INTRODUCTION
Hydroformylation
Hydroformylation reaction is the addition of synthesis gas, consisting of CO and H2, to olefin, enabled by a catalyst to form aldehydes The name
Hydroformylation is a chemical process that involves the addition of hydrogen (hydro) to a double bond, resulting in the formation of a formyl group (H-C=O) This reaction primarily produces aldehydes and requires a catalyst, as well as specific conditions of high temperature and pressure for initiation.
Hydroformylation, a significant chemical process, was discovered accidentally by former soldier and scientist Otto Roelen (1897-1993) After World War I, Roelen resumed his coal research at the Kaiser-Wilhelm Institute in Germany, where he focused on the Fischer-Tropsch process, which converts carbon monoxide and hydrogen into liquid hydrocarbons This process produces a variety of short and long-chain hydrocarbons using different metallic catalysts, including cobalt.
The Fischer-Tropsch process utilizes a mixture of hydrogen (H2) and carbon monoxide (CO), known as synthesis gas or syngas, which is produced from the reaction of carbon and water at high temperatures In the early 1920s, the rising production of coking gas sparked significant interest in the use of syngas, driven by advancements in coking plant applications during that era.
Fig 1.2: Disused Zollverein Coal Mine Industrial Coking plant
Initially, Roelen focused on academic research under the guidance of Professor Franz Fischer, the director of the Institute He successfully defended his dissertation on the decomposition of formaldehyde into carbon monoxide and hydrogen in 1924, earning a "good" grade.
In the following year, Roelen continued his research on the Fischer-Tropsch synthesis During his experiment, he tried to cycle the produced ethylene flow
In a groundbreaking discovery, Roelen found propionaldehyde as a solid white product during a reaction involving cobalt, thorium, and magnesium oxide This unexpected finding, along with the observation that other cobalt salts also triggered the reaction, led him to conclude that the aldehyde was not merely a byproduct of the Fischer-Tropsch process Consequently, Roelen conducted further investigations and, in 1938, published a patent detailing his findings.
“oxo” synthesis [4] Roelen’s discovery lays the foundation of organometallic chemistry and its application on an industrial scale
Fig 1.3: Roelen's accidental discovery of hydroformylation
The first pilot unit for hydroformylation was established at IG Farben Leuna/Merseburg, aiming to scale up the reaction In 1940, Ruhrchemie initiated a plan to produce 10,000 tons of fatty alcohols, but it was halted due to the outbreak of the Second World War Despite this setback, Ruhrchemie successfully commercialized several hydroformylation processes in the subsequent years, utilizing a homogeneous cobalt catalyst.
The initial catalyst used for propylene hydroformylation was HCo(CO)4, operating at high temperatures of 140 to 180°C and pressures between 200 to 450 bar Modifications were made with the introduction of cobalt hydridocarbonyl trialkylphosphine HCo(CO)3PR3, which successfully reduced the pressure requirement to 50 bar This modified phosphine cobalt catalyst demonstrated superior performance compared to the earlier version, achieving a higher selectivity for butyraldehyde and an improved isomer ratio of 7:1, compared to the previous 3:1 ratio.
Fig 1.4: First and second generations of hydroformyl catalyst
During Roelen's era, his significant contributions to homogeneous catalysis went largely unrecognized by the scientific community, primarily due to the ongoing war and his failure to publish his research Despite this oversight, his work was regarded as being on par with that of leading chemists of the time.
Fig 1.5: Otto Roelen, in front of the Ruhrchemie’s plant based on his research
In 2022, the annual production of oxo products reached nearly 10 million metric tons, a remarkable increase from just 300,000 tons of butyraldehydes produced in 1980 This significant growth highlights the rapid advancement and expansion of the oxo chemical industry over a relatively short period.
Formed aldehydes are important intermediates in chemical synthesis, enabling the production of various compounds such as alcohols, esters, carboxylic acids, and amines This reaction effectively converts alkenes into aldehydes while extending the carbon chain by introducing a C1 unit.
Since 1993, the production capacity of aldehydes has significantly increased, with global oxo products capacity rising from over 6 million metric tons per year to more than 9.2 million metric tons per year within just five years This growth has been driven by the use of various alkenes, including propylene, butylene, and long-chain alkenes, in the production of aldehydes.
Table 1.1: Nameplate capacity (* 1000 tons) for productions of aldehydes by hydroformylation in 1998
In 1998, butyraldehydes dominated the hydroformylation aldehyde market, accounting for 75% of the total production The primary producers of these aldehydes were the USA, along with countries from East and West Europe.
The global oxo alcohols market, valued at over 17.5 billion USD in 2020, is projected to reach 25.2 billion USD by 2027, with a growth rate exceeding 5% Asia, Europe, and North America dominate the market, collectively representing 95% of global demand in 2020, with Asian countries like China and Japan expected to experience significant growth.
Fig 1.7: Verbund site in Nanjing, China: A joint industrial plant operated by BASF and
SINOPEC, which houses the Butylene oxo synthesis plant
Initially, industrial hydroformylation focused exclusively on the production of propene However, the use of higher alkenes in hydroformylation presents significant challenges, particularly in the separation of catalysts from the final products Recently, Sasol has successfully commercialized a hydroformylation process for long-chain alkenes, extending up to C18.
1.1.3.1 BASF and the First-generation catalyst
The BASF, Celanese, and Union Carbide process has developed a process for propylene hydroformylation, operated on a scale of 4000 kilotons per year
Fig 1.8: Cobalt tetracarbonyl hydride - First generation of hydroformylation catalyst
The first-generation hydroformyl processes, based on Roelen's research, utilize a cobalt-based catalyst under high-pressure conditions of 30 MPa and a temperature of 150°C This method employs cobalt tetracarbonyl hydride, resulting in a yellow liquid that facilitates the hydroformylation of high olefins However, the process presents challenges in completely separating the catalyst from the resulting higher aldehydes.
In addition, the cobalt salt formation from deprotonation of the process also lowered the catalyst activity, as hydroformylation requires the active
Alternatives development for the current catalyst
1.2.1 Implementation of Ionic Liquid as an organic phase
Biphasic catalysis has evolved to address its limitations, with catalysts dissolving in the aqueous phase while substrates remain in the immiscible organic phase The reaction occurs in a flow reactor where both phases are stirred to enhance interaction Although the catalyst and products are ideally contained in separate phases, preventing metal catalyst leaching remains a significant challenge For instance, TPPTS, a widely used P-donor ligand, has shown decreased activity due to leaching into the organic phase Additionally, high interfacial resistances have hindered mass transfer kinetics between the two phases Recent studies have explored innovative solutions, including the use of ionic liquids as organic solvents, monolithic membrane reactors featuring supported ionic liquid phase materials, and supported ionic liquid phase catalysts specifically for hydroformylation.
Ionic liquids are organic salts that remain liquid at low melting points, making them valuable as organic solvents due to their low vapor pressure, selective compound dissolution, consistent water solubility, polarity, and heat conductivity Their efficiency in product separation and high yields enhances their appeal in catalysis However, the higher costs and limited industrial availability, along with their low-temperature operational range, present significant challenges for widespread use.
Fig 1.14: Ionic liquid cations and anions
Regardless, in hydroformylation, selectivity is always favorable more than conversion, which is the critical feature in hydroformyl homogenous catalyst
To preserve the properties of homogeneous catalysts, a supported ionic liquid phase (SILP) catalyst has been developed This innovative system features the active catalyst dissolved in an ionic liquid, which is uniformly spread as a thin film on a porous support's surface This approach effectively addresses the challenges of product separation and catalyst recovery while enhancing mass transfer within the porous structure.
Fig 1.15: Supported ionic liquid phase catalyst [22]
Choosing the right ionic liquid is crucial for advancing green chemistry in hydroformylation, a process celebrated for its atomic economy and environmental benefits While the ecotoxicity of ionic liquids has been a topic of debate, particularly concerning halogenated variants, our research highlights [BMIM][octylsulfate] as an ideal choice This ionic liquid is cost-effective, non-toxic, and provides optimal conditions for hydroformylation, aligning with sustainable practices in the petrochemical industry.
Fig 1.16: [BMIM][octylsulfate] Ionic liquid
Recent investigations have focused on the use of various supports for heterogeneous catalysts in hydroformylation, particularly silica-modified materials These studies highlight the advantages of using high surface area and stable supports like SBA-15, which exhibit favorable physicochemical properties and catalytic activity for hydroformylation of long-chain alkenes The combination of a large surface area with precise control over pore connections and mesopore size effectively overcomes mass transfer limitations, a key factor in enhancing catalytic activity Additionally, a well-ordered distribution of pore sizes significantly improves gas adsorption and desorption, further boosting transport capabilities.
In 1992, the scientific community became intrigued by the first report of an ordered mesoporous silica material, leading to extensive research across various fields Carbon-based materials, known for their porosity and thermal stability, offer unique advantages in catalysis applications Consequently, ordered mesoporous carbon (OMC) has been developed and utilized in hydroformylation processes The controllable pore size and hydrophobic properties of OMC enhance its interaction with ionic liquids within its porous structure.
Ordered mesoporous carbons (OMCs) can be synthesized through hard-template methods, where a selected porous material is filled with a carbon precursor This process involves carbonization in an inert atmosphere, followed by the removal of the template, leading to the formation of OMCs The characteristics of the pores in the OMCs are primarily influenced by the properties of the initial template material used.
Fig 1.17: Two methods to synthesize ordered mesoporous carbon
This thesis investigates the synthesis of various ordered mesoporous carbons (OMCs) using different templates, including SiO2, SBA-15, and γ-Al2O3 Each type of OMC exhibits unique characteristics, which will be explored in relation to their applications in hydroformylation, highlighting the significance of their porous structures in enhancing the hydroformylation process.
1.2.3 Nano gold catalyst in Hydroformylation
Hydroformylation remains a homogeneous catalyst process, offering advantages such as product selectivity and availability, which facilitate industrial-scale applications However, the primary challenge lies in the separation of the product from the catalyst mixture Developing heterogeneous catalysts presents a viable solution to this issue, prompting extensive research to innovate new oxo technologies Various catalysts, including polymeric heterogeneous catalysts and rhodium metal supported on different materials, have been explored to enhance the efficiency of this process.
Gold nanoparticles (NPs) are among the first nanomaterials synthesized and utilized in various catalytic applications Previous studies have demonstrated that nanogold catalysts exhibit a preference for hydrogenation over hydroformylation in hydroformylation reactions Additionally, propanol, a product formed during these reactions, indicates the hydrogenation of aldehydes derived from hydroformylation.
Fig 1.18: hydroformylation and Hydrogenation under the influence of Gold catalyst
Gold supported on ordered mesoporous carbons (OMCs) offers the benefits of a heterogeneous catalyst while exhibiting minimal interaction between the metal and support This limited interaction may explain the lower activity observed in previous studies Consequently, the hydroformylation activity of gold on OMCs is being compared to that of the supported ionic liquid phase (SILP) catalyst.
The goal of this thesis
This thesis aims to utilize a supported ionic liquid phase and a heterogeneous gold catalyst for hydroformylation, employing high-surface-area, ordered mesoporous carbon to prevent catalyst leaching and mass transfer issues The mesoporous carbon is synthesized via the hard-template method, using templates such as SiO2, synthesized SBA-15, and γ-Al2O3 Previous research indicates that small pore size materials like MCM-41 complicate the filling of ionic liquids, prompting the selection of mesoporous templates with larger pore sizes Additionally, the study investigates the impact of varying amounts of carbon precursor on the behavior of these supports, ensuring the appropriate support is utilized for the process.
The performance of supported ionic-liquid phase catalysts and gold catalysts is evaluated based on C2H4 conversion rates, product selectivity, and turnover frequency (TOF) Additionally, product trends are analyzed through multiple test runs utilizing a self-system Gas Chromatography with Flame Ionization Detector (GC-FID).
Catalysts exhibit distinct characteristics that can be analyzed using various techniques, including Fourier-Transform Infrared (FT-IR) spectroscopy, N2 adsorption-desorption isotherms, Electron Paramagnetic Resonance (EPR), Scanning Electron Microscopy with Energy Dispersive X-ray Analysis (SEM/EDX), and Ultraviolet-Visible (UV-Vis) spectroscopy.
EXPERIMENT
Ordered mesoporous carbon synthesis
Tetraethylorthosilicate (TEOS) and Pluronic P-123 were sourced from Merck Chemical Company, while hydrochloric acid (36%) was obtained from Duc Giang Detergent Company Distilled water was utilized in all experiments, and a 2M hydrochloric acid solution was prepared by diluting the concentrated acid All other chemicals were used as received without further modification.
SBA-15 was synthesized using TEOS as a precursor through a series of steps Initially, 3 g of P123 was dispersed in 45 ml of distilled water at 40°C, followed by the addition of 90 ml of 2M hydrochloric acid while stirring for 2 hours Gradually, 6.375 g of TEOS was incorporated into the mixture, which was then stirred for an additional 24 hours The resulting solution was subjected to autoclaving at 80°C for 8 hours After filtration and multiple washings to achieve a neutral pH of 7, the collected powder was calcined at 550°C for 10 hours, with a heating rate of 1°C/min.
Sucrose (C12H22O11) was bought from Shanghai Zhanyun chemical company, in addition to ethanol C2H5OH (99.7%) and sulphuric acid H2SO4
In this study, SiO2 and γ-Al2O3 (0.063 – 0.200 mm) were sourced from Merck Company, while SBA-15 was synthesized in an earlier section Hydrofluoric acid (HF) at 40% concentration was obtained from RoHS-China and subsequently diluted to a 10% solution All experiments utilized distilled water, with some employing ethanol as a washing agent.
OMCs were synthesized using the hard-template method, beginning with the addition of distilled water, H2SO4, and sucrose into a test tube, which was then mixed thoroughly A 2g template sample (SiO2, SBA-15, or γ-Al2O3) was placed in a 100ml ceramic crucible, and the sucrose solution was added dropwise while grinding the mixture until it reached a toothpaste-like consistency The mixture was caramelized at 100°C for 6 hours, followed by 160°C for 2 hours, with the precursor filling and caramelization repeated until the acidic sucrose solution was exhausted The sample was then transferred to a ceramic crucible boat and subjected to calcination at 400°C for 4 hours and carbonization at 850°C for 2 hours under nitrogen flow, with a heating rate of 5°C/min The resulting black powder underwent treatment with a 1:1 mixture of HF and Ethanol, stirred for 2 hours at 30°C, filtered, and washed with Ethanol, with the etching process repeated once more Finally, the product was dried for 2 hours at 120°C.
Fig 2.1: Ordered mesoporous carbon hard-template method
The table below demonstrates the detail of synthesized OMCs, including sucrose information and used template:
Supported Ionic Liquid Phase Catalyst
Triphenylphosphine TPP (99%), fuming sulphuric acid H2SO4 (SO3)x
(68.0% free SO3, toluene C6H5CH3 were all from Sigma-Aldrich Chemical Cesium hydroxide monohydrate CsOH.H2O (99.5%) and tri-n-octylamine
(N(C8H17)3) were from Acros Organic, where CsOH was dissolved in water and diluted to a 5% wt CsOH solution Methanol was bought from Xilong Scientific
Fig 2.2: Ligand synthesis using Schlenk line
The ligand synthesis procedure was operated in a Schlenk line in Nitrogen atmosphere Diagram Fig 2.2 illustrates the whole procedure
5.2 g Triphenylphosphine was placed in a 1L three-neck round bottom flask with a magnetic stirrer Next, the flask was cooled to 15°C and cycled through a vacuum pump and nitrogen simultaneously using the Schlenk line, avoiding contaminants or water traces in the system Over the next hour, 48g oleum was added slowly under stirring After the addition of oleum and TPP had finished, the mixture was stirred for 150h at 20°C
In the next step, 150g of water was added to dilute the mixture while maintaining a temperature of 10°C After stirring for one hour, 23.9 ml of trioctyl amine and 90 ml of toluene were incorporated, followed by an additional 30 minutes of stirring The mixture was then allowed to separate for 30 minutes, after which the lower phase containing acid and impurities was discarded.
Fig 2.3: The 3-steps ligand separation
A 5% cesium hydroxide monohydrate solution was gradually added to the mixture while stirring until the pH reached 5.5, after which the lower phase containing TPPDS and TPPMS salts was discarded The addition of cesium continued until the pH of the mixture increased to 6.5 The lower phase, which comprised TPPTS sodium salts, was then concentrated using a 100ml mixture of methanol and water in a 10:1 ratio Finally, the mixture was completely dried with a vacuum pump at 80°C, resulting in a white-yellowish crystalline powder of TPPTS-Cs3.
TPPTS-Cs3 ligand was synthesized using the Schlenk line according to the suggestion method in our group previous research [36] and [38], which was also similar to the US patent 4483802 [39]
In section 2.2.1, the TPPTS-Cs3 ligand was synthesized, while OMC supports were prepared in section 2.1.2 The ionic liquid 1-n-Butyl-3-methylimidazolium octysulfate [BMIM][n-C8H17OSO3] was sourced from Merck Chemical, and methanol (99.7%) was obtained from Xilong Scientific Additionally, dicarbonyl (Acetylacetonato) rhodium Rh(acac)(CO)2 was purchased from Sigma Aldrich.
Table 2.2: Supported Ionic liquid phase catalysts component and abbreviations
Support Rh(acac)(CO)2 [BMIM][n- octylsulfate]
Cs3 mass(g)/ L:Rh mass ratio msupport
Finally, ionic liquid, ligand, and dicarbonyl (acetylacetonato) rhodium and
In a double-neck 250ml flask, 60 ml of methanol was added while stirring The flask underwent five cycles through the Schlenk line before and after the mixture was added Following two hours of stirring, the OMC support was introduced into the mixture.
19 using a glass valve The mixture was left to stir for another 2 hours before drying at 80°C to collect the SILP catalysts
All components were precisely weighed using weight paper sheet, and their respective amounts were listed in Table 2.2
Gold catalyst on Ordered mesoporous carbon
Gold catalysts were synthesized using the impregnation method Aurochloric acid HAuCl4 was purchased from Merck company, which was subsequently diluted into a solution concentration of 0.01M Ammoniac NH3
(25%) was also bought from Sigma-Aldrich The weight percent of gold in the catalyst is 1 wt.%
Firstly, 1.5 ml HAuCl4 0.01M was added to a 100ml beaker, in addition to
To prepare the solution, 36 ml of distilled water was stirred for 30 minutes before adding 0.3 g of OMC support, which was stirred overnight After cooling, a 25% NH3 solution was added until the pH reached 10, followed by an additional 5 hours of stirring The mixture was then filtered and thoroughly washed to eliminate Cl- The resulting powder was dried overnight at 110°C and subsequently calcinated under nitrogen flow at 200°C for 2 hours, with a heating rate of 2°C/min.
1 1%wt Au/OMC-γ-Al2O3-1g OMC-γ-Al2O3-1g Cat-Au-2
2 1%wt Au/OMC-SiO2-3g OMC-SiO2-3g Cat-Au-3
Catalyst characteristics
Fourier-Transform Infrared Spectroscopy (FTIR) is a crucial analytical technique employed in both organic and inorganic chemistry This method utilizes electromagnetic radiation within the infrared spectrum, which is categorized into three distinct ranges: Near-Infrared (0.78 to 2.5 µm), Mid-Infrared (2.5 to 50 µm), and Far-Infrared (50 to 1000 µm).
The method operates on a principle akin to absorption spectroscopy, where molecules or functional groups absorb infrared (IR) radiation Unlike ultraviolet or visible light, IR radiation lacks sufficient energy to induce electron transitions Instead, absorption occurs in non-polar molecules, such as O2, N2, and Cl2 The frequency of molecular vibrations is determined by the distance between atoms, and absorption takes place when the IR frequency aligns with the molecular vibration frequency.
Molecular vibrations can be categorized into two types: bending and stretching, which depend on the angle of the bonds The wavenumber of stretching vibrations is determined by the mass of the two atoms involved and the bond stiffness constant, denoted as k.
𝜈 = 5.3 ∗ 10 −12 √𝑘 àWhere ν is the wavenumber of an absorption (cm -1 ), k is the force constant for the bond (N/m), and à is the reduced mass (kg)
Fig 2.5: Stretching and bending vibrations formation [41]
The bending vibrations are the angle-changing between two bonds and consist of four types: scissoring, rocking, wagging, and twisting
Fourier transform infrared (FT-IR) spectra were acquired in transmission mode from KBr pellets at room temperature using a Nicolet iS50 FT-IR spectrometer, achieving a resolution of 4 cm⁻¹ Each sample was characterized through 32 scans per spectrum within the 400 – 4000 cm⁻¹ range, with the measurements conducted at GeVicat laboratory.
11 th floor, D8, Hanoi University of Science and Technology
Fig 2.6: Nicolet iS50 FT-IR
2.4.2 Nitrogen adsorption-de adsorption Isotherm method
The surface area and pore characteristics of supports and catalysts are determined using the Nitrogen adsorption-desorption isotherm method, which is based on the BET (Brunauer–Emmett–Teller) theory of adsorption, distinguishing it from Langmuir's theory.
BET theory expands on Langmuir theory by proposing that gas molecules can adsorb onto a solid in multiple layers without interaction between these layers The quantity of gas adsorbed is influenced by temperature, gas pressure, and the strength of the interaction between the gas and the solid surface Nitrogen is frequently utilized in BET surface area analyses due to its high purity, availability, and robust interaction with solid materials.
The surface area is determined through the BET equation:
The relative pressure (P/P0) and the volume of gas adsorbed at standard temperature and pressure (Xm) are used to calculate the surface area of the solid sample Prior to measurement, all samples underwent a degassing process to eliminate water and contaminants, which involved heating at 80°C for 15 minutes, followed by 150°C and 250°C for 15 minutes each, and finally degassing at 300°C for 2 hours In contrast, the SILP catalysts were only degassed at 80°C for 2 hours to prevent the evaporation of the ionic liquid at higher temperatures.
Fig 2.7: Pre-treatment degassed system for sample
The sample was analyzed using the BET method with 10 data points to assess its surface area and 60 data points to examine pore characteristics Measurements were conducted with the Gemini VII from Micromeritics, USA, at the GeVicat laboratory, Hanoi University of Science and Technology, utilizing Dewars of liquid nitrogen to ensure constant pressure during the analysis.
Fig 2.8: Gemini VII, Micromeritics Surface Area and Porosity
Electron paramagnetic resonance (EPR) spectroscopy uses micro waves to investigate the structure and interactions of paramagnetic centers, from transition-metal ions on the surface to radicals
EPR is a selective technique that reveals details about oxidation states, and when combined with spectroscopic methods like NMR and FT-IR spectroscopy, it enhances the understanding of interactions between electron spins and protons.
The principle of Electron Paramagnetic Resonance (EPR) is similar to nuclear magnetic resonance; however, detecting electron spins in biological materials is more challenging due to the greater magnetic moment of electron spins.
The sample was measured using Bruker EMX-Micro EPR spectrometer at GeVicat laboratory, Hanoi University of Science and Technology
Fig 2.9: Bruker EMX-Micro EPR spectrometer
2.4.4 Scanning electron microscopy/energy-dispersive X-ray spectroscopy
Scanning electron microscopy (SEM) is a powerful imaging technique that utilizes a focused electron beam to generate high-resolution images of samples When the electron beam interacts with the sample, ionized atoms release X-rays or eject electrons These emissions are captured by a secondary electron detector, while an Energy Dispersive X-ray (EDX) detector identifies individual elements based on their unique X-ray emissions.
To visualize the distribution of the element in the specimen, a two- dimensional map is generated by acquiring pixel by pixel This method is called elemental mapping
This thesis utilizes the specimen's net elemental map, derived from an intensity map free of background noise and overlapping peaks The samples were analyzed using the JCM-7000 NeoScope T M Benchtop SEM at the GeVicat laboratory, located within Hanoi University of Science and Technology.
Fig 2.10: JCM-7000 NeoScope TM Benchtop SEM
Ultraviolet/visible spectroscopy employs an energy source that emits wavelengths from 190 to 800 nm, encompassing both ultraviolet and visible light When a compound absorbs this light energy, its electrons are excited to higher energy levels This absorption of energy enables both quantitative and qualitative analysis of the compound.
An ultraviolet/visible spectroscopy mainly consists of a light source, a filter or monochromator (conditional), a detector, and an amplifier to give a result
Fig 2.11: Principle of UV-Vis spectroscopy
The UV-Vis spectroscopy measurement was conducted with an Avaspec 2048L at GeVicat laboratory, Hanoi University of Science and Technology
Fig 2.12: Avaspec 2048L, UV-Vis spectroscopy
The testing-activity system utilized three separate gas cylinders containing H2, CO, and C2H4 in a 1:1:1 ratio, with argon used to flush the system prior to operation The inlet flow was set to 60 ml/min, and the reaction was conducted at a pressure of 7 bar A 60 cm stainless steel tube housed the catalyst, which was packed with 0.15 g of OMCs catalyst and fitted with silica wool at both ends within the oven.
Fig 2.13: Hydroformylation Catalyst’s activity testing system
The system features a bypass line that measures bypass gas prior to testing With a maximum oven temperature of 350°C, the outlet flow is subsequently heated by a heater to prevent condensation along the line before it enters the gas chromatograph (GC) through a sample loop.
Table 2.4: Activity testing system properties
RESULTS AND DICUSSIONS
Synthesis and characterization of ordered mesoporous carbon
3.1.1 Ordered mesoporous carbon on SiO 2 template
Fourier-Transform Infrared Spectroscopy is utilized to analyze the differences among various OMC templates, assessing their coherence with the optimal sucrose concentration for each template and identifying any potential material deflections.
Fig 3.1 shows the FT-IR spectra of OMC-SiO2 and sucrose amount influence on the OMC
Fig 3.1: FT-IR spectra of OMC-SiO 2
The analysis of Fig 3.1 indicates that all ordered mesoporous carbons (OMCs) exhibit a characteristic C-H stretching band at 2910 cm -1, originating from the precursor sucrose, consistent with other sucrose-derived mesoporous carbon materials Additionally, the peaks observed at 1578 cm -1 and 1459 cm -1 correspond to C=O and C-H groups associated with the aromatic ring, respectively This suggests that the synthesized ordered mesoporous carbon features ring functional groups on its surface, as evidenced by the presence of C-H and C- functionalities.
The FT-IR analysis reveals significant spectral features related to the alkoxy group C-O, observed at 1157 cm -1, which is influenced by the asymmetric stretching vibrations of Si-O-Si This indicates that OMCs based on SiO2 contain an acid carboxylic functional group, as evidenced by the presence of C=O bonds In the FT-IR spectra of OMC-SiO2-3g and OMC-SiO2-1.5g, less pronounced peaks at 871 cm -1 correspond to the Si-O bond, while the Si-O bond at 800 cm -1 is primarily derived from the SiO2 template Additionally, the broad band at 1300 cm -1 may be attributed to the OMC-SiO2 material.
The formation of the C-Si bond within the matrix is evident, as indicated by the FT-IR spectra of all OMC/SiO2 materials, which exhibit similar structural characteristics The variation in the amount of sucrose solution does not significantly affect the band intensity, except for the Si-O bands associated with residual SiO2 templates Notably, OMC-SiO2-2g and OMC-SiO2-1g demonstrate slightly superior properties due to their reduced amounts of leftover SiO2 template.
To further select suitable OMC-SiO2 support for the catalyst synthesis, the
N2 adsorption-desorption method is used to measure porosity and surface area, as shown in Table 3.1
Table 3.1: Surface area and porosity of OMC-SiO 2
The average BET surface area for all OMC/SiO2 samples is 1186 m²/g, with a standard deviation of 51.37 m²/g, indicating consistency among the samples In contrast, the average pore volume and average pore size exhibit notable differences across the three samples: OMC-SiO2-.
The average pore volume of OMC-SiO2 samples varies, with values around 3 cm³/g and an average pore size of 100 Å, while OMC-SiO2-3g exhibits a reduced pore volume of 1.777 cm³/g and a smaller average pore size of 70 Å This reduction is attributed to the presence of larger pores in other OMC/SiO2 samples, indicating a broad pore distribution Notably, all OMC/SiO2 samples, except OMC-SiO2-3g, display a secondary pore size, suggesting insufficient sucrose to fully saturate the SiO2 structure, resulting in partially filled template pores and slightly larger pore sizes OMC-SiO2-3g demonstrates a well-ordered structure with most pores measuring 70 Å, while OMC-SiO2-1, OMC-SiO2-1.5, and OMC-SiO2-2g show a gradual increase in the proportion of 70 Å pore widths corresponding to higher sucrose amounts Additionally, a secondary pore at 100 Å is present in all samples, likely corresponding to the SiO2 template.
Consequently, OMC-SiO2-3g is chosen as the support to represent the OMC/SiO2 support, owing to high pore size accuracy, stable functional groups, and no secondary SiO2 template pore
Fig 3.2: Pore distribution of OMC-SiO 2
3.1.2 Ordered mesoporous carbon on SBA-15 template
Next, FT-IR spectra of OMC-SBA-15 were also conducted to investigate the influence of sucrose solution during the precursor filling synthesis step
Fig 3.3: FT-IR spectra of OMC-SBA-15
Figure 3.3 illustrates the impact of sucrose quantity, revealing that both OMC-SiO2 and OMC-SBA-15 exhibit similar spectral bands, albeit with slight variations in their wavenumbers The observed vibrations include C-H stretching, C=O, and C-H in the aromatic region.
The analysis of ring structures at 2914, 1580, and 1573 cm -1 reveals that increasing sucrose amounts significantly enhance band intensity; however, this increase is not directly proportional Notably, OMC-SBA-15-2g and OMC-SBA-15-4g exhibited higher band intensities compared to OMC-SBA-15-3g and OMC-SBA-15-5g, the latter of which lacks acid carboxylic functional groups This absence may be attributed to the sucrose content, as OMC derived from sucrose typically contains a carboxyl functional group Consequently, OMC-SBA-15-5g shows lower Si-O-Si broadening intensity and no detectable Si-O bonds from the SBA-15 template, indicating a complete etching process.
Table 3.2 depicts the surface area and pore characteristics of OMC-SBA-
15 with different carbon content and compared with SBA-15 template In addition, pore distribution is also discussed in Fig 3.4
Table 3.2: Pore characteristics of OMC-SBA-15 and SBA-15 template
All OMC-SBA-15 samples exhibit smaller average pore sizes compared to the SBA-15 template used for synthesis, primarily due to the replication of the SBA-15 pore walls by sucrose during precursor filling Notably, OMC-SBA-15-2;3;4g shows a smaller pore size, while OMC-SBA-15 demonstrates minimal framework shrinkage, as noted by Yunpu Zhai In contrast to OMC-SiO2, OMC-SBA-15 maintains a relatively consistent average pore volume and size, highlighting the importance of selecting an "ordered" mesoporous carbon as a suitable support Additionally, OMC-SBA-15-5g features the largest pore size at 44Å, with no pores detected at 56Å, and a slight presence of 55Å diameter pores.
The 15 pore size of OMC-SBA-15 indicates that the sucrose solution was insufficient to fully saturate the SBA-15 template, leaving residual SBA-15 in the OMC supports This observation is corroborated by the FT-IR analysis, which shows minimal Si-O bands in the spectra of OMC-SBA-15-5g.
Fig 3.4: Pore distribution of OMC-SBA-15
The analysis of pore distribution indicates that OMC-SBA-15-5g is the preferred support due to its precise average pore size In contrast, OMC-SBA-15-4g boasts a higher surface area than OMC-SBA-15-5g, although its pore distribution is slightly less accurate Other variants of OMC-SBA-15 exhibited a lower pore count due to insufficient sucrose solution.
3.1.3 Ordered mesoporous carbon on γ-Al 2 O 3 template
Fig 3.5 depicts the FT-IR spectra of OMC-γ-Al2O3 with different volumes of sucrose acidic solution, resulting in different carbon content on the template
The precursor filling process revealed that γ-Al2O3 required more time to fill due to its smaller surface area and pore volume compared to SBA-15 and SiO2 Consequently, a relatively low volume of sucrose acidic solution, only 1 or 2 mL, was used during preparation Notably, OMC-γ-Al2O3-4g exhibited significantly higher absorbance than other OMC-γ-Al2O3 supports, prompting the inclusion of an additional figure (Fig 3.7) to illustrate the band intensity in samples with lower absorbance.
Fig 3.5: FT-IR spectra of OMC-γ-Al 2 O 3
In Fig 3.5, the OMC-γ-Al2O3-4g exhibits unusual behavior compared to other samples, characterized by a strong, broad band at 3300 cm -1 indicating O-H groups in alcohol, and a sharp peak at 633 cm -1 representing the Al-O bond Additionally, during the etching process of OMC-γ-Al2O3, white crystalline particles began to visibly accumulate upon the introduction of ethanol, suggesting that these particles are Al2O3 precipitated as a result of the reaction with ethanol following the etching.
Fig 3.6: Al-O formation in OMC-γ-Al 2 O 3 -4g
During the OMC washing process, aluminum oxide serves as a template that reacts with HF to generate Al3+ ions However, due to its hydrophilic nature, HF struggles to penetrate the structure of OMCs, as the walls exhibit chemical aversion to HF The inclusion of ethanol in this washing step is crucial, as it facilitates the wetting of the template within the OMC walls Omitting ethanol may lead to the formation of Al2O3 instead.
The addition of ethanol to Al2O3 initiates a continuous two-way reaction, where Al2O3 serves as a catalyst to dehydrate the alcohol and subsequently regenerates itself Notably, γ-Al2O3 is the only aluminum oxide catalyst effective for this process To halt the reaction, one must either raise the temperature to 240°C to produce ethylene or refrain from using ethanol as a washing agent altogether The proposed formation of Al-O is illustrated in Fig 3.6.
Fig 3.7: FT-IR spectra of OMC-γ-Al 2 O 3 , excluding OMC-γ-Al 2 O 3 -4g
In Fig 3.7, asymmetric C-H and alkoxyl C-O’s bands at 2914 cm -1 and
1147 cm -1 , respectively, are much higher than the C-H aromatic band at 1573 cm -
1 The Al-O bond can also be observed, though it is not as high as that of OMC-γ-
Supported Ionic liquid phase catalysts
In the first part, ordered mesoporous carbon with different carbon content was synthesized using the hard-template method, with SiO2, SBA-15, and γ-
Using FT-IR, N2 adsorption methods, and SEM images, suitable ordered mesoporous carbon materials were identified based on pore distribution, surface area, and functional groups Among the SiO2 variants, OMC-SiO2-3g exhibited the best pore distribution, with OMC-SBA-15-4g and OMC-SBA-15-5g also selected due to their properties and the presence of the functional COOH group in OMC-SBA-15 Conversely, the unexpected side reactions of ethanol led to structural collapse in many OMC-γ-Al2O3 samples; however, OMC-γ-Al2O3 was still utilized to synthesize the SILP catalyst, highlighting the unique behavior of this support during formation, which has not been previously reported Additionally, Al2O3 has been recognized as a viable support for hydroformylation processes.
3.2.1 Investigating ionic liquid impregnation in SILP catalyst
The FT-IR spectra of different synthesized SILP catalysts, as illustrated in Fig 3.9, reveal the characteristic bands of the SILP components, including the ionic liquid, ligand, and rhodium active sites.
Fig 3.9: FT-IR spectra of Cat-SILP samples
The SILP catalysts exhibit a distinct triple signature shoulder peak corresponding to S=O stretching vibrations from the sulfonate functional group at 1166, 1056, and 1036 cm-1, although the intensity of these peaks varies These SO3 groups originate from either the ligand TPPTS or the anion of the ionic liquid, octylsulfate Additionally, a medium band at 1632 cm-1 indicates that all SILP catalysts retain their ligand, as only TPPTS contains the P-C bond Furthermore, peaks observed at a wavenumber of 1459 cm-1 further characterize the catalysts.
The C=N stretching vibrations from the organic cation of the ionic liquid [BMIM] are responsible for the peak at 37 cm⁻¹ Additionally, the medium band observed at 477.21 cm⁻¹ is linked to both Rh-O and Rh-P interactions, which arise from residual rhodium precursor and the bonding of rhodium with the ligand in the Rh/TPPTS system It is noteworthy that the intensity of these characteristic bands is affected by the larger amount of KBr pellet during Fourier Transform spectroscopy.
IR measurement However, all SILP components are observed after the catalyst synthesis
The FT-IR KBr pellet method is employed to assess SILP catalysts with varying ionic liquid loading, aiming to determine the potential for increased ionic liquid content within their porous structure.
The SILP catalyst supported on OMC-SBA-15-4g, referred to as Cat-SILP-2, is examined to assess the impact of ionic liquid (IL) loading The primary distinction between the two catalysts lies in their ionic liquid content, with Cat-SILP-1 featuring a 10% IL loading.
In Fig 3.10, the SILP catalyst with increased ionic liquid content (Cat-SILP-2) exhibits significantly higher vibration intensity compared to Cat-SILP-1, particularly with the silanol group O-H at 3400 cm -1, suggesting that the ionic liquid interacts with the support wall through hydrogen bonds As the ionic liquid saturates the porous structure, stronger bonding occurs, while the C-O bond from the OMC-SBA15-4g support diminishes at higher ionic liquid concentrations, indicating that the ionic liquid may cover the entire surface of the support This phenomenon suggests that the functional groups of OMCs could potentially influence product selectivity during hydroformylation.
Fig 3.10: FT-IR spectra of different ionic liquid loading in SILP catalysts
The presence of the ligand signature band P-C at 1632 cm -1 and the rhodium band indicates that both the ligand and active sites are preserved within the ionic liquid layer, regardless of the variations in ionic liquid loading This observation highlights the stability of the components as confirmed by FT-IR analysis.
IR method can also be used to do semi-quantitative analysis on ionic liquid loading, as the Cat-SILP-2 C=N stretching vibrations apparently appear at 1458 cm -1
3.2.2 Surface area and pore distribution of SILP catalysts
3.2.2.1 The influence of degas temperature on SILP catalysts
First, the influence of temperature on the pore structure and surface area is investigated owing to the limited temperature range of ionic liquid
Table 3.4: Surface area and pore characteristics of Cat-SILP-2 at different degas temperature
The Cat-SILP-2 sample underwent degassing at 300°C to remove the ionic liquid from its porous structure, whereas standard Cat-SILP samples were degassed at a lower temperature of 80°C for two hours to prevent the decomposition of the ionic liquid.
The data in Table 3.4 indicates a significant reduction in surface area and average pore volume following the formation of an ionic liquid thin film within the porous structure The lower average pore volume of Cat-SILP-2 confirms the effective filling of ionic liquid through a hierarchical pore-filling mechanism, which progresses from micropores to mesopores and ultimately to macropores Additionally, the emergence of new macropores at 500 Å can be attributed to the ionic liquid spreading to the surface of OMC-SBA-15-4g, resulting in the creation of new macropores.
Heat treatment at 300°C effectively decomposes the ionic liquid on the surface, restoring the average pore size to that of OMC-SBA15-4g Additionally, the evaporation of the ionic liquid enhances the pore volume of Cat-SILP-2-300, resulting in an increased overall BET surface area.
Fig 3.11: Pore distribution of Cat-SILP-2 at different temperatures
At a 30% loading, the ionic liquid volume within the porous structure is approximately 1 cm³/g, which results in the pore distribution shown in Fig 1.1 being overshadowed by the substantial volume of the OMC supports.
3.2.2.2 SILP catalyst on OMC-SBA-15 supports
Table 3.5: Surface area and pore characteristics of Cat-SILP-2 and Cat-SILP-3
In contrast to Cat-SILP-2, the ionic liquid did not exhibit overfilling in the Cat-SILP-3 structure This difference is attributed to the interaction between the acid surface functional groups in Cat-SILP-2 and the ionic liquid, which competes with the ionic liquid's interaction with the walls through hydrogen bonding, leading to overfilling Additionally, similar to other carbon materials, the presence of electron-withdrawing groups resulted in lower ionic liquid uptake by OMC, contributing to the OMC-SBA-15-4g's aversion to the ionic liquid Conversely, Cat-SILP-3, which lacks acid functional groups, demonstrated a higher surface area and slightly larger average pore size than the original support, further supporting this explanation.
Fig 3.12: Pore distribution of Cat-SILP-2 and Cat-SILP-3
Figure 3.12 illustrates the pore distribution for Cat-SILP-2 and Cat-SILP-3 The pore mechanism in the Cat-SILP-2 support at various temperatures indicates that the ionic liquid film has a volume of 0.9 cm³/g and a pore size of 30 Å at 30% ionic liquid loading Additionally, Cat-SILP-2 exhibits a limited number of macropores originating from the surface ionic liquid.
3 slightly shift the average pore size higher, though it is still minimal
3.2.2.3 SILP catalysts on SiO 2 and γ-Al 2 O 3 support
To see if other SILP catalysts' ionic liquid filling behavior corresponds to the SILP on OMC-SBA-15, surface area and pore distribution are measured in Table 3.6
Table 3.6: SILP catalysts on OMC- γ-Al 2 O 3 and OMC-SiO 2 ‘s surface area and pore characteristics
At first, the ionic liquid behavior was thought to be dependent solely on the surface area, so the lowest synthesized OMC surface area, which is OMC-γ-
Al2O3-4g, was used to synthesize a testing SILP catalyst as a reference OMC-γ-
Al2O3-4g structure collapsed owing to the appearance of by-products during
The synthesis process resulted in a significant average pore size, with Cat-SILP-4 displaying a concentration of pores at 700Å and a larger pore at 1500Å due to the presence of ionic liquid This led to a noticeable reduction in surface area and pore volume, consistent with ionic liquid behavior, although some areas of the porous structure remained accessible, as indicated by the FT-IR spectrum Overfilling of the ionic liquid caused higher absorbance in the FT-IR spectra In contrast, Cat-SILP-5 exhibited the highest surface area among SILP catalysts, attributed to the superior OMC surface area used The ionic liquid film is estimated at 1.2 cm³/g, with varying pore sizes observed in Cat-SILP-5's distribution The overall filling of ionic liquid resulted in a significant reduction of pore size to around 70Å, indicating substantial ionic liquid infiltration, while pore sizes ranging from 200 to 300Å were linked to the ionic liquid's presence As the ionic liquid permeated the structure, impurities and water were expelled, leaving open pores for gas adsorption, akin to other ordered mesoporous materials.
Fig 3.13: Pore size distribution of OMC-SiO 2 -3g and OMC-γ-Al 2 O 3 -4g and SILP catalysts
3.2.3 SILP catalysts activity at different temperatures
SILP catalysts activity is divided into three main categories: Ethylene conversion, products selectivity, and products TOF While conversion and
42 selectivity show the catalyst performance several times, TOF measures the instantaneous efficiency of a catalyst towards a specific product
All categories are evaluated based on three groups: the optimal temperature for the hydroformylation using Cat-SILP-2, investigating from 80 to 140°C; all SILP catalysts hydroformylation’s activity at 120°C and 140°C, respectively
Three main products: propanal from hydroformylation and propan-1-ol and propan-2-ol as the hydrogenation derivatives from propanal are the focus of the Ethylene conversion in this thesis
In addition to these products, 2-methyl-2-pentanol and 2-methyl-1- pentanol are also by-products
The ethylene conversion of Cat-SILP-2 in hydroformylation at different temperatures is shown in Fig 3.14
Fig 3.14: Ethylene Conversion with time-on-stream over 0.15g powder Cat-SILP-2, gas flow rate 60mL/min, at 7 bar
Nanogold catalysts on ordered mesoporous carbon
3.3.1 Gold catalysts electron paramagnetic resonance’s spectrum
As a comparison, the EPR spectrum of Cat-Au-1, which was Au catalysts impregnation on OMC-SBA-15, was used to compare the difference between gold catalysts samples
Fig 3.26: Gold catalysts EPR's spectrum
Figure 3.26 shows a pronounced peak signal detected in all gold catalysts both before and after the reaction, with a g factor value of 2.009 indicative of O2 - radicals These radicals may arise from surface defects or functional groups retained from the ordered mesoporous carbon, particularly the COOH groups identified in OMC-4, OMC-7, and OMC-9 Notably, the signal intensity is significantly reduced after the reaction, likely due to temperature effects Given that gold catalysts favor high temperatures for hydroformylation, the reaction temperature was maintained at 300˚C.
3.3.2 Gold catalysts surface area and pore distribution
Catalysts’ surface area and comparison to supports are examined in Table 3.8 Average pore volume and size are also discussed to investigate the aftermath structure following Au impregnation
Table 3.8: Surface area and pore characteristics of Au catalysts and support
Following the impregnation of Au, both Cat-Au-2 and Cat-Au-3 experienced a slight reduction in surface area, likely due to the presence of Au particles within the porous structures However, the average pore size and volume remained consistent, indicating that the impregnation process was successful and did not lead to any significant structural alterations.
Fig 3.27: The pore distribution of Au catalysts and respective support
The pore distribution graph for Cat-Au-2 indicates a shift between two primary pore sizes, 54Å and 34Å, akin to the support material This redistribution of pores to the secondary size of 34Å is attributed to the occlusion of Au particles in the larger pores, leading to a partial fill and subsequent redistribution In contrast, OMC-γ-Al2O3 exhibits two distinct types of pores, whereas OMC-SiO2-3g features only a single pore type.
57 signature pore and the secondary pores from the SiO2 template The distribution shifted slightly lower, but the difference is minimal to be considered
To evaluate catalyst cycle run and structure properties following the reaction at 300°C, N2 adsorption method is implemented to investigate the surface area, pore volume, and average pore size
After the reaction, Cat-Au-2 experienced a reduction of 50% in both surface area and pore volume, although it retained its concentrated pore size This decline can be attributed to the occupation of pores by reaction products, resulting in decreased activity over time Nevertheless, the catalyst maintains a high surface area and precise pore size comparable to the OMC-γ-Al2O3 template.
Fig 3.28: Pore size distribution of Cat-Au-2 pre-reaction and after the reaction
Cat-Au-3 exhibited no changes in surface area or pore properties, maintaining consistency with the support and catalyst prior to the reaction This stability may be attributed to its lower activity compared to Cat-Au-2, resulting in fewer products formed and consequently no entrapment within the catalyst's porous structure While this lower activity may sacrifice immediate performance, it could potentially extend the catalyst's lifespan.
Fig 3.29: Pore distribution of Cat-Au-3 pre-reaction and after the reaction
3.3.3 Gold catalysts UV-Vis spectra
Fig 3.30 depicts the UV/Vis spectra of OMC supports used for the Au catalysts synthesis and Au catalysts
OMC samples exhibit no UV absorption, while all Au catalysts display a significant peak at 524 nm, attributed to the plasmon surface phenomenon of nanogold particles The uniformity in Au particle size results in a distinct sharp peak rather than a broad absorption band Post-reaction analysis shows that catalysts maintain enhanced absorption at 524 nm, although Cat-Au-2 exhibits a lower intensity.
The catalyst demonstrates enhanced absorption, likely due to high-temperature reactions occurring in the absence of oxygen, which may lead to the disappearance of functional groups on the surface of the OMC, either through their involvement in the reaction or as a result of heat treatment.
Fig 3.30: UV-Vis spectra of Au catalysts and support
3.3.4 Gold catalysts elemental mapping and EDX results
The surface characteristics and elemental composition of the gold catalyst are analyzed through the SEM/EDX method, complemented by elemental mapping To ensure a more precise evaluation, the EDX results represent the average of three distinct areas examined at the same magnification scale.
Fig 3.31: Cat-Au-2 elemental mapping at 5,000×magnification
Figure 3.31 illustrates the uniform dispersion of gold particles on the surface of the OMC, with elemental mapping confirming that gold (Au) is evenly distributed alongside carbon atoms across the surface.
Table 3.9: EDS results of Cat-Au-2 before and after the reaction
Sample Element atom% Element mass%
The EDS results in Table 3.9 indicate that Cat-Au-2 was successfully synthesized, with an average gold (Au) mass percentage of 1.33%, which is in close alignment with the theoretical calculation of 1% Minor variations in mass calculations are acceptable within the analysis.
Fig 3.32: Cat-Au-2-after SEM images at 5000×magnification
Following the reaction, Au particles maintained an even distribution on the surface, as illustrated in Fig 3.32 However, EDS analysis indicates a decrease in the mass percentage of Au post-reaction As shown in Fig 3.33, gold particles began to cluster around certain areas of the carbon particles, leading to reduced dispersion across the surface and a lower overall mass percentage.
Accumulation was observed only in specific areas of the surface, while the oxygen element associated with the surface functional group diminished after the reaction, despite an increase in mass percentage This increase is likely due to a decrease in gold mass percentage Additionally, the disappearance of the oxygen element in elemental mapping can be linked to the removal of the surface functional group, which may result from heat treatment or interactions occurring during hydroformylation.
Fig 3.33: Cat-Au-2-after elemental mapping at 5,000×magnification
In the EDS analysis of Cat-Au-3, minimal gold (Au) mass percentage was detected, likely due to undispersed Au particles during impregnation or their excessively small size for detection Although Au particles are visible on the surface, their distribution is less concentrated compared to Cat-Au-2, as illustrated in Fig 3.34.
The EDS results from the respective Cat-Au-3 in Table 3.10 is calculated from different areas
Table 3.10: EDS results of Cat-Au-3
Fig 3.34: Cat-Au-3 elemental mapping at 5,000×magnification
Elemental mapping at 5000× magnification revealed a clear accumulation of Au particles, which may reduce high surface energy as a result of their nanoparticle form This phenomenon is linked to the lower activity observed with Cat-Au-3, which will be further elaborated in the subsequent section.
Fig 3.35: Cat-Au-3-after elemental mapping at 5,000×magnification
Cat-Au-2 and Cat-Au-3 hydroformylation’s activity was tested at 300°C and 7 bar for 5 hours Conversions for both catalysts are shown in Fig 3.36
Fig 3.36: Ethylene Conversion with time-on-stream over 0.15g powder Cat-Au, gas flow rate 60mL/min, at 7 bar and 300°C
Both catalysts demonstrated an increase in conversion over time, stabilizing by the fifth hour of the reaction Cat-Au-3 outperformed Cat-Au-2, achieving a conversion rate of 83%, while Cat-Au-2 reached 74% after the same duration.
powder Cat-SILP-2, gas flow rate 60mL/min, at 7 bar
g powder Cat-SILP-2, gas flow rate 60mL/min, at 7 bar
The selectivity of the SILP-2 catalyst at varying temperatures was assessed, revealing that at 80°C, Cat-SILP-2 exhibited minimal propanal production, as most was converted to propan-1-ol Initially, propanal selectivity was only 0.7% after 30 minutes of reaction, while propan-1-ol selectivity increased steadily during the first hour, achieving 75% after 4.5 hours, before declining after 5 hours.
Propan-2-ol exhibited a significantly lower selectivity, achieving only 35% after the initial 30 minutes, which then dropped to 4% This aligns with the findings from the ethylene conversion section, indicating that lower selectivity at 100°C and 140°C corresponded with reduced conversion rates, as SILP-2 did not produce propan-2-ol or propanal Additionally, propan-1-ol's selectivity was inferior to that observed at 80°C, recording 40% and 25% selectivity, respectively Notably, the selectivity trends for propan-1-ol gradually improved during the initial phase.
4 hours for reaction at 80 and 100°C before dropping to a more stable selectivity
At 120°C, however, the SILP-2 displays a tendency for the propanal formation and propan-2-ol in favor of propan-1-ol, even though the difference is minimal
At 80°C, SILP-2 demonstrates high selectivity for propan-1-ol after 4 hours of reaction, along with minimal propan-2-ol production In contrast, at 100°C and 140°C, neither propanal nor propan-2-ol was detected, resulting in a significantly reduced selectivity for propan-1-ol Throughout the reaction duration, SILP-2 produced all three products; however, the presence of substantial by-products negatively impacted the overall selectivity for each product.
Figure 3.16 illustrates the calculated Turnover Frequency (TOF) for each product Although byproduct formation can affect selectivity during reactions, the TOF provides valuable insights into catalyst performance by reflecting the concentration of the products formed.
Fig 3.16: Propanal, propan-1-ol, propan-2-ol’s TOF with time-on-stream over 0.15g powder Cat-SILP-2, gas flow rate 60mL/min, at 7 bar
The propan-1-ol turnover frequency (TOF) of SILP-2 at 80°C is the highest among tested temperatures, indicating a strong selectivity trend Although propan-1-ol demonstrates superior selectivity, its TOF difference from propan-2-ol is minimal At both 100°C and 120°C, the TOF of propan-1-ol remains consistent at 1.288 h⁻¹ after 4 hours Notably, at 100°C, SILP-2 does not produce any propan-2-ol or propanal, whereas at 120°C, the formation of propanal and propan-2-ol increases steadily over time, reaching 1.443 and 1.412, respectively.
SILP-2 catalysts consistently produce propan-1-ol across all three products; however, the formation of byproducts negatively impacts both selectivity and catalyst conversion rates Optimal catalyst performance is achieved at a temperature of 120°C.
At 80°C, the main products exhibited a notable turnover frequency (TOF) of 45, while SILP -2 closely followed with a slightly lower TOF However, the lower TOF at this temperature effectively prevented byproduct reactions, leading to enhanced selectivity in the overall process.
At an optimal temperature of 120°C, Cat-SILP-2 exhibited the highest conversion rates, with Cat-SILP-3 and Cat-SILP-4 showing similar performance Although Cat-SILP-4 experienced a drop in conversion during the first 30 minutes, it gradually increased to 84% after 5 hours In contrast, Cat-SILP-3 maintained a higher conversion than Cat-SILP-4 for most of the duration but fell from 79% to 32% after 5 hours All catalysts, except Cat-SILP-4, required a minimum of 30 minutes to activate the reaction The unpredictable behavior of Cat-SILP-4 will be further discussed in the selectivity and TOF section.
In addition, Cat-SILP-5 shows stability during the whole run, with conversion ranging from 23 to 25%
Fig 3.17: Ethylene Conversion with time-on-stream over 0.15g powder Cat-SILP, gas flow rate 60mL/min, at 7 bar and 120°C
In the reaction analysis, SILP-4 initially demonstrates an impressive 98% selectivity for propanal, but this significantly declines to 0% after 5 hours, with propan-1-ol and propan-2-ol showing nearly 0% selectivity at 120°C A notable peak at a retention time of 20 minutes emerged during the reaction, impacting the overall selectivity of the products Conversely, SILP-2 shows moderate propanal selectivity of 26% in the first hour, which gradually decreases over the next 4 hours as propan-1-ol and propan-2-ol are produced, albeit with low selectivity Notably, all three products are fully observed during SILP-3's activity testing, which particularly favors the formation of propan-2-ol.
powder Cat-SILP, gas flow rate 60mL/min, at 7 bar and 120°C
g powder Cat-SILP, gas flow rate 60mL/min, at 7 bar and 120°C
The product turnover frequency (TOF) of SILP catalysts at 120°C reveals that SILP-4 exhibits a higher TOF of 4.63 h -1 at the onset of the reaction, despite its nearly 0% selectivity, which decreases to 1.45 h -1 after 5 hours This performance is attributed to the presence of γ-Al2O3 particles, which contribute to hydroformylation alongside the residual OMC, as γ-Al2O3 can also act as a catalyst Additionally, stable production of propan-2-ol and propan-1-ol was noted, with SILP-2 showing a similar TOF for propan-2-ol at 1.42 h -1, while the TOF for propan-1-ol remained constant.
In a reaction test, propan-1-ol was not detected after 4 hours, while SILP-3 demonstrated the highest turnover frequency (TOF) for propan-2-ol among all SILP catalysts, averaging 1.47 h⁻¹ The TOF for both propanal and propan-1-ol decreased gradually during the reaction Conversely, SILP-5 showed a greater preference for propan-1-ol compared to other SILP catalysts, suggesting that while its TOF for propan-2-ol and propan-1-ol increased, SILP-5 required a longer stabilization period, indicating potential for sustained activity over extended reaction times.
Fig 3.19: Propanal, propan-1-ol, propan-2-ol TOF with time-on-stream over 0.15g powder Cat-SILP, gas flow rate 60mL/min, at 7 bar and 120°C
In the hydroformylation process using SILP catalysts at 120°C, distinct behaviors were observed among the catalysts Cat-SILP-2 achieved the highest conversion and stable turnover frequency (TOF) over four hours, yet exhibited low selectivity towards the target products, with selectivity ranging from 15% to 18% In contrast, Cat-SILP-3 demonstrated slightly lower conversion than SILP-2 but favored the formation of propan-2-ol, achieving selectivities of 75% and 55% after five hours, along with the highest TOF for propan-2-ol The performance of SILP-4 was erratic, characterized by rising conversion rates and the emergence of unknown heavy products, resulting in disappointing selectivity despite initially high propanal TOF in the first hour Finally, SILP-5 showed commendable stability, maintaining a relatively constant conversion of 20% with consistent selectivity throughout the reaction.
In the initial two hours, propanal is converted to propan-2-ol, followed by a transition to propan-1-ol over the next three hours SILP-5 demonstrates the highest turnover frequency (TOF) for propan-1-ol, while also achieving slightly improved TOF for both propanal and propan-2-ol This indicates that SILP-5 prioritizes selectivity over conversion, much like SILP-3.
At 120°C, Cat-SILP-3 and Cat-SILP-5 emerge as the most effective catalysts, exhibiting distinct behaviors in product variation and adsorption during reactions The turnover frequency (TOF) for propan-1-ol is higher for Cat-SILP-5 compared to Cat-SILP-3, while the reverse is true for propan-2-ol Additionally, Cat-SILP-5 outperforms Cat-SILP-3 in propanal TOF, correlating with a more favorable n/iso product ratio attributed to the presence of COOH functional groups, which enhance this ratio In terms of selectivity, both catalysts demonstrate low selectivity for propan-1-ol; however, Cat-SILP-3 maintains a relatively high selectivity of 50 to 70%, whereas SILP-5's selectivity improves only after the initial two hours Notably, SILP-5 achieves nearly 80% selectivity for propanal, likely due to the hydrogenation suppression caused by carboxyl functional groups Furthermore, the presence of functional groups influences adsorption, potentially leading to ionic liquid blockage in Cat-SILP-2 and resulting in decreased activity.
Fig 3.20: Ethylene Conversion with time-on-stream over 0.15g powder Cat-SILP, gas flow rate 60mL/min, at 7 bar and 140°C
At 140°C, the conversion trend for SILP catalysts closely resembles that at 120°C, with SILP-2 and SILP-3 achieving 78% and 73% conversion after 5 hours, respectively This represents a slight decrease compared to the conversion rates at 120°C The lower conversion at the higher temperature can be attributed to the preference for hydrogenation over hydroformylation due to the functional groups present in the catalysts.
powder Cat-SILP, gas flow rate 60mL/min, at 7 bar and 140°C
g powder Cat-SILP, gas flow rate 60mL/min, at 7 bar and 140°C
The propanal selectivity graph reveals that SILP-3 initially demonstrated a high selectivity of 71%, attributed to the predominant hydroformylation process producing propanal before hydrogenation occurred, resulting in a subsequent 71% selectivity for propan-2-ol after one hour Throughout the experiment, SILP-3 maintained low propan-1-ol selectivity, remaining below 10% In contrast, SILP-5 exhibited stable conversion patterns similar to those at 120°C, with selectivity favoring propanal after 2.5 hours, transitioning to propan-2-ol after three hours, and finally shifting to propan-1-ol, achieving selectivities of 65%, 77%, and 43%, respectively, indicating an earlier trend shift.
At 120°C, the activation rate is significantly faster, with SILP-4 exhibiting low activity due to the formation of unknown byproducts, leading to nearly 0% selectivity In contrast, SILP-2 demonstrated no production of propanal and propan-2-ol, resulting in an overall propan-1-ol selectivity of 25%, which is slightly higher than that observed in SILP-3 and SILP-5 after 5 hours.
The corresponding TOF of SILP catalysts at 140°C is depicted in Fig 3.22
Fig 3.22: Propanal, propan-1-ol, propan-2-ol TOF with time-on-stream over 0.15g powder Cat-SILP, gas flow rate 60mL/min, at 7 bar and 140°C
At 140°C, neither propanal nor propan-2-ol was detected in SILP-2, resulting in a selectivity and turnover frequency (TOF) of 0 In contrast, propan-1-ol formation plateaued at 1.28 h^-1 over a 5-hour period Interestingly, SILP-4 outperformed SILP-2, exhibiting consistent production of both propan-2-ol and propan-1-ol, despite a slight decrease compared to SILP-2.
5 In propanal TOF graph, SILP-4 started strongly with a 1.57 h -1 but gradually ran its course throughout the next 5 hours It is no doubt that SILP -5 and SILP-3 still came out on top of propanol formation in terms of TOF While both SILP-3 and SILP-5 propan-1-ol TOF ranged around 1.29 h -1 during the first 4 hours, it increased significantly to 1.6 h -1 before slowly dropping to 1.4 h -1 in the last hour for SILP-3 but declined for SILP-5 The same can be said about propan-2-ol and propanal formation for SILP-3, as both respective TOFs are higher than SILP-5,
51 with the only exception being the first few hours of reaction SILP-5 produced marginally more propanal than SILP-3
The primary outcome of the thesis on ethylene conversion is the production of propanal and propanol Among the catalysts tested, only Cat-SILP-3 demonstrated a satisfactory selectivity and production for all three target products, although its lower activity was linked to the formation of byproducts Cat-SILP-2 achieved a notable 25% selectivity for propan-1-ol, with no significant production of propanal or propan-2-ol SILP-5 maintained steady conversion rates, mirroring trends observed at 120°C While SILP-4 exhibited strong selectivity for propanal at 120°C, its chaotic catalyst behavior hindered further performance, resulting in disappointing selectivity at 140°C due to the emergence of byproducts similar to those seen with SILP-3 at the same temperature Additionally, the rates of propanal production with SILP-4 were likely influenced by the interaction between Rhodium and γ-Al2O3 support rather than Rhodium and OMC.
Cat-SILP-3 emerged as the most effective catalyst among SILP variants at both 120°C and 140°C, demonstrating superior propan-2-ol production and a respectable turnover frequency (TOF) for propan-1-ol after 5 hours Its TOF for propanal remained consistent; however, the formation of a heavy, unidentified byproduct negatively impacted its selectivity In contrast, SILP-5 exhibited increased stability only after prolonged exposure.
The reaction conducted for 4 hours at 120°C primarily yielded propanal, while increasing the temperature to 140°C resulted in a more significant production of desirable products However, propanal remained the predominant product during the initial phase of the reaction at both temperatures.
Despite achieving high conversion rates, the SILP-2 products primarily consisted of methyl-1-pentanol and 2-methyl-2-pentanol, which were not the desired outcomes, resulting in disappointing selectivity At 140°C, the reaction did not yield any propanal or propan-2-ol At 120°C, SILP-5 produced slightly more propan-1-ol compared to SILP-3, although SILP-3 remained dominant in propan-2-ol production The presence of carboxyl surface functional groups significantly influenced the n/iso ratio at 140°C Additionally, the turnover frequency (TOF) for propanal in SILP-5 decreased slightly due to more favorable hydrogenation reactions at elevated temperatures, which may have also led to the emergence of an unknown byproduct that hindered both catalysts' activity.
SILP-4 was analyzed to understand its chaotic behavior, demonstrating a turnover frequency (TOF) of nearly 5 h⁻¹ for propanal and maintaining an impressive selectivity of almost 100% during the initial hour However, its performance declined significantly in subsequent hours Despite this drop, SILP-4 still produced a satisfactory amount of the desired product, showing a slight preference for propanal compared to other catalysts.
3.2.7 Catalyst characteristics after the reaction
After the reaction, FT-IR and N2 adsorption methods were used to re- evaluate the catalyst’s characteristics following activation to see if the catalysts could retain their structure
Cat-SILP-2 and Cat-SILP-3 before and after reaction FT-IR spectrum are depicted in Fig 3.23
Fig 3.23: FT-IR spectrum of SILP-2 and SILP-3 catalyst before and after the reaction
Following the reaction, both catalysts maintain their distinctive spectral features, with the P-C bond from the ligand TPPTS detected at 1622 cm -1 and C=N stretching vibrations from the ionic liquid observed at 1459 cm -1 Additionally, three shoulder peaks corresponding to S=O are clearly visible.
The study reveals that the SILP catalysts maintain their structural integrity after 5 hours of reaction, as indicated by the weak rhodium interactions with carbon or oxygen observed at 448 cm-1 The numerical values of 1168, 1109, and 1060 further support this finding, demonstrating the stability of the catalysts throughout the process.
Table 3.7: Surface area and porosity of SILP catalysts before and after the reaction
Cat-SILP-2-after OMC-SBA-
Cat-SILP-3-after OMC-SBA- 45.71 0.139 65.161
The data presented in Table 3.7 indicates that the Cat-SILP-2 structure experienced shrinkage upon exposure to temperature, leading to a decrease in both BET surface area and average pore volume This phenomenon can be attributed to the internal structural shrinkage, which resulted in a reduced pore volume while the average pore size remained relatively unchanged Consequently, the smaller pores expanded, resulting in a broader pore size distribution for the Cat-SILP-2-after, as illustrated in Fig 3.24.
Fig 3.24: Pore distribution of Cat-SILP-2 before and after the reaction
The catalysts for Cat-SILP-4 exhibit minimal surface area before and after the reaction, causing the ionic liquid layer, active sites, and ligands to remain on the catalyst surface and be easily leached out This phenomenon results in a reduced average pore size, as the ionic liquid either evaporates or leaches away, revealing the rough surface of OMC-γ-Al2O3-4g.
Cat-SILP-3 catalysts maintain their surface area and pore volume with only slight changes post-reaction, although the average pore size experiences a minor decrease In contrast, Cat-SILP-5 exhibits a significant increase in surface area, returning to the original value of the OMC-SiO2-3g support, likely due to the evaporation of the ionic liquid.
SILP-5 demonstrated consistent and stable activity across varying temperatures, indicating an ordered internal structure that facilitates unobstructed mass transfer of ethylene during reactions This is further evidenced by the initial hours needed for SILP-5 to reach a steady state, which suggests a process of impurity removal and internal “cleaning” that optimizes ethylene passage.
Fig 3.25: Pore distribution of Cat-SILP before and after the reaction
3.3 Nanogold catalysts on ordered mesoporous carbon
3.3.1 Gold catalysts electron paramagnetic resonance’s spectrum
As a comparison, the EPR spectrum of Cat-Au-1, which was Au catalysts impregnation on OMC-SBA-15, was used to compare the difference between gold catalysts samples
Fig 3.26: Gold catalysts EPR's spectrum