hidrogen production

7/25/2019 Hidrogen Production

http://slidepdf.com/reader/full/hidrogen-production 1/7

Hydrogen production by methane and hydrogen

sulphide reaction: Kinetics and modeling study

over Mo/La2O3eZrO2 catalyst

A.L. Martı́ nez-Salazar a,* , J.A. Melo-Banda a, J.M. Domı́ nguez- Esquivel b,V.H. Martı́ nez-Sifuentes a, Y. Salazar-Cerda a, M.A. Coronel-Garcı́ a a,M.A. Meraz-Melo c

a

Instituto Tecnologico de Ciudad Madero, Los Mangos, 1 de Mayo s/n col., Cd. Madero, Tamaulipas, 89440, M

exicob Instituto Mexicano del Petroleo, Eje Central Lazaro Cardenas 152, San Bartolo Atepehuacan, Gustavo A. Madero,

Ciudad de Mexico, Distrito Federal, 07730, Mexicoc Instituto Tecnologico de Iztapalapa III, Av. Cuitlahuac, Col. Los Reyes Culhuacan, Del. Iztapalapa, Mexico, D.F.,

Mexico

a r t i c l e i n f o

Article history:

Received 8 January 2015

Received in revised form

11 March 2015

Accepted 16 March 2015Available online 22 April 2015

Keywords:

Hydrogen production

Methane reforming with H2S

Kinetics

a b s t r a c t

Actually, catalytic steam reforming method is a widely used to produce hydrogen or syngas

as the first step for the conversion of methane into liquid fuels through the Fischer-

Tropsch synthesis.

The reforming of CH4 by H2S can be considered as an alternative route to produce

hydrogen from methane. With this process, the removal of H2S in natural gas streams is no

longer necessary and sulfur, usually considered as a strong pollutant on liquid fuels in

refineries, is used as a reagent in H2S form of hydrodesulphurization processes. This

alternative produces hydrogen, a valuable chemical and clean energy source, as the

principal product and carbon disulphide as a secondary petrochemical with high com-

mercial value.

The open literature concerning the kinetic study of the methane-hydrogen sulphide

reaction is very poor. In this research the reaction of CH4 and H2S over zirconia catalyst

modified with lanthanum oxide impregnated with molybdenum has been studied in a fixed

bed tubular reactor varying the inlet CH4 flow rate over a temperatures range of 1023

e1323 K and feed molar ratio CH4:H2S of 1:12. The LangmuireHinshelwood theory has been

used to analyze kinetic models. Model discrimination has been performed and the kinetic

parameters and adsorption constants calculated.

Finally, the simulation process of reforming CH4 by H2S has been developed in whichseparation of the produced hydrogen and no reacted H2S involves absorption through

diethanolamine. A tubular reactor from kinetic parameters was dimensioned. Aspen Plus ®

11.1 and Hysys V.8 simulation software was used. Preliminary results showed a high purity

H2 can be produced by this route.

Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights

reserved.

* Corresponding author. Tel.: þ52 8332111039.E-mail address: [email protected] (A.L. Martı́nez-Salazar).

Available online at www.sciencedirect.com

ScienceDirect

j o u r n a l h o m e p a g e : w w w . e l s e v i er . c o m / l o c a t e / he

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 7 3 5 4 e1 7 3 6 0

http://dx.doi.org/10.1016/j.ijhydene.2015.03.074

0360-3199/Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

mailto:[email protected]

http://www.sciencedirect.com/science/journal/03603199

http://www.elsevier.com/locate/he







http://www.elsevier.com/locate/he

http://www.sciencedirect.com/science/journal/03603199

http://crossmark.crossref.org/dialog/?doi=10.1016/j.ijhydene.2015.03.074&domain=pdf

mailto:[email protected]



Introduction

Hydrogen, in addition to be used as essential in the industry of

petroleum for obtaining improved fuels, is considered an

attractive energy supply because it can be burned like gasoline

and natural gas, or converted to electricity without any carbon

emissions at the point of use [1]. Its main advantage is that itscombustion produces only water, which means that does not

emit greenhouse gases, such as fossil fuels and biomass. This

makes it particularly appropriate to replace the petroleum-

based products.

Catalytic reforming of CH4 and H2S to produce H2 and CS2

has poor experimental evidence recorded last years. Mass

catalysts or simple supports that have been used, given their

catalytic properties, do not meet the needs of activation,

selectivity or stability required for this reaction.

Recently, research has been focusing on the structure of

catalysts that have characteristics that generate high potential

for conversion on moderate conditions of operation that could

be the basis of an interesting option of process in the produc-tion of hydrogen. One of these catalysts is zirconia catalyst

modified with lanthanum oxide impregnated with molybde-

num, Mo/La2O3eZrO2, which has shown some advantages in

the improvement of the properties of the catalyst, achieving

high selectivity to hydrogen (64%) in its catalytic performance

[2]. The incorporation of lanthanum oxide in zirconia allowed

the increase of its surface area with enhancedthermal stability

[3]. Also, molybdenum disulfide (MoS2), obtained from the

sulfurization step in the tubular reactor before the reaction

stage, has proven to be a highly effective catalyst for the

decomposition of hydrogen sulfide reaching conversions

higher than 95% at temperatures above 600 C [4].

In this work, kinetics of the reactive system reforming of CH4 by H2S using Mo/La2O3eZrO2 catalyst, based on the

methodology of Langmuir-Hinshelwood-Hougen-Watson was

studied. Pilot plant experimental data of conversion for a

range of temperatures and residence times were obtained

and, by means of them, it was determined which kinetic

model among several possible gave the best setting to the

experimental data. The kinetic parameters and adsorption

constants were obtained for adequate model.

In order to observe an engineering application of the re-

sults of the project, two simulation models were created: one

to propose a process flowsheet that, based on the conditions

of the reaction studied, allows obtain and separate from the

reaction system products with commercially acceptable pu-rity. The other is for calculating the sizing of a tubular reactor

for a given production, based on the kinetic parameters and

adsorption constants calculated.

Proposed kinetic mechanism and models

Catalytic reactions involved in this research work, for the

kinetic study of the reforming of methane with hydrogen

sulfide are as follows:

2H2S þ CH4/CS2 þ 4H2 R1

2H2S/

S2 þ 2H2 R2

At the temperature range in which research took place,

1023e1323 K; and with H2S/CH4 ¼ 12 as a feed molar ratio,

the reaction of pyrolysis of methane can be ignored [5].

The reactions of Table 1 represent the mechanism pro-

posed for this study. Writing rate expressions every step is

considered as an elementary reaction; the only difference is

that the concentration of the species in gas phase isreplaced by its respective partial pressures.

The expressions of rate for adsorption of methane and

hydrogen sulfide are:

rAD1 ¼ kA1

PCH4$CV

CCH2ðsÞ

KCH4

$PH2

(1)

rAD2 ¼ kA2

P4

H2S$C4V

C4SðsÞ

KH2S$P4

H2

! (2)

The expressions of rate for the step of surface reactions

that produce adsorbed carbon disulfide and sulfur and

hydrogen in gaseous phase are:

rs1 ¼ krs1

"CCH2ðsÞ$C2

SðsÞ CCS2ðsÞ$PH2

$C2V

Krs1

# (3)

rs2 ¼ krs2

CSðsÞ$CSðsÞ

CS2ðsÞ$CV

Krs2

(4)

The expressions of rate for desorption of the carbon di-

sulfide and sulfur are:

rD1 ¼ kD1

CCS2ðsÞ

PCS2$CV

KDCS2

(5)

rD2 ¼ kD2

CS2ðsÞ

PS2$CV

KDS2

(6)

For the mechanism postulated in the sequence given by

Table 1, we want to determine the limiting step of the speed.

First, it is supposed that one of thesteps limits speed(controls

rate) and then rate expressions are formulated in terms of

partial pressures of the present species.

a) Assuming surface reactions is the rate controlling step.

rs1 ¼ k1

"PCH4

P2H2S

P3H2

PCS2

$PH2

Ke1

# 1

Q 3r(7)

Table 1 e Steps of a LangmuireHinshelwood kineticsmechanism.

CH4 + (s) CH2(s) + H2

Adsorption4H2S + 4(s) 4S(s) + 4H2

CH2(s) + 2S(s) CS2(s) + H2 + 2(s)Surface reaction

S(s) + S(s) S2(s) + (s)

CS2(s) CS2 + (s)Desorption

S2(s) S2 + (s)

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 7 3 5 4 e1 7 3 6 0 17355







rs2 ¼ k2

"P2

H2S

P2H2

PS2

Ke2

# 1

Q 2r(8)

where

1 þ

KCH4$PCH4

PH2

þKH2 S$PH2 S

PH2

þ KCS2$PCS2

þ KS2$PS2

¼ Q r (9)

b) Assuming adsorption of methane and hydrogen sulfide

over the surface is the rate controlling step.

rAD1 ¼ k1

"PCH4

PCS2$P2

H2

Ke1

# 1

Q r(10)

rAD2 ¼ k2

"PH2S

P0:5S2

$PH2

Ke2

# 1

Q r(11)

where

"1 þ

KCS2$PCS2

$PH2$Krs2

Krs1$KS2$PS2

þK0:5

S2$P0:5

S2

K0:5rs2

þ KCS2$PCS2

þ KS2$PS2

#¼ Q r

(12)

c) Assuming desorption of the carbon disulfide and sulfur

from the surface is the rate controlling step.

rD1 ¼ k1

"PCH4

$P2H2S

P4H2

PCS2

Ke1

# 1

Q r(13)

rD2 ¼ k2

"P2

H2 S

P2H2

PS2

Ke2

# 1

Q r(14)

where

Experimental

Procedure for the analysis of the models and estimation of

the kinetics parameters

A serial of experimental evaluations were carried out at pilot

level using a quartz fixed bed reactor of 1” diameter, for

studying the kinetics of CH4 and H2S reaction. Themass of the

catalyst was fixed as 3 g of Mo/La2O3eZrO2. The experimental

set was constituted by three sections, the first one consisting

of a gas supply (H2S, CH4) unit fitted to a gas control module.

The second section included the reaction system, consisting

of a quartz vertical reactor and an oven (Thermolyne 11000),

which included an internal temperature controller. The third

section was the analytical unit consisting of a gas chromato-

graph Varian Star 3400, with a capillary column Pora-Q bond,

having a length of 25 m, 0.32 mm i.d. diameter, 5 mm phase,

No. CP7361. Before the reaction the catalyst was dried at 393 K

for 2 h under nitrogen (N2), using a flow of 6.84 L/h. The gas

flow carrying the mixture of N2 and H2S (5.95 L/h) was added

while raising thetemperature to about 723K, at a rate of 353K/

h, while keeping it under these conditions for 4 h in order to

ensure the formation of the metal sulfides, i.e., molybdenum

disulfide (MoS2). Then, the flow of N2 gas was substituted by a

flow of helium (He), at a rate of 6.84 L/h, while increasing the

temperature from 723 to 1023 K, at a rate of 353 K/h. Then theCH4:H2S ratio was adjusted to 1:12 to avoid the formation of

carbon; according to previous thermodynamic research [5].

Methane was introduced at the top side of the reactor, at a

flow rate of 0.189 L/h, while H2S was introduced in form of a

gas flow, which was fed at a rate of 2.268 L/h. The reaction

began at 1023 K and it was increased steeply to reach 1123,

1223 and 1323 K, while maintaining the reaction temperatures

for 4 h, respectively, then the analysis of the gaseous products

was performed starting at a time zero and every hour after-

wards. Before releasing the exhaust gases into the atmo-

sphere, the products were sent through a 2 L scrubber

containing a mixture of sodium hydroxide (NaOH, 6 M) solu-

tion with hydrogen peroxide (H2O2), which was proven toremove H2S and other thiols in the gaseous exhaust

effectively.

The previous procedure was repeated varying molar feed

flows for the purpose of obtain experimental data to calculate

the kinetic parameters of the reaction system. The ideal gas

equation has been applied to obtain the profile of volumetric

flows used, shown in Table 2. The initial volumetric flow was

the minimal that the pilot plant can control, 0.189 L/h of CH4

(higher conversion, residence time ¼ 1.46 kg cat.h/mol),

decreasing the residence time in equal intervals according to

the total of experimental tests.

Measurement of the compositions of inlet and outlet of the

reactor allowed to calculate variables to graph results. These

are: the total conversion of methane (xCH4 ) and the conversion

of H2S to S2 (xS2 ) versus residence time, t.

xCH4 ¼ mole of inlet CH4 mole of outlet CH4

mole de inlet CH4(16)

Table 2 e Variation of feed flows.

Volumetric feed flows

Experimentalrun

CH4 (L/h) H2S (L/h)

1 0.189 2.268

2 0.236 2.83

3 0.315 3.78

4 0.4725 5.67

5 0.945 11.34

"1 þ

KCH4$PCH4

PH2

þ KH2 S$PH2S

PH2

þKCH4

$PCH4$K2

H2 S$P2H2 S$Krs1

P4H2

þK2

H2S$P2H2S$Krs2

P2H2

#¼ Q r (15)








xS2 ¼

mole of outlet S2

moles of inlet H2S (17)

t ¼ mcat

.FE

CH4 (18)

a) Conversions were calculated: total conversion of

methane and conversion of H2S t o S2, as a function of inletand

outlet moles of species. Moles were calculated by means of the

experimental data of composition obtained for the differentvalues of temperature in the range of 1023e1323 K and at the

different values of residence times in the range of

0.5e1.46 kg cat.h/mol.

b) For a given temperature, through Excel, constants were

calculated fitting polynomial function to the experimental

data curves conversion versus residence time (t),

xCH4 ¼ a0 þ a1t þ a2t

2 þ a3t3 (19)

and

xS2 ¼ b0 þ b1t þ b2t

2 þ b3t3 (20)

c) By means of the equations

rCH4 ¼

vxCH4

vt¼ a1 þ 2a2t þ 3a3t

2 (21)

and

rS2 ¼ vxS2

vt¼ b1 þ 2b2t þ 3b3t

2 (22)

A table of values of reaction rate versus residence time was

obtained for each temperature.

d) Since

Fig. 1 e Total conversion of CH4 vs residence time at different temperatures, with a feed relation of CH4 :H2S (1:12) and Mo/

La2O3eZrO2 as catalyst.

Fig. 2 e Total conversion of H2S to S2 vs residence time at different temperatures, with a feed relation of CH4 :H2S (1:12) and

Mo/La2O3e

ZrO2 as catalyst.








rCH4 ¼ rs1

and

rS2 ¼ rs2

We have now the information needed to calculate the

values of the kinetic rate constants k1 and k2, as well as the

values of the constants of adsorption equilibrium KCH4, K H2 S,

KCS2 and KS2

, through an iterative calculation procedure,

assuming that they depend only on the temperature. Com-puter program Polymath was used; this program lets you

know so well is the fitting that makes the model to the

experimental data and, as a consequence, compare models.

e) Based on the results obtained in the last step for the best

model, pre-exponential factors of reaction rate, A j, where

j ¼ 1,2 and pre-exponential factors of adsorption, Aadi, where

i ¼ CH4, H2S, CS2, S2, are calculated. Also activation energies,

E j, and adsorption enthalpies, DHi are obtained in the same

calculation. For these calculations, values of k j vs T and Ki vs T,

obtained in the last step are used, as well as the well-known

equations

ln k j ¼ ln A j E jR

1T

(23)

ln Ki ¼ ln Aadi

DHi

R

1

T

(24)

A graph ln k j o ln Ki as a functionof (1/T) is built, according to

we are calculating [6].

Process flowsheet

The proposed process flowsheet was created with the support

of process Simulator Aspen plus and has as basic initial part a

stoichiometric reactor, Rstoic, in which occur the reactions of

the reforming of methane with hydrogen sulfide at 1223 K,

1 atm of pressure and with a feed molar relationship of H2S/

CH4 ¼ 12. Fractional conversions of 0.99 and 0.28 were given as

data, respectively.

Sizing of a tubular reactor

The size of a reactor for the rates of reactants used in process

flowsheet of the last section was calculated for a catalytic

reactor using, process simulator Hysys V8.4 and Plug flow

reactor model [7].

Kinetic methodology calculation

Plots of conversion versus residence time are shown as Figs. 1

and 2.

For each temperature, at a given value of t the reaction

rates rCH4 y rS2 were calculated with the corresponding

quadratic equations. Partial pressures of the species at the

given t value were calculated with the equation P i ¼ (n i /nT )P

were ni are moles of specie i y nT are total moles. Both, ni y nT ,

were calculated as a function of the fractional conversions

xCH4 y xS2 at the given value of t.

The next step was proving the proposed kinetic models

because we already had a table of reaction rate versus resi-

dence time and versus Pi, for each temperature.

Table 3 e Variances and Rmsd of solved models, using the LevenbergeMarquardt algorithm.

Surface reaction Adsorption Desorption

Temperature K Variance Rmsd Variance Rmsd Variance Rmsd

1023 0.5141 0.1179 0.2160 0.0765 0.0290 0.0281

1123 0.0426 0.0340 12.010 0.5762 0.0885 0.0489

1223 0.1588 0.0650 0.5994 0.1273 0.1634 0.0665

1323 0.0055 0.0122 0.0170 0.0215 0.2147 0.0762

The bold values highlight the smallest variance and Rmsd, meaning, the proposed kinetic model with best fit at experimental data.

Table 4 e Values of the kinetic constants and equilibriumadsorption constants as a function of temperature.

Temperature (K)

Constants 1023 1123 1223 1323

k1 139.66 244.56 550.27 581.42

k2 267.40 329.53 586.82 588.82

KCS2 71.27 33.29 7.88 6.90

KS2 6.05 11.00 4.05 3.50

KCH4 102.00 102.00 102.00 102.00KH2S 2.83 3.73 1.62 1.50

Table 5 e Kinetic parameters and adsorption constants.

Reaction A(mol/kgcat.h) E (Kilo Joule/mol)

R1 129,832.44 57.91

R2 13,164.78 33.3

Specie Aad (bar1) DH (Kilo Joule/mol)

CS2 0.00095 95.6

S2 0.3063 27.9

CH4 101.99 bar 0.0005285

H2S 0.1012 bar 29.9

Table 6 e Results of the main streams of the flowsheet.

Component CH4 H2S H2 CS2 S2

Molar flow (kmol/h) 100.0 690.0 890.9 94.5 485.3

Molar fraction 1.0 1.0 0.996 0.993 0.992

Temperature (K) 298 298 299.3 319.4 423

Pressure (at.) 1.0 1.0 10.0 1.0 1.0

Vapor fraction 1.0 1.0 1.0 0.0 0.0








The values of kinetic constants k1 y k2 as well as the values

of the adsorption constants KCH4, KH2S, KCS2 y KS2, were solved

for each proposed kinetic model, by means of the software

Polymath using the Levenberg eMarquardt (LM) algorithm for

non-linear regression. Values of the results of variance and

Rmsd (Root mean square deviation) are shown in Table 3.

These indicators were used for comparison of various models

representing the same dependent variable and thus be able to

make a discrimination among them. The model with the

smaller variance and Rmsd represents data with more

accuracy.

For the present study, the surface reaction model gener-

ates smaller values of indicators. There is a good concor-dance, which implies that the proposed mechanism could be

correct and the controlling step of the speed is the surface

reaction.

In Table 4 can be observed the values of the kinetic con-

stants k1 y k2 and equilibrium adsorption constants KCS2 y KS2,

KCH4, KH2S obtained, taking the surface reactions as the rate

controlling step.

Finally, pre-exponential factors, activation energies and

adsorption enthalpies were calculated. Results are shown in

Table 5.

Results and discussion

Kinetic study of the reactive system

The activation energies of the reforming reactions in the

current study are found smaller than the data reported

by Megalofonos and Papayannakos [8,9] for platinum catalyst

on alpha alumina support and MoS2 catalyst (57.91 kJ/mol

versus 431.66 and 94.62 kJ/mol respectively). This means that

the zirconia catalyst modified with lanthanum oxide

impregnated with molybdenum used in the current study is

more active than catalysts used in previous researches. It

could be attributable a good interaction between active phase

and support, also to high temperatures as consequence a good

dispersion of materials [2].

Process flowsheet

Results of variables in the important currents of input and

output of the proposed flowsheet are observed in Table 6. It

can be observed high purity in all products and significant

production of hydrogen, since it is produced in both reactions.

Additionally, the flowsheet of the process validates the sepa-

ration feasibility of the reaction products and H2S via ab-

sorption with diethanolamine instead of a cryogenic

distillation or the use of membranes [5]. Under those condi-tions the proposal equations demonstrated the possible

thermodynamic process, also an effective separation of main

products.

Sizing of a tubular reactor

Fig. 3 shows the variation of the flows of the components

versus length of the reactor calculated with the hysys

simulator. It can be observed that all flows reach stabiliza-

tion on three meters of length. This is the optimal size to

development the process a commercial level. The final di-

mensions of tubular reactor are: Diameter: 0.30 m and

Length: 3 m.

Summary and perspectives

The best of the models proposed was the one that considered

the surface reaction as rate-limiting step. Values of the kinetic

parameters and the adsorption constants were obtained as a

result of the regression analysis.

A final process flowsheet development by Aspen Plus

Simulator was determinated in this paper. It describe the

possibility to obtain and separate the three products of the

reactive system with high purities and controlling the purity

Fig. 3 e Flux variation of methane reforming components with H2S as a function of the reactor length.








of the main product based on a separation of hydrogen sulfide

by absorption in diethanolamine.

Thesizeof thebasic equipmentof theprocess, which is the

tubular reactor, is reasonable: three meters in length and

30 cm in diameter according with the model equation

proposal.

Acknowledgments

We gratefully acknowledge the financial support of this work

by Mexican Institute of Petroleum.

Abbreviations and variables section

ai, bi constant (i ¼ 0e3)

A j pre-exponential factors of reaction rate

Aadi pre-exponential factors of adsorption of species i

CiðsÞ surface concentration of site occupied by species i

CV molar concentration of vacant sites

E j activation energy (kJ/mol)

FECH4

molar feed flow of methane, mol/h

kAj, krsj, kDj reaction rate constants of proportionality

k j rate constant of reaction j

KDi desorption equilibrium constant of species i

Kej equilibrium constant of reaction j

Ki adsorption equilibrium constant of species i

Krsj surface reaction equilibrium constant of reaction j

mcat catalyst loading, kg cat

ni moles of specie i

nT total moles

Pi partial pressure of species i in the gas phase, bar

rADj

net rate of adsorption of reaction j

rDj rate of desorption of reaction j

ri conversion rates of species i (mol/(kgcat.h))

rSj rate of surface reaction j

R universal gas constant (kJ/(mol K))

T temperature (K)

xCH4 total conversion of CH4, mol/mol

xS2 molar fraction of H2S converted to S2, mol/mol

DHi adsorption enthalpy of species i, kJ/mol

t residence time, kg cat.h/mol

Subscripts

cat catalyst

i gas species

j reaction index (1e2)

r e f e r e n c e s

[1] Simpson AP, Lutz AE. Exergy analysis of hydrogen productionvia steam methane reforming. Int J Hydrogen Energy2007;32:4811e20.

[2] Martı́nez Salazar AL. Produccion de H2 a partir de lareformacion del CH4 con H2S mediante catalizadoresheterog eneos [Ph D. thesis dissertation]. Tamaulipas: Instituto

Tecnologico de Ciudad Madero; 2014.[3] Jong RS, Dong CS, Hae WK. NiO supported on La2O3-ZrO2

modified with WO3 for ethylene dimerization. I Ind Eng Chem2007;13:47e56.

[4] Fukuda K, Dokiya M, Kameyama T, Kotera Y. Catalyticdecomposition of hydrogen sulfide. Ind Eng Chem Fundam1978;17.

[5] Huang C, T-Raissi A. Liquid hydrogen production via hydrogensulfide methane reformation. J Power Sources2008;175:464e72.

[6] Fogler SH. Elementos de ingenierı́a de las reaccionesquı́micas. 4th ed. New Jersey: Prentice Hall PTR; 2006.

[7] Aspen Technology Inc. Aspen hysys: properties and methodstechnical reference. Version Number: V8.4. New Jersey: AspenTechnology Inc; 2013.

[8] Megalofonos SK, Papayannakos NG. Kinetics of the catalyticreaction of methane and hydrogen sulphide over a Pt-Al2O3

catalyst. Appl Catal A 1996;138:39e55.[9] Megalofonos SK, Papayannakos NG. Kinetics of the catalytic

reaction of methane and hydrogen sulphide over MoS2. ApplCatal A 1997;165:249e58.


http://refhub.elsevier.com/S0360-3199(15)00673-4/sref1
































































































hidrogen production

Documents