B IOMASS DEVELOPMENT - "利用液態氫源與雙功能鈀鈷的碳球在常溫常壓下製備2,5-二甲基呋喃"

1. INTRODUCTION

1.2. B IOMASS DEVELOPMENT

The development of biomass can be categorized into three generations. The first generation of biomass utilizes agricultural crops because their composition contains high lipid/oil or sugar contents. Biodiesel and bioethanol can be produced from these crops. Over the last few decades, innumerable works have been published regarding the conversion of this feedstock to energy. However, the competition of using this feedstock for energy and using them as our food supply makes it difficult to rely on this source.

Second generation biomass utilizes agricultural wastes that are also high in lipid/oil and sugar content. These agricultural wastes are often lignocellulosic matter.

Lignocellulose is composed of Cellulose, Hemicellulose and Lignin. Lignocellulosic biomass is a viable option in that (a) it is abundantly available on earth, (b) it does not coincide with our food supply, (c) a positive net energy gain (NEG) can be obtained from the conversion process. Similar products such as biodiesel and bioethanol could also be produced from lignocellulose. The key difference lies in that lignocellulose is

5

composed of strongly polymerized structures that are recalcitrantly sturdy. After going through hydrolysis, the collected sugar units could be used as feedstock to produce fuel.

Third generation biomass utilizes micro-algae as the main source of feedstock.

This feedstock is also largely dependent on the geology and environment, since large masses of micro-algae has to be able to be grown for this alternative to be realized.

Since this research does not focus on third generation biomass, it will be concluded here that second and third generation biomass does not conflict with each other and there is no clear winner of these two, it all depends on the geology and environment of the production site.

6 1.3. Biomass Conversion

As Figure 1.4 has shown, Cellulose (A component of Lignocellulose) can be hydrolyzed into glucose, and then Glucose can be isomerized into Fructose. Lastly, fructose can be dehydrated into 5-Hydroxymethylfurfural (HMF), a reactant in which a lot of the building block chemical synthesis works have been based on. Numerous works have been done on the hydrolysis of Cellulose into glucose. Even direct hydrolysis of raw crops has been researched quite extensively over the past few years.⁷

Figure 1.4 Production of HMF, from cellulose and carbohydrates, serves

as feedstock for a range of chemicals and liquid fuels.

⁸

7

Three generations of biomass pretreatment and hydrolysis methods have be categorized in a review by Sathitsuksanoh et al.⁹ The first generation devises a one-step biomass dissolution and hydrolysis. The second generation goes through biomass dissolution first, followed by enzymatic hydrolysis. The third generation utilizes lignocellulose fractionation. The first generation is mainly based on the use of concentrated acids, which is raw-material independent, however it poses some disadvantages, such as product separation difficulties, acid recovery and acid re-concentration.^10-12 The second generation mainly focuses on the use of enzymes;

the use of enzymes could overcome some of the difficulties faced by the first generation, however, other problems such as high costs, corrosive and toxic properties of pretreatment solvents.^13,14 Third generation methods involve the use of fractionation, overcoming separation difficulties faced by the first generation methods, fractionation could achieve continuous processes through the use of acids and ionic liquids under modest reaction conditions.^15,16

Works on the isomerization of glucose to cellulose or even cellulose oligomers to fructose have been done.¹⁷ Particularly, work on cellulose-to-HMF conversion have been studied extensively over the past few years and high yields could also be achieved.^18-22 Selection of the most suitable method could be made over large quantities of works for the production of HMF, one-step or one-pot synthesis methods

8

are also now available. However, despite large quantities of works and researches on the synthesis of HMF, the research on the synthesis of DMF seems rather neglected. It is therefore what this research aims to search for options and alternatives to add to works on DMF synthesis.

9 2. PAPER SURVEY

2.1. Metal Organic Framework (MOF) based catalysts

Metal Organic Frameworks are inorganic (Metal component) and organic units linked together to form a structure with high degrees of crystallinity. It has recently gained widespread popularity due to its versatility to be shaped differently through the use of different metal nodes and linkers, which often give them different properties or functionalities. Often compared with Zeolites, MOFs show similar characteristics to Zeolites. Generally, MOFs are thermal and chemically stable, and contain micropores which grant them large surface areas of 1000 to 10,000 m²/g. Recently, Postsynthetic Modification (PSM) through covalent bond exchanges of MOFs have also been studied to make MOFs even more versatile in terms of applications.²³ With these attractive characteristics, MOFs can be applied to a wide variety of applications such as gas storage, gas separation, catalysis and sensors.²⁴ Over the last decade, MOFs have been widely explored as potential catalysts.^25-29 Specifically, the class of Zeolitic Imidazolate Frameworks (ZIF) will be discussed here, since the catalyst used here is derived of such origins.

As suggested by the name, ZIF uses imidazolate units as linkers to form a structure with M-Im-M angles of 145°, similar to Si-O-Si angles in Zeolites. ZIF

10

possesses tetrahedral topologies

Figure 2.1 ZIF bond angles corresponding to Zeolite bond angles.

³⁰

The specific material used in this research is ZIF-67, it was chosen because it contains Cobalt as the metal node, which would be used to catalyze the catalysis of hydrogen production from sodium borohydride.³¹ The synthesis of ZIF-67 and its calcination into cobalt supported on carbon has also gained attention quite recently.^32-36 It is convenient to have the cobalt metal homogeneously distributed on the product, rather than using different methods to load the metals onto the material.

11 2.2. Incipient Wetness Impregnation

³⁷

One of the most widely used industrial methods of preparing supported precious metal catalysts combines impregnation and drying. This is a relatively easy method that deposits metals onto supports through physical means. Impregnation is when the liquid phase comes into contact with a solid phase, and the liquid phase is absorbed by the solid phase. A typical impregnation involves allowing a precursor solution containing the metal source to be absorbed into a solid support. The precursor is chosen according to the price and its physicochemical properties. The affinity of the precursor towards the solid support will determine the effect of the impregnation. The particle size of the precursor is also a determining factor, since its interaction with the pore of the solid will determine the effect of the impregnation.

Fick’s Law

Darcy’s Law

When the liquid phase volume exceeds the volume of the pores on the support, wet/diffusional impregnation is given as the name of the method.^38-42 Fick’s law of diffusion, the adsorption capacity of the surface and the adsorption equilibrium constant governs this type of impregnation. When the liquid phase volume is equal to the volume of the pores on the support, the method is named dry impregnation. For

12

this type of impregnation, other effects come into play, such as the pressure driven

capillary flow of the solute inside the pores. Figure 2.1 shows these phenomena in illustration. This phenomenon can be described by Darcy’s Law.

Figure 2.2 Phenomena of transport involved in (a) wet impregnation and (b) dry impregnation. The solute migrates into the pore from the left to

the right of the figures.

³⁷

When the precursor solution is impregnated inside the pores of the support, the mixture is dried to remove the solvent, while the metal stays inside the pores of the support. Typical procedures involve heating in an oven to the boiling point of the solvent with or without selected gas flows. Heating rate and final heating temperatures can all affect the adsorption of the precursor on the pore surfaces. The balance between adsorption, back-diffusion and convection determines the

13

distribution of the precursor inside the pores.^43-46 When the convective flow of the solvent counters the vapor removal flow, a constant-rate period is achieved as shown in Figure 2.2(a). This is usually the case when the heating rate and the drying temperature are high, the solvent front recedes into the pores and the evaporation occurs inside the pores. When the convective flow is slower than the vapor removal flow, a falling-rate period is achieved as shown in Figure 2.2(b). This is usually the case when the heating rate and the drying temperature are low, the solvent front stays at the surface of the support and loss of the precursor particles can occur through microconvection. The drying regime is defined as slow if the constant-rate period predominates, and as fast if it is the falling-rate period.

Figure 2.3 Phenomena of transport involved in (a) the constant-rate period of drying and (b) the falling-rate period of drying. The solvent

migrates from the left to the right of the figures.

³⁷

14

Other methods of metal deposition onto the supports could be explored, such as the stirring of the support in metal precursor solutions all the while adding in reducing agents to deposit the metal onto the supports. A convenient method that could deposit metal with an even distribution onto the supports would be attractive. Deposition to form chemical bonds between support and metals would also help make the metals more firmly attached on the material, which could lead to good recyclability.

15 2.3. Production of 2,5-Dimethylfuran

It has been proposed that the production of DMF from HMF will go through two pathways as shown in Figure 2.4. Both of these pathways involve the hydrogenation and hydrogenolysis of HMF and its intermediates.⁴⁷

Figure 2.4 Synthetic routes to obtain DMF from HMF.

⁴⁸

Since HMF is a widely known chemical platform, most of the works on DMF synthesis has been conducted starting with HMF. HMF is a chemical platform that is often synthesized from hexoses such as fructose through dehydration reactions.

Earlier works of DMF synthesis often start with fructose. Herein, focus will be put on HMF to DMF conversions.

Sudipta and Basudeb et al. synthesized DMF from HMF in a sequential reaction that starts with fructose.⁴⁹ A yield of 32% DMF could be achieved from fructose under conditions stated in Table 2.1. Binder and Raines et al. also synthesized DMF from HMF in a sequential reaction that starts from fructose.⁵⁰ The reaction conditions

HMF

MFAD

BHMF MFM

DMF

16

were a follow up of the work of Román-Leshkov and Dumesic et al.⁴⁷ A yield of 32.5% DMF could be achieved from fructose in this work. Chidambaram and Bell et al. demonstrated tests in ionic liquid systems and obtained a yield of 32% DMF under EMIMCl with addition of acetonitrile.⁵¹ It is mentioned in their work that the solubility of hydrogen inside the solvent is a determining factor for the efficiency of the hydrogenation reaction. Under the same conditions they compared four different metals for DMF synthesis which are Palladium, Platinum, Ruthenium and Rhodium.

They found that Carbon supported Palladium shows the best results.

Thananatthanachon and Rauchfuss et al. used Pd/C as catalyst with formic acid as a liquid hydrogen source and deoxygenation agent at the same time.⁵² A different term was given for aqueous hydrogenations, which is transfer hydrogenation, distinction of this mechanism with traditional hydrogenation can be made by determining whether hydrogen gas is produced or not. Transfer hydrogenation does not need production of hydrogen gas; traditional hydrogenations require production of hydrogen gas. The novelty of this work resides in that no pressured hydrogen gas is needed, which saves a lot of energy. Jungho and Dionisios et al. used Ru/C catalyst with 2-propanol as hydrogen source for transfer hydrogenation.⁵³ It is claimed in their work that no homogeneous acid is used in the reaction as compared to the work of Thananatthanachon’s group. They mentioned that using 2-propanol as a hydrogen

17

donor is attractive due to its green derivation. Alcohols can be readily produced from biomass and the dehydrogenated products can be recycled by hydrogenation, although this requires yet another step with hydrogenation.

Regarding works on using Ru/C as a catalyst, Junhua and Shijie et al.

demonstrated that 60.3% DMF yield could be obtained under the conditions in Table 2.1.⁵⁴ Yanhong and Yanqin et al. used Ru/Co3O4 as catalyst to obtain high yields of 93.4% DMF under the conditions in Table 2.1.⁴⁸ It is reported in their work that Ruthenium plays the role of carbonyl group to hydroxyl group conversion. In fact, most of the works have reported that carbonyl to hydroxyl group conversions can be done easily, where they proposed that C-O bond cleavage (hydrogenolysis) is the determining step. Therefore they used Co₃O₄ as the catalyst for C-O bond cleavage.

One of the earliest works on DMF synthesis dates back to 2007, where Román-Leshkov and Dumesic et al. synthesized DMF starting from either fructose or HMF.⁴⁷ A bimetallic copper and ruthenium catalyst (Cu-Ru/C) was used for the conversion of HMF to DMF. They proposed that due to the copper having lower surface energy than ruthenium, a two-phase system develops on the catalyst in which the copper coats the ruthenium. This was needed because during their sequential reaction from fructose to HMF, chloride ions were present, and poisoning of the copper on the catalyst could occur.⁵⁵ Yields of 71% DMF could be achieved in

18

purified solvents with chloride ions absent. Another work with heterogeneous bimetallic catalyst by Nishimura and Ebitani et al. introduces Palladium and Gold supported on Carbon.⁵⁶ They claim to achieve 96% yields under atmospheric pressure with moderate temperature as shown in Table 2.1. The presence of gold on the catalysts does indeed improve the yield as the reactants were increased and the catalyst amount was decreased. Huang and Fu et al used a bimetallic Nickel and Tungsten Carbide catalyst to catalyze the reaction under the reaction conditions given in Table 2.1.⁵⁷ It was reported in their work that Nickel shows excellent hydrogenation abilities, however its selectivity was bad therefore tests were conducted with tungsten carbide. Tungsten carbide shows excellent selectivity towards the hydrogenolysis of alcohol groups into methyl groups, basically a deoxygenation reaction. This leads them to come up with a catalyst that contains lower amounts of Nickel for the conversion of HMF to an intermediate with both hands switched into alcohol groups then the high amount of Tungsten Carbide would convert the alcohol groups into methyl groups producing DMF. In the last work on bimetallic catalysts to date, Wang and Schuth et al made a catalyst of Platinum and Cobalt nanoparticles encased in hollow carbon nanospheres and conducted the reaction under the conditions listed in Table 2.1.⁵⁸ They have attributed the high yield of 98% to the interaction between the Cobalt and Platinum alloy, because tests with Platinum or Cobalt supported alone

19

resulted in low yields.

Table 2.1 Past Works on DMF Synthesis.

Catalyst Reactants Reaction

20 3. OBJECTIVE

The core idea of this research stems from the functions of our catalyst. Our catalyst contains Palladium and Cobalt, as the name Pd/CoNC implies. It is textbook and common knowledge that precious metals such as Pt, Pd and Ru provide a surface for which hydrogenation reactionsoccur.^61-64 It is therefore suitable that we apply our catalyst to hydrogenation reactions. Hydrogenation reactions require a source of hydrogen whether it comes from directly supplying the system with hydrogen gas or indirect methods such as transfer hydrogenation.^65-67

Sodium Borohydride (NaBH₄) is known to be able to produce hydrogen gas when reacted with water as shown in Eq. 1.^20,21 This means that the production of hydrogen occurs inside the solution, which possibly provides a higher chance of contact between hydrogen atoms and the reaction sites, therefore NaBH₄ has been chosen as our source of hydrogen.

NaBH

+ (2+x)H

O→ NaBO

·xH

O + 4H

(-210 kJ mol

^-1

)

⁶⁸

(Eq. 1)

It is known that Cobalt can catalyze the hydrolysis of NaBH₄ solution to have a higher hydrogen generation rate and produce a larger total amount of hydrogen.^69-73

21

Therefore our catalyst shows bifunctionality towards hydrogenation reactions, producing hydrogen at a faster rate and providing a surface for which the reactions can occur.

The chosen reaction for our research is the conversion of 5-methylfurfural to 2,5-Dimethylfuran. As discussed in the paper survey section. This reaction requires two hydrogenation reactions in total, of which one is the conversion of an aldehyde group to an alcohol group and the other is a hydrogenolysis reaction.

The objective of this research is to achieve a facile method of 2,5-Dimethylfuran synthesis under atmospheric pressure and room temperature while using an aqueous hydrogen source to obtain high yields. The advantages of atmospheric pressure and room temperature are that it provides a safe environment for synthesis, since highly pressured hydrogen gas at high temperatures can be quite dangerous. The reaction is also energy efficient because it does not require additional heating and purging with pressurized hydrogen gas. The use of aqueous hydrogen sources other than transfer hydrogenation on biomass conversion will also be explored here for the first time.

22 Carbon Supported Palladium (5 wt%) Aldrich Cobalt (II) Chloride Hexahydrate Sigma-Aldrich Hydrochloric Acid (HCl) Sigma-Aldrich Magnesium Sulfate (MgSO

) Yakuri Chemicals

Palladium (II) Chloride Aldrich

Sodium Borohydride (NaBH

₄

) Aldrich Sodium Hydroxide (NaOH) Sigma-Aldrich Sulfuric Acid (H

SO

) Sigma-Aldrich

Tetrahydrofuran (THF) Sigma-Aldrich

23 4.2. Equipment

Equipment Type

Centrifugator Sigma 3-30KS

Ultrasonicator Qsonica, Ultrasonic processor Part No.Q700

X-Ray Diffractometer (XRD) Rigaku, Ultima IV

Specific Area and Pore Size

Distribution Instrument Micromeritics ASAP 2010 Scanning Electron Microscope

(SEM) Nova

^TM

NanoSEM 230

TGA PERKIN ELMER PYRIS 1

Syringe Pump KDS-100

Calcination Oven Thermoscientific Lindberg Blue M

24 4.3. Procedure for production of 2,5-Dimethylfuran

Owing to the fact that aqueous hydrogen source that produces hydrogen gas has been used here for the first time in the production of 2,5-Dimethylfuran, the experimental procedure was modified several times over the course of discovering the best system for the production of 2,5-Dimethylfuran. Two of the main reaction systems were discussed here:

Figure 4.1 Reaction systems used in the experiment. (Method 1) Batch Reactions, (Method 2) Semi-Batch Reactions.

4.3.1. Batch Reactions

For a standard experiment procedure, 0.4 mmol of 5-methylfurfural (MFAD),

25

varying amounts of Pd/CoNC catalyst, 4.5 mL of tetrahydrofuran (THF) and a magnetic stir bar were added in a 7 or 10 mL vial to be stirred for 30 min to ensure quality distribution of the reactant throughout the catalyst surface. 0.06 g of Sodium Borohydride (NaBH4) was then added right before addition of 1 mL of de-ionized water into the vial. After the addition of water, there would be intense bubbling, so the cap should be sealed as soon as possible. Parafilm was wrapped around the cap to ensure no loss of hydrogen gas produced. The reaction was placed in a water bath of varying temperatures for varying durations of reaction.

For reactions with addition of Sodium Hydroxide (NaOH), the reactant was 5-Hydroxymethylfurfural (HMF) instead of MFAD. 4.5 mL of THF with 0.1 g of Pd/CoNC catalyst were added as with the procedure of standard tests. After 30 minutes of stirring, de-ionized water was added with NaOH (0.14 mL). It can be observed that there is mild bubbling. The vial was then put in a water bath of varying temperatures for 3 hours.

For reactions with addition of acids, the reactant was also HMF instead of MFAD. 4.5 mL of THF with 0.1 g of Pd/CoNC catalyst were added as with the procedure of standard tests. For the addition of H₂SO₄, 1.6 μL of H₂SO₄ was added with 1 mL of de-ionized water after 30 minutes of stirring. For addition of HCl, 2 μL of HCl was added instead. The reaction was then put in a water bath of 30°C for 3

26

hours.

After the reaction, the catalysts were collected with a magnet and the product solution was poured into a 20 mL vial. Magnesium Sulfate (MgSO4) was added to remove the water from the product solution for the safety of the Mass Spectroscopy.

The MgSO4 solid was filtered with a syringe and a filter disk. Finally, the solution was analyzed by GC-MS with the Agilent HP-5ms Column.

4.3.2. Semi-Batch Reactions

For semi-batch tests, 0.1 g of Pd/CoNC catalyst was prepared in a 10 mL vial with a magnetic stir bar. 0.4 mmol of MFAD was added with 4.5 mL of THF. The aqueous hydrogen source of NaBH₄ (0.06 g) in 30 wt% NaOH solution (1 mL) was prepared by addition of 0.14 mL of NaOH solution with 0.86 mL of de-ionized water in a 4 mL vial to create a basic environment. Then NaBH₄ was added to the solution, slow and minor bubbling can be observed. The solution was quickly transferred into a syringe on a syringe pump, while varying amounts of H₂SO₄ acid was added into the

在文檔中 "利用液態氫源與雙功能鈀鈷的碳球在常溫常壓下製備2,5-二甲基呋喃" (頁 15-0)