Is BBr3 a borderline acid

Studies on the acidity and catalytic activity of unusual LEWIS acids


1 Studies on the acidity and catalytic activity of unusual LEWIS acids Dissertation for obtaining the doctoral degree in natural sciences (Dr. rer.nat.) In the Department of Chemistry at the Philipps University of Marburg presented by the chemist Alexander Roland Nödling from Frankenthal Marburg, 2015

2 Submitted on November 19th by the Chemistry Department of the Philipps University of Marburg (university code number: 1180) accepted as dissertation on. Gerhard Hilt Second reviewer: Prof. Dr. Stefanie Dehnen Day of the Oral Examination: Scientific Career 09 / / 2016 PhD at the Philipps University of Marburg with Prof. Dr. Gerhard Hilt 04 / / 2012 Research stay at the Technical University of Berlin with Prof. Dr. Martin Oestreich 05/2011 Graduated as a qualified chemist 08/2009 Research assistant at the Max Planck Institute for Polymer Research, Mainz, with Dr. habil. Maximilian Kreiter 03 / / 2009 Internship at the Max Planck Institute for Polymer Research, Mainz with Dr. habil. Maximilian Kreiter 04 / / 2011 Study of chemistry at the Philipps-Universität Marburg 03/2005 Abitur at the Albert-Einstein-Gymnasium, Frankenthal (Pfalz)

3 The present work was carried out at the Department of Chemistry at the Philipps University of Marburg and at the Institute for Chemistry at the Technical University of Berlin under the supervision of Prof. Dr. Gerhard Hilt made between September 2011 and November 2015. As part of this work, the following publications have appeared so far: A. Schmidt, A. R. Nödling, G. Hilt, Angew. Chem. Int. Ed. 2015, 54, An Alternative Mechanism for the Cobalt-Catalyzed Isomerization of Terminal Alkenes to (Z) - 2-Alkenes. A. R. Nödling, G. Jakab, P. R. Schreiner, G. Hilt, Eur. J. Org. Chem. 2014, P NMR Spectroscopically Quantified Hydrogen-Bonding Strength of Thioureas and Their Catalytic Activity in Diels-Alder Reactions. A. R. Nödling, K. Müther, V. H. G. Rohde, G. Hilt, M. Oestreich, Organometallics 2014, 33, Ferrocene-Stabilized Silicon Cations as Catalysts for Diels-Alder Reactions: Attempted Experimental Quantification of Lewis Acidity and ReactIR Kinetic Analysis. Poster contributions: A. R. Nödling, K. Müther, V. H. G. Rohde, G. Jakab, M. Oestreich, P. R. Schreiner, G. Hilt Studies on Correlations Between Lewis Acidity and Reactivity of Complex Lewis Acids. ORCHEM 2014,, Weimar A. R. Nödling, F. Pünner, J. Möbus, V. Naseri, G. Hilt NMR-Spectroscopic and kinetic quantification of Lewis acidity: Standard Lewis acids and Lewis acid assisted Lewis acidity. ORCHEM 2012,, Weimar


5 The Lord gave us the atoms, and it s up to us to make em dance. Homer J. Simpson, The Simpsons, Episode 3F23 For Science


7 Acknowledgments My special thanks go to Prof. Dr. Gerhard Hilt for the challenging and original task, the freedom to realize your own ideas and approaches, as well as the excellent support within the scope of this work. I would also like to thank you for the possibility of a research stay at the TU Berlin. For the willingness to make myself available as a second reviewer for this thesis, I would like to thank Prof. Dr. Thank you very much Stefanie Dehnen. I also thank Prof. Dr. Michael Gottfried for willingness to make himself available as an examiner. Many thanks go to Prof. Dr. Martin Oestreich, who enabled me to spend several months in his working group at the TU Berlin as part of a cooperation. With numerous advice and suggestions, he made a major contribution to the successful completion of the project. Prof. Dr. Thanks to Peter Schreiner for helpful discussions and a successful cooperation. I would like to thank all current and former members of the Hilt working group for the great working atmosphere and the great and relaxed cooperation. Thanks to Dr. Marion Arndt, Emre Babaoglu, Dr. Martin Bohn, Florian Braun, Dr. Michael Danz, Dr. Florian Erver, Natalia Fritzler, Dr. Laura Kersten, Corinna Kohlmeyer, Dr. Julian Kuttner, Dr. Anne Miersch, Robert Möckel, Dr. Anna Paul, Dr. Florian Pünner, Maxie Rambow, Philipp Röse, Stefan Roesner, Lars Sattler, Dr. Anastasia Schmidt, Svenja Warratz, Felicia Weber and Sebastian Weber. I would like to thank all of the proofreaders for their efforts and conscientious work. I would like to thank all of the advanced and bachelor students who have given me great help in working on my projects. I would particularly like to thank Lukas Alig for his self-sacrificing work, even though he should perhaps take more coffee breaks. We should also mention all the students who put in a lot of effort during the internship and took numerous synthesis stages off my back. I would like to thank the entire Oestreich working group for their warm welcome during my stay at the TU Berlin. Dr. Kristine Müther, Jens Mohr, Volker Rohde and Dr. Special thanks go to Timo Stahl, not only for the numerous constructive technical discussions and the successful cooperation, but also for the great support during the acclimatization and all the fun away from chemistry. I would like to thank Daniela Bochert for the easy and uncomplicated opportunity

8 to be able to live during the stay in Berlin, her patience with my weekly fluctuating plans and the great atmosphere in the flat share. Dr. I would like to thank Gergely Jakab for providing numerous thioureas as part of our cooperation. I would like to thank Laura Kersten and Anastasia Schmidt not only for their great efforts in proofreading this work, but also for all the advice and all-round support since my thesis. A big thank you goes to Anastasia Schmidt for the great and relaxed time in the laboratory. Without you, the work would have been a lot less fun. Robert Möckel should also be mentioned, thanked for the extensive proofreading as well as for the very chic collaboration on the CBS project and the broadest musical taste. Apart from the order in the laboratory, it was also nice in the new building thanks to you. I would like to thank Martina Pfeiffer for her friendliness, helpfulness and the good cooperation. I would like to thank the staff in the service departments of the Chemistry Department for their excellent work, which enabled the research to run smoothly. The NMR department should be emphasized here, without which a large part of this work would not have been possible. Miss Dr. Xiulan Xie, Dr. Ronald Wagner, Gert Häde, Cornelia Mischke and Klaus Pützer were always very patient and helpful despite my unusual requests. Not-yet-Doctor Raufer, Dr. I thank Kapper and the Knuckleheads for the chemistry-free time and everything else. Thanks to all of the roommates of my former flat share for the great 4 years, especially Aru and Schlimmbo, who still owe me 99 lives. My greatest thanks go to my family for their trust and continued support. I am glad that you are always there for me. Lastly, I'd like to thank my wife Hannah, not only for taking me, but also for making me a better person.



11 Contents 1 INTRODUCTION LEWIS acidity The definition of acids and bases according to G.N.

12 5.

131 List of Abbreviations 210 A.2 NMR Spectra of Selected Compounds 215 A.3 Kinetic Profiles of Selected Reactions 223 A.4 JOB Diagrams 229 A.5 Selected HPLC Chromatograms 231 A.6 Bibliography 233 iii


15 1 Introduction 1 Introduction 1.1 LEWIS acidity The definition of acids and bases according to GN LEWIS GILBERT N. LEWIS laid fundamental building blocks in his 1916 published article [1] and his 1923 published work [2] Valence and the Structure of Atoms and Molecules for his best-known work, published in 1938, [3] in which he elaborated his now known definition of acids and bases: 1 ... a basic molecule is one that has an electron pair which may enter the valence shell of another atom to consummate the electron-pair bond; an acid molecule is one which is capable of receiving such an electron pair into the shell of one of its atoms. It was only in this work that he carried out the more general ideas of his previous work in more detail and showed that they correspond to an acid-base concept. In the meantime, similar concepts have been developed by SIDGWICK, [8] LAPWORTH [9] and INGOLD [10]. The term electron pair donors and electron pair acceptors, which LEWIS adopted, comes from SIDGWICK. Scheme 1: Acid-base reaction according to LEWIS. A = electron pair acceptor, D = electron pair donor. Scheme 1 shows the simplified concept of an acid-base reaction according to LEWIS. Here an acceptor, hereinafter referred to as LEWIS acid (LS), reacts with a donor, hereinafter referred to as LEWIS base (LB), in a neutralization reaction to form an acid-base adduct. Because of this very simple concept, a large number of atoms and molecules are included in the acid-base theory. LS can therefore include protons, cations of all kinds, neutral molecules with electron-poor atoms that can increase their coordination number, or electron-poor systems such as 1 and 2 (see Figure 1); the list for LB is similarly extensive. 1 In his work Valence and the Structure of Atoms and Molecules, he developed the theory of proton donors as acids and proton acceptors as bases independently of J. N. BRØNSTED [4], but did not develop it any further. [2,5] Today, LOWRY is often ascribed the independent development of the same theory, [6] however whether this was just as mature is not entirely clear. [7] 1

16 1.1 Lewis acidity Figure 1: Examples of acids according to LEWIS. AM = alkali metal, EAM = alkaline earth metal, M = metal, X = halogen. Due to the broad definition of acids and bases, there were numerous attempts in the following decades to develop theoretical models that describe the behavior of LEWIS acids and bases. Important contributions from quantum chemistry come from MULLIKEN, [11] derived from his studies on charge transfer complexes and donor-acceptor reactions in general. [12] KLOPMAN tried in his work on the reactivity of organic compounds, based on PEARSON's HSAB concept (hard and soft acids and bases), LEWIS acids and bases with the help of the pertubation theory of molecular orbitals ( MOs). In doing so, he developed the terms known today, charge-controlled and orbital-controlled, for donor-acceptor reactions. [13] A fertile field of theoretical chemistry emerged from these foundations, which, based on the FUKUI function, [14] deals with the prediction of reactivity using concepts such as maximum hardness [15] or global and local hardness and softness. [16] The origin of this theoretically rather demanding concept, the HSAB concept by PEARSON, [17] will be discussed below. PEARSON's concept of hard and soft acids and bases One problem with the definition of acids and bases according to LEWIS is the lack of a general reference particle , like the proton in BRØNSTED's theory, which can be used to scale the strength of an acid or base. The choice of a reference turned out to be unsuitable because the strength of a LEWIS acid does not only depend on its inherent strength but also on the LEWIS base. [3] Based on his own preliminary work with the EDWARDs equation, [18] the work of BJERRUM, [19] SCHWARZENBACH, [20] as well as AHRLAND, CHATT and DAVIES [21] regarding the stability of metal complexes in aqueous solution, PEARSON developed in 1963 [17] the HSAB concept. This semi-quantitative approach did not represent a scale for LEWIS acidity or basicity, but divided acids and bases into different classes with regard to their reactivity with one another. [22] Both acids and bases were divided into the classes hard, borderline or moderate and soft (see Figure 2). A common distinction is based on the polarizability of the 2nd

17 1 Introduction Particles made, hard acids and bases are difficult to polarize, while soft ones are easy. [23] According to KLOPMAN's theory, this corresponds to energetically high-lying LUMOs (lowest unoccupied molecular orbital) for hard acids and low-lying HOMOs (highest occupied molecular orbital) for hard bases. The opposite is the case with soft acids and bases. [23] From this, PEARSON derived two rules according to which hard acids react preferentially and quickly with hard bases and soft acids react preferentially and quickly with soft bases. This should make it possible to estimate reactivities, especially in the case of ambident particles such as CN, which can bond with either the carbon atom or the nitrogen atom with a LEWIS acid (or generally with an electrophile). Figure 2: Selection of some acids and bases that have been classified according to the HSAB principle. [5] In addition to its practical use for the rationalization of selectivities in organic synthesis, [16] the HSAB concept is subject to some criticism. [24] This also includes conceptual aspects, such as the use of complex stabilities in an aqueous environment instead of in the gas phase. For example, the hard F in the gas phase would form more stable complexes with almost all cations, regardless of their hardness, than with I, which is in contrast to the HSAB concept. Harsh criticism of the transfer of the HSAB concept to reactivity in organic transformations and the KLOPMAN theorem was recently expressed by MAYR. He came to the conclusion that the HSAB principle is not suitable for predicting the reactivity of organic, ambident reagents and is only used retrospectively to rationalize selectivities. [25] According to MAYR, a better description in the case of reactions under kinetic control is the MARKUS theory. Despite the problem that the strength of LEWIS acids and bases depends on the respective partner, [26] as 3

18 1.2 Quantification of Lewis acidity in the HSAB principle, many attempts have been made to establish acidity and basicity scales. In the following, the important studies on the quantification of LEWIS acidity within the scope of this work are discussed. Quantification of LEWIS acidity Basic information on the quantification of LEWIS acidity Before some of the most important studies on the quantification of LEWIS acidity are presented, an important conceptual distinction should be made . A study of the literature reveals three different approaches to quantifying LEWIS acidity. The first and oldest approach is based on the determination of thermodynamic data such as reaction and dissociation enthalpies or complex formation constants of the LEWIS acids with several bases or reference bases. The level of the enthalpies or the constants is then regarded as a measure of the acidity. Closely linked to this are modern, quantum chemical methods with which such data can now be calculated accurately. [26,28] This approach has its origins in inorganic and theoretical chemistry and primarily tries to determine the inherent acidity of LEWIS acids. In the second approach, the LEWIS acidity of acids is determined through their influence on a reference base, usually using spectroscopic methods. This is to determine the effect of the acids on the electronic structure. Low-electron LEWIS acids influence the electron density in the reference base, which can be detected using suitable spectroscopic methods. These methods increasingly find their origin in organic and organometallic chemistry, as they simulate the effect of LEWIS acids on LEWIS bases and enable digestions, e.g. can be obtained via catalytic activity. A third approach from theoretical chemistry deals with the calculation of the electron deficiency or the valence deficit of the LEWIS acids, but is not as extensive as those described so far. Therefore it is only briefly referred to here. [29,30] 2 Reference should be made to the extremely extensive work by LAURENCE and GAL, which deal with the creation of LEWIS basicity scales. [27] Therein several scales with different reference acids are presented and discussed, which can serve as an aid for the estimation of reactivities and synthesis planning. 4th

19 1 Introduction In the following some thermodynamic as well as spectroscopic methods are described, the latter will be dealt with in greater detail. Thermodynamic quantification methods After the LEWIS acid-base theory was established, some groups devoted themselves to determining the strength of LEWIS acids. As a rule, this work was not very extensive or only very specific. The first extensive studies were by D. P. N. SATCHELL and R. S. SATCHELL. [31] The complex formation constants of LEWIS acids MXn with a large number of LEWIS bases, including anilines, ketones or amines, were determined. The studies compared either LEWIS acids of an element (e.g. BX3) or a main group (e.g. MCl4). Although this meant that no extensive comparisons were possible, but the acidity trends that are generally known today such as BF3

20 1.2 Quantification of Lewis acidity, DRAGO continued to develop the D&O equation with MARKS, in which mono- and divalent cations are treated. [33] The currently most important and also very extensive thermodynamic method for quantifying LEWIS acidity is the fluoride ion affinity (FIA) established by CHRISTE. The FIA ​​corresponds to the enthalpy of the addition of a fluoride ion to a monomeric LEWIS acid in the gas phase (see Scheme 2). As can be seen from Scheme 2, the FIA ​​consists of an experimentally accessible part, the addition of fluoride to carbonyl fluoride (4), and a quantum-chemically determined term, the transfer of a fluoride ion from COF3 (3) to the LEWIS acid corresponds, together. This method was used because at the time of CHRISTE's studies, quantum chemical methods still provided imprecise results in the calculation of charged particles, such as the fluoride ion. [34] Scheme 2: Determination of the fluoride ion affinity (FIA), which is used as a measure of the LEWIS acidity. This method was used to determine the FIAs of over 100 LEWIS acids, some of which are shown in Table 1. 3 The FIAs of the LEWIS acids given in Table 1 correspond to expectations within the groups, i.e. the acidity increases with the period of the halide substituents (entries 1 4, F

21 1 Introduction Table 1: Selection of some FIAs as well as chloride ion affinities (CIAs), hydride ion affinities (HIAs) and methyl ion affinities (MIAs). Entry LS FIA / kj mol 1 a) CIA / kj mol 1 c) HIA / kj mol 1 c) MIA / kj mol 1 c) 1 BF BCl BBr BI AlCl3 457 (332) b) AlBr3 494 (393) b) AlI3 499 (393) b) GaF GaCl GaBr AsF5 430 c) SbF5 493 c) [531] d) 607 [544] d) a) Values ​​from Ref. [36], MP2 / PDZ level. b) Values ​​in brackets indicate the FIAs for the LS in standard conditions, i.e. H. fixed for AlX3. c) Values ​​from Ref. [30], G3 level. d) SbF5H and SbF5Me are unstable and would break down into SbF4 + HF and MeF, respectively. The values ​​of the formation of these decay species are given in square brackets. Furthermore, the reaction enthalpies are assessed as a whole and different effects such as reorganization energies in the fluoride ion addition are disregarded. Nonetheless, the FIA ​​is a simple method for estimating acidity if other values ​​are not available and is now widely used. [37] Due to the ease of access, other scales have been developed, such as the chloride-ion affinity (CIA), hydride-ion affinity (HIA) and methyl-ion affinity (MIA). [30] These take into account findings from the HSAB principle in order to evaluate the affinity of LEWIS acids with respect to soft bases in addition to the hard FIA. 5 This shows, for example, that BI3 (entry 4), according to its FIA, is less acidic than all aluminum-based LS (entries 5 7), but has a higher MIA than all of these and accordingly represents a softer but nevertheless very strong LEWIS acid Spectroscopic quantification methods In addition to the previously explained thermodynamic methods for quantifying LEWIS acidity, the spectroscopic methods offer some advantages. On the one hand, spectroscopic investigations are usually carried out in solution, i.e. solvent effects are taken into account. On the other hand, they quantify the acidity of an acid based on its effects on a LEWIS base. This makes it easier to estimate how the 5 The CIA, HIA and MIA from KROSSING are no longer based on COF2 data but are derived from the respective trimethylsilyl compounds. [30] 7

22 1.2 Quantification of Lewis acidity LEWIS acids behave as catalysts. Most of these methods therefore come predominantly from groups that can be assigned to organic chemistry. Scheme 3: Principle of the spectroscopic LEWIS acidity determination with a probe R D 6. = Wavelength of the radiation of the spectroscopy used. The principle of the spectroscopic methods is based on the fact that the complexation of a functional group of a reference base 6 causes a stronger lowering of the HOMO or the LUMO in complex 7 (see Scheme 3). Given a given spectroscopic method, this then leads to a hypsochromic or bathochromic shift of an observed signal. [38] The magnitude of the shift should correlate with the acidity of the acid. The first approaches come from LAPPERT and COOK from the 1960s and were carried out by means of IR spectroscopy (see Table 2), further studies used NMR, UV-Vis or ESR spectroscopy. The most extensive studies with the more common LEWIS acids and their scales are summarized in Table 2; further probes are listed in Figure 3. Figure 3: Further spectroscopic probes for the quantification of LEWIS acidity from D. P. N. SATCHELL and R. S. SATCHELL, [39] FUKUZUMI [40,41] and YOON. [42] The probes listed in Figure 3 do not appear in Table 2 for several reasons. D. P. N. SATCHELL and R. S. SATCHELL, who have already dealt with thermodynamic methods of investigating LEWIS acidity (see Chapter 1.2.2), tried to correlate complex formation constants of LS with perinapthenone (13) with the IR shift differences. No correlation was found and IR spectroscopic measurements were found to be unsuitable. [39] FUKUZUMI was able to show that the LEWIS acidity of acids can be quantified by ESR spectroscopic investigations of superoxide metal complexes. 8th

23 1 Introduction Table 2: Summary of some spectroscopic studies to quantify LEWIS acidity. Entry Developer Method Reference probe Examined LEWIS acids Sequence Reference 1 LAPPERT IR spectroscopy BF3, BCl3, BBr3, AlCl3. 9

24 1.2 Quantification of Lewis acidity These data showed a correlation with the fluorescence maxima of the LS complexes with 10-methylacridone (14). FUKUZUMI was able to show that 14 is suitable as a fluorescence probe for the determination of LEWIS acidities. [40,41] FUKUZUMI's studies mainly investigated metal perchlorates, which cannot be compared with the studies listed in Table 2. The alizarin probes 15 and 16, on the other hand, were used as solid-state probes by being vapor-deposited onto a substance to be examined. The height of the shift in the UV-Vis bands of the Alizarine 15 or 16 was then taken as a measure of the solid-state acidity. [42] Of the studies listed in Table 2, the methods of CHILDS with crotonaldehyde (12) as the 1 H-NMR probe [49] and those of BECKETT [47,48] with triethylphosphine oxide (11) as the 31 P- NMR probe 6 extensive use. [53 62] The prioritization of NMR spectroscopy over IR spectroscopic methods may be due to the fact that measurements under strictly anhydrous conditions in IR spectroscopy and the more complex IR spectra have been difficult to carry out for a long time. With the GUTMANN-BECKETT method, e.g. only the shift of a signal can be observed, the assignment of the bands in IR spectroscopic methods is not always easy. [39] A detailed comparison of the data in Table 2 is dispensed with, as the results are easy to interpret and, as a rule, similar sequences of the LEWIS acids were obtained. However, it is interesting to compare the spectroscopic methods with the FIA ​​from Table 1. Thus, with all spectroscopic methods, the boron halides are shown to be more acidic than the respective aluminum halides. This illustrates the difference between the thermodynamic acidity of an acid and its effect on a LEWIS base. Both the thermodynamic, the theoretical and the spectroscopic studies made useful contributions to the understanding of LEWIS acidity and basicity. Most of these studies, however, neglected one practical aspect that LEWIS linked with the acid-base term in his 1938 work [3]: the catalytic activity. Although many LFERs (linear free energy relationship) were developed in the course of the LEWIS theory, which linked physico-chemical parameters such as nucleophilicity with reactivity, only a few such studies were carried out on LEWIS acidity, although one of the main areas of application of LEWIS acids is their use as catalysts. [63] Studies such as that by PIERS, in which the less acidic PhB (C6F5) 2 (19) catalyzes an allyl stannylation more quickly than the more acidic B (C6F5) 3 (20), [64] illustrate, 6 The method is mostly used as GUTMANN -BECKETT method. [50 52] Other phosphine oxides such as triphenylphosphine oxide (17) or tri-n-butylphosphine oxide (18) are also often used. 10

25 1 Introduction that a link between acidity and kinetic studies is necessary. A series of studies that examined a link between the two parameters acidity and activity will be discussed in the next chapter Linking quantified LEWIS acidity with activity The only quantitative study that links LEWIS acidity with catalytic activity is a small study by LASZLO the 90s. [65] This investigated the theoretical basis of the use of a spectroscopic LEWIS acidity probe based on the work of CHILDS. [49,66] LASZLO was able to show that the shift difference (1 H) of the H-3 used by CHILDS as a measure of the LEWIS acidity in crotonaldehyde-LEWIS acid complexes correlates with the calculated energies of the lowest * orbital of the complexes having. [66] LASZLO then investigated the relationship between the * orbital energies and the catalytic activity of six LEWIS acids in the ene reaction between -pinene (21) and methyl acrylate (22) shown in Scheme 4. There was a clear correlation between the * - orbital energies and the catalytic activities of the LEWIS acids investigated (BCl3> AlBr3> AlCl3> EtAlCl2> Me2AlCl> Me3Al). Scheme 4: En reaction investigated by LASZLO to determine the activity of some LEWIS acids. [65] This means that spectroscopic methods for the quantification of LEWIS acidity are in principle suitable for predicting the catalytic activity of simple LEWIS acids. In addition to this singular study, some work has been carried out in recent years that qualitatively compared the activity of catalysts with their acidity. This includes work such as by CARLSON, who examined the relationship between various LEWIS acid parameters and their reactivity in alkylation, DIELS-ALDER and FRIEDEL-CRAFTS reactions. [67] BERKE investigated the influence of the acidity of boron-based LEWIS acids in mechanistic studies on the hydrogenation of alkenes. An inverse dependency was found, less acidic boranes gave higher TONs (catalytic productivity, English: turnover number). [68] In addition to this work, there are studies that deal not with the acidity as a parameter but with the selectivity and activity of LEWIS acids in catalyzed 11

26 1.2 Quantification of Lewis Acidity Reactions Deal. [69] For example, it is quite extensive. the work of KOBAYASHI, who examined about 40 metal chlorides for their activity and their selectivity with regard to the LEWIS-basic functionalities of the substrates in the reaction in Scheme 5. [70] Scheme 5: Reaction carried out by KOBAYASHI to study the activity and selectivity of metal chlorides. Ne = not specified, possibly neo-pentyl. [70] Based on the total yield of 27 and 28 after the given reaction time, the activity was divided into active (> 40%), weak (<40%) and inactive. From the ratio of 27 to 28, the chlorides were designated as aldehyde-selective (27:28 2: 1), aldimine-selective (28:27 2: 1) or neutral. Above all, metal chlorides in high oxidation states such as BCl3, TiCl4, ReCl5, FeCl3, MoCl5 or WCl6 were found to be active. Chlorides of the form MCl2 were particularly weak or inactive. It appears interesting that many of the active main group metal chlorides were rather aldehyde-selective, highly oxidized transition metal chlorides were neutral. The metal chlorides found to be weak were exclusively aldimine-selective or neutral. Finally, the quite new studies by SIGMAN should be mentioned, which are in the tradition of the classic LFER work, but in which new theoretical concepts and methods were used. [71,72 76] Various reactions and catalysts were investigated for the relationship between their selectivities and common physical-organic-chemical parameters, including acidities of the catalysts. Due to the high complexity of many modern synthetic methods and the progressive support of theoretical chemistry in organic-synthetic questions, research into the investigation of LEWIS acidity as a parameter has increased. Due to new areas of application of complex LEWIS acids, e.g. activation of small molecules like H2 by frustrated LEWIS pairs, this trend is likely to continue in the future. 12th

27 1 Introduction Preliminary work in the HILT group In the last few years, the HILT group has also undertaken work on the link between LEWIS acidity and the catalytic activity of acids. To quantify the acidity of LEWIS acids, the deuterated quinolizidine probe 29-d1 was reacted with an excess of the respective acid (see Scheme 6). 7 Scheme 6: Principle of using the deuteroquinolizidine probe 29-d1 to determine the LEWIS acidity of a LS based on the displacement difference (2 H). As a measure of the acidity of a LEWIS acid, the 2 H-NMR spectroscopically detectable shift difference (2 H) between complexed 30-d1 and free probe 29-d1 was chosen, analogous to the probes in Section 1.2.3. [78,79] As with the GUTMANN-BECKETT method, only one signal has to be observed, which is an advantage with more complex, chiral LEWIS acids. The acidities obtained for common LEWIS acids coincided with the results of CHILDS. [49,80] Furthermore, the investigated LEWIS acids were used as catalysts in three different reactions. MÖBUS was able to show that from a threshold acidity a correlation of the shift differences with the catalytic activity of the LEWIS acids in the DIELS-ALDER reaction between 1,4-naphthoquinone (31) and 2-methyl-1,3-butadiene (32, isoprene ) exists (see Scheme 7). [78] LEWIS acids with low acidity catalyzed the reaction extremely weakly or not at all. SnCl4 and TiCl4 showed a deviation from this correlation, which catalyzed less than the (2 H) values ​​suggested. In two further reactions, a POVAROV reaction and an intramolecular cyclization of a diazoester, PÜNNER was also able to show a correlation between the quantified LEWIS acidity and the catalytic activity of the acids. [81] With this work it could be shown that in the case of simple LEWIS acids there is a connection between spectroscopically determined acidity and catalytic activity. 7 The probe 29-d1 was already used by BOHLMANN in IR spectroscopic investigations into the structure of alkaloids. [77] 13

28 1.2 Quantification of Lewis acidity Scheme 7: Reaction investigated by MÖBUS to quantify the catalytic activity of various LEWIS acids. [78] As part of my diploma thesis, the acidity of silyl triflates was to be investigated in addition to the development of easily accessible deuterated 2 H-NMR probes. [82] The deuteroquinolizidine probe 29-d1 was found to be unsuitable because the adducts were unstable. Deuteropyridine 34-d5 could be used as an alternative probe. The reactivity of the silyl triflates was determined between 31 and 32 in the DIELS-ALDER reaction described by MÖBUS. A moderate correlation between the (2 H) values ​​and the catalytic activity was found here. Me3SiOTf (35) showed a significantly increased reactivity compared to the other investigated silyl triflates, which could not yet be explained. [83] However, a problem in this study was the influence of possible traces of H2O in the dried solvents. [84] The hydrolysis of the silyl triflates can produce HOTf, which is also able to catalyze the reaction. Initial investigations revealed unclear tendencies as to whether there was a hidden proton catalysis, but the results that spoke against it predominated. [82] 14

29 2 Task 2 Task Based on the studies of our working group [78,80 83] already presented in the previous chapter on the quantification of LEWIS acidity and its connection with the catalytic activities of LEWIS acids, more extensive studies should be carried out within the scope of this work. In contrast to the previously used, simple metal halide and silyl triflate-based LEWIS acids, such as AlCl3, BF3, ZnI2, Me3SiOTf or t-bume2siotf, it was planned to investigate more unconventional and also new LEWIS acidic catalysts used in modern organic synthesis. It was of interest whether these systems can be quantified in a similarly simple manner and, in particular, whether correlations with their catalytic activity can be found. This should also try to address specific issues relating to these LEWIS acid classes (see following chapter). 2.1 Quantification of the LEWIS acidity of silylium ions Through previous studies on silicon-based LEWIS acids as part of our own diploma thesis, [82] we became aware of the ferrocenylsilylium ions 36 presented by OESTREICH [85,86]. These highly reactive, trivalent silicon cations (hereinafter referred to as silylium ions) are known as potent LEWIS acid catalysts in DIELS-ALDER reactions (see Scheme 8). a) b) Scheme 8: a) Basic structure of the class of ferrocenylsilylium ions presented by OESTREICH 36. b) Example of a demanding DIELS-ALDER reaction catalyzed by 36a. Due to their hitherto difficult synthesis and manageability, silylium ions are a class of highly reactive LEWIS acids that is little used in catalytic organic synthesis and therefore little studied. In cooperation with the OESTREICH working group, it should be discussed whether this class of LEWIS acids can be quantified analogously to known, simple LEWIS acids from our previous studies [78,80 83] and correlated with their catalytic activity. The experimental quantification of the LEWIS acidity should be undertaken with one of our tried and tested 2 H-NMR probes [80, 83] (see Figure 4). The catalytic activities were determined by means of in situ Fourier transform infrared spectroscopy (isftir-15

30 2.2 Carbenium ions as Lewis acid spectroscopy) using a silylium ion-catalyzed DIELS-ALDER reaction established in the OESTREICH working group. Figure 4: Deuterated 2 H-NMR probes used in previous studies to quantify LEWIS acidity. After obtaining the two parameters LEWIS acidity and catalytic activity, possible correlations between them should be sought and, in the absence of correlations, potential causes should be investigated. The durability of the current practice of using the 29 Si-NMR shifts of silicon-based LEWIS acids (SLS) as a measure of their LEWIS acidity [87,88] should be further examined. 2.2 Carbenium ions as LEWIS acids In contrast to the silylium ions already mentioned, trivalent carbon-based cations (hereinafter referred to as carbenium ions), in particular triphenylcarbenium ions (hereinafter referred to as trityl ions, see Figure 5) have been known for over 100 years. [89] Due to the positive charge, these organic cations have LEWIS acidic properties and, like the homologous silylium ions, should be able to act as catalysts. Nevertheless, apart from a few examples, they have received little attention in organic catalysis. [90] A successful application of tritylions as Figure 5: General structure of tritylions and their features. WCA = weakly coordinating anion. Catalysts in DIELS-ALDER reactions were not yet known at the beginning of this work. Inspired by the similarity and high reactivity of the homologous silicon compounds, the aim of the present work was to investigate the catalytic activity of a number of differently substituted trityl ions with regard to DIELS-ALDER reactions. When using trityl ions as LEWIS acids, there is often the problem that very strong BRØNSTED acids formed by hydrolysis of the cations act as additional catalysts. [91,92] Therefore, the question regarding the actual catalytically active species cation or proton should be examined. 16

31 2 Task 2.3 LEWIS Acid Activation of Oxazaborolidines In addition to the projects planned so far, the subject of which would be the investigation of rather unconventional LEWIS acids, systems that have become established in practice should also be studied. A relatively new method, the use of LEWIS and BRØNSTED acids to activate chiral, weak LEWIS acids, was used by E. J. COREY a few years ago. He first used BRØNSTED [93] and later LEWIS acids, [94] to create formidable catalysts for enantioselective DIELS-ALDER reactions by activating chiral, proline-based oxazaborolidines 41, better known as CBS catalysts (CBS: COREY-BAKSHI- SHIBATA) (see Scheme 9). In the case of the LEWIS acids, however, only AlBr3 proved to be effective. A number of other LEWIS acids, from the moderately strong BF3 Et2O to SnCl4 to the very strong BCl3, were found to be poorly activating or not very promising. These observations were not explained in more detail, which is why the activation of oxazaborolidines by LEWIS acids should be investigated in this work. The aim was to check whether the activation of oxazaborolidines 41 by LEWIS acids could be determined quantitatively. This would take place in analogy to the quinolizidine-based probe 29-d1 2 H-NMR spectroscopy using a deuterated CBS catalyst 41-d1. Scheme 9: Planned method for quantifying the activation of a Deutero-CBS catalyst 41-d1 by LEWIS acids by 2 H-NMR spectroscopy. If activation values ​​of different levels could be measured for different LEWIS acids, kinetic measurements of catalyzed DIELS-ALDER reactions should be used to check whether there is actually a threshold acidity as observed by COREY, or whether there are other parameters for the different levels of activation caused by different ones LEWIS acids are responsible. Since the LEWIS acid activated CBS catalyst is a chiral catalyst, an additional attempt should be made to include the enantioselectivity of the investigated reactions in a correlation between LEWIS acid strength and catalytic activity. 17th

32 2.4 Quantification of the hydrogen bond activation strength of thioureas 2.4 Quantification of the hydrogen bond activation strength of thioureas A class of catalysts that at first glance do not appear to be LEWIS acids are the thioureas, which have been used in asymmetric catalysis for a number of years. These organocatalysts are able to activate functional groups for transformations through hydrogen bonds. SCHREINER was able to show that thioureas can behave analogously to classic LEWIS acids. [95] If, -unsaturated carbonyls such as 42 are reacted with a thiourea, adducts 43a 42 are formed, which are similar in structure and reactivity to the adducts of 42 with classic LEWIS acids such as SnCl4 (see Scheme 10). [96,97] This adduct formation leads to a lowering of the LUMO of 42 and activates the substrate, e.g. for DIELS-ALDER reactions. [95,98] Scheme 10: Complexation of oxazolidinone 42 with thiourea 43 or the LEWIS acid SnCl4. Based on our experience with classic LEWIS acids, it was of interest whether this similarity of thioureas to classic LEWIS acids is reflected in a possible quantifiability of their hydrogen bridge activation strength. The aim of this work was therefore to develop a method to be able to easily quantify the activation strength of the thioureas and to subsequently investigate whether a correlation between this parameter and the catalytic activity of the thioureas could be found. 18th