Carbon analysis and supplementary data 

Get Complete Project Material File(s) Now! »

High performance of noble metals

Among the DRM catalysts, researchers agree that noble metals (Rh, Ru, Pt, Ir and Pd) are the most active and resistant to carbon deposition.31,33,36,37,38,39,40 Table A-1 of the Appendix reports few bibliographic data on catalysts based on noble metals. The catalytic activity and stability of different noble metals decreased in the following order: Rh~Ru>Ir> Pd >Pt.37,41 Although Rh and Pt are among the most studied elements,41,42,43,44,45,46,47,48 the most efficient catalysts were those containing Rhodium (Rh) and Ruthenium (Ru). However, despite their excellent activity, stability and selectivity, the high cost of noble metals and their restricted availability hinder their industrial application. In fact, the price of noble metals does not only depend on the nature of the metal, but also on the type of precursor. In comparison to nickel, the price of noble metals can be 1500-15000 times higher.
Consequently, noble metals are most often used in bimetallic systems, as shown in Table A-3 of the Appendix. Doping of catalysts with noble metals in small quantities remains more economic than the use of these precious metals as active phases. The addition of Pt43, Pd49, Ru50 or Rh51,52 to Ni-based catalysts improved carbon resistance during the dry reforming reaction. The Ni-Rh clusters formed on mesoporous Al2O3,53 SiO254 and boron nitride55 supports enriched the surface of the catalysts with Ni and enhanced the reforming activity. The main advantage of these noble metals is associated with their H2 spillover effect, when H2 molecules dissociate on the surface of the noble metal to hydrogen species, which are highly mobile and can diffuse easily on adjacent ones via the support surface.48,56,57 This improves the reducibility of metal oxides and ameliorates the catalytic stability by limiting heavy coke formation.

Use of transition metals and importance of nickel

Non-noble metals like Ni and Co became attractive alternatives due to their lower cost and higher availability.58,59 Indeed, studies on methane dry reforming over Ni and Co catalysts started as early as the 1920s by Fischer and Tropsch.11 However, these catalysts were subjected to severe deactivation due to carbon deposition. Thus, people quickly dropped the research on such application. Later, few studies in the 1980s reported that these active phases were relatively active on SiO2 supports, in opposition to CuO and Fe3O4 catalysts that deactivated rapidly.60 Succeeding studies used silica61,62 and alumina25,62,63 as supports. Nevertheless, the conversions obtained did not meet the equilibrium values and side reactions continued to affect the final conversions.
At the beginning of the 21st century, the increased concerns about global warming and the depletion of petroleum oil reserves promoted renewed interests in dry reforming of methane. Since then, the importance of this topic of research continuously increased with time. We statistically considered the articles obtained by looking for either the words “CO2 dry reforming” using the Web of Science search engine, or “Methane dry reforming” using Google Scholar (Fig. I-3). Some variations are observed between the two search engines, however, the trend is the same. The resulting bar-charts reveal a significant increase in the number of publications after 2010, up to mid-2016.
Furthermore, statistics were done on some of the most studied elements in dry reforming of methane obtained by looking for “Methane dry reforming” and the corresponding metal using Google Scholar (Fig. I-3). The plot shows that Ru, Rh and Pt are no longer greatly studied, due to their high cost, as discussed earlier. In opposition, extensive studies are concerned with the use of Co and more importantly Ni.
In this literature, Co-based catalysts usually show lower coke deposition in comparison to Ni-based systems;64,65,66,67 however they remain less performing in DRM.46 Therefore, Ni-based catalysts are considered as the most promising alternatives among the non-noble metals. Indeed, close resemblance in the catalytic activity was obtained on Rh and Ni-based catalysts;41 however, Ni-based catalysts are subjected to heavy coke deposition and consequent catalytic deactivation. Therefore, several approaches are being studied in literature in order to increase the stability of these catalysts and enhance their coke-resistance. The studies mainly focus on the nature of supports, catalyst preparation methods and addition of promoters. These topics will also be important concerns of this thesis.

Ni-based catalysts for methane dry reforming

Active phase sintering causes an increase in the metal particle size and a significant reduction of its surface area, leading to lower catalytic activity. The reduction of nickel particles size promotes carbon resistance. In literature, different critical particle sizes have been selected for the inhibition/reduction of carbon formation, like 7 nm on Ni/alumina aerogel,71 10 nm on Ni-Mo/SBA-1572 and Ni/Al2O3-La2O3,73 and 15 nm on Ni/meso-Al2O374 and Pt/Ni-MCM-41.46 Baudouin et al. studied the impact of nickel particle size in the range of 1.6 to 7.3 nm on dry reforming of methane at 500°C.75 However, this range remained below the threshold determined by other authors and this explains why they found an independent relationship between catalytic activity and particle size in their study. Thus, it is generally agreed that the dispersion of active metal (Ni) is of particular importance for an efficient nickel reforming catalyst. When nano-sized nickel particles were obtained with narrow distribution in Ni/SBA-15 catalysts, the catalytic activity was even higher than that on Ru-catalysts for ammonia decomposition.76 Nonetheless, the formation of well-dispersed active metals in supported catalysts is greatly affected by metal precursors, loading techniques and supporting materials.
In general, nickel nitrate is the most commonly used precursor due to its commercial availability, low cost, high solubility in water and easy decomposition at moderate temperatures.77,78 Nevertheless, other precursors like nickel chloride, acetate, citrate or formate are tested as well.27,79,80,81,82,83,84,85 The use of chloride precursors generates large particles, as observed on Ni/Al2O3,84 Ni/TiO2,83 Ni/CeO2,27 and Fe/SBA-1579 whereas nitrate and acetate precursors appear to be efficient for small particles formation, good metal dispersion and subsequent high catalytic activities.
The catalyst preparation procedure can also play an important role in the dispersion of the active phase. For instance, different preparation techniques such as impregnation (co-impregnation or sequential, in the case of bimetallic catalysts), precipitation, direct synthesis (also known by single-step or one-pot synthesis), ion exchange, plasma treatment and many more are reviewed.86,87,31 While plasma treatment is quite expensive, other procedures are preferred due to their simplicity and their possible implementation in laboratory as well as in industry. The incipient wetness impregnation, in particular, is the most widely used method but it does not always lead to good metal dispersion.
Furthermore, the selection of catalytic supports plays a critical role in enhancing the catalytic activity and suppressing carbon formation.87 In dry reforming of methane, non-porous supports mainly consisting of Al2O3, MgO, CeO2 and/or ZrO2, microporous supports like zeolites and mesoporous supports including Al2O3, CeO2 and/or ZrO2, MCM-41, SBA-15 and MgO were tested. The differences of reactivity of Ni-species on these supports are detailed below.

N2 adsorption-desorption isotherms

Principle: Some of the textural properties of the samples can be evaluated by nitrogen sorption. Fig. II-4 illustrates the ASAP 2020 (Micromeritics) apparatus used in this work, with both degas and analysis schematics of the equipment. Degas, or vacuum outgassing, is first performed in order to remove all physisorbed species from the surface of the sample. Then, the N2 isotherms are collected using point-by-point procedure. Nitrogen gas is introduced in successive amounts, the system is allowed sufficient time to reach equilibrium pressure, and then each point on the isotherm is determined by measuring the volume of nitrogen gas (Vexp)adsorbed (or desorbed). This procedure is done at a constant temperature (liquid N2, ~77K or -196.15°C) over a range of relative pressures(P/P0) between 0 and 1.
Different types of isotherms can be obtained, depending on the textural properties of the samples. Detailed description of each type can be found in the recently updated technical IUPAC report published in 2015.
In our work, the isotherms mainly encountered are (i) type II isotherm given by physisorption of gases on nonporous or macroporous adsorbents and (ii) type IV(a) isotherm given by mesoporous adsorbents (Fig. II-5a). The inflection point (marked as point B on the isotherms) corresponds to monolayer coverage of the surface. Further increase in relative pressure results in multi-layer coverage. A typical feature of type IV(a) isotherms is the final saturation plateau up to P/P0 = 1. Furthermore, when adsorption and desorption plots do not coincide, a hysteresis loop is observed and is attributed to mesoporous structure (type IV(a) isotherms). The hysteresis loops are also classified according to their shapes into 6 characteristic types (Fig. II-5b). These will be further described in the following chapters, depending on the results obtained.

Scanning Electron Microscopy (SEM)

Principle: Micrographs are produced by scanning the specimen with a focused beam of electrons. The electrons interact with atoms in the sample; secondary ones are generated and emitted from the surface of the specimen by inelastic scattering interactions with beam electrons, generating images that can be used for information about the surface topography and composition. In addition to this secondary electron detection mode, SEM can also operate in backscattered mode, for a better visualization of external particles. By collecting retro-diffused or back-scattered electrons (emitted by elastic scattering interactions), contrast between areas with different chemical compositions can be observed since heavy elements backscatter electrons more strongly than light elements, and thus appear brighter in the image.
Procedure: The SEM images are observed on a Hitachi SU-70 Scanning Electron Microscope with a Field Emission Gun (SEM-FEG). A low voltage of 1 keV is employed at a maximum distance up to 2.4 mm between the electron source and the sample, without the need to cover the sample with a carbon or gold conducting film previous to observations. The images are recorded in a mixed mode to obtain simultaneous information on the surface morphology of the grains (70% of secondary electrons signal) and on the external Ni-based species, if present (30% of retro-diffused signal). Type of information: The shape of the grains (width, height and thickness) and the external particles are observed by SEM. Any modification of grain morphology upon thermal treatments can be identified, using this technique. On the spent samples, after reaction, carbon deposition (in the form of single or multi-walled carbon nanotubes) can be also viewed by SEM.

READ  Palladium Catalyzed Carbonylation and Related Reactions 

Transmission Electron Microscopy (TEM)

Principle: A beam of electrons goes through an ultra-thin specimen. Depending on the thickness, density and chemical composition of the sample, electrons are absorbed. An image can thus be formed from the interaction of the electrons transmitted with the sample. When high-resolution images are obtained, the reticular distance can be measured and used to identify the nature of the particle, based on the reticular distance found on the ICDD (International Center for Diffraction Data) sheets. TEMs can also be equipped with energy dispersive spectroscopy (EDS) that can reveal the chemical composition of the specimen. If the incident beam has enough energy, it may eject a core-level electron energy (Ea) from an atom. The electron vacancy at the core level is filled by a higher-level electron energy (Eb). As a result of this electron transition, an X-ray of energy (Eb-Ea), characteristic of the element, is emitted. The resulting spectra may contain different peaks at various energies, indicative of the presence of different elements.
Procedure: TEM images are collected using a 2010 JEOL JEM-200 electron microscope operating at 200 keV (LaB6 gun) equipped with an energy dispersive spectroscopy (EDS). Few milligrams of powder are diluted in ethanol and deposited on a copper grid coated with a 10 nm carbon membrane.
Type of information: The morphology of the support (pore widths and distribution) and the dispersion of the supported particles are studied by TEM. Light and dark areas on the images are influenced by the thickness of the observed specimen. When the electron beam is parallel to the main axis of the silica grains, the 2D hexagonal structure of the SBA-15 can be observed (Fig. II-6). When the beam is perpendicular to the main axis of the grains, the mesoporous channels of the SBA-15 are seen as an alternating pattern of light and dark grey. The supported particles appear in dark on the grey silica walls owing to their higher electronic density. Their distribution along the channels of the support or on the external surface can thus be easily distinguished. A statistical counting of at least 500 particles is done to estimate the average particles diameter size by using a software named Comptage de particules, LRS.

Therm ogravimetric analysis coupled to mass spectrometry (TGA-MS)

Principle: Thermogravimetric differential thermal analysis (TGA) measures both heat flow & weight changes in a material either as a function of increasing temperature, or isothermally as a function of time, under a controlled atmosphere. As the temperature increases, various components of the sample are decomposed and the corresponding weight loss is calculated. Changes in weight can be more clearly distinguished in the first derivative TGA curve. The effluent gases from the TGA equipment are transferred to the MS where the components can be identified. The results from thermogravimetric analysis may be presented by mass versus temperature (or time) curve, referred to as the thermogravimetric curve, or rate of mass loss versus temperature curve, referred to as the differential thermogravimetric curve.
Procedure: Measurements are performed on a SDT Q600 (TA Instruments) having a horizontal dual beam balance design, coupled to a Thermostar GDS 301T3 mass spectrometer (Pfeiffer). The samples are heated from room temperature to 900°C at a rate of 10°C.min-1 under air (50 ml.min-1).
Type of information: This technique is mainly used in this work for the study and quantification of carbon deposition. Higher weight loss indicates higher carbon deposition. In all cases, as seen in Fig. II-11, the major weight loss takes place at high temperatures (~600°C) due to the oxidation of carbon: C (s) + O2 (g) → CO2 (g).
The exothermic aspect of the reaction is reflected by the positive increase in heat flow curve. The oxidation is verified by the formation of CO2 through the increase of the MS signal corresponding to the mass 44 (Fig. II-11).

Table of contents :

General Introduction
CHAPTER I : Literature review 
I.1 Methane dry reforming: general context
I.1.1 Methane resources
I.1.2 Pathways for methane-to-syngas conversion
I.1.3 Advantages of dry reforming of methane
I.1.4 Main and side reactions during dry reforming of methane
I.2 Methane dry reforming catalysts: general overview
I.2.1 High performance of noble metals
I.2.2 Use of transition metals and importance of nickel
I.2.3 Existing industrial processes
I.3 Ni-based catalysts for methane dry reforming
I.3.1 Structural approaches to limit nickel sintering and coke resistance
I.3.2 Chemical approaches to enhance carbon resistance
I.4 Aim of this work
I.5 References
CHAPTER II : Experimental part 
II.1 Materials preparation
II.1.1 Supports
a) SBA-15 silica supports syntheses
b) Mesoporous CeO2 support synthesis
c) Commercial supports
II.1.2 Addition of metal salts
a) Two solvents impregnation (2S)
b) Incipient wetness impregnation (IWI)
c) Direct synthesis (DS)
II.2 Characterization techniques
II.2.1 N2 adsorption-desorption isotherms
II.2.2 Scanning electron microscopy (SEM)
II.2.3 Transmission electron microscopy (TEM)
II.2.4 X-ray diffraction (XRD)
II.2.5 Temperature-programmed reduction (TPR)
II.2.6 Thermogravimetric analysis coupled to mass spectrometry (TGA-MS)
II.2.7 Raman spectroscopy
II.2.8 Temperature programmed hydrogenation (TPH)
II.2.9 X-ray photoelectron spectroscopy (XPS)
II.3 Catalytic test
II.3.1 General description of the equipment
II.3.2 Operating procedure
II.3.3 Effluent gas analysis
II.3.4 Expression of results
II.3.5 Validation and reproducibility of the test
II.4 Thermodynamics of the reaction
II.4.1 Effect of dilution
II.4.2 Effect of carbon deposition
II.4.3 Effect of pressure
II.5 References
CHAPTER III : Active and stable Ni/SBA-15 catalysts at 500°C 
III.1 Material preparation
III.2 Physico-chemical properties of supports and calcined samples
III.2.1. Textural properties of the SBA-151 and SBA-152 supports
III.2.2. Porosity of the calcined impregnated samples
III.2.3. Identification and size of supported nanoparticles
III.2.4. Location of the supported nanoparticles
III.2.5. Reducibility of calcined samples
III.3 Catalytic performance in dry reforming of methane
III.3.1 Catalytic activity
III.3.2 Catalytic stability and selectivity
III.4 Physico-chemical properties of spent catalysts
III.5 Conclusion
III.6 References
CHAPTER IV: Influence of textural properties 
IV.1 Comparison of small and larger SBA-15 syntheses
IV.2 Comparison between synthesized SBA-A3 and commercial supports
IV.2.1 Porosity of calcined samples
IV.2.2 Size, dispersion and reducibility of supported nanoparticles
IV.2.3 Catalytic activity and stability
IV.3 Improvement of the synthesized SBA-15 support
IV.3.1 Physico-chemical properties of the samples
IV.3.2 Catalytic activity and stability
IV.3.3 Effect of hydrothermal treatment
IV.4 Conclusion
IV.5 References
CHAPTER V: Influence of preparation procedure 
V.1 Effect of nickel addition method
V.1.1 Catalytic activity and stability
V.1.2 Porosity of calcined samples
V.1.3 Reducibility of the samples
V.1.4 Size and dispersion of the supported nanoparticles
V.2 Effect of nickel precursor and pre-treatment
V.2.1 Catalytic activity and stability
V.2.2 Porosity of calcined samples
V.2.3 Reducibility of the samples
V.2.4 Size and dispersion of the supported nanoparticles
V.3 Effect promoters/dopants
V.3.1 Catalytic activity and stability
V.3.2 Porosity of calcined samples
V.3.3 Reducibility of the samples
V.3.4 Structural ordering of reduced and spent catalysts
V.3.5 Size and dispersion of supported nanoparticles
V.4 Influence of the nature of support – preliminary results
V.4.1 Catalytic activity and stability
V.4.2 Porosity of calcined samples
V.4.3 Reducibility of the samples
V.4.4 Identification and size of the supported nanoparticles
V.5 Conclusion
V.6 References
CHAPTER VI : Carbon analysis and supplementary data 
VI.1 Carbon analysis on spent catalysts
VI.1.1 Effect of stability test duration
VI.1.2 Carbon quantification on spent catalysts
VI.1.3 Carbon structure on spent catalysts
VI.2 Study of more severe reaction conditions
VI.2.1 Effect of gas hourly space velocity (GHSV)
VI.2.2 Effect of pressure
VI.2.3 Effect of regeneration
VI.3 Conclusion
VI.4 References
Conclusion and perspectives 


Related Posts