Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black

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 x x x ( 2 0 1 7 ) 1 e1 7

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black M. Gautier, V. Rohani, L. Fulcheri* MINES-ParisTech, PSL e Research University, PERSEE e Centre proc
ed
es,
energies renouvelables et systemes
energ
etiques, 1 Rue Claude Daunesse, 06904 Sophia Antipolis, France

article info

abstract

Article history:

In the prospect of a large scale deployment of Renewable Energy for electricity production,

Received 3 March 2017

plasmas will definitively be a major option to get tuneable, high temperature enthalpy

Received in revised form

sources without direct CO2 emissions. This paper focuses on the direct decomposition of

29 August 2017

methane for the simultaneous synthesis of hydrogen and high value-added carbon black.

Accepted 3 September 2017 Available online xxx

After a review of gas phase carbon particle nucleation and growth physico-chemical phenomena, a new original model for the plasma decomposition of methane is presented. The model solves a reactive turbulent flow in a 3D geometry. The nucleation is

Keywords:

based on a detailed reaction mechanism and the particle growth is handled by a sectional

CFD

method. This model opens the way towards a better understanding of carbon particles gas

Nucleation

phase nucleation and growth and consequently to a fine control of high value-added

Growth

carbon black grades.

Population balance model

© 2017 Published by Elsevier Ltd on behalf of Hydrogen Energy Publications LLC.

Introduction In the current worldwide energy context characterized by climate change and fossil fuel depletion, the massive use of renewable energy appears essential for the future. The large scale deployment of renewables faces major limitations, among which: 1. the resource variation and its predictability, 2. the supply-demand matching, 3. storing the electricity produced from the renewables in electro-chemical form (i.e. batteries) can be challenging,

especially for mobile applications where high energy densities are required, 4. high initial investment costs (CAPEX), even though major cost reductions have been achieved in the past few years, particularly in the PV and wind turbines areas. In this prospect, energy storage issues will probably be decisive. Green synthetic fuel, including H2, are expected to become a major primary energy storage player [1,2], provided the fact that the conversion of electricity into green synthetic fuel can be done efficiently and cost effectively [3e6]. The conversion of low carbon content electricity into green synthetic fuel is particularly suitable if it can be achieved

* Corresponding author. E-mail address: [email protected] (L. Fulcheri). https://doi.org/10.1016/j.ijhydene.2017.09.021 0360-3199/© 2017 Published by Elsevier Ltd on behalf of Hydrogen Energy Publications LLC. Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

2

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 x x x ( 2 0 1 7 ) 1 e1 7

together with a partial (or total) fuel decarbonization. In this perspective, three main options are especially interesting: 1. the use of biomass as raw material, 2. the CO2 retro-conversion, 3. the direct decarbonization of fossil fuels. While, the first two options have been extensively studied all over the world, the direct decarbonization pathway has been seldom explored [7e14]. Thermal plasmas offer many benefits for the conversion of electrical to chemical energy, providing a flexible, controllable and tunable heating source without direct CO2 emissions [15]. In addition, they are particularly suitable for endothermic processes and when high temperatures are required; such conditions being often encountered in most thermochemical processes, especially for combustion-based processes. First efforts related to the decomposition of hydrocarbons in a plasma were published in 1920 by Rose [16]. The method involved a vessel with two electrodes for the pyrolysis of a gaseous feedstock flow. From 1920 to 1990 many carbon black companies followed Roses footsteps, among which are Goodyear [17], Ashland Oil and Refining [18], Continental Carbon [19], Phillips Petroleum [20,21]. In his paper, Gonzalez et al. [22] provides a comprehensive survey of studies carried out in the field. In the 90's, the Norwegian engineering company KVAERNER (now AKKER-KVAERNER), in collaboration with SINTEF-NTNU Prof. BAKKENs group in Trondheim-Norway [23], has been intensively operating on the setup of a DC plasma system dedicated to the co-production of carbon black and hydrogen from natural gas and several patents were issued by Lynum S. et al. [24e30]. The plasma torch technology was based on the generation of a high velocity blown arc plasma rotating at the tip of two concentric cylindrical hollow graphite electrodes under the combined effect of a strong axial gas flow associated with an external magnetic coil. In 1992, a first 3 MW plasma pilot was successfully installed and tested in Sweden at ScanArc HOFORS facilities [30]. In 1997, KVAERNER started the construction of a first industrial plant in Karbomont (Canada) with a 20,000 tons carbon black and 70 Million Nm3 hydrogen capacity. Unfortunately, the development of the process was stopped in 2003 due to technological issues in the carbon black quality control. Almost concurrently, Fulcheri and Schwob started in 1993, the development of a 3-phase Alternative Current plasma prototype [31]. This technology led to four patents [32e35] and number of papers [23,36e46]. A comprehensive review of the main three phase AC plasma torch achievements is given in Ref. [47]. Other works based on inductively coupled plasma and arc technologies were led in Canada by Soucy's group [48], Meunier et al. [49e52], in the US by Fincke et al. [53], in Korea by Kim et al. [54,55], Cho et al. [56] and in China by Zhang et al. [57], Sun et al. [58]. Finally, in the 2010's, the company ATLANTIC HYDROGEN, located in New Brunswick (Canada), developed the CarbonSaver™ concept based on a

non-thermal plasma technology. This concept was centered on the partial decarbonization of grid gas to produce natural gas enriched with hydrogen and re-introduce it into the grid. Carbon black being a byproduct of the process. The concept was developed up to the pilot scale, but this development stopped in 2016 when the company was assigned into bankruptcy [59e61].

Co-production of carbon black and H2 Hydrogen is the ultimate form of decarbonized fuel. 98% of the hydrogen production comes from fossil fuels reforming (mostly by methane Steam Reforming) [62,63] with huge direct impacts on CO2 emissions corresponding (world average, source DOE 2013) to 12 kg CO2 eq. per kg of H2. Giving rise to 720 Million tons CO2 eq. emissions per year representing by itself no less than 2.25% of the total worldwide CO2 emissions. Since the industrial revolution, carbon black production steadily increased to reach current a production of 12 million tons per year [64]. Last news on the Worldwide Carbon Black Market says the demand will continue to grow and it is expected a production of 20 million tons by the year 2020. 90% of the production being used in the tire industry and 95% being produced with the furnace process [65]. This process relies on the incomplete combustion of different heavy carbonaceous feedstocks. It is also characterized by direct CO2 emissions with an average amount around 4 kg CO2 eq. per kg of carbon black, giving rise to more than fifty million tons carbon dioxide emissions annually and accounting around 0.125% of the Total Worldwide Emissions [66]. The principle of the new plasma process aims to replace these two existing processes, both characterized by high environmental impacts, by a single environmental friendly process allowing the simultaneous synthesis of pure hydrogen and solid carbon directly splitting natural gas thanks to an outer electric energy source. The major objectives are the replacement of two combustion methods thanks to the use of a flexible external energy supply and the co-synthesis of two valuable chemicals without direct CO2 emissions with 100% carbon yields [67]. Regarding thermodynamic figures, let's notice that the standard enthalpy reaction of methane dissociation related to hydrogen production (Eq. (1)) is thermodynamically less costly than steam reforming (Eq. (2)) with 37.8 kJ versus 63 kJ per H2 mole respectively. It is also worth noticing when comparing plasma cracking of methane with water splitting (Eq. (3)), that the production of H2 by direct methane cracking is thermodynamically 8 times less costly than the production of H2 from liquid water splitting with 37.8 kJ=ðH2 Þmol against 285 kJ=ðH2 Þmol respectively. CH4 /C þ 2H2

DH ¼ þ75:6kJ=mol

CH4 þ 2H2 O/CO2 þ 4H2 H2 O/H2 þ O2

DH ¼ þ252kJ=mol

DH ¼ þ285kJ=mol

(1) (2) (3)

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

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 x x x ( 2 0 1 7 ) 1 e1 7

3-Phase AC plasma technology The plasma process developed at MINES-Paristech is composed of a 3-phase AC plasma power source that provides electricity to three electrodes in graphite set at the upper of the reactor, a high temperature vessel outfitted with a graphite nozzle specially designed for the fast mixing of the feedstock with the plasma flow, and an injector. The equipment is terminated by a tail bag filter. Water and gas networks supply water cooling and process gases. A schematic view of the overall process is represented in Fig. 3. The three phase power creates a thermal arc revolving at very high velocity at the bottom of three cylindrical bulk graphite electrodes under the combined effect of the gas flow, arc current, voltage and frequency and self-induced electromagnetic forces. Each electrode acting alternately as anode and cathode with twice the current frequency. The electrode erosion is limited by the fast velocity motion of the arc between the three electrodes. Despite this limited erosion, the electrode consumption is compensated by a continuous feeding of electrodes rods. The plasma facility is equipped with a series of diagnostics and analyses tools including: digital oscilloscope, Emission Optical Spectroscopy, optical and classical pyrometry, calorimetry, ultra-fast imaging, GC and GS-MS.

What is carbon black? Carbon black is a nanostructured material having a high carbon content, commonly higher than 97 weigh percent. It is generally composed of fine quasi spherical particles, called primary particles, linked together to form aggregates or agglomerates having more or less complex structures. Aggregates differ from agglomerates owing to the fact they can hardly be broken under mechanical constraints. The average primary particle diameter varies depending on the quality of the carbon black between few tenths to few hundred nanometers [65,68]. The structure relates to the primary particles organization into the aggregate. Wide entwining or splitting gives a high structure, while less noticeable interlacing or

3

forking denotes a low structure. Fig. 1 represents a TEM photograph of a typical aggregate of carbon black together with a schematic representation of a primary particle cross section. Carbon black particles are generally composed of small crystallites showing a turbostratic arrangement. In a way, certain confusion may exist in the literature between the terms carbon black and black carbon. Indeed the first refers to a material (product) while the second refers to a fatal, unavoidable hazardous combustion residue (waste). We will see in the following that they generally originate from very similar growth mechanisms. Carbon black properties generally result from the processing conditions that can be, to a certain extent, controlled to produce different grades. The applications of carbon black materials depend on a wide number of physicochemical properties that will not be developed here. Regarding their main applications, it is however important to point out the two most important parameters, widely used in carbon black industry and playing a key role for processing and vulcanization properties of rubber compounds which are: (i) mean particle diameter and (ii) aggregate morphology currently referred as carbon black structure. Straight measurement of particle size is impractical however, native carbon black (without post treatment), having almost zero porosity, a simple method to estimate the mean particle diameter is to measure the specific surface area which directly depends on the mean particle diameter. Aggregate morphology can be assessed with different methods based on absorption ability.

Gas phase nucleation and growth Carbon black formation can be seen as a continuous dehydrogenation process of a carbon system, which is initially in the form of hydrocarbons fuel. The thermodynamics drive the formation of CeC bondings at the expense of CeH bonds, creating at first alkene and alkyne precursors. The main species is usually the acetylene [70,71]. Then precursors react to create progressively aromatic

Fig. 1 e TEM image of a typical carbon black aggregate together with a schematic representation of a primary particle cross section [69].

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

4

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 x x x ( 2 0 1 7 ) 1 e1 7

carbon molecules, named polycyclic aromatic hydrocarbons (PAHs) [65]. The mechanism creating PAHs highly depends on the hydrocarbons fuel [72]. The PAHs continuously grow by chemical reaction and form bigger PAHs, up to 10e18 cycles depending on the operating conditions. The main growing path is the H-Abstraction-C2 H2 -Addition (HACA) mechanism [73]. The kinetic of HACA mechanism decreases with the growth [74] due to the increase of the thermodynamic barrier. The growth toward bigger carbon compounds is then initiated by PAH collisions. The main sticking causes of two colliding PAHs remain unclear but it involves Van der Waals forces [75] and/or chemical bond formation [76]. Recent experimental evidences suggest the nucleation occurs as a circumpyrene (14 cycles) staking [77e79] which tends to prevail Van der Waals hypothesis. Nevertheless, chemical bonds hypothesis still remains valid especially for very high temperature process (>2000 K) where the Van der Waals forces cannot be strong enough to resist the thermal stress [74,80]. This step of first PAH collision is known as the nucleation process. After the nucleation, the large resulting PAHs continuously coalescence and grow to create nanodroplets of highly viscous tar. The solidification of these droplets, also called maturation process, arises by an internal rearrangement of the constituting PAHs according to graphitic layers [81,82]. Eventually, primary carbon nanoparticles emerge from this solidification. In parallel, growth by chemical reactions on the surface of carbon structure and growth by coagulation continue, except the coagulation does not result in spherical structures anymore but in fractal forms. This process is known as the aggregation. Aggregate are usually formed even though the maturation and the surface growth are not yet finished. This results in irregular smoothed fractals with very strong bond between primary particles. Aggregates are thus very resistant to mechanical stress. As the maturation proceeds and the surface growth rate decreases, weaker fractal forms appear. These last structures are called agglomerates. Like quenching process in metallurgy, experimental conditions characterize the final product. In particular, the thermal and chemical histories have a major influence regarding the carbon system evolution and they define its final state. The following parameters are established as highly determinant for the quality of the solid carbon material produced: temperature [80,83,84], type of hydrocarbons fuel and its concentration [72], hydrogen concentration, oxygen presence and concentration [85], and operating pressure [86]. Fig. 2, sums up the carbon structure evolution during hydrocarbons thermal cracking.

Modeling of gas phase particles formation: population balance model The state of a system constituted of particles is usually described by its size distribution density function nV ðx; t; VÞ where nV ðx; t; VÞdV is the number of particle per unit of fluid volume with the particle's volume comprised between V

and V þ dV at a spatial position x and at the time t. V is the particle volume and is an internal variable of the size distribution density. Even though other internal variables can be considered and added, V remains the only internal variable of the size distribution density in this study. The dynamics of the system, in which particles grow by coagulation and heterogeneous reaction on their surfaces, are represented by the general dynamic equation (Eq. (4)) [87,88], also known as the population balance equation (PBE) [89].
 vnV ðx; t; VÞ ¼ V$½unV ðx; t; VÞ þ V$ Dp ðVÞVnV ðx; t; VÞ vt v ½GðVÞnV ðx; t; VÞ þ IdD ðV  V Þ  vV ZV       ~ nV x; t; V  V ~ dV ~ ~ VV ~ nV x; t; V þ b V;

(4)

0

Z∞ 

    ~ nV ðx; t; VÞn x; t; V ~ dV ~ b V; V

0

V is the critical particle volume where the nucleation begins, u is the gas speed, Dp is the particle diffusion coefficient. I is the nucleation rate, dD is the Dirac delta function, G is the growth rate by heterogeneous reaction, b is the collision frequency function for coagulation. The class method, also called discrete method, is a sectional method which was developed by Gelbard and Seinfeld [88]. It is founded on representing the continuous particle size distribution (PSD) in terms of discrete bins within which the particle density is homogeneous. By discretizing and integrating Eq. (4), and considering the steady state, Eq. (4) becomes Eq. (5).    GðVi1 ÞNi1 ðxÞ V,ðuNi ðxÞÞ ¼  V,  Dp;i VNi ðxÞ þ Vi  Vi1  GðVi ÞNi ðxÞ  þ Si;coag ðxÞ þ dD ðiÞSnuc ðxÞ; Viþ1  Vi i ¼ 0; …; M  1

ð5Þ

where Ni ðxÞ characterizes the particle concentration of size index i [#:m3 ] at the x position and is defined as follow: ZViþ1 ci2f 0; 1; …; M  1g; Ni ðxÞ ¼

nðx; VÞdV

(6)

Vi

The benefits of this approach are numerical strength and the aptitude to give the PSD directly without any hypothesis on the distribution profile. The disadvantages are that bins must be expressed a priori and that a large number of bins is generally required. The discrete technique is therefore time computing expensive but necessitates less assumptions. In the present work, the particle size distribution function is discretized with a geometric progression based on the volume. Eq. (7) shows this discretization. M different particle sizes are considered. ci2f1; 2 ; … ; M  1 g; Vi ¼ rVi1

(7)

For the next equations, the spatial dependence on N is dropped out for the sake of clarity, but the spatial dependence

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

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 x x x ( 2 0 1 7 ) 1 e1 7

5

Fig. 2 e Chronological order of the carbon particle formation during thermal conversion of hydrocarbons.

surely remains. The volume fraction ap;i and the mass fraction Yp;i of particle in the bin i, i.e. size i, can be expressed by: ci2f0; 1; …; M  1 g; ap;i ¼ Vi Ni ci2f 0; 1 ; …; M  1 g; Yp;i

(8)

r ap;i ¼ c rm

(9)

The nucleation and the growth source term in Eq. (5) are discussed in the next section, the coagulation source term is expressed Eqs. (10) and (11). This expression of the coagulation source term is valid only for r < ¼ 2. 0 S0;agg ¼ rc V0 @ 

r b N0 N0  2ðr  1Þ 0;j

M1 X

1 b0;j N0 Nj A

(10)

j¼1

  i1 X   Vj þ Vi1  Vi1 1  0:5di1;j bi1;j Ni1 Nj Si;coag ¼ Vi  Vi1 j¼0   i  X   Viþ1  Vi þ Vj þ 1  di;M1 bi;j Ni Nj 1  0:5di;j Viþ1  Vi j¼0 þ di;M1

M1 X j¼0



M1 X j¼0



1  0:5dj;M1

CFD modeling of particles formation in a 3-phase AC plasma reactor

 VM1 þ Vj bM1;j NM1 Nj VM1

bi;j Ni Nj ; ci2f1 ; … ; M  1g

ð11Þ

In order to gain more understanding of the process, a CFD model was built, solving the previously presented equations. 3D simulation of a 55 kWe plasma process for methane cracking is performed. The models takes into account detailed chemistry mechanism coupled with a population balance model for particle formation and growth.

Geometry The geometry used relates to Deme et al. work [90]. The geom
of etry represents the plasma torch developed at center PERSEE MINES-ParisTech. Fig. 3 illustrates the geometry. The experimental trials of Deme et al. were performed using nitrogen as plasma carrier gas. The carrier gas enters at the top of the reactor where it is heated by an electrical arc. The plasma flow is then accelerated by a nozzle and then enters in a cylindrical reactive chamber where the methane is injected, through circular aperture situated on the side of the reactive chamber wall.

Models Source model The unsteady state behavior of the arc is simplified using time-averaged source terms of thermal energy and

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

6

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 x x x ( 2 0 1 7 ) 1 e1 7

value. More information about this method are available in the ANSYS Fluent user guide [92].

Chemical mechanism During previous investigations [86], the chemical mechanism of fuel decomposition was modeled according to a simple first order reaction, Eq. (12). (12) CH4 /CðsÞ þ 2H2 In order to increase the chemical mechanism accuracy, different literature mechanisms were reviewed [53,82,93e96]. After a test on the accuracy versus computational cost of the different mechanisms, the authors chose the Holmen mechanism [94]. The Holmen mechanism is presented in Table 1 and considers 19 different species: H2 , H, CH4 , CH3 , C2 H, C2 H2 , C2 H3 , C2 H4 , C2 H5 , C2 H6 , a  C3 H4 , n  C3 H7 , C3 H6 , p  C3 H4 , a  C3 H5 , C4 H4 , C4 H5 , C4 H6 , C6 H6 . The authors then added two reactions of solid carbon production from acetylene and benzene precursors, precursors which are generated by the Holmen mechanism. The Eqs. (13) and (14) present these two added reactions. The kinetic parameters of this reaction are based on Fincke et al. research on methane cracking by plasma process [53].

Fig. 3 e Scheme of the geometry used for the modeling of the MINES-ParisTech reactor. The plasma chamber describes the part of the reactor where the plasma is generated. The reactive chamber is the part of the reactor where methane dissociation and carbon particles formation occurs.

momentum in an alleged constant arc volume. This average results from previous Fulcheri's group research [91]. The arc volume is approximated to a torus with a circular section and with a volume of2:473  105 m3 . The thermal energy provided by the electrical arc is estimated to 50.4 kW. The momentum source is set to 420 N/m along the axial direction and pushes the flow toward the outlet of the reactor.

Turbulence model The turbulence is simulated using the RNG-k  ε model with an enhanced wall treatment close to the wall. This method consists in the division of the boundary layer in two sublayers: a turbulent layer and a viscous-type layer close to the wall. The viscosities are not the same and are specified in the near-wall cells according to the turbulent Reynolds number

C6 H6 /6Cs þ 3H2

(13)

C2 H2 þ Cs /3Cs þ H2

(14)

Eqs. (13) and (14) are the two sources of carbon mass in the model. This mass of carbon produced is included within the nucleation source term in the population balance model (PBM). The nucleation source term, used in the PBM, is described Eq. (15). Snuc ¼

     MC EA;1 EA;2 þ 2k2 ½C2 H2 ½Cs exp 6k1 ½C6 H6 exp rC V0 RT RT (15)

rC represents the solid carbon density, MC is the carbon molecular weight, R the ideal gas constant, T the temperature, (k1 , EA;1 ) and (k2 , EA;2 ) are respectively the kinetic parameters of the reactions 13 and 14. Their values are presented in Table 2. The nucleation source term embeds the surface growth by the reaction 14. The heterogeneous reaction term in the PBM is thus set to zero here (G ¼ 0). By doing this, the authors assume that the surface growth mainly occurs on the smallest carbon particles. This assumption was already corroborated by Leung and Lindsted [97]. A not yet published study by Fulcheri's group, on the surface growth influence on the PSD, also reaches the same conclusion. Nevertheless, an error estimation of this hypothesis is expected in the near future.

Thermodynamic and transport properties The ideal gas law is considered. The conductivity, viscosity and mass diffusivity of the mixture are determined according to the gas kinetic theory [98]. Thermodynamic properties and

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

7

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 x x x ( 2 0 1 7 ) 1 e1 7

Table 1 e Holmen mechanism [94], k ¼ ATn expðE=ðRTÞÞ, E in cal/mol and k uses cm3 , moles, seconds. No

Reaction

A

n

E

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

CH4 /CH3 þ H CH4 þ H/CH3 þ H2 CH3 þ CH3 /C2 H6 C2 H6 þ H/C2 H5 þ H2 C2 H6 þ CH3 /C2 H5 þ CH4 C2 H5 /C2 H4 þ H CH3 þ CH3 /C2 H4 þ H2 C2 H4 þ CH3 /C2 H3 þ CH4 C2 H4 þ CH3 /n  C3 H7 C2 H4 þ H/C2 H3 þ H2 C2 H3 /C2 H2 þ H CH3 þ C2 H3 /C3 H6 n  C3 H7 /C3 H6 þ H C3 H6 /a  C3 H5 þ H C3 H5 /C2 H2 þ CH3 C3 H5 /a  C3 H4 þ H C3 H5 þ H/a  C3 H4 þ H2 C3 H6 þ H/a  C3 H5 þ H2 C2 H3 þ C2 H3 /C4 H6 C2 H3 þ C2 H4 /C4 H6 þ H C2 H2 þ H/C2 H þ H2 C2 H2 þ CH3 /C2 H þ CH4 C4 H6 þ H/C4 H5 þ H2 C4 H5 /C4 H4 þ H C2 H þ H/C2 H2 C2 H3 þ C2 H2 /C4 H5 CH3 þ CH3 /C2 H5 þ H C4 H5 þ C2 H2 /C6 H6 þ H C2 H4 /C2 H3 þ H C2 H5 þ C2 H2 /C2 H6 þ C2 H C2 H5 þ H/C2 H6 C2 H4 /C2 H2 þ H2 C2 H3 þ H/C2 H2 þ H2 C2 H2 þ CH3 /p  C3 H4 þ H C3 H6 /p  C3 H4 þ H2 C3 H6 þ CH3 /a  C3 H5 þ CH4

3:51  1015 2:25  104 1:01  1015 5:54  102 5:50  101 2:00  1013 1:00  1016 6:62  100 2:00  1011 1:32  106 1:93  1028 1:00  1013 1:58  1016 1:00  1015 1:16  1010 5:00  1013 1:00  1013 5:00  1012 1:26  1013 5:00  1011 6:02  1013 1:81  1011 1:00  1014 1:00  1015 1:81  1014 1:10  1012 1:80  1012 6:02  1012 1:00  1016 2:71  1011 3:07  1013 7:94  1012 9:64  1013 6:20  1011 8:00  1012 1:58  1012

0 3 0.64 3.5 4 0 0 3.7 0 2.53 4.783 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.44 0 0 0.44 0

104,000 8768 0 5174 8296 39,700 32,000 9512 7170 12,258 51,123 0 38,000 88,000 43,200 35,000 0 1500 0 7315 22,300 17,300 15,000 41,400 0 4000 10,400 9000 108,000 23,400 0 88,760 0 20,000 81,150 8800

transport coefficients for every species come from CHEMKIN database from the following references [53,94].

reactor with a flow of 9.00 Nm3 =h and at 300 K. Boundary conditions for species and particles transport equations followed a null Neumann condition at the wall.

Boundary conditions Model hypothesis and limits The boundary conditions for the energy equations are represented in Fig. 4. In this figure, h represents a heat transfer coefficient which was determined by experimental calorimetric balances for the different walls of the reactor [90]. Pure methane flow enters at the fuel inlet with a flow of 0.56 Nm3 =h and at 300 K. The nitrogen enters at the top of the

Table 2 e Kinetic parameter values of Eqs (13) and (14), index 1 corresponding to Eq. (13) and index 2 corresponding to Eq. (14). Kinetic parameter k1 EA;1 k2 EA;2

Value

Unit

0:75  10 174.6 5:00  107 100.6 5

1

s kJmol1 m3 mol1 s1 kJmol1

1. Due to the high computation cost of the chemical mechanism implemented in the CFD code, no radiative model could be used presently. Consequently, temperature gradients are overestimated and the loss of heat energy at the wall is underestimated. 2. The very fast chemistry occurring when the methane flow meets the plasma flow “forced” the used of an implicit transient solver. The permanent is reached using a significant time step which might underestimate the kinetics of the fastest reactions among the mechanism used. Results on the plasma/hydrocarbons mixing zone have to be dealt with cautiously. 3. The maturation of the particles is not considered. A slight over-estimation of the mean particle diameter may result from this approximation.

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

8

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 x x x ( 2 0 1 7 ) 1 e1 7

Fig. 4 e 3D schematic view of a quarter of the MINES-ParisTech reactor with thermal boundary conditions used for the modeling. The injection plane can been seen as the plane (yz).

4. The formation of fractal carbon structures (aggregation) is not modeled since the carbon black quality mainly depends on the primary particle size. 5. Carbon deposition on the wall is not modeled. 6. No possible effects of the turbulence on the reaction rates are modeled. 7. Gravity is taken into account (gravity vector toward negative z, Fig. 4).

Mesh and computation ANSYS Fluent v16.0 software has been used to solve the system of transport equations describing the flow in the geometry. The geometry is divided into two structured

meshes in order to benefit of the axial symmetry of the plasma chamber part, see Fig. 3 for the part definition. A 2D mesh of 132 050 cells is used for the plasma chamber with a fine discretization close to the electric arc. Output results from the plasma chamber are used as input for a 3D mesh of 442 818 cells describing the reactive chamber. The plasma flow at the outlet of the plasma chamber is assumed to be independent of the reactive chamber results since the methane jet is very small compared to the plasma jet. The simulation has been run on a Linux Intel cluster, architecture x86_64, using 18 processors at 2.4 GHz. Simulation took about 4 days to converge. Most of the computation time is due to the methane dissociation mechanism used in such a geometry.

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

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 x x x ( 2 0 1 7 ) 1 e1 7

9

Fig. 5 e Velocity field at the beginning of the reactive chamber within the injection plane.

Fig. 6 e Temperature field in the injection plane of the reactive chamber of the reactor.

Result analysis and discussion Results are focused on the reactive chamber of the reactor since this is where the chemical reactions and the particle formation and growth occur.

Fig. 7 e Temperature field in a horizontal section at axial position of the methane injection, the black circle represents the plasma injection position at the top of the reactor.

Fig. 5 represents the velocity field at the beginning of the reactive chamber in the injection plane (plane (yz) of Fig. 4). The left side of Fig. 5 represents the top of the reactive chamber and the right side represents the bottom. Fig. 5 shows a high mixing zone. This mixing zone, driven by fluid recirculation, is beneficial for the hydrocarbons cracking because the bigger the residence time is the higher the methane cracking rate is. Fig. 6 represents the temperature field in the injection plane. This figure shows a large decrease of the temperature when the plasma gas meets the hydrocarbons jet, going from 6000 K to 3000 K in few centimeters. This figure exhibits the high thermal energy consumption by chemical reactions and the rapidity of these reactions. Fig. 7 shows the temperature field in a horizontal part of the reactive chamber where the methane is injected. It illustrates how hydrocarbon jet goes deep in plasma flow. Fig. 7 exposes how the hydrocarbon jet goes deep in the plasma flow. According to this figure the hydrocarbon flow rate is well defined since it stops at the reactive chamber center. Fig. 8 gives the mass fractions of the following species CH4 , C2 H2 , C6 H6 and Cs in the injection plane of the reactive chamber (plane (yz) of Fig. 4). Fig. 8 reveals how quickly the methane is dissociated and that acetylene (C2 H2 ) constitutes the main precursors for solid carbon production. The rate of methane dissociation is 100%. An interesting observation is the presence of a small amount of benzene (C6 H6 ) at the end of the methane jet. This small amount of benzene is associated to the initiation of solid carbon production since

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

10

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 x x x ( 2 0 1 7 ) 1 e1 7

Fig. 8 e Mass fractions of CH4 , C2 H2 , C6 H6 and Cs in the injection plane of the reactive chamber.

acetylene reaction cannot produce solid carbon without the presence of solid carbon. Once the carbon production is initiated by the benzene dissociation, see Eq. (13), the main path for solid carbon production swiftly becomes the acetylene reaction, Eq. (14). Figs. 9 and 10 give the mass fraction of carbon particles for different particle sizes, going from nanoparticles (top Fig. 9) to macroparticles (bottom Fig. 10), in the injection plane of the reactive chamber. These figures show the growth of the particle along with the flow. The first coagulation processes are extremely fast and, consequently, the particle growth preliminary occurs in counterflow. The preliminary growth actually follows more the nuclei1 gradient than the flow. 1 Nuclei can be define as the smallest identifiable carbon particles. In this simulation nuclei are particles which have a volume V0 .

Once the particles' size reaches few nanometers, the coagulation process gets slower and the growth finally follows the flow. Same observation can be seen in the normal plane of the injection plane, Figs. 11 and 12, at the difference that the particle concentrations, in this plane (plane (xz) of Fig. 4), are lower for the first half of the reactor. This difference attenuates when getting close to the outlet due to a progressive homogenization of the fluid mixture by diffusion and convection. These figures of size particle mass fraction show a trend to find, at the outlet, bigger particles close to the wall. From these results, the average particle diameter at the outlet can be computed and is about 72.4 nm. A good agreement is obtained by comparing this mean global mean diameter to the mean diameter of the particles observed on the scanning electron microscope images from the experiment.

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

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 x x x ( 2 0 1 7 ) 1 e1 7

11

Fig. 9 e Mass fractions for different particle sizes in the injection plane of the reactive chamber of the reactor. Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

12

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 x x x ( 2 0 1 7 ) 1 e1 7

Fig. 10 e Mass fractions for different particle sizes in the injection plane of the reactive chamber of the reactor. Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

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 x x x ( 2 0 1 7 ) 1 e1 7

13

Fig. 11 e Mass fractions for different particle size in the normal plane of the injection plane of the reactive chamber. Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

14

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 x x x ( 2 0 1 7 ) 1 e1 7

Fig. 12 e Mass fractions for different particle size in the normal plane of the injection plane of the reactive chamber. Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

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 x x x ( 2 0 1 7 ) 1 e1 7

Conclusion In 1995, in an article published in IJHE entitled: From methane to hydrogen, carbon-black and water [31] Fulcheri and Schwob presented a new concept for the cracking of methane allowing the co-production of carbon black and hydrogen using a plasma. This process has been under development at MINES-ParisTech for more than 20 years and today the idea is about becoming an industrial reality since an industrial pilot plant based on this concept is currently under development in the USA. The present article exposes the last advances of MINESParisTech in such a process. A detailed theoretical vision of carbon black particle production during hydrocarbon cracking is presented and gives a deep understanding of the different phenomena involved during the formation and the growth of carbon particles. The MINES-ParisTech plasma process prototype is described and its associated numerical modeling is performed. The model solves a reactive turbulent flow in a complex 3D geometry. In this modeling, a detailed reaction mechanism is used to produce acetylene and benzene, both identified as carbon particle precursors. The growth of carbon particle is handled by a sectional method. The numerical results give a complete vision of the different steps in the process. A broad validation of the numerical model is given by the measure of out-process mean particle diameter. The present numerical part is a preliminary step of a larger numerical campaign which will consist in parametric studies on carbon black formation by plasma process in order to optimize the carbon black grade. The thermal plasma direct methane decarbornization presents a prodigious potential. Nevertheless, the economic viability of the process currently depends on the aptitude to simultaneously give pure hydrogen and high value-added carbon black with well-controlled characteristics, especially the particle size distribution. Consequently, a deep understanding of gas phase nucleation and growth phenomena is decisive for the successful development of this process at short term. Regarding longer term prospects, the direct methane decarbornization by thermal plasma could present a truly alternative to carbon capture and sequestration by sequestering solid carbon instead of CO2. By doing so, the environmental risks are significantly reduced. As a matter of fact, when regarding the hydrogen production solely, direct methane decarbornization is ideally 8 times less costly than the water electrolysis. When compared to the current steam reforming process, the direct methane decarbornization can potentially reduce the CO2 emissions from 12 kg of CO2 per kg of H2, to zero emission. In this way, the direct methane decarbornization by thermal plasma is one of the most reliable alternative technology compatible with the 2$/(kg H2) objective of the United States Department of Energy.

references

[1] Dincer I. Green methods for hydrogen production. Int J Hydrogen Energy 2012;37(2):1954e71. 10th International Conference on Clean Energy 2010.

15

[2] Dincer I, Acar C. Review and evaluation of hydrogen production methods for better sustainability. Int J Hydrogen Energy 2015;40(34):11094e111. [3] Muradov N. Hydrogen via methane decomposition: an application for decarbonization of fossil fuels. Int J Hydrogen Energy 2001;26(11):1165e75. [4] Muradov NZ, Veziroglu TN. From hydrocarbon to hydrogencarbon to hydrogen economy. Int J Hydrogen Energy 2005;30(3):225e37. [5] Muradov NZ, Veziroglu TN. “Green” path from fossil-based to hydrogen economy: an overview of carbon-neutral technologies. Int J Hydrogen Energy 2008;33(23):6804e39. [6] Muradov N. Low to near-zero CO2 production of hydrogen from fossil fuels: status and perspectives. Int J Hydrogen Energy 2017;42(20):14058e88. [7] Steinberg M. Fossil fuel decarbonization technology for mitigating global warming. Int J Hydrogen Energy 1999;24(8):771e7. [8] Hirsch D, Epstein M, Steinfeld A. The solar thermal decarbonization of natural gas. Int J Hydrogen Energy 2001;26(10):1023e33. [9] Hirsch D, Steinfeld A. Radiative transfer in a solar chemical reactor for the co-production of hydrogen and carbon by thermal decomposition of methane. Chem Eng Sci 2004;59(24):5771e8. [10] Hirsch D, Steinfeld A. Solar hydrogen production by thermal decomposition of natural gas using a vortex-flow reactor. Int J Hydrogen Energy 2004;29(1):47e55. [11] Abanades S, Flamant G. Production of hydrogen by thermal methane splitting in a nozzle-type laboratory-scale solar reactor. Int J Hydrogen Energy 2005;30(8):843e53. [12] Rodat S, Abanades S, Sans J-L, Flamant G. A pilot-scale solar reactor for the production of hydrogen and carbon black from methane splitting. Int J Hydrogen Energy 2010;35(15):7748e58. the 10th Chinese Hydrogen Energy Conference. [13] Abanades S, Flamant G. Experimental study and modeling of a high-temperature solar chemical reactor for hydrogen production from methane cracking. Int J Hydrogen Energy 2007;32(10e11):1508e15. [14] Patrianakos G, Kostoglou M, Konstandopoulos AG. Effect of seeding on hydrogen and carbon particle production in a 10 mw solar thermal reactor for methane decomposition. Int J Hydrogen Energy 2012;37(21):16570e80. [15] Mohsenian S, Esmaili MS, Fathi J, Shokri B. Hydrogen and carbon black nano-spheres production via thermal plasma pyrolysis of polymers. Int J Hydrogen Energy 2016;41(38):16656e63. [16] JR Rose. Process of and apparatus for producing carbon and gaseous fuel. United States Patent Office (1,352,085). [17] WK Lewis. Method of producing carbon black. US Patent 1536612 A; 1923. [18] Jakowsky J. Process and apparatus for manufacture of carbon-black unsaturated gases and hydrogen. 1926. [19] Eisenberg RS. How can you patent genes? Am J Bioeth 2002;2(3):3e11. [20] D Gardner, Production of carbon by electrical discharge, US 2572851 A; 1951. [21] C Sheer, S Korman, Production of carbon black, US 3009783 A; 1961. [22] Gonzalez-Aguilar J, Moreno M, Fulcheri L. Carbon nanostructures production by gas-phase plasma processes at atmospheric pressure. J Phys D Appl Phys 2007;40(8):2361e74. [23] Bakken JA, Jensen R, Monsen B, Raaness O, Waernes AN. Thermal plasma process development in Norway. Pure Appl Chem 1998;70(6):1223e8.

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

16

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 x x x ( 2 0 1 7 ) 1 e1 7

[24] S Lynum, K Hox, J Hugdahl, Production of carbon black. Patent WO 9320154; 1993. [25] S Lynum, K Haugsten, K Hox, J Hugdahl. Plasma torch device for chemical processes. US Patent 5846674; 1996. [26] S Lynum, K Hox, K Haugsten, J Langoy. System for the production of carbon black. WO 93/20153; 1993. [27] S Lynum, K Hox, J Hugdahl. Production of carbon black. WO1993020154 A1; 1993. [28] S Lynum, K Hox, N Myklebust, J Hugdahl. Electrode consumption in plasma torches. US 5951960 A; 1999. [29] S Lynum, N Myklebust, K Hox. Decomposition of hydrocarbon to carbon black, US 6068827 A; 2000. [30] Gaudernack B, Lynum S. Hydrogen from natural gas without release of CO2 to the atmosphere. Hydrogen Energy Prog XI 1996;1e3:511e23. [31] Fulcheri L, Schwob Y. From methane to hydrogen, carbonblack and water. Int J Hydrogen Energy 1995;20(3):197e202. [32] Schwob Y, Fischer F, Fulcheri L, Willemez P. Conversion de carbone ou de composes carbones dans un plasma. 1994. [33] Y Schwob. Method and device for producing fullerenes. US 6358375 B1; 2002. [34] Fabry F, Grivei E, Probst N, Smet R, Peroy J, Flamant G, et al. Device and method for converting carbon containing feedstock into carbon containing materials, having a defined nanostructure. 2008. [35] Charlier J, Fabry F, Flamant G, Gonzales J, Grivei E, Gruenberger T, et al. Carbon nanostructures and process for the production of carbon-based nanotubes, nanofibres and nanostructures. 2004. [36] Fulcheri L, Schwob Y, Flamant G. Comparison between new carbon nanostructures produced by plasma with industrial carbon black grades. J De Phys III 1997;7(3):491e503. [37] Ravary B, Fulcheri L, Flamant G, Fabry F. Analysis of a 3-phase ac plasma system. High Temp Mater Process 1998;2(2):245e60. [38] Ravary B, Fulcheri L, Bakken JA, Flamant G, Fabry F. Influence of the electromagnetic forces on momentum and heat transfer in a 3-phase ac plasma reactor. Plasma Chem Plasma Process 1999;19(1):69e89. [39] Fabry F, Flamant G, Fulcheri L. Carbon black processing by thermal plasma. analysis of the particle formation mechanism. Chem Eng Sci 2001;56(6):2123e32. [40] Fulcheri L, Probst N, Flamant G, Fabry F, Grivei E, Bourrat X. Plasma processing: a step towards the production of new grades of carbon black. Carbon 2002;40(2):169e76. [41] Ravary B, Bakken JA, Gonzalez-Aguilar J, Fulcheri L. Cfd modeling of a plasma reactor for the production of nanosized carbon materials. High Temp Mater Process 2003;7(2):139e44. [42] Gonzalez-Aguilar J, Deme I, Fulcheri L, Gruenberger TM, Fabry F, Flamant G, et al. 3d modelling of carbon black formation and particle radiation during methane cracking by thermal plasma. High Temp Mater Process 2003;7(1):51e6. [43] Abanades S, Badie JM, Flamant G. On-line temperature measurement in a plasma reactor for fullerene synthesis. High Temp Mater Process 2003;7(1):43e9. [44] Gonzalez-Aguilar J, Deme I, Fulcheri L, Flamant G, Gruenberger TM, Ravary B. Comparison of simple particleradiation coupling models applied on a plasma black process. Plasma Chem Plasma Process 2004;24(4):603e23. [45] Fulcheri L, Gruenberger TM, Aguilar JG, Fabry F, Grivei E, Probst N, et al. Plasma processing of carbon nanomaterials. High Temp Mater Process 2004;8(1):119e38. [46] Gonzalez-Aguilar J, Fulcheri L, Flamant G. Particle contribution to the radiative heat transfer in a plasma reactor. High Temp Mater Process 2006;10(1):39e46. [47] Fulcheri L, Fabry F, Takali S, Rohani V. Three-phase ac arc plasma systems: a review. Plasma Chem Plasma Process 2015;35(4):565e85.

[48] Merlo-Sosa G. Laure; Soucy, Dodecane decomposition in a radio-frequency (rf) plasma reactor. Int J Chem React Eng 2005;3(4):2e26. [49] Pristavita R, Roy R, Moran B, Meunier JL, Berk D. Volatile compounds present in carbon blacks produced by thermal plasmas. Plasma Chem Plasma Process 2011;31(6):839e50. [50] Pristavita R, Meunier JL, Berk D. Carbon nano-flakes produced by an inductively coupled thermal plasma system for catalyst applications. Plasma Chem Plasma Process 2011;31(2):393e403. [51] Pristavita R, Mendoza-Gonzalez NY, Meunier JL, Berk D. Carbon blacks produced by thermal plasma: the influence of the reactor geometry on the product morphology. Plasma Chem Plasma Process 2010;30(2):267e79. [52] Pristavita R, Mendoza-Gonzalez NY, Meunier JL, Berk D. Carbon nanoparticle production by inductively coupled thermal plasmas: controlling the thermal history of particle nucleation. Plasma Chem Plasma Process 2011;31(6):851e66. [53] Fincke JR, Anderson RP, Hyde TA, Detering BA. Plasma pyrolysis of methane to hydrogen and carbon black. Ind Eng Chem Res 2002;41(6):1425e35. [54] Kim KS, Seo JH, Nam JS, Ju WT, Hong SH. Production of hydrogen and carbon black by methane decomposition using dc-rf hybrid thermal plasmas. IEEE Trans Plasma Sci 2005;33(2):813e23. [55] Kim SC, Lim MS, Chun YN. Hydrogen-rich gas production from a biomass pyrolysis gas by using a plasmatron. Int J Hydrogen Energy 2013;38(34):14458e66. [56] Cho W, Lee SH, Ju WS, Baek Y, Lee JK. Conversion of natural gas to hydrogen and carbon black by plasma and application of plasma carbon black. Catal Today 2004;98(4):633e8. [57] Zhang YP, Liu CJ, Zhang MH. Carbon black and carbon nanotubes produced from acetylene using non-thermal plasmas at atmospheric pressure. Chem Lett 2000;29:1204e5. [58] Sun DL, Wang F, Hong RY, Xie CR. Preparation of carbon black via arc discharge plasma enhanced by thermal pyrolysis. Diam Relat Mater 2016;61:21e31. [59] Boutot T, Buckle K, Collins F. High-concentration hydrogen production from natural gas using a pulsed dielectric barrier discharge. In: Hydrogen and fuel cells conference, Toronto, Canada; 2004. [60] Boutot T, Liu Z, Niu W, Whidden T, Yang Y. Hydrogen and nanostructured carbon by plasma decomposition of natural gas. In: 214th ECS meeting, abstract N 477, Hawaii, USA; 2008. [61] Liu Z, Whidden T, Boutot T, Yang Y. Apparatus and method for effecting plasma-based reactions. 2010. [62] Turner JA. Sustainable hydrogen production. Science 2004;305(5686):972e4. [63] Abbas HF, Daud WW. Hydrogen production by methane decomposition: a review. Int J Hydrogen Energy 2010;35(3):1160e90. [64] Lucintel. Global carbon black market 2015e2020: trends, forecast, and opportunity analysis. 2015. [65] Donnet J. Carbon black: science and technology. 2nd ed. Taylor & Francis; 1993. [66] Fulcheri L. HDR : nanostructures de carbone par plasma. University of Perpipgnan; 2003. [67] Fulcheri L. Direct decarbonization of methane by thermal plasma for the co synthesis of carbon black and hydrogen. In: The 14th High-tech plasma processes conference; 2016. [68] Chadwick SS. Ullmann's encyclopedia of industrial chemistry. Ref Serv Rev 1988;16(4):31e4. [69] Heidenre R, Hess WM, Ban LL. A test object and criteria for high resolution electron microscopy. J Appl Crystallogr 1968;1:1e19. [70] Porter G. Combustion researches and reviews. London: Butterworth Publishing Co.; 1955. p. 108e24. [71] Bonne U, Homann K, Wagner H. Carbon formation in premixed flames. In: Tenth symposium (international) on

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021

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 x x x ( 2 0 1 7 ) 1 e1 7

[72]

[73] [74] [75]

[76]

[77]

[78]

[79]

[80] [81]

[82]

[83]

[84]

combustion, the combustion institute, Pittsburgh; 1965. p. 503e12. Richter H, Howard JB. Formation of polycyclic aromatic hydrocarbons and their growth to soot e a review of chemical reaction pathways. Prog Energy Combust Sci 2000;26(4e6):565e608. Frenklach M. Reaction mechanism of soot formation in flames. Phys Chem Chem Phys 2002;4(11):2028e37. Wang H. Formation of nascent soot and other condensedphase materials in flames. Proc Combust Inst 2011;33:41e67. Frenklach M. Computer modeling of infinite reaction sequences e a chemical lumping. Chem Eng Sci 1985;40(10):1843e9. D'Anna A. Detailed kinetic modeling of particulate formation in rich premixed flames of ethylene. Energy Fuels 2008;22(3):1610e9. Grotheer HH, Pokorny H, Barth KL, Thierley M, Aigner M. Mass spectrometry up to 1 million mass units for the simultaneous detection of primary soot and of soot precursors (nanoparticles) in flames. Chemosphere 2004;57(10):1335e42. Herdman JD, Connelly BC, Smooke MD, Long MB, Miller JH. A comparison of raman signatures and laser-induced incandescence with direct numerical simulation of soot growth in non-premixed ethylene/air flames. Carbon 2011;49(15):5298e311. Botero ML, Chen DP, Gonzalez-Calera S, Jefferson D, Kraft M. Hrtem evaluation of soot particles produced by the nonpremixed combustion of liquid fuels. Carbon 2016;96:459e73. D'Anna A. Combustion-formed nanoparticles. Proc Combust Inst 2009;32:593e613. Hurt RH, Crawford GP, Shim HS. Equilibrium nanostructure of primary soot particles. Proc Combust Inst 2000;28:2539e46. Kholghy MR, Veshkini A, Thomson MJ. The core-shell internal nanostructure of soot e a criterion to model soot maturity. Carbon 2016;100:508e36. Blanquart G, Pitsch H. Analyzing the effects of temperature on soot formation with a joint volume-surface-hydrogen model. Combust Flame 2009;156(8):1614e26. Dewa K, Ono K, Watanabe A, Takahashi K, Matsukawa Y, Saito Y, et al. Evolution of size distribution and morphology of carbon nanoparticles during ethylene pyrolysis. Combust Flame 2016;163:115e21.

17

[85] Frenklach M, Wang H. Detailed mechanism and modeling of soot particle formation. Berlin, Heidelberg: Springer Berlin Heidelberg; 1994. p. 165e92. [86] Gautier M, Rohani V, Fulcheri L, Trelles JP. Influence of temperature and pressure on carbon black size distribution during allothermal cracking of methane. Aerosol Sci Technol 2016;50(1):26e40. [87] Friedlander S. Smoke, dust, and haze: fundamentals of aerosol dynamics. Oxford University Press; 2000. [88] Gelbard F, Seinfeld JH. Simulation of multicomponent aerosol dynamics. J Colloid Interface Sci 1980;78(2):485e501. [89] Ramkrishna D. Population balances: theory and applications to particulate systems in engineering. Elsevier Science; 2000. [90] Deme I. Contribution la modlisation de l'coulement dans un racteur plasma pour la fabrication de noirs de carbone: Influence du rayonnement des particules de carbone. Thesis. 2002. [91] Rehmet C. tude thorique et exprimentale dune torche plasma triphase arcs libres associe un procd de gazification de matire organique. Thesis. 2013. [92] ANSYS. ANSYS FLUENT user's guide, https://www.sharcnet. ca/Software/Ansys/17.0/. [93] Dors M, Nowakowska H, Jasinski M, Mizeraczyk J. Chemical kinetics of methane pyrolysis in microwave plasma at atmospheric pressure. Plasma Chem Plasma Process 2014;34(2):313e26. [94] Holmen A, Olsvik O, Rokstad OA. Pyrolysis of natural-gas e chemistry and process concepts. Fuel Process Technol 1995;42(2e3):249e67. [95] Eaves NA, Dworkin SB, Thomson MJ. The importance of reversibility in modeling soot nucleation and condensation processes. Proc Combust Inst 2015;35:1787e94. [96] Ranzi E, Frassoldati A, Grana R, Cuoci A, Faravelli T, Kelley AP, et al. Hierarchical and comparative kinetic modeling of laminar flame speeds of hydrocarbon and oxygenated fuels. Prog Energy Combust Sci 2012;38(4):468e501. [97] Leung KM, Lindstedt RP, Jones WP. A simplified reactionmechanism for soot formation in nonpremixed flames. Combust Flame 1991;87(3e4):289e305. [98] Bird R, Stewart W, Lightfoot E. Transport phenomena. Wiley; 2007.

Please cite this article in press as: Gautier M, et al., Direct decarbonization of methane by thermal plasma for the production of hydrogen and high value-added carbon black, International Journal of Hydrogen Energy (2017), https://doi.org/10.1016/ j.ijhydene.2017.09.021