DEGREE PROJECT IN CHEMICAL ENGINEERING AND TECHNOLOGY, FIRST LEVEL
STOCKHOLM, SWEDEN 2019
KTH ROYAL INSTITUTE OF TECHNOLOGY
KTH ENGINEERING SCIENCES IN CHEMISTRY, BIOTECHNOLOGY AND HEALTH
Catalyzed synthesis of aromatic esters
Oscar Dalla-Santa
DEGREE PROJECT
Bachelor of Science in
Chemical Engineering and Technology
Title: Catalyzed synthesis of aromatic esters Swedish title: Katalyserad syntes av aromatiska estrar
Keywords: Catalysis, metallocene, zirconium, zirconocene triflate, aromatic esterification, etherification, bis(cyclopentadienyl)zirconium(IV)
bis(trifluoromethanesulfonate), electrophilic aromatic substitution, ester, ether
Workplace: KTH, Royal Institute of Technology Department of organic chemistry Supervisor at
the workplace: Helena Lundberg Supervisor at
KTH: Kaye Stern
Student: Oscar Dalla-Santa
Date: 2019-07-01
Examiner: Kaye Stern
Foreword
I would like to thank Helena Lundberg for giving me the opportunity to do this scientific work at the department of organic chemistry at KTH. It has been a great learning experience and amongst the most enjoyable times I have had during my education. I would also like to thank Kaye Stern, my supervisor from the program, for always taking her time and being such a great support. I would also like to thank Giampiero Proietti and Björn Blomkvist for helping me find my way around the
laboratory.
Abstract
In this project the possibilities of using a metallocene catalyst for an aromatic esterification has been studied. The purpose of this scientific work was to study and develop a method with less environmental impact than the conventional for producing aromatic esters, as well as to minimize the use of hazardous chemicals. This method was then to be optimized for synthesizing an aromatic ester.
Benzyl alcohol and benzoic acid were used as model substrates throughout this project. They react to form the ester benzyl benzoate, a reaction that is slow. Several metallocene complexes were therefore tried as catalysts for this aromatic esterification. The reaction was also studied in different solvents and under different conditions, such as under inert atmosphere. High performance liquid chromatography (HPLC) was used to sample and study the results of these experiments, with the use of both an internal standard and reference standard.
The catalyst that was found to work best for the esterification was zirconocene triflate. The choice of solvent also greatly affects the formation of ester, with hexane found to have the highest positive influence on the yield. We chose to use toluene instead, since it is less toxic and works for a wider variety of substrates than hexane. Using twice the amount of alcohol compared to the acid was also found to be very effective for increasing the yield. The addition of a small amount of water to the mixture under nitrogen gas showed a positive impact on the yield of benzyl benzoate.
The esterification did not only form benzyl benzoate as a product but also formed dibenzyl ether and both 2-benzyl toluene and 4-benzyl toluene. These by-products are formed from the benzyl alcohol and they were identified with both 1H-NMR and 13C-NMR, as well as GC-MS.
The optimized reaction conditions were to use twice the amount of alcohol in comparison to the acid, with 2 mol% of zirconocene triflate. Half an equivalent of water is also added to the mixture in this method. Toluene is used as solvent and the reaction is conducted at 80℃, resulting in a 74% yield of benzyl benzoate.
Sammanfattning
Möjligheten att använda en metallocen som katalysator i en aromatisk förestring har undersökts i detta projekt. Syftet med detta vetenskapliga arbete var att studera och utveckla en metod med mindre miljömässig påverkan än de konventionella metoderna för att producera aromatiska estrar, samt för att minimera användandet av farliga kemikalier.
Bensylalkohol och bensoesyra har använts som modellsubstrat i detta projekt. De bildar tillsammans estern bensylbensoat genom en långsam reaktion. Flera metallocenkomplex prövades därför som katalysatorer till denna aromatiska förestring. Reaktionen studerades också i olika lösningsmedel och under olika betingelser, såsom under inert atmosfär. HPLC har använts för att ta prover och för att studera resultaten från dessa experiment, med användandet av både intern standard och referensstandard.
Den katalysator som fungerade bäst för förestringen var zirkonocentriflat. Valet av lösningsmedel påverkar starkt bildningen av ester, varvid hexan visade sig att ha störst positiv inverkan på utbytet.
Vi valde dock att använda toluen, eftersom det är mindre toxiskt och fungerar för fler substrat än hexan. Att använda dubbelt så mycket alkohol som syra, visade sig att vara väldigt effektivt för att få högre utbyte. Att addera en liten mängd vatten till reaktionsblandningen under kvävgas visade en positiv inverkan på utbytet av bensylbensoat.
Det bildades inte bara bensylbensoat i förestringen utan även dibensyleter, samt både 2-bensyltoluen och 4-bensyltoluen. Dessa biprodukter bildas från bensylalkoholen och de identifierades med både
1H-NMR and 13C-NMR, samt GC-MS.
De optimerade reaktionsbetingelserna är att använda dubbel mängd alkohol i förhållande till syra, med 2 mol% av zirkonocentriflat. I metoden tillsätts även en halv ekvivalent vatten till blandningen.
Toluen används som lösningsmedel och reaktionen sker vid 80℃, med ett utbyte av bensylbensoat på 74%.
Table of contents
List of abbreviations 1
1. Introduction 2
1.1 Purpose 2
1.2 The aim of the project 2
1.3 Methodology 2
2. Theoretical framework 3
2.1 Organic esters 3
2.1.1 The esterification process 3
2.2 Metallocene catalysts 4
2.3 Green chemistry 5
2.4 Reaction Progress Kinetic Analysis 6
3. Results and discussion 7
3.1 Catalysts 7
3.2 Solvents 8
3.3 The effects of an inert atmosphere and water 8
3.4 Reaction kinetics 9
3.4.1 Reaction progress kinetic analysis 9
3.4.2 Kinetics for reactions with added water under nitrogen gas 13
3.5 Optimized reaction 15
3.6 Identifying the by-products of the esterification 16
3.7 Proposed catalytic mechanism 17
4. Conclusion 19
5. Experimental 20
5.1 Standard procedure 20
5.2 Testing different reaction conditions 21
5.2.1 Catalysts 21
5.2.1.1 Synthesizing the fluorinated complexes. 21
5.2.2 Solvents and temperature 22
5.2.3 Reactions under inert and added water 22
5.3 Kinetic studies 23
5.4 Optimization 23
5.5 Large scale reaction without benzoic acid 24
5.5.1 Separation and identification 24
References 25
Appendix 1. Mass spectra 30
Appendix 2. NMR spectra 27
1
List of abbreviations
THF – Tetrahydrofuran
zirconocene triflate – bis(cyclopentadienyl)zirconium(IV) bis(trifluoromethanesulfonate) HPLC – High performance liquid chromatography
NMR – Nuclear magnetic resonance
GC-MS – Gas chromatography-mass spectrometry
Internal standard – abbreviated for the 4,4’-di-tert-butyl-biphenyl Cp – Cyclopentadienyl
OTf – triflate or trifluoromethanesulfonate (CF3SO3−) PFBS – Perfluorobutanesulfonate (C4F9SO3−)
PFHS – Perfluorohexanesulfonate (C6F13SO3−)
2
1. Introduction
The focus of this project was to catalytically synthesize an aromatic ester in an environmentally friendly way. The results from this scientific work could, depending on the outcome, be used by Helena Lundberg in further research within environmentally friendly synthesis. Helena works as a researcher at the Division of Organic Chemistry at the Department of Chemistry at KTH, Royal Institute of Technology.
1.1 Purpose
The overall purpose of this project was to study and develop a method for environmentally friendly organic synthesis of aromatic esters, in order to minimize the environmental impact of conventional methods as well as to reduce the use of hazardous chemicals.
1.2 The aim of the project
This project aimed to develop and optimize a method for synthesizing aromatic esters using an organometallic catalyst. This was done with respect to environmentally friendly organic synthesis. The kinetics for this catalyzed reaction were studied using High Performance Liquid Chromatography (HPLC), with the end-product also being analyzed with Nuclear Magnetic Resonance (NMR) and Mass Spectrometry (MS).
1.3 Methodology
The project started with a review of relevant scientific literature about aromatic esterification, metallocene catalysts and green chemistry. After gathering a decent amount of information about these subjects the experimental work begun. The first experiments were about testing different catalysts and then subsequentially moving forward by testing different solvents, temperatures and other conditions for the reaction. After developing an understanding for the reaction and the conditions that favor the catalyzed formation of ester, it was time to identify the products in the reaction mixture. This was done by separating the contents of the reaction mixture and analyzing the different compounds with 1H-NMR, 13C-NMR and GC-MS. The gained spectra were then compared with reference spectra to determine the compounds found.
3
2. Theoretical framework
2.1 Organic esters
Esters are a group of organic compounds with the formula R1COOR2, as seen in Figure 1. They are usually derived from a carboxylic acid and an alcohol that has undergone a condensation, with the loss of water. Esters are commonly found in nature in the forms of fats, fatty oils, waxes and fruity fragrances. The ones with lower molecular weight are very volatile and often have pleasant scents.
With increasing molecular weight, the volatility decreases and the esters start forming waxes or solids.
Esters with longer carbon chains are insoluble in water, but they are on the other hand soluble in many organic solvents. This goes with the exception of esters that are formed of short chained carboxylic acids and alcohols, that are miscible with water to some extent. (Riemenschneider and Bolt 2005) Organic esters are fine chemicals of great importance for the manufacturing of pharmaceuticals, plasticizers, food additives, emulsifiers, polyesters and more (Liu et al. 2006). The estimated annual production of polyesters is several million tons. The most important of these are polyethylene terephthalate for PET-bottles, acrylate esters for plastics and cellulose acetate for synthetic fibers and films. (Riemenschneider and Bolt 2005)
Figure 1. Ester with the formula R1COOR2.
2.1.1 The esterification process
One of the most common ways to form an ester is through the Fischer esterification process. In this process a carboxylic acid is mixed with an alcohol and the reaction is then catalyzed by the addition of a strong acid, such as sulfuric acid. This reaction will produce an ester and water. However, this reaction will only go towards an equilibrium between the product and the reactants, which can be seen in Figure 2. To shift the equilibrium towards the products one can either add more of a reactant, often the less expensive alcohol, or remove some of the product. Constantly removing some product can be achieved by boiling the mixture or by adding a water binding agent such as calcium hydride or molecular sieves. The esterification process is usually quite slow and the formation of aromatic esters is even slower. (Riemenschneider and Bolt 2005)
The conventional catalysts for esterification are usually sulfuric acid, hydrofluoric acid, hydrochloric acid and phosphoric acid (Sadanandan and Bhaskaran 2014). These are often used in large amounts and the remaining acidic waste must be neutralized before it can be disposed of into the environment (Liao et al. 2011). It is therefore of great importance to find environmentally friendly catalysts for ester
4
synthesis. These catalysts need to be good Lewis acids in order to be able to activate the carboxylic acid, so that the carbonyl carbon becomes more susceptible to a nucleophilic attack (Barbosa et al.
2006). The oxygen atom of the alcohol will then attack the partially positive carbonyl carbon of the carboxylic acid to produce an ester and water. (The) water becomes the leaving group after a proton shift and it consists of the hydroxyl group of the carboxylic acid, as well as a proton of the alcohol (McNaught and Wilkinson 1997).
Figure 2. Reversible formation of esters.
2.2 Metallocene catalysts
A metallocene compound is an organometallic complex that consists of at least one metal ion and two cyclopentadienyl anions (C5H5−). These cyclopentadienyl rings are often abbreviated as Cp and they have a conjugated system of p-orbitals where the electrons are delocalized. It is with all conjugated p- orbitals that the cyclopentadienyl anions bind to the metal ion’s d-orbitals. In these complexes the metal ion has an oxidation state of +II and the complexes are often written as [MCp2] (Chirik 2010).
Many of the metallocenes are non-hazardous and commercially available, making them excellent catalysts (Khanapure et al. 2016).
For metallocenes with higher states of oxidation, additional ligands will bind to the metal ion. Very common are the tetravalent metallocenes that have two bound cyclopentadienyl anions as well as two other ligands. They are often written as [Cp2M X2]. These complexes exhibit different catalytic properties depending on what ligands are bound. (Hays and Hanusa 1996)
A common and stable complex is the dichloride metallocene [Cp2M Cl2], which can be seen in Figure 3. It possesses a catalytic activity since it is a weak Lewis acid that can form cationic species, which in turn are very reactive. The dichloride complexes of especially group 4 transition metals are often used as catalysts for the acetylation of phenols, alcohols and amines, using acetic anhydride (Kantam et al.
2006). They are also often used as catalysts for olefin polymerization (Hays and Hanusa 1996). The Lewis acidity of the complex can be increased if the two chloride ligands are substituted with for example two trifluoromethanesulfonate (triflate) groups. This leads to a much stronger electron- drawing capability of the metal ion, which can activate functional groups so that they become more susceptible towards a nucleophilic attack. (Zhang et al. 2012)
5
Figure 3. Structure of a tetravalent metallocene dichloride complex (Khanapure et al. 2016).
Metallocenes with a higher Lewis acidity have therefore been getting more attention in organic synthesis recently. The search for air- and water-stable complexes with strong acidity is highly prioritized since the catalytic activity increases with acidity. When working with metallocenes that are not air- and water-stable, one should handle the complexes under anhydrous conditions or consider to perform the reaction under an inert atmosphere. (Qiu et al. 2009)
2.3 Green chemistry
The concept of green chemistry is described as environmentally friendly chemical synthesis. This area of chemistry and chemical engineering focuses on developing products and processes that minimizes or eliminates the use and generation of hazardous substances. Green chemistry encourages the use of technological advances to prevent pollution and to reduce the necessity of nonrenewable resources, in order to minimize the environmental impact of chemical processes. The overall goals of green chemistry are thus to use resources more efficiently and to incorporate safer production processes. These goals have been turned into the following twelve basic principles in order to promote the practice of green chemistry (Anastas and Warner 1998):
• Prevention of waste. It is better to prevent the generation of waste rather than treating or cleaning up waste material.
• Atom economy. A synthetic method should use all the materials in the process to produce the final product. This will generate less waste and more product.
• Less hazardous chemical synthesis. The generation or use of toxic substances should be avoided in synthetic methods.
• Designing safer chemicals. Chemical products should be designed so that they fulfill their desired function while being as non-toxic as possible.
• Safer solvents and auxiliaries. The use of auxiliary substances should be avoided if possible, otherwise it should be as non-hazardous as possible.
• Design for energy efficiency. The energy requirements for a chemical process should be minimized and if possible, the process should be conducted at room-temperature and atmospheric pressure.
• Use of renewable feedstocks. The use of renewable feedstocks and materials is preferred over non-renewable ones if it is practically possible.
• Reduce derivatives. The generation of chemical derivatives should be minimized or avoided if possible as such steps require additional reagents and produce additional waste.
• Catalysis. Small amounts of a catalytic reagent should be used to increase the yield of a chemical reaction instead of adding more stoichiometric reagents.
• Design for degradation. When a chemical product has fulfilled its function, it should break down into non-harmful substances that do not pollute the environment.
6
• Real-time analysis for pollution prevention. In-process monitoring and control should be developed to prevent the production of hazardous substances.
• Inherently safer chemistry for accident prevention. Chemicals in a process should be chosen to minimize the risk of explosions, fires and accidental releases.
2.4 Reaction Progress Kinetic Analysis
Reaction progress kinetic analysis is a methodology that helps understand the driving forces of a chemical reaction. It enables the analysis to be conducted through a minimal number of experiments that may also give a better understanding of the reaction mechanism. These experiments are carried out under the same conditions that the reaction is typically performed, with minimal manipulations.
This process is significantly faster than the traditional and classical kinetic measuring techniques, as well as providing information about the reaction mechanism. (Blackmond 2005)
A method for acquiring accurate and continuous experimental data and a computational software for managing the collected data is needed to conduct the reaction progress kinetic analysis. The use of reference standards and/or internal standards should be considered depending on the method of analysis, to ensure that the error in the collected experimental data is kept to a minimum. (Blackmond 2005)
In this work, the concentration of reactants or formed product is plotted on the y-axis as a function of time on the x-axis. The rate of the reaction is then represented by the slope in the diagram. However, there are plenty of other ways to graphically analyze collected data and depending on the reaction, it might even be necessary to use another method (Blackmond 2005). By running a set of reactions where the concentration of the reactants and catalyst is increased compared to standard conditions while keeping the other parameters constant, the reaction component that has the highest influence on the reaction rate and conversion can be identified. This is done by plotting the collected data from each of the experiments on the y-axis against time on the x-axis. A steeper slope compared to standard conditions indicates that the reaction is positive order in the concentration of the component that was increased in that particular experiment. This information can in turn be used in order to optimize the reaction and to study the mechanism. By determining which components have a positive effect on the reaction rate, one can draw the conclusion that the rate determining step is correlated to that/those component(s) (Blackmond 2005).
7
3. Results and discussion
3.1 Catalysts
Catalysts based on zirconium, hafnium, niobium and titanium were studied under the same conditions. All reactions had a starting concentration of 1 M of benzoic acid and benzyl alcohol, along with a concentration of 0.02 M of a catalyst. The reactions were carried out in 1 ml of dried THF at 65℃. The amount of benzyl benzoate was analyzed with HPLC after either 18 or 24 hours. The metallocene complex that showed the greatest catalytic activity was the zirconocene triflate, with a yield of roughly 10% benzyl benzoate. To some catalysts molecular sieves (3Å) were added, they are denoted with “MS” in Figure 4 below. The addition of molecular sieves resulted in lower yields for the triflate catalysts, but increased the yields for the catalysts with chloride ligands. This could mean that water has a significant role in the catalytic function of the triflate complexes. The dichloride complexes are however sensitive to water, which is why their catalytic properties increased with the addition of molecular sieves. The perfluorinated hexane sulfonate and butane sulfonate catalysts showed little to no activity, probably because we had trouble forming the right complexes. Due to the timeframe we therefore went forward with the commercially available zirconocene triflate catalyst in order to optimize the reaction.
Figure 4. Yields of benzyl benzoate using different metallocene catalysts.
Concentration of 1 M for benzoic acid and benzyl alcohol, along with 0.02 M of a catalyst. The temperature was kept at 65℃ and the solvent was dry THF.
01 23 45 67 89 10
Yield benzyl benzoate (%)
Catalyst
Yields using different metallocene catalysts
8
3.2 Solvents
When studying the reaction in various dry solvents, a different reaction procedure was used. The temperature was changed to 80℃ for all experiments, except for the reaction in diethyl ether and one reaction in toluene where the temperature was set to 30℃ (table 1, entries 3 and 4). The concentrations of benzoic acid and benzyl alcohol remained at 1 M, along with the concentration of zirconocene triflate remaining at 0.02 M for all experiments. Samples were taken after 24 hours and analyzed using HPLC. A correlation between nonpolar solvents and higher yields can be seen in table 1, where toluene and hexane give significantly higher yields at 80℃. This could be because the more polar compounds in the reaction mixture, such as the alcohol and acid, are closer to each other in the non-polar solvent. The higher yields when using non-polar solvents could also depend on the ethers being polar solvents that might coordinate to the catalyst to some extent , which in turn could sterically hinder the benzoic acid from coordinating. The attempt at using a lower temperature in toluene resulted in a significantly lower yield. Using hexane as a solvent gave the highest yield of 64%
benzyl benzoate, but due to its toxicity we decided to continue with toluene.
Table 1. Yields of benzyl benzoate using different solvents, measured with HPLC.
Entry Solvent Yield of benzyl benzoate (%)
1 Tetrahydrofuran 18
2 1,4-Dioxane 16
3 Diethyl ether 30℃ 2
4 Toluene 30℃ 4
5 Toluene 40
6 Hexane 64
Concentrations of benzyl alcohol and benzoic acid were 1 M, the concentration of zirconocene triflate was 0.02 M. The temperature was set to 80℃ except for the experiments that mentions a temperature of 30℃. Samples were taken after 24 hours.
3.3 The effects of an inert atmosphere and water
To study the significance of water, different amounts of distilled and purified milli-Q water was added to the reaction mixtures. The experiments were carried out at 80℃ in 1 ml of dried toluene under a nitrogen atmosphere. The concentration of benzoic acid and benzyl alcohol were kept at 1 M, along with the concentration of the catalyst remaining at 0.02 M. To different reaction mixtures were then added 0.5 mmol (0.5 equivalents), 1 mmol (1 equivalent) or 5 mmol (5 equivalents) of distilled and purified water before heating it for 24 hours under a nitrogen atmosphere. One experiment was conducted under nitrogen gas without the addition of water, to study whether the yield would change compared to the standard reaction. Another experiment was carried out under normal atmosphere with the addition of molecular sieves (3Å) to a mixture with the same concentrations as the standard.
9
In table 2 the yields of the different experiments are shown next to the standard reaction in toluene that was run under normal atmosphere (table 2, entry 1). The difference between normal and inert atmosphere is very small but having 0.5 mmol of purified and distilled water gave a higher yield after 24 hours (table 2, entry 3). However, adding 5 mmol of water to the reaction mixture resulted in a very low yield, probably due to the degeneration or inhibition of the catalytic complex (table 2, entry 5). The yield of benzyl benzoate when adding molecular sieves was also a lot lower than the standard reaction (table 2, entry 6). This could mean that water is significant in the catalytic function of the zirconocene triflate complex.
Table 2. The yields of benzyl benzoate for experiments under inert with added water or sieves (3Å), measured with HPLC.
Entry Experiment Yield of benzyl benzoate (%)
1 Standard conditions 40
2 Standard conditions under inert 41
3 Inert + 0.5 mmol water 48
4 Inert + 1 mmol water 41
5 Inert + 5 mmol water <1
6 Standard conditions + sieves 16
The concentration of benzyl alcohol and benzoic acid was 1 M, with the concentration of zirconocene triflate being 0.02 M. Th e solvent was dry toluene and the temperature was set to 80℃, with samples being taken after 24 hours.
3.4 Reaction kinetics
3.4.1 Reaction progress kinetic analysis
The kinetics for these reactions were studied with samples being taken at the start, then every hour for 6 hours and lastly after 24 hours. The concentration of benzyl benzoate was then plotted as a function of time, thus generating a slope corresponding to the rate of the reaction. The impact of doubling the concentration of either benzoic acid, benzyl alcohol or catalyst was studied while keeping the other concentrations the same as in the standard procedure. The reactions were carried out in 1 ml of dried toluene at 80℃. The results of these experiments can be seen in Figure 5 next to the standard reaction. Surprisingly, the formation of benzyl benzoate is very close to the standard reaction for all the other experiments. One would expect that at least the catalyst would have a greater impact on the yield and rate of the reaction. More catalyst actually gave less ester over the course of 24 hours as compared to the standard reaction, even though the reaction rate did increase a bit. Doubling the amount of benzyl alcohol or benzoic acid did however give a yield of 67% and 53% respectively.
10
Strangely enough, the increased concentration of benzoic acid, benzyl alcohol and the zirconocene triflate all seem to influence the rate of the reaction in a very similar positive order.
Figure 5. Concentration of benzyl benzoate as a function of time.
The concentrations in the standard reaction is 1 M for benzyl alcohol and benzoic acid, along with 0.02 M of zirconocene triflate. For the other experiments one concentration was doubled. The temperature was set to
80℃ and the solvent was dry toluene. Samples were taken every hour.
In Figure 6, the concentration of benzyl alcohol is plotted against time. The experiment with twice the concentration of benzyl alcohol is adjusted to start from 1 M for easier comparison with the other experiments. The declining concentration of alcohol is not proportional to the amount of benzyl benzoate that has formed for these experiments. This has to do with the formation of by-products in the reacting mixture. These by-products consume alcohol as that is the only concentration that declines at another rate than the rate of formation of benzyl benzoate. This is discussed in more detail in section 3.6.
Yield after 24 hours 2*[acid] 53%
2*[alcohol] 67%
2*[catalyst] 37%
Standard 40%
11
Figure 6. Concentration of benzyl alcohol as a function of time.
The concentrations in the standard reaction is 1 M for benzyl alcohol and benzoic acid, along with 0.02 M of zirconocene triflate. For the other experiments one concentration was doubled. The temperature was set to
80℃ and the solvent was dry toluene. Samples were taken every hour.
To further study which factors influence the rate of reaction for the formation of benzyl benzoate a diagram where the concentration of benzoic acid is plotted against time was drawn. In Figure 7, the experiment with the double amount of benzoic acid has its concentration adjusted by subtracting 1 M from the measured concentration to enable easier comparison with other experiments and the standard reaction. Figure 7 shows that the concentration of benzyl alcohol is in fact the more influential factor for the esterification. Therefore, the rate determining step is most likely the step when the benzyl alcohol performs a nucleophilic attack on a benzoic acid molecule that is coordinated to the catalyst. For these experiments, there is no clear first order reaction for any of the reactants or the catalyst, doubling the concentration does not double the rate of the reaction. This observation could be explained by the formation of by-products and the possibility that the catalyst exists in several forms in the solution, where one or more forms are active in the esterification process and some are not.
Yield after 24 hours 2*[acid] 53%
2*[alcohol] 67%
2*[catalyst] 37%
Standard 40%
12
Figure 7. Concentration of benzoic acid as a function of time.
The concentrations in the standard reaction is 1 M for benzyl alcohol and benzoic acid, along with 0.02 M of zirconocene triflate. For the other experiments the concentration of one of the reagents or the catalyst was doubled. The temperature was set to 80℃ and the solvent was dry toluene. Samples were taken every hour.
The influence of the alcohol becomes clearer in Figure 8 below, were the different lines have been enlarged in order to confirm that it is in fact a significant difference between them.
Figure 8. Concentration of benzoic acid as a function of time, enlarged.
The concentrations in the standard reaction is 1 M for benzyl alcohol and benzoic acid, along with 0.02 M of zirconocene triflate. For the other experiments the concentration of one of the reagents or the catalyst was doubled. The temperature was set to 80℃ and the solvent was dry toluene. Samples were taken every hour.
0,5 0,6 0,7 0,8 0,9 1
0 1 2 3 4 5 6
Benzoic acid [M]
Hours
Concentration of benzoic acid over time - RPKA
Adjusted Benzoic acid Benzyl alcohol 2 M Catalyst 0.04 M Standard reaction Yield after 24 hours
2*[acid] 53%
2*[alcohol] 67%
2*[catalyst] 37%
Standard 40%
Yield after 24 hours 2*[acid] 53%
2*[alcohol] 67%
2*[catalyst] 37%
Standard 40%
13
To further study the kinetics of the catalyst, some experiments with different concentrations of catalyst were performed. The studied concentrations were 0.01 M, 0.02 M (which is the standard reaction) and 0.04 M (which is the RPKA experiment from above). In Figure 9 the concentration of ester has been plotted over time for each of these experiments. There we can see that the esterification is of positive order for the catalyst concentration, since the rate increases with increasing concentration.
Figure 9. Concentration of benzyl benzoate as a function of time.
The concentrations in the standard reaction is 1 M for benzyl alcohol and benzoic acid, along with 0.02 M of zirconocene triflate. The other experiments have twice or half the concentration of catalyst. The temperature
was set to 80℃ and the solvent was dry toluene. Samples were taken every hour for six hours and then after 24 hours.
3.4.2 Kinetics for reactions with added water under nitrogen gas
The formation of benzyl benzoate over time is plotted for the experiments that were carried out under inert atmosphere and with the addition of water. The concentrations of benzyl alcohol and benzoic acid were 1 M, with the zirconocene triflate holding a concentration of 0.02 M. The temperature was set to 80℃ and the solvent was dry toluene. Samples were taken at the start and then every hour for 6 hours, with the final sample being taken after 24 hours. In Figure 10, the concentration of benzyl benzoate is plotted against time for the standard reaction along with the reactions under nitrogen gas. One of the mixtures under inert atmosphere had no added water, while two of the mixtures had been added 0.5 mmol and 1 mmol of water respectively. The standard reaction and the corresponding reaction under inert atmosphere without added water had a very similar rate of reaction and yield.
However, when adding 0.5 equivalents (0.5 mmol) of water under inert conditions the rate of the reaction decreased while the overall yield of benzyl benzoate increased. After 24 hours a yield of 40%
benzyl benzoate was seen for the standard reaction, while the experiment with 0.5 equivalents of added water gave a yield of 48%. This could mean that the water coordinates to the catalyst, inhibiting its catalytic activity somewhat, but also affecting the catalyzed formation of by-products.
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1
0 5 10 15 20
Benzyl benzoate [M]
Hours
Concentration of benzyl benzoate over time - further kinetic studies
Catalyst 0.04 M Standard reaction Catalyst 0.01 M
14
Figure 10. Concentration of benzyl benzoate as a function of time.
The concentration was 1 M for benzyl alcohol and benzoic acid, along with 0.02 M of zirconocene triflate. The temperature was set to 80℃ and the solvent was dry toluene. Samples were taken every hour.
In Figure 11, the concentration of benzyl alcohol is plotted against time for the same experiments.
Roughly twice the amount of benzyl alcohol remains in the mixture that was added 0.5 mmol of water, compared to the standard reaction after 6 hours. This shows that the formation of ester and by- products are lowered by adding water, with the by-products being affected the most. This over the course of 24 hours leads to a higher yield of benzyl benzoate.
Figure 11. Concentration of benzyl alcohol as a function of time.
Yield after 24 hours Standard 40%
Inert 41%
Inert + 0.5 mmol water 48%
Inert + 1 mmol water 41%
Yield after 24 hours Standard 40%
Inert 41%
Inert + 0.5 mmol water 48%
Inert + 1 mmol water 41%
15
The concentration was 1 M for benzyl alcohol and benzoic acid, along with 0.02 M of zirconocene triflate. The temperature was set to 80℃ and the solvent was dry toluene. Samples were taken every hour.
3.5 Optimized reaction
From the experiments conducted in this work, a higher yield of benzyl benzoate was seen when using twice the amount of benzyl alcohol compared to benzoic acid. An increase in the yield of ester was also seen when using inert atmosphere and adding 0.5 mmol (0.5 equivalents) of water. When combining these factors, we get an optimized reaction where the yield of benzyl benzoate is significantly higher than the standard reaction. In the optimized reaction the concentration of benzyl alcohol is 2 M, the concentration of benzoic acid is 1 M and the concentration of zirconocene triflate is 0.02 M. The components were dissolved/diluted in 1 ml of dried toluene and the temperature was set to 80℃, samples were taken every hour for 6 hours. The rate of reaction for the esterification is lower than in the standard reaction, but over the course of 24 hours the final yield is higher. In Figure 12 the concentration of benzyl benzoate is plotted against time. The final yield after 24 hours is roughly 74% with respect to benzoic acid as compared to 40% under standard conditions.
Figure 12. Concentration of benzyl benzoate as a function of time.
The reactions were studied at 80℃ in 1 ml of dried toluene with samples being taken every hour. Starting concentrations were set to 1 M of benzoic acid and 2 M of benzyl alcohol, along with 0.02 M of zirconocene
triflate.
The concentration of benzyl alcohol does not decrease in the same way that it did in the standard reaction, nor like it did during the experiment with 2 M of alcohol, see Figure 13. Since less alcohol is used to form by-products, more benzyl benzoate can form over time which leads to an increase in yield for the ester even though the rate is lower.
Yield after 24 hours Standard 40%
Optimized reaction 74%
16
Figure 13. Concentration of benzyl alcohol as a function of time.
Reactions studied at 80℃ in 1 ml of dried toluene with samples being taken every hour. Concentrations of benzyl alcohol adjusted to start at 1 M, except for the standard reaction.
3.6 Identifying the by-products of the esterification
The by-products were first seen in the UV-spectra of the HPLC, where two large peaks could be seen eluating after the benzyl benzoate ester. These signals varied a lot in size, but after studying the concentrations it was clear that they originated from the benzyl alcohol. Without the catalyst, no by- products formed and with the addition of water, the rate of formation was significantly lower. A possible explanation for this could be that the catalytic function is inhibited by coordinating water molecules, which affects the formation of by-products more than the formation of the ester. The method of separation and identification is described in detail in section 5.5.1 in the experimental part of this report.
The first by-product has a 1H- and 13C-NMR spectra that matches with the reference spectra of dibenzyl ether. These NMR spectra can be seen in appendix 1 and they are a very close match with the reference spectra for this compound (National Center for Biotechnology Information). Dibenzyl ether was also a suggested molecule when analyzing the reaction mixture with GC-MS. The probability that this was the correct molecular structure was 78% according to the mass spectrometer’s library search.
The mass spectra can be seen in appendix 2.
The second by-product showed a lot of signals in both the 1H- and 13C-NMR spectra. This is due to the second compound actually being a mixture of two isomers. In the 1H- and 13C-NMR spectra the two compounds match the reference spectra of 2-benzyl toluene and 4-benzyl toluene (Srimani et al.
2010). These are two regioisomeric compounds that were in a ratio of roughly 1:3. The NMR spectra can be seen in appendix 2 and they match quite well. The compound 2-benzyl toluene did show up on the GC-MS, with a probability of 18%. This spectrum can be seen in appendix 1 and the low probability could be explained by the fact that both of these regioisomeric compounds have very similar retention times in the GC, which certainly would make the search function less correct.
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2
0 1 2 3 4 5 6
Benzyl alcohol [M]
Hours
Concentration of benzyl alcohol over time - optimized reaction
Adjusted Optimized reaction
Adjusted Benzyl alcohol 2 M Benzyl alcohol 2 M
Standard reaction
Optimized
Yield after 24 hours Standard 40%
Optimized reaction 74%
2*[alcohol] 67%
17
The isolated by-products were used to calibrate a reference standard so that measurements could be made with HPLC. Using this reference standard, the concentrations could be calculated for all experiments for the dibenzyl ether and the 2-/4-benzyl toluene mixture. The yields of these compounds after 24 hours can be seen below in table 3. The yields are calculated with respect to the limiting reagent which is benzyl alcohol. The experiments listed here were all carried out in 1 ml dried toluene respectively, at 80℃.
Table 3. Calibrated amounts of products after 24 hours for some experiments, with the limiting factor set as alcohol.
Concentration of reactants (M)
Dibenzyl ether
2-/4-benzyl toluene
Benzyl benzoate Experiment Benzyl
alcohol
Benzoic acid
Catalyst Conc.
[M]
Yield
%
Conc.
[M]
Yield %
Conc.
[M]
Yield %
Standard reaction 1 1 0.02 0.21 42 0.18 18 0.40 40
Benzyl alcohol 2 M 2 1 0.02 0.53 53 0.28 14 0.67 33
Benzoic acid 2 M 1 2 0.02 0.14 28 0.16 16 0.53 53
Catalyst 0.04 M 1 1 0.04 0.22 44 0.19 19 0.37 37
Inert + 0.5 mmol
H2O 1 1 0.02 0.20 40 0.13 13 0.48 48
No acid 1 0 0.02 0.31 62 0.07 7 0 0
Optimized reaction 2 1 0.02 0.53 53 0.14 7 0.74 37
3.7 Proposed catalytic mechanism
The proposed catalytic mechanism for the esterification can be seen in Figure 14 below. The cycle starts with a benzoic acid molecule coordinating to the catalytic complex. The carbonyl group is then activated, which makes the carbonyl carbon more susceptible towards nucleophilic substitution from a nearby nucleophile, such as the benzyl alcohol. An intermediate state is formed, where the benzyl alcohol is bound to the benzoic acid. The rate determining step is proposed to be the part where the alcohol reacts with the acid and the ester forms, which is supported by the results of the RPKA- experiments. The produced ester and water finally let go of the catalytic complex, which allows the cycle to start over.
18
Figure 14. Proposed catalytic mechanism for the esterification
19
4. Conclusion
The catalysts with high Lewis acidity worked best for synthesizing benzyl benzoate, with zirconocene triflate showing the highest yield. The non-polar solvents that were tested showed that the more non- polar the solvent, the more ester was produced. It was also showed that having twice the amount of benzyl alcohol compared to the acid, gave more of the desired product. The addition of water under an inert atmosphere slows down the rate of the reaction, but the final yield after 24 hours is higher.
When putting together these findings into an optimized reaction, the yield of benzyl benzoate reaches 74% and the amount of by-products gets minimized to some extent.
Benzyl benzoate was confirmed to be synthesized with the use of HPLC and GC-MS. Since a known sample of the ester was used to make the reference standard, the retention time was also noted in the process which in turn confirmed its presence. The by-products from this reaction have been identified with 1H-NMR, 13C-NMR and GC-MS. They also match the retention times seen in the HPLC throughout the project. These products are dibenzyl ether and 2-/4-benzyl toluene. The formation of these products is catalyzed by the zirconocene triflate complex. The catalyst can activate both the benzoic acid and benzyl alcohol, which gives rise to different products.
20
5. Experimental
5.1 Standard procedure
A model reaction was set as a starting point for carrying out the different experiments. In this model the benzoic acid and benzyl alcohol both had concentrations of 1 M. The internal standard that was used in the reactions had a concentration of 0.05 M and the catalyst had a concentration of 0.02 M.
The volume for this model was set to be 1 ml. The internal standard that was used for the entirety of the project was 4,4’-di-tert-butylbiphenyl. It was chosen as an internal standard since it is nonpolar and bulky, which made it a good internal standard for our HPLC analysis.
The benzoic acid and internal standard were weighed and inserted into a dried 10 ml volumetric flask.
The amounts inserted of acid and standard were 1.221 grams (10 mmol) and 0.1332 grams (0.5 mmol) respectively. The flask was then filled with a dry solvent to produce a mixture containing 1 M of acid and 0.05 M of standard. Depending on the solvent the flask sometimes needed to be heated to dissolve the solids. 1 ml of this solution was then added to a dried 4 ml glass vial along with a teflon stirring bar. Of the catalyst, 0.02 mmol was added to the vial before sealing it with teflon-tape and attaching a septum cap. Lastly, 0.103 ml (1 mmol) of benzyl alcohol was added through the septum.
The glass vial was then placed in an oil bath at 65℃ with samples usually being taken after 24 hours.
Samples of 20 µl were taken from the sealed glass vial through the septum cap and placed in a 0.5 ml filter vial. The filter vials contained 0.45 ml of HPLC grade acetonitrile and 0.05 ml of distilled water, in which the 20 µl were diluted to have a concentration no higher than 0.04 M of either reactants or product. These kinds of filter vials were used as the reaction mixtures produce solid particles over time, which may clog the HPLC column.
The HPLC setup that was used in this project consists of a C18 column as stationary phase with a mixture of acetonitrile and water as the mobile phase. The ratio between acetonitrile and water started at 40% acetonitrile and 60% milli-Q purified water, before successively increasing to 95%
acetonitrile and 5% water. The increase in acetonitrile started after 2 minutes and took roughly 5 minutes to reach its peak. After a few minutes the amount of acetonitrile decreased and slowly went back to the starting ratio. The volume of mobile phase was set to a constant 1 ml per minute. The detector that was used for the HPLC was a UV-detector set to a wavelength of 265 nm and 255 nm.
The conjugated compounds in this project were very good at absorbing the UV-light and gave strong signals. Furthermore, in this project the use of both an internal standard as well as a reference standard for the HPLC was used. The internal standard was used specifically to correct for the error in dilutions and injection volumes, while the reference standard made it possible to calculate concentrations. The reference standard was made through a series of created standard solutions for the benzoic acid, benzyl alcohol, benzyl benzoate and the internal standard.
21
5.2 Testing different reaction conditions
5.2.1 Catalysts
The same approach as used in the model reaction was used in order to test the different catalysts. The concentrations in these reactions were set to 1 M of benzoic acid and benzyl alcohol, 0.05 M of internal standard and 0.02 M of catalyst. Dried tetrahydrofuran was used as solvent for these experiments and the temperature was kept at 65℃. Samples of 20 µl were taken and diluted to 0.5 ml, like in the standard procedure, before being analyzed with HPLC after 18 or 24 hours. A list of the tested catalysts is shown in table 4, along with how they were obtained. How some of the catalysts were made is explained below the table.
Table 4. Catalysts and how they were obtained
Catalyst Formula How the complex was obtained
Zirconocene dichloride [Cp2ZrCl2] Commercially bought Zirconocene triflate [Cp2Zr(OTf)2] Commercially bought
Zirconocene PFBS [Cp2Zr(PFBS)2] Synthesized from dichloride complex Zirconocene PFHS [Cp2Zr(PFHS)2] Synthesized from dichloride complex Hafnocene dichloride [Cp2HfCl2] Commercially bought
Hafnocene triflate [Cp2Hf(OTf)2] Synthesized from dichloride complex Niobocene dichloride [Cp2NbCl2] Commercially bought
Niobocene triflate [Cp2Nb(OTf)2] Synthesized from dichloride complex Titanocene triflate [Cp2Ti(OTf)2] Commercially bought
5.2.1.1 Synthesizing the fluorinated complexes.
The triflate complexes of hafnium and niobium were synthesized from their respective dichloride complexes. This was done by taking 0.25 mmol (1 equivalent, 0.0949 grams) of hafnocene dichloride or 0.25 mmol (1 equivalent, 0.0735 grams) of niobocene dichloride and dissolving it in a volumetric flask with 5 ml of dried THF. To another 5 ml volumetric flask 0.5 mmol (2 equivalents, 0.1285 grams) of silver triflate was added, before it being filled with dried THF. The two solutions were mixed and stirred in the dark, at room temperature, for one hour. The precipitated silver chloride was filtered off and washed with 2 ml of dried THF. The filtrate was mixed with 10 ml of hexane and put in the freezer for 24 hours. The formed precipitate was then filtered and vacuum dried, with a yield of roughly 50%.
(Tang et al. 2017)
An attempt at synthesizing the complexes with perfluorobutane- and perfluorohexane sulfonates from the hafnocene- and niobocene dichloride complexes was also made after reading an article from a group that had a slightly different starting point. These fluorinated compounds did not come in the form of a silver salt, but in the form of a potassium salt. So, the first step was to try and replace the potassium with silver. This was done by inserting 1 mmol (1 equivalent, 0.3382 grams) of perfluorobutanesulfonate or 1 mmol (1 equivalent, 0.4382 grams) of perfluorohexanesulfonate into a glass vial with a stirring bar and 0.75 ml of distilled water. While the potassium salt was stirring, 0.55 mmol (0.55 equivalents, 0.1275 grams) of silver(I)oxide (Ag2O) was added slowly. The mixture was left to stir in the dark at room temperature for one hour. The precipitate was then filtered and washed
22
with 0.25 ml of distilled water, before being vacuum dried. The bubbly precipitate was then dissolved in dried acetone, with some sodium sulfate to remove the last bit of water. The mixture was then filtered, and the filtrate was dried using a rotary evaporator. A rubbery solid was collected in roughly 15% yield. This solid was dissolved in 1.3 ml of THF along with 0.04 mmol (0.04 equivalents, 0.0151 grams) of zirconocene dichloride and left to stir for one hour in the dark at room temperature. The precipitate was then filtered off and washed with 1 ml of dried THF, with the filtrate being dried and tested as a catalyst in the esterification. (Whitesides and Gutowski 1976)
5.2.2 Solvents and temperature
The standard procedure was followed when testing the different solvents, except for the temperature that was set to either 30℃ or 80℃, instead of 65℃. The concentration was 1 M of benzoic acid and benzyl alcohol, along with 0.02 M of zirconocene triflate. The experiments used 1 ml of dried solvent each. Samples were taken after 24 hours and analyzed with HPLC to study the yield of benzyl benzoate.
See table 5 for the different experiments and their temperature.
Table 5. Tested solvents at 30℃ or 80℃.
Entry Solvent Temperature 1 Tetrahydrofuran 80℃
2 1,4-Dioxane 80℃
3 Diethyl ether 30℃
4 Toluene 30℃
5 Toluene 80℃
6 Hexane 80℃
5.2.3 Reactions under inert atmosphere and with added water
To test the effects of added water and inert conditions the experiments were conducted in 1 ml of dry toluene at 80℃. Samples were taken at the start and then every hour for 6 hours, and then after 24 hours. The concentrations of the reactants were still 1 M and the concentration of zirconocene triflate remained at 0.02 M. The water used in these experiments was milli-Q ultrapure water that was bubbled with nitrogen gas for roughly 20-30 minutes to remove dissolved oxygen.
Firstly, the solids were weighed up in amounts according to the standard procedure. They were then placed in a glass vial that was sealed with teflon-tape under the septum cap, so they were not dissolved in the solvent this time. The glass vial was then attached to a vacuum pump by inserting a needle through the septum that was connected to the pump. The air in the vial was then removed by the pump and replaced with nitrogen gas. This process was repeated several times. To the vial was then added 1 ml of dry toluene, 1 mmol (1 equivalent, 0.103 ml) of benzyl alcohol and lastly the amount of water for that experiment. The amount of water added to each of these vials can be seen in table 6 below.
