. Cech, T.D. Tiemann, TMS AIME 245 (1969) 1727–1733.
[30] M.A. Fahim, J.D. Ford, Can. J. Chem. Eng. 54 (1976) 578–583.
T
It was found that the reaction rate was first order with respect
to hydrogen gas concentration. This work is used as a foundation
for further studies in the formation of Mo from molybdenite with
hydrogen as the reducing agent and CaO as the sulphur scavenging
agents.
[
31] T. Chida, J.D. Ford, Can. J. Chem. Eng. 55 (1977) 313–316.
[32] H.Y. Sohn, S. Won, Metall. Trans. B 16 (1985) 831–839.
[33] F. Ullmann, Ulmann’s Encyclopedia of Industrial Chemistry, vol. 22, Wiley-VCH,
Weinheim, 2003, p. 26.
[
34] J. Szekely, J.W. Evans, H.Y. Sohn, Gas–Solid Reactions, Academic Press Inc., New
York, 1976.
References
[35] S. Vyazovkin, C.A. Wight, Thermochim. Acta 340 (341) (1999) 53–68.
36] S. Vyazovkin, Thermochim. Acta 355 (2000) 155–163.
[37] S. Vyazovkin, C.A. Wight, Chem. Mater. 11 (1999) 3386–3393.
[
[
1] L. Chang, H. Yang, J. Li, W. Fu, Y. Du, K. Du, Q. Yu, J. Xu, M. Li, Nanotechnology 17
(
2006) 3827–3831.