Pd/S,O-ligand Catalysed Regioselective C–H Olefination of Anisole Derivates

(1)

1

Bachelor Thesis Scheikunde

Pd/S,O-ligand Catalysed Regioselective C–H

Olefination of Anisole Derivates

door

Rianne van Diest

14 december 2020

Studentnummer

11677635

Onderzoeksinstituut

Verantwoordelijk docent

Van ’t Hoff Institute for Molecular Sciences

Dr. M.A. (Tati) Fernández Ibáñez

Onderzoeksgroep

Begeleider

(2)

2

LIST OF ABBREVIATIONS

Ac-Leu-OH = N-acetyl-L-leucine

AcOH = acetic acid

AMLA = ambiphilic metal-ligand activation

Ar = aryl

BDE = bond dissociation energy

BIES = base-assisted internal electrophilic substitution

BM = σ-bond metathesis cHex = cyclohexane CMD = concerted metalation-deprotonation DCE = dichloroethane DCM = dichloromethane DMF = N,N-dimethylformamide EA = electrophilic addition

EDG = electron donating group

EI = electron ionization

Eq = equivalent

EtOAc = ethyl acetate

EWG = electron withdrawing group

FG = functional group

FI = field ionization

FTIR = fourier transform infrared

Hep = heptane

Hex = hexane

HFIP = hexafluoroisopropanol

HRMS = high resolution mass spectrometry KIE = kinetic isotope effect

M = metal

Me = methyl

MeOH = methanol

NMR = nuclear magnetic resonance

OA = oxidative addition OAc = acetate on. = overnight PPh3 = triphenylphosphine RDS = rate-determining step RF = retardation factor Rt. = room temperature S,O- = sulfur,oxygen-

tAmOH = tert-amyl alcohol (2-methyl-2-butanol)

tBu = tert-butyl

TFA = trifluoroacetic acid

TFE = trifluoroethanol

THF = tetrahydrofuran

TLC = thin layer chromatograpy

(4)

4

ABSTRACT

A general methodology for para-selective C–H olefination of ortho-substituted anisole derivates is described. The reaction proceeds via the Fujiwara-Moritani catalytic cycle and can be performed under mild conditions. It makes use of a S,O-ligand, which improves both reactivity and regioselectivity of the reaction. Various anisole derivates have been screened, with donating as well as electron-withdrawing side groups on the ortho position. The optimal reaction conditions differ for anisole derivates bearing electron donating or withdrawing substituents to ensure high yield and selectivity. This research is a step towards a general methodology for para-selective C–H olefination of all kinds of anisole derivates.

Scheme 1: General procedure for regioselective C–H activation with anisole derivates.

POPULAR SCIENTIFIC SUMMARY

Natural compounds are regularly used, for all kinds of applications. For example, many medicines are isolated from plants. However, this isolation process is not always very sustainable and often costly. Therefore, nature is a great inspiration source for chemists. If we can procedure these medicines in the lab, there is no need for difficult isolation processes from plants.

Anisoles are building blocks, that can often be found in medicine. An example is shown in Figure 1. This research focussed on synthesis with these anisoles. These structures contain unreactive carbon-hydrogen bonds, which can be functionalized via a process called C–H activation. In this process, the unreactive carbon-hydrogen bond is replaced by a more functional bond, like carbon-carbon, or carbon-oxygen. In this way, other molecules can be attached to anisole, just like putting two Lego blocks together. A problem is the selectivity of the reaction: the new molecule can be attached to every free corner position of the anisole.

This research focussed on selective C–H activation of anisole derivates. We developed a general procedure with mild conditions for the C–H activation of these building blocks, which is shown in Scheme 2. C–H activation needs a metal catalyst to improve the reaction. In this research, a palladium catalyst is used. A bio-organic molecule is coordinated to the catalyst, to increase the reactivity of the catalyst and to direct the selectivity towards a certain site of the anisole. This molecule is called a S,O-ligand, because it binds with its sulphur and oxygen atom to the palladium. We used the general method to test various anisole derivates, which provided the desired products in good yields. This efficient and selective functionalization of anisole derivates could provide a useful strategy for the synthesis of medicine and other natural compounds.

Scheme 2: General procedure for selective C–H activation with anisole derivates.

Figure 1: Medicine (Licochalcone C), with the anisole structural moiety is shown in red.

(5)

5

1. Introduction

1.1 C–H activation as attractive strategy for green chemistry

In the last decade, metal-catalysed C-H activation has seen a significant growth as strategy in manipulating the reactivity of C–H bonds.1_{C–H activation is applied to replace unreactive C–H bonds}

by functional groups, i.e. formation of C–C, C–N or C–O bonds. Transition metals like Ru, Rh, Pt, or Pd from which the latter one is mostly used, can insert into C–H bonds and thereby activate the bond.2

A general diagram for this C–H functionalization is shown in Scheme 3. First, the C–H bond is replaced by a C–M bond, which represents the C–H activation step. Next, the metal complex can be functionalized via e.g. transmetallation or oxidative addition. Via a termination step the two organic groups can be connected and thereby resulting in a functionalization of the molecule. A detailed discussion of the mechanism for C–H activation is given in the background information.

C–H functionalization can be used to synthesize various organic compounds, like pharmaceutical drugs.3–7_{Nature namely produces many complex organic molecules, that are used in all}

kinds of medicine. In some cases, synthesizing these compounds in the lab turned out to be challenging or even impossible. For example, Licochalcone C (Figure 1) is a medicine that can be extracted from licorice, but the yield of this extraction process is only 15 mg / 2 kg.8_{C–H activation could provide an}

easy, cheap and efficient strategy to synthesize such medicines, so that there is no more need for expensive extraction procedures. In this way, C–H activation can not only help in synthesizing (new) medicines, but also in making more medicines commercially available.

Scheme 3: C-H functionalization via metal-catalysed C–H activation (adapted from ref 2).2

Figure 1: Licochalcone C

An important advantage for this strategy over traditional synthetic methods, is that there is no need to pre-functionalize the reactant. Instead, the reaction can be performed directly with the C–H bond that is already present in the molecule. This makes C–H activation an efficient and cheap strategy, because no extra synthesis steps have to be used, and as a result, the atom economy is relatively high. Next to that, there is less waste formation. Therefore, C–H activation can also be an attractive strategy for green chemistry. Scheme 4 shows this advantage in a comparison of C–H activation and the traditional methods for C–H functionalization.

Scheme 4: C–H functionalization via metal-catalysed C–H activation (a) or via pre-functionalization of the reactant (b).2_{In this scheme, FG represents any functional group (adapted from ref 2).}

(6)

6

1.2 Improving the selectivity of C–H activation with ligands

Although much research into C–H activation has been performed in the last decades, a major problem that remains is the regioselectivity of the reaction, which is crucial for applications as medicine synthesis.1_{One way to tackle this issue, is by making use of directing groups, for which many methods}

have been reported.9,10_{In this directed C–H functionalization, a Lewis base is covalently attached to the}

substrate, to direct the metal to a specific C–H bond.

However, methods for direct functionalization of simple arenes without directing groups are limited. For these compounds, the selectivity is mainly directed by the native reaction site of the arene: electron withdrawing groups (EWGs) direct substitution to the meta position and electron donating groups (EDGs) to the ortho and para position. Due to this reason, it is difficult to achieve high regioselectivity in C–H functionalization with these simple arenes, especially for arenes with EDGs, because both the ortho and para position is favoured. Additionally, it is difficult to functionalize sites different to the native reaction site, because of the unreactive nature of these C–H bonds.2

For these nondirected arenes, an alternative is the use of ligands or catalytic systems that promote regioselective C–H functionalization, but this research is still in its infancy.9_{Studies performed}

by the groups of Yu, Sanford, Stahl and Glorius have made a great contribution to the development of these systems: they found pyridine-based ligands that could efficiently direct various Pd-catalysed C– H functionalization reactions.9,11,12

One of these reactions was the Fujiwara-Moritani reaction, which is the Pd-catalysed C–H bond olefination of arenes. The detailed mechanism of this reaction is discussed in the background information. For this reaction, Yu’s group synthesized a pyridine ligand, that could effectively direct C–H functionalization of electron-deficient arenes to the meta position.11_{Their ligand not only}

promoted the reactivity of these arenes, but also enabled reoxidation of Pd(0) by oxygen, as shown in Scheme 5a. Sanford’s group later used a slightly less bulky pyridine ligand to improve reactivity and selectivity for electron-rich arenes (Scheme 5b).12

Scheme 5: a) General reaction scheme of Yu’s methodology (adapted from ref 11).11_{b) General}

reaction scheme of Sanford’s methodology (adapted from ref 12).12

Five years later, the research group of Fernández-Ibáñez developed a bidentate S,O-ligand, which showed high efficiency in the Pd-catalysed C–H olefination of both poor and electron-rich arenes (see Scheme 6a).9_{Besides improving the reactivity of the arenes, the ligand also altered the}

site selectivity of the reaction, compared to the reaction with no ligand.

Another two years later, an efficient method was developed for nondirected palladium catalysed C–H functionalization of anilines under mild conditions in the same research group.13_{Using the ligand}

resulted in highly para-selective reactions and it was found that their Pd/S,O-ligand based catalyst was applicable to a broad substrate scope of different anilines (see Scheme 6b).13_{Even for anilines with}

electron withdrawing substituents, the para-product was afforded in good yields. Additionally, competition experiments with and without ligand showed that both the conversion and reaction speed

(7)

7 were improved by using the ligand. This suggests that the ligand can improve the rate-determining step of the reaction. They also compared these results with the site selectivity by using a pyridine ligand reported by Sanford et al. The results showed that the pyridine ligand did not alter the site selectivity of the reaction.

Currently, the exact mechanism of this Pd/S,O-ligand based catalyst in the Fujiwara-Moritani reaction is still under investigation. This could provide more insight in the role of this ligand in the catalytic cycle. Understanding the mechanism can help to improve the catalytic system further. Because of the recent discovery of this ligand, the number of tested substrates is relatively small and limited to specific substrate groups. Moreover, the results in these papers are already promising, but they show large variations in yield for different substrates and ligands. Therefore, it is also important to test this methodology for a broader range of substrates and gain more insight in the influence of the S,O-ligand in this reaction.

Scheme 6: a) General reaction scheme of S,O-ligand promoted C–H olefination of arenes (adapted from ref 9).9_{b) General reaction scheme of S,O-ligand promoted C–H olefination of anilines (adapted}

from ref 13).13

1.3 C–H olefination of anisole derivates

Anisoles are, like anilines, important building blocks in many pharmaceutical drugs. Therefore, one type of arenes tested via the methodology of Scheme 6a was anisole derivates.9_{For anisole, this resulted}

in a good yield (80%), with moderate selectivity (o:p = 1.5:1). A disadvantage of this method is the high temperature and the excess of arene, which are both not preferable considering costs and environmental impact.

The good results achieved with the S,O-ligand were also noticed by other research groups. Research by Zhu et al. developed a methodology in which a S,O-ligand was used, together with an additional ligand (Ac-Leu-OH), for highly para-selective C–H olefination of anisole derivates (see Scheme 7).14_{For anisole, they reported a good yield of 82% with o:p = 71:29. However, this method}

uses two oxidants, including expensive silver acetate, and two ligands, resulting in a larger waste stream. Besides these ligand-promoted methods, also methods without any ligands have been reported. You et al. developed a general methodology for C–H olefination of heteroarenes at room temperature.15

Remarkably, for anisole they reported 63% yield with only para-product. Although the reaction is performed at room temperature and no ligand is needed, the moderate yield and excess of arene give this reaction a disadvantage.

(8)

8 Scheme 7: General reaction scheme developed by Zhu et al. for the C–H olefination of anisole derivates with (methylsulfonyl)ethene (adapted from ref 14).14

Scheme 8: General reaction scheme developed by You et al. for the C–H olefination of anisole with methyl acrylate (adapted from ref 15).15

Concluding, these methodologies all have advantages and disadvantages. This shows that more research is needed to optimize the C–H olefination of anisole derivates. A comparison of the different reaction schemes mentioned above, shows that the S,O-ligand has some advantages over using pyridine-based ligands or no ligands; the S,O-ligand improves reactivity, selectivity and increases the reaction speed.9

Therefore, this research focussed on the use of the Pd/S,O-ligand catalytic system in anisole olefination, based on the results of the research group of Fernández-Ibáñez9,13_{. Specifically, the goal of}

this research is to optimize the palladium-catalysed regioselective C–H olefination of anisole derivates. In addition, an attempt will be made to keep the methodology as green as possible, as long as the yield is not negatively affected. Because anisoles are electron-rich arenes, just like anilines, it is expected that using a S,O-ligand would improve the selectivity and reactivity of this reaction as well.

To achieve this goal, the reaction will first be optimized for 2-methylanisole. Because of the extra methyl group on the ortho position, the selectivity would be more to the para position; one ortho position is already blocked, and the other ortho position will become more sterically hindered by the methoxy group. For this optimization, focus will be on temperature, solvent, ligand, catalyst and oxidant. After optimization, the methodology was evaluated on other anisole derivates with both electron-donating and electron-withdrawing side groups on the ortho position.

With these results, hopefully more insight in the reactivity and selectivity of the S,O-ligands will be provided. Furthermore, the substrate scope of this catalytic system can be broadened further. Together with other research into C–H activation, this research will improve green synthesis methodologies of medicine and natural compounds.

(9)

9

2. Background information

2.1 C-H activation mechanisms

The apolar nature of a C-H bond makes the bond dissociation energy (BDE) relatively high.1_{The BDE}

increases in the series C(sp3_)–H→C(sp2_{)–H→C(sp)–H; for benzene rings, the molecule class used in}

this project, the BDE is 473 kJ/mol. C–H activation is used to functionalize these unreactive bonds in an efficient and environmentally friendly way, in general by using a transition metal catalyst. In this process, the transition metal is inserted into the C–H bond, resulting in a more reactive bond, due to increased polarization of the bond. Subsequentially, the complex can react with other molecules, resulting in a functionalized C–H bond.

There are different mechanisms proposed for the C–H activation step. In these steps, two processes contribute to the weakening and breaking of the C–H bond: donation from the dπ-orbital of the metal into the empty σ* orbital of the C–H bond, and donation from the filled σ-orbital of the C-H bond into the empty dσ-orbital of the metal. Based on these processes, the C–H activation can be classified as electrophilic or nucleophilic.1_{Electron-deficient, late transition metals like Pd}II_have

low-energy orbitals and as a result, electron donation will be mainly from the C–H bond towards the metal (electrophilic). On the other hand, electron-rich metal complexes will mainly donate electrons to the C– H bond (nucleophilic). This difference is shown in Scheme 9.

Scheme 9: Difference between electrophilic and nucleophilic metal-catalysed C–H activation (adapted from ref 1).1

The mechanisms for C–H activation can be divided in three categories: oxidative addition (OA), σ-bond metathesis (BM), and, more general, electrophilic activation (EA). OA mainly occurs for low-valent late transition metals, BM for electron-poor early transition metals, and EA for electron-poor late transition metals.2,16_{As follows, the often-used palladium complexes usually react via EA mechanisms.}

In the specific case of Pd-assisted C-H activation of arenes, no other mechanisms have been observed.2

A variant of the EA mechanisms is ambiphilic metal-ligand activation (AMLA), or concerted metalation-deprotonation (CMD), in which a Lewis-basic heteroatom or a bridging ligand deprotonates the C–H bond in a concerted mechanism, forming a C–M bond. Especially electron-poor arenes react via this mechanism. A similar mechanism is base-assisted internal electrophilic substitution (BIES); this mechanism is more common for electron-rich arenes. Another EA mechanism is electrophilic metalation, or electrophilic palladation in the case of a Pd catalyst. Here, a sEAr mechanism is followed,

in which a Wheland intermediate is formed. An overview of these EA mechanisms is shown in Scheme 10.

(10)

10 Scheme 10: Overview of different EA mechanisms for Pd-catalysed C–H activation of arenes

(adapted from ref 2).2

2.2 Fujiwara-Moritani reaction and the role of the S,O-ligand

In this research, a S,O ligand was used in the C–H functionalization of anisole derivates. To describe the mechanism of this C–H activation, first the catalytic cycle of the reaction must be understood. This reaction followed the Fujiwara-Moritani catalytic cycle (Scheme 11).2_{The reaction starts with a}

Pd(II)-catalyst, that catalyses the C–H activation of the arene. Next, the olefin is inserted by 1,2-migratory insertion. After β-hydride elimination, the product dissociates from the catalyst. This step is followed by reductive elimination, which leaves a Pd(0) catalyst. The catalyst is then oxidized, so that the Pd(II) complex is retrieved.

Scheme 11: General catalytic cycle of Fujiwara-Moritani reaction (adapted from ref 2).2_{X represents}

(11)

11 When using a ligand, the first step of the cycle is not C–H activation, but coordination of the ligand to the palladium. Naksomboon studied the formation of these catalyst/ligand complexes and their role in the C–H activation of arenes. She used PPh3 as extra ligand to obtain the complex with only one

S,O-ligand. This ligand did not influence the overall reaction. She found that, upon reacting Pd(OAc)2

with 1 eq. of S,O-ligand and 1 eq. of PPh3, a palladium complex with one acetate (OAc) ligand, one

PPh3 ligand and one S,O-ligand (bidentate) was formed.2,13

By reacting an excess benzene with the Pd/S,O-ligand complex, a new complex was formed, in which the OAc group was replaced by the benzene ring as ligand (Scheme 12). Moreover, kinetic experiments with the Pd/S,O-ligand complex showed that this is the resting state of the catalyst, indicating that the C–H activation is the rate determining step (RDS).

Scheme 12: Formation of complex after C–H activation (adapted from ref 2).2

To study the mechanism of this C–H activation, competition experiments were performed with and without S,O-ligand, by using both electron-rich and electron-poor arenes. In both cases, the electron-rich arenes showed to be preferable. This already excluded the CMD/AMLA mechanism. Next, a series of kinetic tests with an electron-poor arene revealed that electrophilic aromatic palladation was unlikely. With and without ligand, large kinetic isotope effect (KIE) values were observed, while for formation of the Wheland complex, low KIE values would be expected.

Consequently, the BIES mechanism was proposed for C–H activation using the S,O-ligand. The S,O-ligand probably activates the formation of a cationic palladium complex, which in turn stimulates the C–H cleavage step of the BIES mechanism. However, more research is needed to confirm the mechanism of this C–H activation.

Based on the previously discussed mechanistical studies of the C–H olefination of anilines, a mechanism of the Pd/S,O-ligand catalysed C–H olefination of anisole derivates can be proposed. The proposed catalytic cycle is shown in Scheme 13, using 2-methylanisole as model arene and 3-methyl-2-[(perfluorophenyl)thio]butanoic acid as S,O ligand; the perfluorophenyl-group is depicted as ‘R’ for clarity of the cycle. The reaction starts with ligand coordination, by a reaction with the S,O-ligand and the catalyst. Afterwards, C–H activation follows, as proposed via the BIES mechanism. The olefin inserts via migratory insertion and the product is obtained after β-hydride elimination. Next, the ligand dissociates, and the palladium catalyst has to be oxidized to Pd(II) to close the catalytic cycle.

(12)

12 Scheme 13: Proposed catalytic cycle, with 2-methylanisole as model substrate, of Fujiwara-Moritani reaction, using a S,O-ligand (adapted from ref 2).2_{R = perfluorophenyl.}

2.3 Selectivity of the reaction

The C–H activation and migratory insertion are both steps that influence the selectivity of the reaction, although both in a different way. Besides steric hindrance, the regioselectivity of C–H activation is mainly directed by electronics.17_{Arenes with an electron-donating side group will be mainly ortho and}

para directing, while arenes with electron-withdrawing side groups are mainly meta directing. For electron-donating side groups, the regioselectivity arises from resonance structures, as shown in Scheme 14. Because the negative charge can only be on the ortho and para position, these sites are available for coordination to the palladium. The negative charge stimulates the reaction and therefore, electron-donating groups are also activating groups. For electron-withdrawing side groups, there is resonance and also an inductive effect. Together, these effects create a slightly positive charge on the ortho and para positions, making these sites less reactive than the meta position. Therefore, electron-withdrawing groups are also called deactivating groups; they do not stimulate reactions on certain sites, but only inhibit them.

Scheme 14: Resonance structures of anisole.17

(13)

13 Halogens are a special case: the inductive effect is withdrawing, but they can also donate electrons by resonance (mesomeric effect). These effects compete, but the mesomeric effect is stronger than the inductive effect. This is because the inductive effects are strongest in near proximity, while the mesomeric effect influences the whole ring. Consequently, these compounds are mainly para directing. However, the competition between the effects results in lower overall reactivity.

Then follows the next step in the catalytic cycle: migratory insertion. The arene can coordinate to both sides of the double bond. For cationic Pd-complexes, as in this catalytic cycle, the selectivity is regulated by electronic effects; the nucleophilic attack of the arene focusses on the side of the alkene with the least electron density.18_{In the case of ethyl acrylate, this means the nucleophilic attack is on}

the end of the molecule. When using, for example, methoxyethene, the attack would be on the other site The S,O-ligand helps to improve this selectivity; the bulkiness of the ligand helps to direct C– H activation to the para position, and the ligand enables formation of a cationic Pd-complex, resulting in a nucleophilic attack at the end of the alkene.2

3. Experimental procedure

All chemicals used for synthesis and optimization were bought from commercially available sources, unless stated otherwise. Detailed procedures are described in the Supporting Information.

First, the reaction conditions for C–H activation were optimized for a model substrate, 2-methylanisole (S1), starting with the general procedure shown in Scheme 15. 1 eq. arene, 1.5 eq. ethyl acrylate, x mol% Pd(OAc)2, x mol% S,O-ligand and 1 eq. PhCO3tBu as oxidant in 0.2M solvent were

added into a pressure tube and stirred overnight at a certain temperature. The reaction was then optimized in terms of ligand, temperature, solvent, concentration, catalyst, and oxidant (see Supporting Information). Yield and selectivity were determined with 1_{H-NMR, by using 1 eq. of CH}

2Br2 as internal

standard. S,O-ligands L1 and L2 (Figure 2) were prepared following a procedure in literature (see Supporting Information).19_{S,O ligands L3 and L4, shown in Figure 2, were already present in our lab.}

They were synthesized following a similar procedure.

Scheme 15: General procedure used for optimization of C–H activation.

Figure 2: Different S,O-ligands used for optimization.19

S OH O L1 F5 S O OH L2 S O OH F5 L3 S O OH L4

(14)

14 The optimized procedure was used for C–H activation with other anisole derivates, shown in Figure 3. Substrates S220_{and S3}21_{were synthesised following procedures in literature. For the}

substrates with electron withdrawing side groups, the reaction was optimized further due to low yield, with S6 as model substrate (see Supporting Information). Purification of the products was performed by silica column chromatography. Eluents were chosen based on TLC measurements. UV and KMnO4

-staining were used for visualization of the TLC. The purified products were characterized with RF

-value, 1_H-NMR,13_C-NMR,19_{F-NMR (if applicable), HRMS, and FTIR (see Supporting information).}

Additionally, two literature procedures with similar C–H activation methodologies were repeated, to gain more insight in the reactivity and selectivity. The first paper, by You et al., reported a reaction with anisole without any ligands at room temperature. They wrote that only para product was formed, with 63% yield (Scheme 8).15_{The second paper by Zhu et al. reported a reaction in which two}

ligands were used: Ac-Leu-OH and a S,O-ligand (see Scheme 7).14_{For 2-methylanisole, they reported}

82% yield with a 71:29 ortho:para ratio of product.

(15)

15

4. Results and discussion

4.1 Substrate scope of Pd/S,O-ligand catalysed C

–

H olefination of anisole derivates

The results of all optimization reactions are shown in the Supporting Information. Yield is reported as NMR-yield and selectivity as ratios of this yield. Para, ortho and ‘other’ products were formed: other is most likely a combination of the meta product and the disubstituted (both on ortho and para position) product.

The optimization of the general procedure was started with ligand optimization (see Table 1). Both L1 and L2 were tested in a reaction with DCE as solvent. Next to that, a reaction without ligand was performed, to see if the ligand has any influence on the reaction at all. The results showed that without ligand, no reaction took place. Although both ligands promoted the reactivity, L1 provided a higher yield. Therefore, this ligand was used for further optimizations. The higher yield of L1 is probably caused by electronics. The fully fluorinated arene group at the sulphur enables the ligand to withdraw electrons from the palladium, creating a more reactive, cationic palladium complex.

Table 1: Ligand optimization.

Next, the reaction was tested with different temperatures (see Table 2). Even at room temperature, some reactivity was observed, but only in small amounts. However, selectivity was good with a 7:1 para:ortho ratio. At 40 and 60 °C, the yield increased, while the selectivity did not change. At 80 °C, the yield was highest, but the selectivity dropped. Due to the higher temperature, more side products were formed. At 60 °C, the yield was still high with 82% and the selectivity was also good. Furthermore, lower temperatures are preferable considering costs and environmental impact.

Table 2: Temperature optimization

Subsequentially, different solvents were screened (see Table 3). This parameter showed to be an important factor for both the reactivity and selectivity for the reaction. A total of eleven different solvents were tested. THF, MeOH, tAmOH and acetone all gave low yield and moderate selectivity. DMF also resulted in low yield and even lower selectivity. TFA and HFIP caused low yields as well, but with these solvents, another side product was formed. This side product is shown in Figure 4. For HFIP, this product was formed with 22% yield. It seems that TFA and HFIP promoted the reactivity too much, causing the product to react further to this side product. So, the low yields for these solvents were not caused by lack of reactivity, but by excess of reactivity.

Entry T (°C) Solvent

(0.2M) Ligand NMR total yield Para Ortho Others

1 40 DCE - 0% - - - 2 40 DCE L1 54% 0.76 0.13 0.11 3 60 DCE L1 82% 0.73 0.1 0.17 4 60 DCE L2 64% 0.78 0.13 0.09 Entry T (°C) Solvent (0.2M) NMR total yield

Para Ortho Others

1 23 DCE 21% 0.72 0.14 0.14

2 40 DCE 54% 0.76 0.13 0.11

3 60 DCE 82% 0.73 0.1 0.17

(16)

16 Table 3: Solvent optimization

Figure 4: Side product formed with TFA and HFIP. In the end, four solvents showed good reactivity and selectivity: TFE, acetic acid, ethyl acetate and DCE. Especially the high yield of ethyl acetate is worth noticing, since this solvent is considered as one of the ‘greenest’ solvents.22_{In comparison with TFE, which showed similar yield and selectivity,}

ethyl acetate is a better choice both for the environment and the costs. Initially, both acetic acid, DCE and ethyl acetate were used for next optimizations, because of yield and environment. Later, ethyl acetate was eliminated because of lower yields. Nevertheless, it can still be a nice and ‘green’ solvent for this reaction.

The concentration of the reaction mixture was optimized by testing three different concentrations: 0.1M, 0.2M and 0.3M (see Table 4). Concentration variations did not influence the selectivity, but the lower concentration resulted in lower yield. The best yield was achieved with a concentration of 0.2M.

Table 4: Concentration optimization

Entry Solvent Concentration NMR total yield

Para Ortho Others

1 EtOAc 0.1M 55% 0.67 0.11 0.22 2 AcOH 0.1M 63% 0.73 0.11 0.16 3 EtOAc 0.2M 76% 0.68 0.12 0.20 4 AcOH 0.2M 98% 0.73 0.11 0.16 5 EtOAc 0.3M 73% 0.70 0.11 0.19 6 AcOH 0.3M 89% 0.70 0.12 0.18

For the oxidant optimization, different equivalents of the initial oxidant (PhCO3tBu) were

tested. In addition, the reaction was tested with oxygen and hydrogen peroxide as oxidant. However, the initial oxidant still worked better (see Table 5). Using more equivalents did not alter the yield and selectivity, but 1_{H-NMR spectra showed that, when using 1.5 eq., there was still some starting material}

and oxidant present after 24h. Therefore, this reaction was repeated with 48h reaction time. This increased the yield but lowered the selectivity. Thus, the initial amount of 1 eq. was used for further optimizations.

Table 5: Oxidant optimization

Entry Oxidant Equivalents Solvent

(0.2M) NMR total yield Para Ortho Others

1 PhCO3tBu 1.0 EtOAc 76% 0.68 0.12 0.20

2 PhCO3tBu 1.1 EtOAc 70% 0.67 0.11 0.22

3 PhCO3tBu 1.5 EtOAc 77% 0.69 0.12 0.19

4 (48h) PhCO3tBu 1.5 EtOAc 95% 0.62 0.13 0.25

Entry Solvent

1 THF 20% 0.60 0.15 0.25

2 MeOH 3% 0.33 Traces Traces

3 tAmOH 20% 0.6 0.15 0.25 4 Acetone 37% 0.62 0.14 0.24 5 DMF 21% 0.47 0.29 0.24 6 EtOAc 76% 0.68 0.12 0.20 7 AcOH 98% 0.73 0.11 0.16 8 TFA Traces 0.1 0.1 0.8 9 TFE 76% 0.67 0.15 0.18 10 HFIP 54% 0.35 0.09 0.56 11 DCE 82% 0.73 0.1 0.17

(17)

17

5 O2 - EtOAc 2% Traces Traces Traces

6 H2O2 1.0 AcOH 44% 0.59 0.16 0.25

7 H2O2 1.0 DCE 7% 0.57 0.14 0.29

Lastly, the catalyst was optimized (see Table 6). Two different palladium(II) sources were tested, and with different mol%. Again, the initial catalyst gave better yields. Furthermore, the yield and selectivity were almost the same for 5 and 10 mol% of catalyst. Even 2 mol% showed 61% yield, but this was significantly lower than with 5 and 10 mol%. Consequently, both 5 and 10 mol% were used to test the substrates.

Table 6: Catalyst optimization

At this point, a solvent was chosen; initially, acetic acid was used, because it gave better yield and is a halogen-poor solvent. However, when different substrates were tested, acetic acid resulted in very low yields, as shown in Table 7. This is because the olefin reacted with the acetic acid to a new side product, shown in Figure 5. In entry 3, for example this side product was formed with 31% yield. To prevent this from happening, DCE was chosen for substrate testing.

Table 7: Testing of substrates S1, S2 and S6 in AcOH and DCE (10 mol% catalyst + ligand) at 60 °C.

With these optimized conditions, anisole derivates S1–S11 were tested. The results are shown in Table 8, with NMR yield and ratios of para (p), ortho (o) and other (r) products. For S1–S5, this procedure resulted in good yields. For S1–S4, the selectivity for the para product was also around 70-80%. The products of these reactions were isolated and characterized (see Supporting Information). However, for the anisole derivates with electron withdrawing groups, the yield varied between 14 and 31%, and for S9, no product was observed at all.

Entry Catalyst Mol% Solvent

1 Pd(TFA)2 10 AcOH 72% 0.72 0.11 0.17 2 Pd(TFA)2 10 DCE 43% 0.74 0.12 0.14 3 Pd(OAc)2 2 DCE 61% 0.71 0.13 0.16 4 Pd(OAc)2 5 AcOH 59% 0.75 0.12 0.13 5 Pd(OAc)2 10 AcOH 98% 0.73 0.11 0.16 6 Pd(OAc)2 5 DCE 85% 0.71 0.12 0.17 7 Pd(OAc)2 10 DCE 82% 0.73 0.1 0.17

Entry Substrate Solvent

1 S1 AcOH 98% 0.73 0.1 0.26 2 S1 DCE 82% 0.73 0.1 0.17 3 S2 AcOH 54% 0.72 - 0.28 4 S2 DCE 68% 0.75 - 0.25 5 S6 AcOH 17% 0.59 0.18 0.23 6 S6 DCE 24% 0.5 0.17 0.33

Figure 5: Side product formed in reaction with acetic acid.

(18)

18 Table 8: Substrate screening via general procedure, using 0.2M DCE. a_{Isolated yield of para product.}

To improve the yield for the anisole derivates with electron-withdrawing substituents, the reactions with S6–S11 were performed at 80 °C. However, this only increased the yield by a few percent. Despite this, the NMR spectra showed something special. In significant amounts, products with the double bond in cis conformation were formed. This is unexpected because the cis conformation is energetically more unfavourable than the trans conformation. The peaks of these cis-products were visible in many spectra, but only if zoomed in. At this higher temperature, larger amounts were formed. Because the higher temperature showed no improvement, the reaction was optimized again with S6 as model substrate. Different solvents, based on the results of the earlier optimizations, were screened (ethyl acetate, TFA and HFIP). The latter two had shown to promote reactivity, by forming the side product in Figure 4. However, with substrate S6, using ethyl acetate and TFA only lowered the yield. HFIP, on the other hand, doubled the yield and increased the selectivity, as shown in Table 9 (63%, p:o:r = 6:2:2).

Table 9: Solvent optimization Entry Solvent NMR total

yield Para Ortho Others

1 DCE 31% 0.45 0.32 0.23

2 EtOAc 15% 0.47 0.2 0.33

3 TFA 12% 0.5 0.17 0.33

(19)

19 Still, the reaction could be improved further. S,O-ligands L3 and L4 (see Figure 2) were tested, but they gave no improvement. Next, the concentration of the reaction mixture was varied from 0.1 to 1.0M. The higher concentrations resulted in higher yield, with the optimum at 0.4M (see Table 10). Table 10: Concentration optimization

Therefore, the anisole derivates with electron withdrawing substituents (S6–S11) were tested again, following the same general procedure as for the anisole derivates with electron donating substituents, except for the use of HFIP (0.4M) instead of DCE (0.2M). The results of the second substrate screening are shown in Table 11. Substrate S9 was not tested anymore, because it did not show any reactivity before. For S6, the yield turned out to be a bit lower than during the optimization. Probably the reaction is a little inaccurate, because of the low amounts of solvent used. For S7, S10 and S11, the reactivity and selectivity were similar (75-70% yield with 70-80% para-product). However, for S8, the yield was halved, compared to the reaction in 0.2M DCE. For S6, S7, S10 and S11, the products of the reaction were isolated and characterized.

Table 11: Substrate screening via general procedure, using 0.4M HFIP. a_{Isolated yield of para}

product.

With the screening of four electron-rich anisoles, four electron-poor anisoles and anisole itself, this general procedure already proved to be applicable for different anisole derivates with substituents on the ortho position. Nevertheless, more, similar anisole derivates could be tested to prove this further. Next to that, anisole derivates with substituents on the meta and para position could be tested. In

Entry Concentration

(M) NMR total yield Para Ortho Others

1 0.1 67% 0.52 0.15 0.33 2 0.2 63% 0.59 0.22 0.19 3 0.3 84% 0.55 0.19 0.26 4 0.4 91% 0.53 0.2 0.27 5 0.7 79% 0.56 0.19 0.25 6 1 83% 0.54 0.21 0.25

(20)

20 addition, the reaction could be tested with different olefins. For every new substrate type, probably some new optimizations will be needed.

4.2 Comparison of results with similar literature procedures

In the paper by You et al., a general procedure for the C-H olefination of heteroarenes was developed.15

In their methodology, an excess of anisole was reacted with methyl acrylate. They also used Pd(OAc)2

as catalyst, however without an additional ligand. As solvent, TFA was used, and the reaction was performed at room temperature. Remarkably, they reported that only the para product was formed, with 63% yield. Especially the high para selectivity was surprising, due to the reason that we were not able to achieve high para selectivity.

Therefore, this procedure was repeated and 35% yield was achieved, with a para:ortho ratio of 6:4. This yield is way lower than the yield reported in the paper, and the reaction is certainly not para-selective. The crude 1_{H-NMR spectrum of this reaction is shown in the Supporting Information and}

shows the formation of two regio-isomers. The yield was calculated by adding CH2Br2 as internal

standard. Because of overlapping signals, not all peaks could be integrated, but the peaks of the double bond already show the yield of the separate products.

The paper by Zhu et al., used a methodology with a S,O-ligand for the C-H olefination of 2-methylanisole (see Scheme 8).14_{Ac-Leu-OH was used as additional ligand. Instead of ethyl acrylate,}

they used (methylsulfonyl)ethene. Moreover, they used Pd(TFA)2 as catalyst and two different oxidants

(silver acetate and K2S2O8). The reaction was performed in acetic acid at 80 °C. They reported a yield

of 82%, which seems comparable to the results achieved in this project. However, they mentioned a regioselectivity of 71:29 ortho:para product for ortho methyl anisole. This would be remarkable, as ortho methyl anisole should be mainly para-directing, because one ortho position is blocked and the other one is more sterically hindered.

Hence, this procedure was also repeated. The same conditions and chemicals were used, but instead of ligand L5, as shown in Scheme 7, ligand L2 (see Figure 2) was used, because this one was already present on our lab. The differences in electronics and steric hindrances are negligible. The results were difficult to interpretate, because of overlapping signals in the 1_{H-NMR spectrum. After a}

workup of evaporation, dilution in water and EtOAc and extraction with EtOAc, a clear crude 1_H-NMR

spectrum could be measured. This showed a yield of 73%, with a good selectivity of para:others 72:28 (see Scheme 16). ‘Others’ included a side product similar to that in Figure 4. Moreover, about 6% other isomers were formed.

Noticeable, this selectivity of this experiment for the para-product is almost exactly the same as the selectivity reported in the article for the ortho-product. This was expected, because 2-methylanisole is para-directing. It appears that a small error has been made in the article when reporting the results. Nevertheless, the selectivity of this reaction is worth noticing, as with our procedure, 80 C° resulted in a moderate selectivity (p:o:r = 6:1:7).

Scheme 16: Reaction scheme and results of repeated literature procedure of Zhu et al (adapted from ref 14).14a_{Isolated yield of para product.}

(21)

21 Whether the results of the paper of Zhu et al. are correct, our general procedure has some advantages over this one. Our procedure works at a lower temperature, and only one ligand is needed. Moreover, there is no need for two oxidants and the oxidant that we use, is relatively ‘green’. Additionally, two equivalents and anisole derivates are used, which is less favourable, if the methodology is also considered for more complex and thereby more expensive anisole derivates. These results show that there is still much to be discovered in the field of C-H activation. Different catalysts, solvents or starting materials: all parameters can all influence the reaction.

5. Conclusion

An efficient methodology with mild conditions was developed for the palladium-catalysed C–H olefination of anisole derivates. The procedure was based on the general procedure for C–H olefination of anilines, previously developed in our group.13_{It was optimized for various anisole derivates, resulting}

in one general procedure with two variants. One is for anisole derivates with electron-donating groups and uses 0.2M DCE, and the other one is for anisole derivates with electron-withdrawing groups and uses 0.4M HFIP. This general procedure is shown in Scheme 17. The reaction mechanism follows the Fujiwara-Moritani catalytic cycle and makes use of a S,O-ligand, that forms a cationic Pd-complex in the catalytic cycle. The ligand probably helps to increase reactivity during the C–H activation step and to direct regioselectivity.2_{In total, ten various anisole derivates were screened, with electron-donating}

and electron-withdrawing groups on the ortho position. This resulted in para-selective C–H olefination with moderate to good yields. The products could be isolated in moderate yields. During the optimization of the reaction, the use of a S,O-ligand and the solvent were found to have the most influence on the reactivity.

With this research, the overall substrate scope of the Pd/S,O-ligand catalysed regioselective C– H olefination of arenes is broadened further. This shows that S,O-ligands can be applied in reactions with various types of arenes to stimulate reactivity and regioselectivity. Nevertheless, more research is needed to broaden the substrate scope of anisole derivates even further; until now, only anisole derivates with substituents on the ortho position have been screened. Future research can test the general procedure for anisole derivates with substituents on the meta or para position, and with multiple substituents. Up to now, the results of this project can offer more understanding in the possibilities of C–H activation and are another small step towards the sustainable synthesis of organic compounds.

(22)

22

6. References

(1) Roudesly, F.; Oble, J.; Poli, G. Metal-Catalyzed C–H Activation/Functionalization: The Fundamentals. J. Mol. Catal. A Chem. 2017, 426, 275–296. DOI: 10.1016/j.molcata.2016.06.020

(2) Naksomboon, K. Bidentate Ligand Promoted Palladium-Catalyzed C–H Olefination of Aromatic Compounds, University of Amsterdam, 2019.

(3) Chen, K.; Lei, X. Recent Applications of C–H Functionalization in Complex Molecule Synthesis. Curr. Opin. Green Sustain. Chem. 2018, 11, 9–14. DOI: 10.1016/j.cogsc.2018.01.001

(4) Liu, C.; Chen, R.; Shen, Y.; Liang, Z.; Hua, Y.; Zhang, Y. Total Synthesis of Aplydactone by a Conformationally Controlled C−H Functionalization. Angew. Chemie - Int. Ed. 2017, 56 (28), 8187– 8190. DOI:10.1002/anie.201703803

(5) Chapman, L. M.; Beck, J. C.; Wu, L.; Reisman, S. E. Enantioselective Total Synthesis of (+)-Psiguadial B. J. Am. Chem. Soc. 2016, 138 (31), 9803–9806. DOI: 10.1021/jacs.6b07229

(6) Quinn, R. K.; Könst, Z. A.; Michalak, S. E.; Schmidt, Y.; Szklarski, A. R.; Flores, A. R.; Nam, S.; Horne, D. A.; Vanderwal, C. D.; Alexanian, E. J. Site-Selective Aliphatic C-H Chlorination Using N-Chloroamides Enables a Synthesis of Chlorolissoclimide. J. Am. Chem. Soc. 2016, 138 (2), 696–702. DOI: 10.1021/jacs.5b12308

(7) Leal, R. A.; Bischof, C.; Lee, Y. V.; Sawano, S.; McAtee, C. C.; Latimer, L. N.; Russ, Z. N.; Dueber, J. E.; Yu, J.-Q.; Sarpong, R. Application of a Palladium-Catalyzed C−H Functionalization/Indolization Method to Syntheses of Cis -Trikentrin A and Herbindole B . Angew. Chemie 2016, 128 (39), 12003– 12007. DOI: 10.1002/ange.201605475

(8) Wang, Z.; Cao, Y.; Paudel, S.; Yoon, G.; Cheon, S. H. Concise Synthesis of Licochalcone C and Its Regioisomer, Licochalcone H. Arch. Pharm. Res. 2013, 36 (12), 1432–1436. DOI: 10.1007/s12272-013-0222-3

(9) Naksomboon, K.; Valderas, C.; Gómez-Martínez, M.; Álvarez-Casao, Y.; Fernández-Ibáñez, M. Á. S,O-Ligand-Promoted Palladium-Catalyzed C-H Functionalization Reactions of Nondirected Arenes. ACS Catal. 2017, 7 (9), 6342–6346. DOI: 10.1021/acscatal.7b02356

(10) Lyons, T. W.; Sanford, M. S. Palladium-Catalyzed Ligand-Directed C-H Functionalization Reactions. Chem. Rev. 2010, 110 (2), 1147–1169. DOI: 10.1021/cr900184e

(11) Zhang, Y. H.; Shi, B. F.; Yu, J. Q. Pd(II)-Catalyzed Olefination of Electron-Deficient Arenes Using 2,6-Dialkylpyridine Ligands. J. Am. Chem. Soc. 2009, 131 (14), 5072–5074. DOI: 10.1021/ja900327e (12) Kubota, A.; Emmert, M. H.; Sanford, M. S. Pyridine Ligands as Promoters in Pd II/0-Catalyzed C-H

Olefination Reactions. Org. Lett. 2012, 14 (7), 1760–1763. DOI: 10.1021/ol300281p

(13) Naksomboon, K.; Poater, J.; Bickelhaupt, F. M.; Fernández-Ibáñez, M. Á. Para-Selective C-H Olefination of Aniline Derivatives via Pd/S,O-Ligand Catalysis. J. Am. Chem. Soc. 2019, 141 (16), 6719–6725. DOI: 10.1021/jacs.9b01908

(14) Yin, B.; Fu, M.; Wang, L.; Liu, J.; Zhu, Q. Dual Ligand-Promoted Palladium-Catalyzed Nondirected C-H Alkenylation of Aryl Ethers. Chem. Commun. 2020, 56 (22), 3293–3296. DOI: 10.1039/d0cc00940g (15) She, Z.; Shi, Y.; Huang, Y.; Cheng, Y.; Song, F.; You, J. Versatile Palladium-Catalyzed C-H

Olefination of (Hetero)Arenes at Room Temperature. Chem. Commun. 2014, 50 (90), 13914–13916. DOI: 10.1039/c4cc05827e

(16) Rothenberg, G. Catalysis. In Applied Organometallic Chemistry; WILEY-VCH Verlag GmbH & Co. KGaA: Mörlenbach, 2008; Vol. 22, pp 77–88.

(17) Clayden, J.; Greeves, N.; Warren, S. Organic Chemistry; Oxford University Press Inc.: New York, 2012; pp 479–492.

(18) Cabri, W.; Candiani, I.; Bedeschi, A.; Penco, S.; Santi, R. α-Regioselectivity in Palladium-Catalyzed Arylation of Acyclic Enol Ethers. J. Org. Chem. 1992, 57 (5), 1481–1486. DOI: 10.1021/jo00031a029 (19) Naksomboon, K.; Valderas, C.; Gómez-martínez, M.; Álvarez-casao, Y.; Fernández-ibáñez, M. Á.

Supporting Information S , O-Ligand-Promoted Palladium- Catalyzed C − H Functionalization Reactions of Nondirected Arenes. ACS Catal. 2017, 7 (9), 6342–6346. DOI: 10.1021/acscatal.7b02356 (20) Gruza, M. M.; Chambron, J. C.; Espinosa, E.; Aubert, E. Synthesis and Stereochemical Properties of

“Extended” Biphenols Bridged by Ortho-, Meta-, and Para-Phenylene Spacers. European J. Org. Chem. 2009, 36, 6318–6327. DOI: 10.1002/ejoc.200900837

(21) Parrish, C. A.; Buchwald, S. L. Palladium-Catalyzed Formation of Aryl Tert-Butyl Ethers from Unactivated Aryl Halides. J. Org. Chem. 2001, 66 (7), 2498–2500. DOI: 10.1021/jo001426z

(22) Welton, T. Solvents and Sustainable Chemistry. Proc. R. Soc. A Math. Phys. Eng. Sci. 2015, 471 (2183). DOI: 10.1098/rspa.2015.0502

(23)

23

Supporting information

Pd/S,O-ligand Catalysed Regioselective C-H

Olefination of Anisole Derivates

(24)

24

1.0 General information

Chromatography: Macherey-Nagel Silica P60 size 40-63 µm, TLC: Macherey-Nagel silica gel 60 (0.20mm). Visualization of the chromatogram was performed by UV and KMnO4. 1H, 13C and 19F were recorded on Bruker

AMX 400 and Bruker DRX 300 using CDCl3 as solvent. Chemical shift values are reported in ppm with the

solvent resonance as the internal standard (CDCl3: δ 7.26 for 1H, δ 77.16 for 13C). Data are resported as follows:

chemical shifts, multiplicity (s = singlet, d = doublet, dd = doublet of doublets, t = triplet, q = quartet, m = multiplet), coupling constants (Hz), and integration. Mass spectra were recorded on a AccuTOF GC v 4g, JMS-T100GCV mass spectrometers. IR spectra were recorded on a Bruker Alpha FTIR machine and wavelengths are reported in cm-1_.

2.0 Synthesis of S,O-ligands

2

L1: 3-Methyl-2-[(perfluorophenyl)thio]butanoic acid

2,3,4,5,6-pentafluorothiophenol (1.33 mL, 10 mmol, 1 eq) was added to a mixture of 2-bromo-3-methylbutanoic acid (1.81 g, 10 mmol, 1 eq) and NaOH (0.1 g, 20 mmol, 2 eq) in tBuOH (35 mL, 0.29M) at room temperature. The reaction was refluxed overnight and concentrated under reduced pressure. The resulting pale, yellow crude was dissolved in water (100 mL) and acidified (6M aq. HCl solution) until pH = 1. The aqueous layer was extracted with Et2O (3x100 mL) and the combined organic layers were washed with NaHCO3 (saturated, 3x10

mL). The resulting aqueous layers were combined and acidified (6M aq. HCl solution) until pH = 1. The aqueous layer was extracted with Et2O (3x200 mL). The organic layers were combined and dried over anhydrous MgSO4.

Then, the mixture was filtered and concentrated under reduced pressure. The resulting yellow oil was purified with silica column chromatography, using 8:2 Hep:EtOAc (product RF = 0.24), which provided the product as

yellow oil that later crystallized (36% yield). Because the first time the solution was not acidified enough, the yield is lower than expected. Synthesis and purification of the ligand was confirmed with 1_H-NMR.

L2: 3-Methyl-2-(phenylthio)butanoic acid

Thiophenol (1.03 mL, 10 mmol, 1 eq) was added to a mixture of 2-bromo-3-methylbutanoic acid (1.81 g, 10 mmol, 1 eq), NaOH (0.8 g, 20 mmol, 2 eq) in EtOH (30 mL) at room temperature. The reaction was refluxed overnight and concentrated under reduced pressure. The resulting pale, yellow crude was dissolved in water (100 mL) and acidified (6M aq. HCl solution) until pH = 1. The aqueous layer was extracted with Et2O (3x100 mL)

and the combined organic layers were washed with NaHCO3 (saturated, 3x100 mL). The resulting aqueous layers

were combined and acidified (6M aq. HCl solution) until pH = 1. The aqueous layer was extracted with Et2O

(3x200 mL). The organic layers were combined and dried over anhydrous MgSO4. Then, the mixture was filtered

and concentrated under reduced pressure. The resulting yellow oil was purified with silica column chromatography, using 8:2 Hep:EtOAc (product RF = 0.23), which provided the product as yellow oil (53% yield).

(25)

25

3.0 Optimization of general procedure – Electron-rich substrates

All optimization reactions were performed in a 0.1 mmol scale. S1 was used as model substrate for initial optimization. A general procedure was followed: S1 (12.4 μL, 1 eq, 0.1 mmol), ethyl acrylate (16.4 μL, 1.5 eq, 0.15 mmol), Pd(OAc)2 (2.3 mg, 0.1 eq, 0.01 mmol), S,O-ligand (0.1 eq, 0.01 mmol) and PhCO3tBu (19.0 μL, 1

eq, 0.1 mmol) as oxidant in solvent (0.2M, 0.5 mL) were added into a pressure tube. The tube was closed with a screw cap and placed in a pre-heated oil bath to react overnight at a certain temperature. The reaction mixture was filtered over celite and concentrated under reduced pressure. Yield and selectivity were determined with 1_H-NMR

by adding CH2BR2 (7.08 μL, 1 eq, 0.1 mmol) as internal standard.

3.1 Ligand optimization

The general procedure was followed, with S1 (12.4 μL, 1 eq, 0.1 mmol), ethyl acrylate (16.4 μL, 1.5 eq, 0.15 mmol), Pd(OAc)2 (2.3 mg, 0.1 eq, 0.01 mmol), S,O-ligand L1 (3.0 mg, 0.1 eq, 0.01 mmol) / S,O ligand L2 (2.1

mg, 0.1 eq, 0.01 mmol) / no ligand, and PhCO3tBu (19.0 μL, 1 eq, 0.1 mmol) as oxidant in DCE (0.2M, 0.5 mL).

The reaction mixture was heated in an oil bath at 40 or 60 °C.

3.2 Temperature optimization

The general procedure was followed, with S1 (12.4 μL, 1 eq, 0.1 mmol), ethyl acrylate (16.4 μL, 1.5 eq, 0.15 mmol), Pd(OAc)2 (2.3 mg, 0.1 eq, 0.01 mmol), S,O-ligand L1 (3.0 mg, 0.1 eq, 0.01 mmol), and PhCO3tBu (19.0

μL, 1 eq, 0.1 mmol) as oxidant in DCE (0.2M, 0.5 mL). The reaction mixture was heated in an oil bath at 23, 40, 60 or 80 °C.

3.3 Solvent optimization

μL, 1 eq, 0.1 mmol) as oxidant in solvent (0.2M, 0.5 mL). The reaction mixture was heated in an oil bath at 60 °C.

(0.2M) Ligand NMR total yield Para Ortho Others

1 40 DCE - 0% - - -

2 40 DCE L1 54% 0.76 0.13 0.11

3 60 DCE L1 82% 0.73 0.1 0.17

4 60 DCE L2 64% 0.78 0.13 0.09

1 23 DCE 21% 0.72 0.14 0.14 2 40 DCE 54% 0.76 0.13 0.11 3 60 DCE 82% 0.73 0.1 0.17 4 80 DCE 94% 0.57 0.11 0.32 Entry Solvent (0.2M) NMR total yield

Para Ortho Others

1 THF 20% 0.60 0.15 0.25

2 MeOH 3% 0.33 Traces Traces

3 tAmOH 20% 0.6 0.15 0.25 4 Acetone 37% 0.62 0.14 0.24 5 DMF 21% 0.47 0.29 0.24 6 EtOAc 76% 0.68 0.12 0.20 7 AcOH 98% 0.73 0.11 0.16 8 TFA Traces 0.1 0.1 0.8 9 TFE 76% 0.67 0.15 0.18 10 HFIP 54% 0.35 0.09 0.56 11 DCE 82% 0.73 0.1 0.17

(26)

26

3.4 Concentration optimization

μL, 1 eq, 0.1 mmol) as oxidant in solvent (0.1M, 1.0 mL / 0.2M, 0.5 mL / 0.3M, 0.3 mL). The reaction mixture was heated in an oil bath at 60 °C.

Entry Solvent Concentration NMR total

1 EtOAc 0.1M 55% 0.67 0.11 0.22 2 AcOH 0.1M 63% 0.73 0.11 0.16 3 EtOAc 0.2M 76% 0.68 0.12 0.20 4 AcOH 0.2M 98% 0.73 0.11 0.16 5 EtOAc 0.3M 73% 0.70 0.11 0.19 6 AcOH 0.3M 89% 0.70 0.12 0.18

3.5 Oxidant optimization

μL, 1 eq, 0.1 mmol) / (20.9 μL, 1.1 eq, 0.11 mmol) / (28.5 μL, 1.5 eq, 0.15 mmol)as oxidant in solvent (0.2M, 0.5 mL). The reaction mixture was heated in an oil bath at 60 °C. The same reaction was tried with an oxygen balloon as oxidant, and with hydrogen peroxide(10.2 μL, 1 eq, 0.1 mmol) as oxidant.

Entry Oxidant Equivalents Solvent

1 PhCO3tBu 1.0 EtOAc 76% 0.68 0.12 0.20

2 PhCO3tBu 1.1 EtOAc 70% 0.67 0.11 0.22

3 PhCO3tBu 1.5 EtOAc 77% 0.69 0.12 0.19

4 (48h) PhCO3tBu 1.5 EtOAc 95% 0.62 0.13 0.25

5 O2 - EtOAc 2% Traces Traces Traces

6 H2O2 1.0 AcOH 44% 0.59 0.16 0.25

7 H2O2 1.0 DCE 7% 0.57 0.14 0.29

3.6 Catalyst optimization

The general procedure was followed, with S1 (12.4 μL, 1 eq, 0.1 mmol), ethyl acrylate (16.4 μL, 1.5 eq, 0.15 mmol), Pd(OAc)2 (2.3 mg, 0.1 eq, 0.01 mmol) / (1.1 mg, 0.05 eq, 0.005 mmol) / (0.4 mg, 0.02 eq, 0.002 mmol)

or Pd(TFA)2 (3.3 mg, 0.1 eq, 0.01 mmol), S,O-ligand L1 (3.0 mg, 0.1 eq, 0.01 mmol), and PhCO3tBu (19.0 μL, 1

eq, 0.1 mmol) as oxidant in solvent (0.2M, 0.5 mL). The reaction mixture was heated in an oil bath at 60 °C. Entry Catalyst Mol% Solvent

1 Pd(TFA)2 10 AcOH 72% 0.72 0.11 0.17 2 Pd(TFA)2 10 DCE 43% 0.74 0.12 0.14 3 Pd(OAc)2 2 DCE 61% 0.71 0.13 0.16 4 Pd(OAc)2 5 AcOH 59% 0.75 0.12 0.13 5 Pd(OAc)2 10 AcOH 98% 0.73 0.11 0.16 6 Pd(OAc)2 5 DCE 85% 0.71 0.12 0.17 7 Pd(OAc)2 10 DCE 82% 0.73 0.1 0.17

(27)

27

4.0 Synthesis of substrates

S2: 1-(tert-butyl)-2-methoxybenzene20

A solution of 2-tert-butylphenol (0.46 mL, 1 eq, 3 mmol) in THF (0.75 mL) was slowly added to a suspension of freshly powdered KOH (0.59 g, 3.5 eq, 10.5 mmol) in THF (1.5 mL), while the temperature was kept below 10 °C. The mixture was stirred for 2 hours at room temperature. Then, CH3I (0.56 mL, 3 eq, 9 mmol) was added

dropwise at 0 °C. The mixture was stirred for 18 hours at room temperature. Then, the mixture was filtered and concentrated under reduced pressure, after which the product was obtained as yellow oil (69% yield). 1_{H NMR}

(300 MHz, CDCl3) δ 7.29 (dd, J = 7.7, 1.6 Hz, 1H), 7.19 (td, J = 7.7, 1.6 Hz, 1H), 6.90 (m, 2H), 3.84 (s, 4H), 1.38

(s, 9H). 13_{C NMR (101 MHz, CDCl}

3) δ 158.68, 127.14, 126.65, 120.41, 111.68, 55.10, 29.86. HRMS (FI) calcd

for C11H16O[M]+: 164.1201; found: 164.1312. IR (neat): νmax (cm-1): 2953, 1489, 1461, 1449, 1234, 1095, 1030,

744.

(28)

28 S3: 1-(tert-butoxy)-2-methylbenzene21

NaO-tBu, (1.2493 g, 1.3 eq, 13 mmol), Pd(OAc)2 (56.1 mg, 0.025 eq, 0.25 mmol) and

2’-(dicyclohexylphosphaneyl)-N,N-dimethyl-[1,1’-biphenyl]-2-amine (0.1181 g, 0.03 eq, 0.30 mmol) were added to an oven-dried resealable flask. The flask was evacuated, backfilled with argon and sealed with a septum. Toluene (10 mL), 1-chloro-2-methylbenzene (1.17 ml, 1 eq, 10 mmol), and additional toluene (10 mL) were added sequentially via syringe. The flask was closed with a stopper, sealed and placed in a 100 °C oil bath. The mixture was stirred for 17 hours. Then, the mixture was cooled and diluted in ether (45 mL). The mixture was filtered through celite and rinsed with ether. The solution was concentrated under reduced pressure. The product was purified by silica column chromatography, with 2% EtOAc in Hex. The product was obtained as yellow oil (18% yield). RF = 0.50 (Hex / EtOAc = 9:1). 1H NMR (300 MHz, CDCl3) δ 7.16 (d, J = 6.8 Hz, 1H), 7.13 – 7.05 (m,

1H), 7.03 (m, 1H), 6.98 – 6.91 (m, 1H), 2.25 (s, 3H), 1.39 (s, 8H). 13_{C NMR (101 MHz, CDCl}

3) δ 154.53, 132.22,

131.02, 126.19, 122.70, 122.36, 79.06, 29.36, 17.38. HRMS (FI) calcd for C11H16O[M]+: 164.1201; found:

164.1375. IR (neat): νmax (cm-1_{): 2977, 1488, 1458, 1390, 1365, 1239, 1165, 1114, 925, 900, 778, 765, 729.}

(29)

29

5.0 Substrate testing 1

st

_time

A general procedure was followed: arene (1 eq, 0.25 mmol), ethyl acrylate (41.4 μL, 1.5 eq, 0.38 mmol), Pd(OAc)2

(2.8 mg, 0.05 eq, 0.0125 mmol) / (5.6 mg, 0.1 eq, 0.025 mmol), S,O-ligand L1 (3.8 mg, 0.05 eq, 0.0125 mmol) / (7.5 mg, 0.1 eq, 0.025 mmol) and PhCO3tBu (47.6 μL, 1 eq, 0.25 mmol) as oxidant in DCE (0.2M, 1.25 mL) were

added into a pressure tube. The tube was closed with a screw cap and placed in a pre-heated oil bath to react overnight at 60 °C. The reaction mixture was filtered over celite and concentrated under reduced pressure. Yield and selectivity were determined with 1_{H-NMR by adding CH}

2BR2 (17.5 μL, 1 eq, 0.25 mmol) as internal standard.

6.0 Optimization of general procedure – Electron-poor substrates

All optimization reactions were performed in a 0.1 or 0.25 mmol scale. S6 was used as model substrate for initial optimization. A general procedure was followed: S6 (11.2 μL, 1 eq, 0.1 mmol), ethyl acrylate (16.4 μL, 1.5 eq, 0.15 mmol), Pd(OAc)2 (2.3 mg, 0.1 eq, 0.01 mmol), S,O-ligand (0.1 eq, 0.01 mmol) and PhCO3tBu (19.0 μL, 1

eq, 0.1 mmol) as oxidant in solvent (0.2M, 0.5 mL) were added into a pressure tube. The tube was closed with a screw cap and placed in a pre-heated oil bath to react overnight at a certain temperature. The reaction mixture was filtered over celite and concentrated under reduced pressure. Yield and selectivity were determined with 1_H-NMR

by adding CH2BR2 (7.08 μL, 1 eq, 0.1 mmol) as internal standard.

6.1 Temperature optimization

μL, 1 eq, 0.1 mmol) as oxidant in DCE (0.2M, 0.5 mL). The reaction mixture was heated in an oil bath at 60 or 80 °C.

Entry Substrate Catalyst / ligand

mol% NMR total yield Para Ortho Others

1 S1 5 85% 0.71 0.12 0.17 2 S1 10 82% 0.73 0.1 0.17 3 S2 5 21% 0.67 - 0.33 4 S2 10 68% 0.75 - 0.25 5 S3 5 52% 0.79 - 0.21 6 S3 10 70% 0.77 - 0.23 7 S4 5 68% 0.76 0.12 0.12 8 S4 10 92% 0.75 0.13 0.12 9 S5 5 49% 0.43 0.41 0.16 10 S5 10 63% 0.43 0.40 0.17 11 S6 5 14% 0.5 0.14 0.36 12 S6 10 24% 0.5 0.17 0.33 13 S7 5 20% 0.55 0.15 0.3 14 S7 10 24% 0.63 0.17 0.2 15 S8 5 19% 0.58 0.16 0.26 16 S8 10 31% 0.55 0.13 0.32 17 S9 5 - - - - 18 S9 10 - - - - 19 S10 5 9% 0.56 0.22 0.22 20 S10 10 14% 0.64 0.14 0.2 21 S11 5 25% 0.6 0.2 0.2 22 S11 10 31% 0.68 0.16 0.16

Entry Substrate Temperature NMR total yield

Para Ortho Others

1 S6 60 24% 0.5 0.17 0.33

2 S6 80 31% 0.45 0.32 0.23

(30)

30

6.2 Solvent optimization

The general procedure was followed, with S6 (28.1 μL, 1 eq, 0.25 mmol), ethyl acrylate (41.4 μL, 1.5 eq, 0.38 mmol), Pd(OAc)2 (5.6 mg, 0.1 eq, 0.025 mmol), S,O-ligand L1 (7.5 mg, 0.1 eq, 0.025 mmol), PhCO3tBu (47.6

μL, 1 eq, 0.25 mmol) as oxidant in solvent (0.2M, 1.25 mL). The reaction mixture was heated in an oil bath at 60 °C.

6.3 Ligand optimization

The general procedure was followed, with S6 (28.1 μL, 1 eq, 0.25 mmol), ethyl acrylate (41.4 μL, 1.5 eq, 0.38 mmol), Pd(OAc)2 (5.6 mg, 0.1 eq, 0.025 mmol), S,O-ligand L1 (7.5 mg, 0.1 eq, 0.025 mmol) / S,O-ligand L3 (7.2

mg, 0.1 eq, 0.025 mmol) / S,O-ligand L4 (6.7 mg, 0.1 eq, 0.025 mmol), and PhCO3tBu (47.6 μL, 1 eq, 0.25 mmol)

as oxidant in HFIP (0.2M, 1.25 mL). The reaction mixture was heated in an oil bath at 60 °C.

6.4 Concentration optimization

μL, 1 eq, 0.1 mmol) as oxidant in HFIP (0.2M, 0.5 mL) / (0.1M, 1.0 mL) / (0.3M, 0.33 mL) / (0.4M, 0.25 mL), (0.7M, 0.14 mL) / (1.0M, 0.1 mL). The reaction mixture was heated in an oil bath at 60 °C.

4 S7 80 33% 0.61 0.18 0.21 5 S8 60 31% 0.55 0.13 0.32 6 S8 80 37% 0.59 0.16 0.09 7 S9 60 - - - - 8 S9 80 - - - - 9 S10 60 14% 0.64 0.14 0.2 10 S10 80 21% 0.62 0.19 0.19 11 S11 60 31% 0.68 0.16 0.16 12 S11 80 34% 0.68 0.21 0.11

Entry Solvent NMR total

1 DCE 24% 0.5 0.17 0.33

2 EtOAc 15% 0.47 0.2 0.33

3 TFA 12% 0.5 0.17 0.33

4 HFIP 63% 0.59 0.22 0.19

Entry S,O

ligand NMR total yield Para Ortho Others

1 L1 63% 0.59 0.22 0.19

2 L3 58% 0.53 0.26 0.21

3 L4 32% 0.5 0.28 0.22

Entry Concentration

(M) NMR total yield Para Ortho Others

1 0.1 67% 0.52 0.15 0.33 2 0.2 63% 0.59 0.22 0.19 3 0.3 84% 0.55 0.19 0.26 4 0.4 91% 0.53 0.2 0.27 5 0.7 79% 0.56 0.19 0.25 6 1 83% 0.54 0.21 0.25

(31)

31

7.0 Substrate testing 2

nd

_time

A general procedure was followed: arene (1 eq, 0.25 mmol), ethyl acrylate (41.4 μL,1.5 eq, 0.38 mmol), Pd(OAc)2

(5.6 mg, 0.1 eq, 0.01 mmol), S,O-ligand L1 (7.5 mg, 0.1 eq, 0.01 mmol) and PhCO3tBu (47.6 μL, 1 eq, 0.25 mmol)

as oxidant in HFIP (0.4M, 0.63 mL) were added into a pressure tube. The tube was closed with a screw cap and placed in a pre-heated oil bath to react for 17-24 hours at 60 °C. The reaction mixture was filtered over celite and concentrated under reduced pressure. Yield and selectivity were determined with 1_{H-NMR by adding CH}₂_BR₂

(17.5 μL, 1 eq, 0.25 mmol) as internal standard.

8.0 Characterization of purified products

General procedure for S1–S5: arene (1 eq, 0.25 mmol), ethyl acrylate (41.4 μL, 1.5 eq, 0.38 mmol), Pd(OAc)2

(5.6 mg, 0.1 eq, 0.025 mmol), S,O-ligand L1 (7.5 mg, 0.1 eq, 0.025 mmol) and PhCO3tBu (47.6 μL, 1 eq, 0.25

mmol) as oxidant in DCE (0.2M, 1.25 mL) were added into a pressure tube. The tube was closed with a screw cap and placed in a pre-heated oil bath to react overnight at 60 °C. The reaction mixture was filtered over celite and concentrated under reduced pressure. Yield and selectivity were determined with 1_{H-NMR by adding CH}

2BR2

(17.5 μL, 1 eq, 0.25 mmol) as internal standard. The products were purified by silica column chromatography and eluent was chosen based on TLC measurements, which were visualized by UV and staining with KMnO4. Isolated

yield was determined by weight measurement.

General procedure for S6–S11: arene (1 eq, 0.25 mmol), ethyl acrylate (41.4 μL,1.5 eq, 0.38 mmol), Pd(OAc)2

(5.6 mg, 0.1 eq, 0.01 mmol), S,O-ligand L1 (7.5 mg, 0.1 eq, 0.01 mmol) and PhCO3tBu (47.6 μL, 1 eq, 0.25 mmol)

as oxidant in HFIP (0.4M, 0.63 mL) were added into a pressure tube. The tube was closed with a screw cap and placed in a pre-heated oil bath to react overnight at 60 °C. The reaction mixture was filtered over celite and concentrated under reduced pressure. Yield and selectivity were determined with 1_{H-NMR by adding CH}₂_BR₂

(17.5 μL, 1 eq, 0.25 mmol) as internal standard. The products were purified by silica column chromatography and eluent was chosen based on TLC measurements, which were visualized by UV and staining with KMnO4. Isolated

yield was determined by weight measurement.

8.1 Ethyl (E)-3-(4-methoxy-3-methylphenyl)acrylate & methyl

(E)-3-(2-methoxy-3-methylphenyl)acrylate (3.1)

The general procedure was followed, with S1 (31.0 μL, 1 eq, 0.25 mmol). The products were purified by silica column chromatography, using 5:5 Hep:DCM, resulting in a total isolated yield of 51% (isomers mixed). The product was obtained as pale-yellow oil. The isomers were isolated from the optimization reactions, using the same eluent.

Ethyl (E)-3-(4-methoxy-3-methylphenyl)acrylate: RF = 0.54 (Hep / DCM = 3:7). 1H NMR (400 MHz, CDCl3) δ

7.62 (d, J = 15.9 Hz, 1H), 7.34 (m, 2H), 6.81 (d, J = 9.0 Hz, 1H), 6.30 (d, J = 15.9 Hz, 1H), 4.25 (q, J = 7.1 Hz, 2H), 3.86 (s, 3H), 2.22 (s, 3H), 1.33 (t, J = 7.1 Hz, 3H). 13_{C NMR (101 MHz, CDCl}

3) δ 167.59, 159.78, 144.70,

130.20, 127.91, 127.37, 126.85, 115.54, 110.07, 60.41, 55.57, 16.40, 14.52. HRMS (FI) calcd for C13H16O3 [M]+:

220.1099; found: 220.1113. IR (neat): νmax (cm-1_{): 1704, 1632, 1604, 1503, 1304, 1250, 1218, 1174, 1157, 1132,}

982.

Entry Substrate NMR total yield

Para Ortho Others

1 S6 77% 0.55 0.19 0.26

2 S7 75% 0.75 0.11 0.15

3 S8 16% 0.63 0.13 0.3

4 S10 79% 0.8 0.11 0.09

(32)

(33)

33 Methyl (E)-3-(2-methoxy-3-methylphenyl)acrylate: RF = 0.25 (Hep / DCM = 1:1). 1H NMR (400 MHz, CDCl3)

δ 7.97 (d, J = 16.1 Hz, 1H), 7.40 (d, J = 7.4 Hz, 1H), 7.21 (d, J = 7.4 Hz, 1H), 7.04 (t, J = 7.1 Hz, 1H), 6.48 (d, J = 16.1 Hz, 1H), 4.27 (q, J = 7.1 Hz, 2H), 3.75 (s, 3H), 2.31 (s, 3H), 1.34 (t, J = 7.1 Hz, 3H). HRMS (EI) calcd for C13H16O3 [M]+: 220.1099; found: 220.1103. IR (neat): νmax (cm-1): 2938, 1712, 1633, 1605, 1503, 1467,

Pd/S,O-ligand Catalysed Regioselective C–H Olefination of Anisole Derivates

Bachelor Thesis Scheikunde

Pd/S,O-ligand Catalysed Regioselective C–H

Olefination of Anisole Derivates

door

Rianne van Diest

14 december 2020

Studentnummer

11677635

Onderzoeksinstituut

Verantwoordelijk docent

Van ’t Hoff Institute for Molecular Sciences

Dr. M.A. (Tati) Fernández Ibáñez

Onderzoeksgroep

Begeleider

TABLE OF CONTENTS

LIST OF ABBREVIATIONS

ABSTRACT

POPULAR SCIENTIFIC SUMMARY

1. Introduction

2. Background information

3. Experimental procedure

4. Results and discussion

–

5. Conclusion

6. References

Supporting information

Pd/S,O-ligand Catalysed Regioselective C-H

Olefination of Anisole Derivates

1.0 General information

2.0 Synthesis of S,O-ligands

3.0 Optimization of general procedure – Electron-rich substrates

3.1 Ligand optimization

3.2 Temperature optimization

3.3 Solvent optimization

3.4 Concentration optimization

3.5 Oxidant optimization

3.6 Catalyst optimization

4.0 Synthesis of substrates

5.0 Substrate testing 1

time

6.0 Optimization of general procedure – Electron-poor substrates

6.1 Temperature optimization

6.2 Solvent optimization

6.3 Ligand optimization

6.4 Concentration optimization

7.0 Substrate testing 2

time

8.0 Characterization of purified products

8.1 Ethyl (E)-3-(4-methoxy-3-methylphenyl)acrylate & methyl

(E)-3-(2-methoxy-3-methylphenyl)acrylate (3.1)

_time

_time