ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2019

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1810

Host-Guest Interactions forStructural Analysis of OrganicMolecules

Development of new Tools for StereochemicalCharacterization

SANDRA OLSSON

ISSN 1651-6214ISBN 978-91-513-0660-5urn:nbn:se:uu:diva-382219

Dissertation presented at Uppsala University to be publicly examined in Room A1:107, BMC,Husargatan 3, Uppsala, Friday, 14 June 2019 at 09:15 for the degree of Doctor of Philosophy.The examination will be conducted in English. Faculty examiner: Professor Kari Rissanen(University of Jyväskylä, Department of Chemistry).

AbstractOlsson, S. 2019. Host-Guest Interactions for Structural Analysis of Organic Molecules.Development of new Tools for Stereochemical Characterization. Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty of Science and Technology 1810. 63 pp.Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-513-0660-5.

The focus of this thesis is on the development of molecular tweezers as host-guest systems forstereochemical characterization of small organic molecules.

There are two central problems to stereochemical characterization of small molecules. Thefirst is that there are few methods for the determination of absolute stereochemistry and thecomplexity of the task increases with the number of chiral centres in the molecule. The secondis the flexibility of small molecules. The data received from NMR spectroscopy, the standardmethod for structural characterization is an average of all conformers present in solution, makingexact determination of the structure challenging if stereocenters are involved.

This research group has previously developed two bis-porphyrin molecular tweezers thatcan be used as hosts for relative stereochemical determination by NMR (NAMFIS) by limitingthe flexibility of the guest molecules and for absolute stereochemical determination Exciton-Coupled Circular Dichroism (ECCD) by providing suitable chromophores. However, the use ofthese tweezers was limited to diamine molecules.

To extend the application of both tweezer host systems, a range of metalloporphyrins havebeen investigated both computationally and experimentally to identify candidates bindingstrongly to oxygen containing functional groups (Paper I). One of the porphyrins identified hasbeen used to synthesise a new system with the potential to be used for relative stereochemicaldetermination by NMR (NAMFIS) of molecules with a wide variety of functional groups (PaperIV).

To further develop the systems, the photo-isomerization properties of the stiff stilbenelinker of one of the tweezers has been investigated. Macrocyclic diether model compounds,incorporating the stiff stilbene linker have been used to show that the photo-isomerizabilitystrongly is affected by molecular strain (Paper II). The results suggest that if paired with asuitable porphyrin (with no UV/Vis absorption overlap) a photo-switchable tweezer might beconstructed.

To map the guest range of the stiff stilbene linked tweezer it was found that this tweezer givesstrong to moderate signals for monoamines with a rigid structure (Paper III). As ditopic bindingis usually a requirement for an ECCD signal when using bis-porphyrin molecular tweezers, thisis a discovery that should be of general interest.

Keywords: metalloporphyrins, host-guest systems, bis-porphyrin tweezers, stiff stilbenemacrocycles, organic synthesis, porphyrin binding studies, NMR, ECCD

Sandra Olsson, Department of Chemistry - BMC, Box 576, Uppsala University, SE-75123Uppsala, Sweden.

© Sandra Olsson 2019

ISSN 1651-6214ISBN 978-91-513-0660-5urn:nbn:se:uu:diva-382219 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-382219)

Is it impossible, or is it impossiblefor you?

PhD-comic

To Jakob and Sonja

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Olsson, S., Dahlstrand, C., Gogoll, A. Design of oxophilic met-

alloporphyrins: an experimental and DFT study of methanol binding. Dalton Trans., 2018, 47, 11572-11585.

II Olsson, S., Perez, O. B., Blom, M., Gogoll, A., Effect of Ring Size on Photoisomerization Properties of Stiff Stilbene mac-rocycles. Submitted.

III Olsson, S., Schäfer, C., Blom M., Gogoll, A. Exciton-Coupled Circular Dichroism Characterization of Monotopically Bind-ing Guests in Host-Guest Complexes with a Bis-(zinc porphy-rin) Tweezer. ChemPlusChem, 2018, 83, 1169-1178.

IV Olsson, S., Gogoll, A. Synthesis of a novel bis-Mg-porphyrin

tweezer. Manuscript.

V Appendix

Reprints were made with permission from the respective publishers. Publications not included in thesis:

VI Bunrit, A., Dahlstrand, C., Olsson S. K., Srifa P., Huang G., Or-thaber A., Sjöberg P. J. R., Biswas S., Himo F., Samec J. S M. Brønsted Acid-Catalyzed Intramolecular Nucleophilic Sub-stitution of the Hydroxyl Group in Stereogenic Alcohols with Chirality Transfer. J. Am. Chem. Soc. 2015, 137, 4646-4649.

VII Bunrit A., Dahlstrand C., Srifa P., Olsson S. K., Huang G., Biswas S., Himo F., Samec J. S. M. Nucleophilic Substitution of the Hydroxyl Group in Stereogenic Alcohols with Chiral-ity Transfer. Synlett, 2016, 27, 173-176.

Contribution

The author wishes to clarify her contribution to the research presented in the present thesis.

I Planned the study, performed the synthesis, measurements, eval-

uations and part of the DFT calculations. Wrote the first draft of the paper and contributed to the revised final manuscript.

II Contributed to the planning of the study. Screened synthetic routes to the target compounds. Supervised the synthesis and photochemical investigations. Performed the computational study. Wrote the manuscript.

III Contributed to the planning of the study, performed part of the ECCD, UV/Vis measurements and evaluations. Performed the binding studies and molecular modelling studies. Supervised the synthesis and ECCD and UV/Vis measurements. Wrote the first draft of the paper and contributed to the revised final manuscript.

IV Planned the study and performed the synthesis, characterization

and preliminary measurements. Wrote the manuscript.

Contents

1  Introduction ......................................................................................... 11 1.1  Working with small organic molecules ........................................... 11 

1.1.1  Conformation, they just won’t be still ................................... 11 1.1.2  Configuration, the evil twin problem ..................................... 12 

1.2  Stereochemical determination ......................................................... 13 1.2.1  The use of NMR spectroscopy ............................................... 14 1.2.2  The use of Exciton-Coupled Circular Dichroism .................. 15 

1.3  Enter molecular tweezers! ............................................................... 16 1.3.1  Design of the systems ............................................................ 17 1.3.2  The problem with alcohols .................................................... 20 1.3.3  The problems with size and flexibility ................................... 20 

1.4  To solve the problems, aims of this thesis ...................................... 21 1.4.1  Porphyrins vs alcohols ........................................................... 21 1.4.2  Stiff stilbene linker................................................................. 22 1.4.3  Tweezers ................................................................................ 23 

2  The toolbox, overview of methods ...................................................... 25 2.1  Synthesis and complexations .......................................................... 25 

2.1.1  Synthesis ................................................................................ 25 2.1.2  Complexations ....................................................................... 26 

2.2  Binding studies ................................................................................ 27 2.2.1  Binding stoichiometry ............................................................ 28 2.2.2  Binding constants ................................................................... 29 

2.3  Nuclear magnetic resonance ........................................................... 29 2.3.1  Chemical shifts ...................................................................... 30 2.3.2  Scalar coupling constants ....................................................... 30 2.3.3  Nuclear Overhauser effects .................................................... 31 

2.4  ECCD .............................................................................................. 31 2.5  Computational methods .................................................................. 31 

2.5.1  Molecular mechanics ............................................................. 31 2.5.2  Density functional theory ....................................................... 32 

3  Porphyrins and guest interaction (Paper I) .......................................... 33 3.1  The plan .......................................................................................... 33 3.2  Synthesis ......................................................................................... 34 3.3  Complexation experiments ............................................................. 35 

3.4  Computational predictions .............................................................. 36 3.5  Conclusions ..................................................................................... 38 

4  Linker dynamics (Paper II) .................................................................. 41 4.1  The plan .......................................................................................... 41 4.2  The cyclic compounds .................................................................... 41 4.3  Switching study ............................................................................... 42 4.4  Conclusions ..................................................................................... 43 

5  Molecular tweezers (Papers III and IV) ............................................... 45 5.1  A plan of three parts ........................................................................ 45 

5.1.1  A Twisted Tweezer (i) ........................................................... 45 5.1.2  The elusive trans-Tweezer (ii) ............................................... 48 5.1.3  The new and exciting “Fleezer” (iii) ...................................... 50 

5.2  Conclusions ..................................................................................... 52 

6  Conclusion and outlook ....................................................................... 53 

7  Sammanfattning på svenska ................................................................ 55 

8  Acknowledgements.............................................................................. 57 

9  References ........................................................................................... 59 

Abbreviations

CD Circular Dichroism ECCD Exciton-Coupled Circular Dichroism DCM Dichloromethane DFT Density Functional Theory DMF Dimethylformamide DMP Dess-Martin Periodinane HATU (1-[Bis(dimethylamino)methylene]-1H-1,2,3-tria-

zolo[4,5-b]pyridinium 3-oxid hexafluorophosphate HBTU (2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethylu-

ronium hexafluorophosphate IR Infrared spectroscopy MALDI Matrix Assisted Laser Desorption/Ionization MM Molecular Mechanics MS Mass Spectrometry NAMFIS NMR Analysis of Molecular Flexibility in Solution NBS N-Bromosuccinimide NMR Nuclear Magnetic Resonance NOE Nuclear Overhauser Effects SS Stiff Stilbene TLC Thin Layer Chromatography TPP Tetraphenylporphyrin TPPBr8 Octabromo tetraphenylporphyrin TPFPP Tetra(pentafluorophenyl)porphyrin TPFPPBr8 Octabromo tetra(pentafluorophenyl)porphyrin UV/Vis UV-Visible spectroscopy XRC X-Ray Crystallography

11

1 Introduction

1.1 Working with small organic molecules Small molecules are at the core of organic chemistry. These have a variety of use in both research and industry; natural products, drug candidates, building blocks for polymers and functional materials, and as precursors in organic synthesis. Whether they are isolated natural products or a synthetic target, the challenge of characterization is always present. Too small for visual inspec-tion, a variety of techniques have been developed to help answer the tricky question of what the structure of an organic compound actually is. To under-stand the nature of the difficulties there are a couple of fundamental properties that need to be discussed.

1.1.1 Conformation, they just won’t be still Whenever discussing small molecules their structure is usually neatly drawn as a skeletal formula. This might be a good way of representing how the atoms are connected but gives the false impression that these compounds are flat and rigid. As organic chemists usually work with solutions, this is a poor assump-tion. The molecules will be in constant intramolecular motion involving bend-ing, rotating and stretching of covalent bonds. The different forms these mol-ecules can adopt through motion are called conformers. The effect of this on the analytical techniques used to characterize compounds largely depends on the property measured, the time scale and the temperature. For the recording of UV-Visible spectroscopy (UV/Vis), mass spectrometry (MS) and infrared spectroscopy (IR) spectra this is a minor problem as the properties measured (the UV/Vis absorption, molecular mass and IR absorption) are not greatly affected by the change in conformation. However, the main technique for de-termining the three-dimensional structure of small molecules is Nuclear Mag-netic Resonance (NMR) spectroscopy. This technique has a comparably long time scale.1 The result is that the spectrum received will be an average of all the conformations present in solution. As the sample needs to stay in solution for the technique to work properly, lowering the temperature is a very limited solution to this problem. The information that can be extracted about a mole-cule will therefore be partly dependent on its flexibility, with more cyclic and more rigid structures being easier to fully characterize.

12

1.1.2 Configuration, the evil twin problem The configuration of a small molecule is best represented by the before men-tioned skeletal formula. This is by far the most popular way to represent or-ganic molecules as the positions of atoms and bonds are rendered very com-prehensive. From methods such as UV/Vis, IR and MS information about the size of the molecule and the kind of bonds and functional groups present can be gained. However, to determine the way these are arranged in space NMR spectroscopy or x-ray crystallography (XRC) are commonly used. As small flexible molecules often do not crystalize well options might be limited. There are advances being made using crystalline molecular sponges that might widen the scope of XRC to include more of these challenging systems.2,3 In the meantime NMR spectroscopy might be the only option for some of them. NMR usually yields the basic bonding pattern without major difficulty. 1H and 13C NMR spectroscopy utilizing various two-dimensional methods are rou-tinely used to elucidate the three-dimensional structure of small molecules. There are limitations however, as the time scale for NMR measurements al-lows for extensive molecular motion. The spectra received are averages of a large number of conformers which makes the finer details of molecular struc-ture difficult to determine.

The class of compounds where the characterization process runs in to major problems is chiral molecules. These are pairs of molecules with the same con-figuration, close to identical in all respects but one: they are non-superimpos-able mirror images of each other (Figure 1).

Figure 1. Example of three enantiomeric pairs (1,2-Dichlorocyclopentane, alanine and thalidomide). These molecules are non-superimposable mirror images of each other.

These compounds are called enantiomers and as a result of their high similar-ity they have the same UV/Vis, IR, MS and NMR spectra.4 One clear differ-ence between enantiomers is that they bend plane polarized light in opposite

13

directions. As of yet, there has not been any spectroscopic methods developed that can relate optical rotation to absolute stereochemistry. In organic mole-cules the major source of chirality are stereogenic centers. These are sp3-hy-bridized carbon atoms binding four different groups. Stereogenic centers are characterized as R or S according to the Cahn-Ingold-Prelog system.5 Another important source of chirality is helical or axial chirality. The enantiomers dis-playing helical chirality are characterized as either P or M by using a variation of the Cahn-Ingold-Prelog system.6,7 Examples of both types are shown in Figure 2.

Figure 2. In molecules with stereogenic centers (e.g. alanine) the centers are charac-terized as R or S. In molecules with axial chirality (e.g. heptahelicane) the enantio-mers are characterized as P and M.

Now to the obvious question: Why bother? It is easy to assume (and has been assumed historically) that the similarity of the enantiomers will result in iden-tical properties in most contexts. The main fault of this reasoning is the as-sumption that chirality is an unusual property. Chirality is very prevalent in biological systems which therefore interact differently with different enantio-mers the same way a right foot will comfortably fit a right shoe but not a left one.8 Among the examples in Figure 1 the infamous thalidomide is shown. This compound was used to treat morning sickness in pregnant women in the late 1950s and early 1960s but it turned out that while one of the enantiomers had the advertised pharmaceutical effects the other caused birth defects.9 Be-cause of the possibility of enantiomers having radical different interactions within biological systems, the need to identify, control, and be able to deter-mine chirality, has become a very central part of organic and pharmaceutical chemistry.10

1.2 Stereochemical determination The process of elucidating the chirality of molecules is referred to as stereo-chemical determination. When discussing this subject, it is important to be

14

clear about what is actually being referred to. Stereochemistry is often dis-cussed as either absolute or relative. Absolute stereochemistry refers to how the atoms are arranged in space. When this information is known the stereo-genic centre/centres can be assigned as R or S. The relative stereochemistry is when a compound has several stereocentres and only their configuration rela-tive to each other is known. For example, it might be determined that the rel-ative configuration of a compound with two stereocentres is either (RS) or (SR) but it cannot be determined which. Relative stereochemistry is denoted with a star (R*) or with rel/relative written in front of the compound name.11

Determination of stereochemistry is preferably done by XRC analysis.12-14 This produces a 3D image of the molecule where the stereocentres can be as-signed as R or S absolutely. However, not all compounds crystalize well. Ad-ditional methods (besides XRC) used for stereochemial characterization in-clude Exciton-Coupled Circular Dichroism (ECCD) and methods based on NMR spectroscopy. Stereochemical determination is always tricky, especially so relative stereochemical determination of molecules with multiple stereo-centers. There are a number of cases where compounds have been wrongly assigned.15,16 In these cases molecular flexibility plays a big part, making spec-tra difficult to interpret.

1.2.1 The use of NMR spectroscopy As NMR spectroscopy is the go-to method for characterization of small mol-ecules it is natural to try to make it work for stereochemical determination as well. The most straightforward way to achieve this is by reactions that convert enantiomers into diastereomers.17,18 As these will have different chemical shifts it is possible to determine which one has formed and from that deduce which enantiomer was present in the starting material. The clear downside of this is that it requires the derivatization of the molecule of interest.

There are several methods to determine the relative stereochemistry of small molecules by NMR spectroscopy. These include the use of chemical shift re-agents and the analysis of J-couplings and computational fitting of experi-mental data to similar known compounds.19-22 These methods work reasonably well for rigid molecules but give less robust results with flexible compounds. A method developed to fill this gap is NMR Analysis of Molecular Flexibility in Solution (NAMFIS).23,24 This is a method used to identify the most abun-dant conformers of a molecule in solution. The method involves matching li-braries of conformers, calculated by molecular mechanics (MM), to experi-mentally measured NMR data such as nuclear Overhauser effects (NOE) and scalar J-couplings. However, the conformational space can be too big for NAMFIS to successfully determine the relative stereochemistry for flexible

15

molecules without some restrictions of the molecular movement. A clear ad-vantage of the NMR based methods is that compounds with several stereocen-ters can be analysed.

1.2.2 The use of Exciton-Coupled Circular Dichroism

Exciton-Coupled Circular Dichroism (ECCD) is a technique for the determi-nation of the absolute stereochemistry of chiral molecules with two or more chromophores.25 The interaction of the electric transition moments of the chro-mophores will give rise to Cotton effects and the spectrum will display char-acteristic bisignate curves.26 The signs of the couplets reflect the chirality of the investigated molecule. As not all molecules of interest have two interact-ing chromophores supramolecular complexes have been used to make ECCD a more general method.27 The molecule of interest is complexed with a host-molecule that contains the chromophores and will adapt its structure to the chirality of the guest molecule (Figure 3).

Figure 3. Top: tweezer conformation when free (left) and when complexing chiral guests (middle and right). Bottom: Corresponding CD signal. When molecular twee-zers are used for ECCD analysis the tweezer itself give no CD signal. When a chiral guest is complexed, the tweezer adopts the chirality and the complex will generate a CD signal. The enantiomer will give the opposite signal.

With the help of computational methods, the M- and P-enantiomers formed by the host-guest system together with a chiral guest can be related to either the positive or negative Cotton effects. By checking in which complex (M or P) a chiral guest would be most stable and comparing this to the experimen-tally measured CD signal the absolute stereochemistry of the guest can be elu-cidated.28 A downside of this method is that it is limited to compounds with one or two stereocentra.

16

1.3 Enter molecular tweezers! In order to address both the issues with flexibility and the need for chromo-phores, molecular tweezers have been used (Figure 4). They have successfully worked as an aid for determination of both relative and absolute stereochem-istry by NMR and ECCD.29-31

Figure 4. Molecular tweezers consist of two tips containing binding sites connected by a linker. Guest molecules coordinate ditopically to the binding sites.

A molecular tweezer can be divided into parts which fill different functions. They generally consist of two tips whose function is to be able to coordinate ions or functional groups. The tips are connected by a linker which keeps them at a certain distance. The linker can provide additional functionalities such as rigidity, a secondary binding site or a dynamic function for changing the size of the tweezer. Molecular tweezers are part of host-guest systems that can bind guest molecules ditopically and in that way restrict molecular movement (Fig-ure 5). If a tweezer contains two or more chromophores it also meets the re-quirements for ECCD and UV/Vis investigations.

Figure 5. Host-guest complex formed between a molecular tweezer and guest mole-cule.

17

1.3.1 Design of the systems There are many reported molecular tweezers.32,33 They are usually designed with a certain application in mind and therefore vary a lot in components and properties. However, the basic template of a molecular tweezer remains the same; you need two tweezer tips that can coordinate guest molecules and a linker in between (Figure 4).

1.3.1.1 Porphyrin tweezer tips Metalloporphyrins are well known for their ability to coordinate to guest mol-ecules in the axial position.34 A lot of them are commercially available and the required reactions for linker attachments are well established. Because of this they are a popular choice for the tips of molecular tweezers. Depending on the intended application, the free base or different metalloporphyrins have been used (Figure 6).28,35

Figure 6. A basic porphyrin consists of four linked pyrrole units. The positions on the pyrroles are called -positions and the positions on the linkers are called meta-posi-tions. A porphyrin that is not coordinating a metal cation is called a free base (e.g. P or TPP), a porphyrin that is coordinating a metal cation is called a metalloporphyrin (e.g. MP or MTPP).

Porphyrins also have strong UV/vis absorption and the use of bis-porphyrin tweezers for determination of the absolute stereochemistry of guest molecules by ECCD is an established method.29

Metalloporphyrins also have interesting properties regarding NMR spectros-copy. Porphyrins are aromatic systems and aromatic systems have delocalized -electrons that are free to move around. In the presence of a magnetic field these electrons will start to circulate, generating a ring current. This creates an anisotropic local magnetic field in the direction opposite to the inducing magnetic field (Figure 7).

18

Figure 7. The porphyrin is an aromatic system. In the presence of a magnetic field (B0) the electrons of this system will start to circulate, generating a ring current. This current generates a local magnetic field in the opposing direction to the one applied, i. e. B0. Above and below the plane of the ring the applied magnetic field will be partially canceled causing shielding ( < 0) and in the plane of the ring reinforcement between the magnetic fields results causing deshielding ( > 0).

This kind of anisotropy effect is called the ring current effect.36-38 This effect is strong for metalloporphyrins and guest molecules coordinated to the metal will have lowered chemical shifts. In molecular tweezers the guests are posi-tioned in between two metalloporphyrins, resulting in a large combined effect of two porphyrin rings. Thus, the chemical shifts for the entire guest molecule may move into the negative region of the NMR spectrum. This can be very advantageous as this area is usually free of signals from the tweezer and con-taminants. Additionally, the effect will depend on the individual nuclei’s phys-ical distance to the porphyrin so signals that might otherwise overlap might become separated from each other.

1.3.1.2 Tweezer linkers Many different linkers have been used in molecular tweezers, from simple carbon chains to more complex crown ether and isophthalic acid linkers.28,29,35 The choice of linker is central for the application of the tweezer. If the intent is ECCD, a flexible linker might be able to accommodate a larger range of guests. However, if the intent is to limit the flexibility of the guest a more rigid linker structure is necessary. Based on previous studies in our group, two link-ers with very interesting properties have been selected (Figure 8).39,40

19

Figure 8. Linkers for molecular tweezers that have been shown to have interesting properties in previous studies by this group. The glycoluril linker (top) has a semi-rigid structure. Because it is covalently bound to two -positions on the porphyrin no rotation of the porphyrin is possible. The Z-stiff stilbene linker (bottom) has a slightly pre-twisted structure that produces a tweezer with inherent axial chirality. This linker is attached to one of the -positions of the porphyrin, allowing for rota-tion of the porphyrin tips.

The first is the glycoluril linker. This is a semi-rigid linker that is attached to two of the porphyrin -positions, thereby preventing rotation of the porphyrin units. The glycoluril linker has some flexibility when it comes to the gap size.39 This research group has previously shown that NAMFIS, combined with a glycoluril-based bis-porphyrin tweezer can be used to determine the relative stereochemistry of small flexible diamines with multiple stereocen-ters.41 The other interesting linker is the Z-stiff stilbene linker. Due to steric hindrance, this linker is slightly skewed in the Z-conformation resulting in ax-ial chirality. The tweezer synthesized with this linker is a racemic mixture but adapts to chiral guests and initial studies on diamines showed very promising results in ECCD.40 The inherent twist of the linker seems to enhance the CD response.

1.4 Presenting the problems There are several good proofs of concept both from this group and others for the use of bis-porphyrin molecular tweezers for determination of both absolute and relative stereochemistry.29-41 The main limitations are centered on the range of guests that can be analyzed. The factors central to this point are the guests ability to bind to the metalloporphyrin, the size of the guest and the flexibility of the guest. The first problem is mainly porphyrin related whereas the second and third are linker related.

20

1.4.1 The problem with alcohols Most of the bis-porphyrin tweezers reported are based on the ZnTPP unit (Fig-ure 6).28,29,35 This porphyrin binds very well to amine guests and consequently diamines have been extensively studied, especially by ECCD.42-44 However, the ZnTPP binds poorly to oxygen containing functional groups.45 A simple web search on the catalogue of, for example Sigma Aldrich will show that hydroxyl groups are a lot more common than amines, ca 2:1. Add carboxylic groups and ketones and the number of possible oxygen-containing guests is huge. Nature is a great source of chiral compounds and here the hydroxyl group is also very prevalent.46-48 With this in mind, a natural next step in im-proving the versatility of bisporphyrin tweezers is to modify the porphyrins to accommodate oxygen containing functional groups as well as amines. A strong binding to the tweezer is critical for the NMR-based methods. There will usually be an exchange between bound and free guest in the analyzed sample and if this is rapid, it will make the NMR-signals very broad. Another problem with low affinity to the tweezer is that a huge excess of guest might have to be added for enough guest to be bound to the tweezer. In ECCD this will not affect the data as the signal of the tweezer is studied but in NMR were the signal of the guest is followed the signal of the bound guest might drown in the signal of the free guest. To facilitate both ECCD and NMR studies a metalloporphyrin binding oxygen-containing functional groups on a level similar to ZnTPP binding amines would be ideal.

1.4.2 The problems with size and flexibility The size of the guests is a limitation in several ways. First of all, ditopic bind-ing requires two binding sites on the guest molecules. For small chiral guests two functional groups that can bind to the tweezer might not be available. To solve this, derivatization of the guest to provide a second binding site has sometimes been used.44 But as modification of the guest itself should always be a last option other solutions should be sought. So far all reports of ECCD analysis of monotopically binding guests have included some kind of second-ary binding to the linker or functional groups on the porphyrin rim.49-53 Since relatively few studies have looked into this there might be room for improve-ments. Another limiting factor regarding the size is if the guest is larger than the twee-zers gap size. A tweezer that could either adapt itself via some mechanism or exist in different sizes would be desirable to widen the range of guests it could accommodate. Different sizes of tweezers could also help to limit the flexibil-ity of some guest molecules. Some example of switchable tweezers have been reported and the concept shows promise (Figure 9).28

21

Figure 9. Example of a switchable molecular tweezer with two configurations. A long flexible guest could potentially be stretched out by such a system, facilitating more detailed analysis.

However, a point in favor of a more modular system is that you need a tweezer of different rigidity depending on guest and analysis. For the NMR-based methods little molecular movement is better whereas the ECCD and UV/Vis applications are less affected by this, due to the different in time scale of these methods. The ability to choose linker size and rigidity depending on the guest and intended analysis method would increase the utility of these kinds of sys-tems.

1.5 To solve the problems: aims of this thesis The main aim of this thesis is to address the problems with binding to oxygen-containing functional groups and improve the range of guest sizes from the perspective of linker dynamics. The plan is to work with the tweezer designs previously studied by the group as a starting point and make the necessary adjustments in order to be able to accommodate as large a range of guest mol-ecules as possible.

1.5.1 Porphyrins vs alcohols For the porphyrin tweezer tips it is desirable to keep the TPP (Figure 6) frame-work as the solubility of this porphyrin in organic solvents is unusually good.54 The easiest solution to the alcohol binding problem is to exchange the zinc(II) cation for another metal with a higher affinity for oxygen. Both the literature and computer calculations can be used to find suitable metal ions. A previous study by Li et al. also indicated that adding electron withdrawing groups to the porphyrin increases the binding affinity towards alcohols.45,55 With this in mind porphyrins halogenated at the phenyl rings and/or -positions might

22

show better affinities towards oxygen (Figure 6). Combining these two prop-erties gives a large number of possible candidates (Figure 10).

Figure 10. Starting with four free base porphyrins (top) and combining them with nine metal cations (middle) results in 36 different metalloporphyrins (bottom) to evaluate.

The standard way of evaluating the strength of binding is by measuring the binding constant. Porphyrins have a strong UV/Vis absorption that shows a clear red-shift when complexing another molecule.53 From the measurement of this red-shift at different host:guest rations the binding constant can be cal-culated.56-60 There are other parameters that can be followed in order to deter-mine the binding constant, such as NMR chemical shifts, but that method is more time consuming and requires larger quantities of porphyrin and guest.

1.5.2 Stiff stilbene linker Amongst the previously studied linkers the stiff stilbene (SS) and glycoluril linkers (Figure 8) have shown interesting properties. Especially the SS linker has several properties that make it a good target for further studies. This linker exists in two conformers with E and Z geometry and can be photo-isomerized. The wavelengths for the “photo-switching” are 300 nm and 360 nm respec-tively which is close to the absorption of the TPP type porphyrins.61,62 Possible problems with photo-switching the SS-tweezer might arise from this overlap.

23

However, there is the possibility that the absorption of some of the new por-phyrins might be sufficiently shifted compared to the ZnTPP to allow for photo-switching of a new SS-linked tweezer. Another factor that would be interesting to investigate is if the force of the isomerization is sufficient to stretch a coordinated guest molecule or if a tweezer:guest complex would pre-vent the switching. A combination of calculations with experimental studies of synthesized cyclic SS-systems should answer this question.

Another option using the SS-linker is to use the E- and Z-isomers combined with porphyrin units as modules to design tweezers appropriate for different guest sizes. From previous calculations, it is known that the gap size of the Z-SS-tweezer is approximately 6.1 Å and that of the E-SS-tweezer is approxi-mately 18.4 Å. As the synthesis of the Z-SS-tweezer is already known, the synthesis of its E isomer by the same method appears to be straightforward.

1.5.3 Tweezers The perhaps most interesting property of the Z-SS-tweezer is the axial chiral-ity it displays. Preliminary studies indicated that CD signals are enhanced by this feature. The main question is if this tweezer could work for monotopically bound guests. This would represent a significant improvement of the range of guests accessible by this method. Using the Zn-Z-SS-tweezer to investigate a library of monoamines with no secondary binding sites and different amounts of flexibility would be a natural starting point.

To make the systems as utilitarian as possible the linkers need to allow for covalent attachment to two porphyrins that possess good binding affinities to-wards both nitrogen and oxygen containing functional groups. Combining the SS-linker or the glycoluril linker with the porphyrins selected in this study results in a number of potential candidate structures (Figure 11).

It should be noted that the SS-tweezer could potentially also be either switched to the E isomer or synthesized this way in order to accommodate larger guests. All tweezer syntheses are challenging as the resulting structures are large and have poor solubility. Additional challenges include that the final coupling re-actions typically give low yields and some tweezers are sensitive to light and possibly air.39,40 How the change of porphyrin will affect the reactions in-volved is difficult to predict as these porphyrins have only rarely been studied.

After successful synthesis and characterization of a new potentially oxophilic molecular tweezer, binding studies would be conducted in order to confirm this property. Both ECCD and NMR studies could be performed depending on the linker used.

24

Figure 11. Combining the Z-SS-linker and the glycoluril linker with the selected porphyrins results in eight different tweezers. The estimated difficulty of the synthe-ses involved is indicated by color (green-straightforward, orange-challenging, red-very challenging).

25

2 The toolbox, overview of methods

2.1 Synthesis and complexations There are different types of chemical bonds. Organic synthesis is mainly fo-cused on making and breaking covalent bonds. These are the bonds making up organic molecules by atoms sharing electrons. In this thesis organic syn-thesis is used to functionalize porphyrins, make molecular tweezers and guest molecules. However, the main focus is on another kind of bond, the metal-ligand coordination bond. This is the kind of interactions that make up metal-ligand complexes and that is the basis for the host-guest interactions in our tweezer systems.

2.1.1 Synthesis The synthetic work of this thesis is mainly centered on two areas, porphyrin functionalization and tweezer coupling reactions. Porphyrin functionalization can take place either at the meta- or -positions (Figure 6). In this study all porphyrins used are of the TPP motif and have either phenyl or pentafluoro-phenyl groups in the meta-position. The porphyrins have been functionalized either to change their properties as hosts for metal cations or to add a func-tional group for coupling to a linker. The tweezer coupling reactions used to produce the two different kinds of tweezers require either a free amine group at a -position or a dione at one of the pyrrole units (Figure 12).

Figure 12. The coupling reactions used to produce two of the previously reported tweezers, namely the glycoluril tweezer (top) and the Z-SS-tweezer (bottom).39,40

26

The linkers are either tipped by acid chlorides or by double free amine groups. One of the major challenges of the coupling reactions is the sensitivity of all of the components. The high reactivity of most of the compounds makes pu-rification impractical and the reactions are performed with mostly crude start-ing materials. This naturally adds to the difficulty of the purifications of the tweezers, as several bi-products can form, and negatively affect the yields.

A number of cyclic SS-systems have also been synthesized in order to study the photodynamic properties of the SS-unit. The syntheses of these systems are similar to that of the tweezers but the linker is added before the stiff stil-bene is formed via the McMurry reaction (Figure 13).63 Another difference is that the couplings to the linkers are via Williamson’s ether synthesis rather than an amide coupling reaction.64

Figure 13. The synthetic pathway to the SS-macrocycles studied in paper II.

2.1.2 Complexations Complexation reactions are at the heart of this thesis. The ones used here are of two kinds, porphyrin metalations and the formation of host-guest com-plexes. Porphyrins are generally synthesized as free bases and even when this is not the case the intended metal is generally not used in the synthesis.65-67 Porphyrins can be both metalated and demetalated, so that a metal used in a specific step can later be removed and replaced by another (Figure 14).

27

Figure 14. Free base porphyrins can be metalated with a variety of metal ions. The metalloporphyrin can be demetalated by exposure to a strong acid to regain the free base. Most metalations and demetalations are associated with high yields.

Many porphyrin metallation reactions are carried out by dissolving the por-phyrin in chloroform, dichloromethane (DCM) or dimethylformamide (DMF), adding the metal cation as a salt in great excess and refluxing the mixture until full conversion is observed by UV/Vis or thin layer chromatog-raphy (TLC).68-71 A demetalation is achieved by exposure to strong acid, usu-ally sulfuric acid. Whether a solvent is used or the porphyrin is dissolved in the acid itself varies depending on the metal and porphyrin involved.72,73 The complexation of guest molecules to porphyrins or molecular tweezers is based on the principle of self-assembly.74 When dissolved and mixed the por-phyrin and guest will start interacting with each other forming a complex (Fig-ure 15).

Figure 15. A metalloporphyrin has the ability to complex one or two guest mole-cules in an axial position. The number and nature of complexed guests depends on the porphyrin and metal in question.74

Different metalloporphyrins will form complexes with different guest families making it a very interesting field of study. With the possibility to change both metal ion and to functionalize the porphyrin in different ways one could po-tentially tune the system to complex the guest (or guests) of choice.

2.2 Binding studies When studying host-guest systems there exist a couple of parameters that are used to characterize the composition and dynamics of the complex. The first

28

one is the binding stoichiometry which shows the system’s composition. The second one is the binding constant (Ka) which is a measurement of the strength of the host-guest interaction.

2.2.1 Binding stoichiometry If host-guest systems involve hosts and/or guests with more than one binding site the stoichiometry of the formed complex is not necessarily predetermined. To elucidate the stoichiometry of a specific complex the “method of continu-ous variation” (known as Job plot) can be used.75,76 In this method a series of measurements with different host-guest ratios but constant combined (host + guest) concentration is performed. When plotted against the molar fraction of host or guest a characteristic curve is obtained (Figure 16).

Figure 16. Job plot, showing the theoretical curves for 1:1(purple), 1:2(blue) and 1:3(red) binding stoichiometry, with maxima at the indicated molar fractions.

Depending on the stoichiometry of the complex the maximum of the curve will be at different molar fractions. For 1:1 stoichiometry the maximum will be at 0.5, for 1:2 at 0.66 and for 1:3 at 0.75. Any analytical method that mon-itors a parameter dependent on the concentration of host-guest complex can be used to construct Job plots, such as UV/Vis, NMR, IR, fluorescence spec-troscopy or ECCD.

29

2.2.2 Binding constants The binding constant (K) is a thermodynamic measurement of the strength of the host-guest interaction (Equation 1 and 2). It is related to the free energy and thus also to entropy and enthalpy (Equation 3 and 4).

⇄ Equation 1

Equation 2

∆G RTlnK Equation 3

∆G ∆H ∆S Equation 4

This is a useful relationship as a negative free energy will indicate a favored process whereas a positive free energy will indicate that the reverse process is favored. As the free energy of a complex versus the free energy of the free components can be calculated using the density functional theory (DFT) method the stability of a complex can potentially be predicted prior to its syn-thesis.

The binding constant for a host-guest complex can be experimentally deter-mined by monitoring a changing property of the host or guest when the host:guest ratio is changed. The standard way of doing this is in a titration experiment where the guest concentration is gradually increased. A change in UV absorption or in NMR chemical shift is monitored and the binding con-stant is calculated via one of a number of available methods.56-59 The method used in this thesis is an iterative non-linear regression analysis.60

2.3 Nuclear magnetic resonance spectroscopy Nuclear magnetic resonance (NMR) spectroscopy is a very important analyt-ical technique for the structural characterization of organic molecules at an atomic level. It is one of the few methods available to determine the 3D mo-lecular structure of a compound. It is also a powerful tool in the study of com-plexation dynamics. In this application it has the advantage over UV/Vis spec-troscopy, where signals of host, guest and complex usually overlap, by poten-tially showing separate signals for the host and guest and their relative ratio. It is also possible to observe unwanted ligands that might cause interference.

30

Below are described three parameters of special significance to the present study: chemical shifts, scalar coupling constants and the nuclear Overhauser effects.

2.3.1 Chemical shifts The chemical shift is proportional to the resonant frequency of a nucleus in a magnetic field. It is measured in ppm and determined relative to a standard compound, such as tetramethylsilane for 1H and 13C. The chemical shift is re-lated to (amongst other factors) the electron density around a nucleus, which most importantly is depending on inductive effects, resonance effects and hy-bridization.77 In addition, anisotropy effects are usually present.77 These arise from local magnetic fields generated by moving electrons in chemical bonds in the vicinity of the nucleus. Thus, a nucleus will experience a local magnetic field Blocal that may be different from the magnetic field B0 generated by the magnet of the NMR instrument. The nucleus is considered to be either shielded (Blocal < B0) or deshielded (Blocal > B0) depending on its position in space. The position in space may vary over time in non-rigid molecules. In addition, interactions with other molecules or ions generate additional effects which also may be orientation-dependent. Since the exact position of a nucleus and the details of intermolecular interactions are normally not known pre-cisely, it may be difficult to relate the chemical shift directly to a chemical structure. However, chemical shifts can be used specifically to shed further light on structural mysteries.41,78,79 When studying the complexation to por-phyrins or bis-porphyrins the ring current effect will have a dramatic effect on the chemical shifts of the guest molecules (Figure 7).

2.3.2 Scalar coupling constants Neighboring atoms in a molecule will affect the NMR signal of a nucleus by generating fine structure, so-called signal splitting. This phenomenon is known as scalar coupling. By analyzing the splitting pattern (multiplicity) and the size of the splitting (quantified by the scalar coupling constant, J) infor-mation on the number of coupling spins and their proximity as well as dihedral angles can be extracted. As this is a coupling by spins through covalent bonds these parameters give valuable information about the molecular structure. Dif-ferent conformational parameters such as bond angles and the dihedral angle between two atoms can be derived.80,81 The scalar coupling constant has long been one of the most valuable tools in structure determination, and fitting of experimental data to calculated coupling constants have been used in order to determine the structure of small organic molecules.20,22 This method has also been attempted to determine relative stereochemistry of small organic mole-cules, with mixed results.15,16,20,22

31

2.3.3 Nuclear Overhauser effects The Nuclear Overhauser Effects (NOE) is an important parameter for eluci-dation of the three-dimensional structure of a molecule. The NOE is based on an interaction between nuclei through space rather than through bonds. It can therefore give information about distance between nuclei. The NOE arises from dipolar cross relaxation. The effect is distance dependent and protons usually need to be within 5 Å for an observable effect.82 The effect is also dependent on molecular motion and will represent an average of various con-formers present. This is a very useful parameter when studying supramolecu-lar complexes as the parts making up the complex, i. e. host and guest, might be close in space but not connected by covalent bonds. It can also be used to study the possible tweezer conformers.

2.4 ECCD The basics of this method is described in the introduction (section 1.2.2).

2.5 Computational methods Computational methods are a powerful tool in the study of molecular systems. Depending on the size and complexity of the system there is a wide array of available methods for the calculation of geometry, energy and dynamics. For small and simple systems more parameters can be calculated at a lower com-putational cost. For larger systems, simpler methods are usually used to study the approximate structure and dynamics of the system. This kind of data can be used to evaluate potential target systems to study or as a compliment to understand experimental results and observations. There are a number of methods pairing calculated data and structures to experimentally measured data for the determination of structures and stereochemistry.83-85

2.5.1 Molecular mechanics Molecular mechanics (MM) is a method based on classical mechanics. It treats all atoms as particles and uses force-fields to calculate their positions and from this the energy of the system. As this method has a low computational cost it can be used for the calculation of geometries, transition states and confor-mations even for large systems such as proteins.86 MM has also become an important tool in the study of supramolecular systems.87-89 The energies re-ceived can be used to compare different conformations but it is not considered a correct reflection of bond formation and dissociation energies.86 The force-

32

field used for the investigations of the conformers in Paper II and different conformations of the host-guest systems in Paper III was OPLS3.90

2.5.2 Density functional theory Density functional theory (DFT) is a quantum mechanical method that is used to calculate the electronic structure of atoms and molecules in the ground state. It is based on the Hohenberg-Kohn theorems which states that the energy of a system in the ground state is determined by the electron density.91,92 In DFT, functionals are used in order to calculate the energy of the system. Even though it is a lot more expensive in terms of computational cost compared to MM the DFT method is a relatively cheap method to calculate the electronic structure of molecules compared to other quantum mechanical methods. One downside is that it does not describe dispersion interactions well, but there are dispersion-corrected methods which improve the accuracy of the calcula-tions.93,94 To further improve the method and give a more accurate description of e.g. bond lengths, vibrational frequencies and atomization energies there are several hybrid-DFT methods. In these the approximations for the ex-change-correlation energy include parts of exact exchange from Hartree-Fock theory.95,96 The binding constant, energy and geometry calculations carried out were using either the dispersion corrected hybrid-DFT functional B3LYP-D3(Paper I) or standard B3LYP (Paper II and III).95,97

33

3 Porphyrins and guest interaction (Paper I)

3.1 The plan To facilitate the use of molecular tweezers for stereochemical characterization of molecules lacking amino groups we attempted to solve the problem of bind-ing guests with oxygen-containing functional groups to metalloporphyrins. As mentioned in the introduction (section 1.5.1), the focus was on changing the metal cation and on using a different porphyrin with a more electron poor cen-ter. Based on previous studies by Borhan and coworkers, showing enhanced binding to isopropanol and isopropylamine for 5-(4-methylcarboxyphenyl)-10,15,20-tri(pentafluorophenyl)porphyrin compared with the corresponding ZnTPP-monoester, halogenated porphyrins were selected as potential targets (Figure 17).45,55

Figure 17. The previously reported 5-(4-methylcarboxyphenyl)-10,15,20-tri(pen-tafluorophenyl)porphyrin (left) and 5-(4-methylcarboxyphenyl)-10,15,20-tri-phenylporphyrin (right).45,55

Metal cations were chosen based on their oxophilicity and on their size.98 As the central coordination cavity of porphyrins is approximately 4 Å (N to N) this sets a rough size limit for the metal cations. Another selection was made based on metal cations that were known to bind to TPP. Based on these criteria nine metals were chosen. As shown in Figure 10 four porphyrins combined with nine metals would potentially yield 36 metalloporphyrins for the study. In order to limit this number the simpler MTPP:s were studied first and the most promising metals were then complexed to the more exotic porphyrins.

Another important factor is the model guest molecules. To keep the studies simple small molecules representing alcohols, carboxylic acids and ketones were chosen, i. e. methanol, acetic acid and acetone. As a reference, pyridine

34

was also included. Pyridine is not a typical amine, but it is well studied as a ligand to metalloporphyrins.99-101 As can be easily perceived this kind of study is fairly time-consuming. A com-putational study was therefore included to investigate if standard computa-tional methods could be used in order to accurately predict the metalloporphy-rin:guest interactions. To limit the computational cost only methanol was used as a guest.

3.2 Synthesis The TPP and TPFPP are commercially available and the -positions are free for reactions facilitating coupling with a linker. Therefore, they were acquired commercially and used as received. The TPPBr8 is commercially available but its synthesis from TPP and MTPP is well documented. As the TPFPPBr8 is not commercially available the bromination of the porphyrin -positions was investigated starting with the bromination of TPP.

There are many cases of the -position bromination of TPP in the literature. They mainly fall into two categories, the ones using bromine and the ones using N-Bromosuccinimide (NBS).102-108 In the process of investigating this reaction a large number of the reported methods were tried with only one method (NiTPP reacted with NBS in 1,2-dichlorobenzene) resulting in the product in high yields. All other methods resulted in decomposition of the starting material or in very low yields (Figure 18).

Figure 18. The different synthetic pathways tested to synthesize MTPPBr8.

35

Following the bromination, treatment with a strong acid (H2SO4) results in demetalation and the free base is received. The same method was used to pro-duce the NiTPFPPBr8. However, all attempts to demetalate this metallopor-phyrin returned the starting material (Figure 19). As all other pathways to the brominated species yielded no product the TPFPPBr8 was removed from the study.

Figure 19. Synthesis of the NiTPFPPBr8 (middle) and the attempted synthesis of the TPFPPBr8 (right).

The complexation reactions of the porphyrin free bases to the metal cations followed established methods, as is described in the supplementary material to Paper I.68-71

3.3 Complexation experiments The fastest and most common way to measure porphyrin:ligand binding con-stants (Ka) is by UV/Vis titrations. When the metalloporphyrin complexes a ligand a clear red shift is observed and the changes in absorption at different wavelengths can be used to calculate the binding constant. Upon titration of some metalloporphyrins it was observed that no (or very small) red-shifts oc-curred. Instead, a decrease followed by an increase of absorption was noted (Figure 20).

Figure 20. Showing the UV/Vis titrations of ZnTPP and CoTPPCl with pyridine. The ZnTPP:pyridine titration shows the characteristic red-shifts (left) whereas the CoTPPCl:pyridine titration only displays differences in intensity (right).

36

The reason for the lack of red shift was revealed by 1H NMR studies. The “free” porphyrins not displaying red-shifts when complexing guests were al-ready complexing water. Thus, upon titration with another guest an exchange of guest was observed rather than a free porphyrin binding to a free guest. For most systems drying at 50oC under high vacuum for 24 h would remove the water, with the CoTPPCl being an exception. With this complication the bind-ing constants were measured by proton NMR rather than UV/Vis. The ad-vantage of this method is that it is possible to observe both host and guest separately and also keep track of any impurities and other complexing mole-cules. Due to the porphyrin ring current effect a guest forming a complex with a porphyrin will have a different chemical shift compared to the free guest (Figure 7). With free and bound guests exchanging, the average chemical shift will change with the porphyrin:guest ratio. This change can be used to calcu-late the binding constant. Using NMR titration data the binding constants of fourteen metalloporphyrins to four guest molecules (methanol, acetic acid, ac-etone and pyridine) were determined (Table 1).

Table 1. The measured binding constants (Ka) of methanol, acetic acid, acetone and pyridine to fourteen metalloporphyrins.

Binding constant, Ka (M-1)

Porphyrin Methanol Acetic acid Acetone Pyridine

AlTPPCl 3.8×10-3 2.8×101 nb nb

CoTPPCl 3.2×10-3 1.7×102 nb cc CoTPFPPCl 5.4×101 1.8×104 nb cc FeTPP nb nb nb nb MgTPP 5.7×101 D 7.4 1.7×101 MgTPFPP 1.5×102 2.3×103 3.1×102 1.3×104 MgTPPBr8 1.6×102 8.2×102 3.8×102 1.3 MnTPPCl nb nb nb nb NiTPP nb nb nb nb SnTPPOH2 nb nm nm 1.3×104 O=TiTPP nb 4.8×102 nb nb ZnTPP 2.4×10-3 3.6×102 nb 6.8×103 ZnTPFPP 1.5×101 1.2×103 1.1×10-2 2.7×104 ZnTPPBr8 2.0×101 2.5×102 6.0×10-3 3.2×102

nb = no binding; nm = not measured; nc = cannot calculate, binding too weak; cc = cannot calculate, binding too strong; D = demetallation; a very broad signals; b stoichiometry 1:2.

3.4 Computational predictions The theoretical binding constant (Ka,theo) can be calculated from the free en-ergy of binding (G) by breaking out the binding constant in Equation 3 (Equation 5):

∆

Equation 5

37

The free energy of binding is defined as the free energy of the complex minus the free energies of the porphyrin and the guest (Equation 6): ∆ Equation 6 A negative free energy of binding indicates that the complex is energetically favorable compared to the porphyrin and guest as free entities. Thus, by using computational methods to calculate the free energies of the porphyrin, guest and complex, information about the stability of the complex can be gained. As reliable energies for various systems were needed, DFT calculations were used. There are different DFT methods and the literature shows that it can differ which one gives the best representation of different metalloporphy-rins.109-111 For this study both B3LYP-D3 and M06 with the 6-31G(d,p) basis set for the H-, C-, O-, F-, Cl-atoms and SDD basis set and pseudo potential for all metals were tried and compared.95,97,112-114 Of these B3LYP-D3 gave the best correlation between calculations and experiments over the full set of por-phyrins. Another factor to take into account is that metals have different spin states. To find the one with the lowest energy in each case the three lowest spin states were calculated for each metal. In addition to the free energy, the metal-ligand bond length and the system charge distribution were investi-gated.

To reduce the computational cost the calculations were limited to one guest (methanol) and the majority of the calculations were performed on MP sys-tems (Figure 21). As indicated in Figure 21 some of the substituted TPP-type systems were included in order to investigate the effect of halogenation on the charge distribution.

Figure 21. Structures of the porphyrins included in the computational study.

38

Comparison of the calculated data with the experimental results showed that a combination of two factors was needed in order to predict stable complexes. The free energy of binding should be high and the methanol ligand should display a negative change in charge upon complexation.

3.5 Conclusions The computational study resulted in a relatively fast and easy method to de-termine the potential stability of a porphyrin:guest complex without having to do the synthesis and complexation experiments. As the experiments show that relatively few of the metalloporphyrins actually bind to the ligands this is a valuable tool that could save a lot of time in the lab.

For this thesis, the most valuable results are the experimental binding con-stants (Table 1). From these, four interesting porphyrins emerged (Figure 22).

Figure 22. Metalloporphyrins in the experimental study that showed the strongest binding to the largest number of guests. *Cannot calculate, binding too strong.

The CoTPFPPCl and ZnTPFPP have reasonable binding constants to metha-nol, acetic acid and pyridine but no or weak binding to acetone. Most interest-ing are the ZnTPPBr8 and the MgTPFPP that have suitable binding constants for all the tested guests. All of these could potentially be used in the construc-tion of new tweezers. From a practical point of view, the TPFPP porphyrin is theoretically easier to couple to a linker as the -positions are free (Figure 11). To couple the TPPBr8, the bromination would have to take place after the

39

functional group used in the coupling has been added. This synthesis is sus-pected to be more challenging than the coupling of the TPFPP for which the same synthetic pathway as for TPP should be accessible. Based on these fac-tors the metalloporphyrin of choice for future construction of tweezers would be the MgTPFPP.

40

41

4 Linker dynamics (Paper II)

4.1 The plan As mentioned in the introduction (section 1.5.2) the stiff stilbene linker exists as two geometric, i. e. E and Z isomers and so far, the Z-isomer has been used to construct bis-porphyrin tweezers. A tweezer based on the E-isomer would have a larger gap size and potentially be able to bind to guests that are too large for the Z-SS-tweezer. The two tweezers could be synthesized separately but a more intriguing prospect would be to use the SS linker’s ability to isomerize photochemically (“photo-switch”). Not only could the E-isomer be generated from the Z-isomer when needed, it would also open up other fea-tures such as the stretching of an already bound guest (Figure 9). The stiff stilbene is photo-switchable with λmax 300 and 360 nm for the E and Z isomer respectively, but whether the force is strong enough to switch the tweezer and most importantly stretching a long chain is as yet undetermined.61,62 Thus for clarification, a series of cyclic model compounds were investigated both ex-perimentally and computationally.

4.2 The cyclic compounds The stiff stilbene macrocycles were kept as simple as possible. A Williamson ether synthesis was used to insert a linker between the two indanone units.64 The stiff stilbene was then formed via the McMurry reaction (Figure 13).63 Four different linker lengths were used, six, eight, ten and twelve carbons long (Figure 23). Note that the Z-isomer was received from the McMurry reaction.

Figure 23. The four SS-macrocycles with linker lengths of 6, 8, 10 and 12 carbons can theoretically be photo-switched between Z (left) and the E (right) isomer.

42

The distance, r, between the two oxygens of the Z-stiff stilbene unit is 5.4 Å. In a fully stretched all-trans O-(CH2- CH2)n-O chain the distance between the oxygen atoms increases with the number of carbon atoms with r = 7.5 Å (n = 3), 9.4 Å (n = 4), 9.9 Å (n = 5) and r = 10.3 Å (n = 6). Therefore, the shortest linker (n = 3), should make switching almost impossible as this would result in a highly strained structure. In contrast, the twelve-carbon linker (n = 6) should result in a strain-free stiff stilbene macrocycle.

4.3 Switching study As a reference for the switching behavior of the SS-macrocycles a SS-diester previously studied in the group was used (Figure 24).62

Figure 24. The reference SS-diester previously studied in the group.62

The photo-stationary mixture of the reference compound contains an isomer ratio of 55:45 E:Z. Irradiation of the SS-macrocycles show the expected drop in E-isomer as the linker length is decreased (Figure 25).

Figure 25. The E/Z composition of the photo-stable mixture of the cyclic com-pounds compared to that of a non-cyclic reference.

43

The switching was followed both by UV/Vis and NMR spectroscopy. Inter-estingly the amount of E-isomer never hit zero. The decreased linker length clearly increased the photo-stability of the Z-isomer, but never to the point of making it fully photo-stable.

The energy differences between the E- and Z-isomers were calculated by DFT (B3LYP, 6-31G(d,p), SCRF-SMD with DCM as solvent) and shows that the energy difference between the cis and trans isomers dramatically increase as the length of the linker shortens (Figure 26).95,112,115-117 For the longest (12 carbon) linker the trans isomer is energetically favored as compared to the cis analogue but not to the same extent as in the unlinked reference.

Figure 26. Showing the difference in free energy (G) in kJ/mol between the Z and E isomers of the SS-macrocycles and the unlinked reference compound.

As indicated by both the computational and experimental switching study the linker seems to somewhat affect the photo-stability even for longer linker lengths.

4.4 Conclusions The switching study of the SS-macrocycles shows that the switching mecha-nism is strong enough to stretch even tight carbon chains. However, compar-ing the calculated energy differences between the E and Z isomers to the ex-perimental results strongly suggest that the linker affects the switching. For the longest linker the E-isomer is energetically favored but even then the Z-isomer dominates the photo-stationary mixture.

44

Since the switching still works until the linker gets very short it should be possible to use this mechanism in order to photo-switch a SS-based tweezer. One downside is that the switching results in a roughly 1:1 E/Z mixture even when acyclic. However, the huge difference in structure between the E and Z tweezers should make the two clearly distinguishable by NMR spectroscopy if the concentration of the E-isomer is high enough. Another problem is that the UV/Vis absorption of the SS unit and the absorption of the TPP overlap. With the very high absorption of the porphyrins preliminary studies on the Z-SS-tweezer have indicated that the switching will be ineffective due to this aspect. However, the previously presented porphyrin studies (Paper I) opens up the possibility to use other porphyrins as tweezer tips. For example, the MTPFPPs do not have any significant absorption overlap with the SS unit and might therefore be a good starting point for an improved photo-switchable SS-tweezer. If a good tweezer design is found the stereochemical characterization of guests varying from roughly, 6-18Å (distance between to metal centers in a corresponding bis-metalloporphyrin tweezer with SS linker) in length could be covered with the same system.

45

5 Molecular tweezers (Papers III and IV)

5.1 A plan of three parts The main focus of this thesis is to study how bis-porphyrin tweezers can be improved for use in stereochemical characterization. As mentioned in the in-troduction (section 1.5) a central part is to widen the range of available guests that can be characterized using the established NMR and ECCD methods. The plan to achieve this consists of three parts: (i) To investigate if the Z-SS-tweezer previously studied in our group can also produce CD signals for monotopically binding guests; (ii) to synthesize and investigate the properties of the E-SS-tweezer; (iii) to use the results from paper I to design and synthe-size a new tweezer with the potential to bind both oxygen and nitrogen con-taining functional groups.

5.1.1 A Twisted Tweezer (i) As mentioned in the introduction (section 1.5.3) the main limitation of the bis-porphyrin tweezers are that ditopic binding is required. The goal is to investi-gate if a signal can be detected from the Z-SS-tweezer with monotopically binding guests. Previous studies have shown that the Z-SS-tweezer (Figure 12) performs very well for ECCD characterizations.40 The main feature that sets the Z-SS-tweezer apart from other tweezers used for stereochemical char-acterization by ECCD is that instead of adopting a twist after coordinating a chiral guest this tweezer has an inherent twist. The source of this is the axial chirality of the Z-SS linker due to steric interaction between two aromatic pro-tons. When not coordinating a chiral guest molecule the tweezer exists as a racemic mixture (Figure 27).

Figure 27. The Z-SS-tweezer has inherent axial chirality and exists as a racemic mixture, with the M- and P-enantiomers in equal amounts when not coordinating a chiral guest.

46

That the M- and P-enantiomers would form in equal amounts from synthesis is intuitive, but the question remains if they exist at equilibrium and can switch freely. This is suggested by the fact that the coordination of a guest leads to excess of one of the enantiomers, but then again the guest might bind to one of them in a higher degree. To answer this question the M to P transition of the linker was studied by DFT (B3LYP, 6-31G(d,p), SCRF-SMD with DCM as solvent). 95,112,115-117 The dihedral angle was changed stepwise with geome-try optimization, frequency and stability calculations. The results showed that the energy needed for the transition is 1.2 kJ/mol, which is significantly lower than the approximately 2.5 kJ/mol representing the thermal energy at room temperature. From this we conclude that the SS linker exists as an equilibrium mixture rather than a static mixture between the E and Z-enantiomers.

To test if the Z-SS-tweezer also works for monotopically binding guest mole-cules a number of monoamines with varying size and flexibility was chosen (Figure 28).

Figure 28. The guest molecules used to test if the Z-SS-tweezer could give a CD signal for monotopically binding guests. Colors indicate the strength of the CD sig-nal received.

47

Of the guests investigated, three gave strong CD signals and another eight gave moderate signals. As the guests only have one functional group for co-ordination to the porphyrin this is a very good result. Evaluation of the struc-ture of the guests shows that the small and flexile guests do not give a CD signal whereas the more rigid ones do. In some cases, a clear distinction is difficult to make. For example, why G19 and not G8 and G13 but not G10 give a signal is hard to explain. To further verify that the signals arise from monotopically binding guests the binding constants were investigated. From comparing the binding constants of free TPP binding to pyridine and 1,6-diaminohexane with the Z-SS-tweezer binding the same diamine as well as guests resulting in no, moderate and strong CD signals some interesting indications were given. The guests with no or moderate CD signals have binding constants to the Z-SS-tweezer of the same order of magnitude as the free TPP binding to pyridine. The guests giv-ing a strong CD signal have binding constants to the tweezer of the same order of magnitude as the free TPP binding to the diamine, but one order of magni-tude lower than the tweezer binding the diamine. This could be interpreted as the guests with the strong CD signal represent something between monotopic and ditopic binding. There are two intuitive possible explanations for this ob-servation. The first is that the guests with a strong CD signal might be binding ditopically, but a lot weaker at the secondary site. The other explanation is that there exist stabilizing forces between two guest molecules binding simultane-ously inside the tweezer cavity, making them behave as a diamine rather than two monoamines. To determine which explanation is more likely to be correct the binding sto-chiometry was investigated by the construction of Job plots based on ECCD for one of the guests giving a strong CD signal (D-G9) and a chiral diamine reference (Figure 29).75,76 The reason for constructing the Job plots using ECCD rather than UV/Vis even though the signal strengths are low at these host:guest ratios is that all the guests bind to the tweezer. Thus, the results from UV/Vis or NMR spectroscopy need not be related to the complex pro-ducing the CD signal. However, when using ECCD only CD active complexes contribute to the Job plot and all other complexes are effectively invisible. As shown by the results in Figure 29 the D-G9 is binding with a 1:2 stoichiometry which would fit to simultaneous monotopical binding of two guest molecules with interaction between them. From this it can be deduced that for the guests giving a strong CD signal some interaction between two monotopically bound must be responsible for the stronger binding constants.

48

Figure 29. Job plot of (2S,3R,4R,5R)-1,6-diamino-1,6-dideoxy-2,3,4,5-tetra-O-me-thyl sorbitol41 binding with a 1:1 stoichiometry, thus ditopically (left). D-tyrosine methylester D-G9 binding with a 1:2 stoichiometry, thus monotopically (right). The theoretical curves for 1:1 and 1:2 binding stoichiometry, respectively are shown in red.

Molecular modeling (MM) was used to investigate the guest:guest interactions that might stabilize the CD active complexes (OPLS3 force field).90 All guests giving a CD signal can -stack. The guest D-G9 can also form guest:guest hydrogen-bonds. The number and strength of these interactions seems to be central to understanding the differences in binding constants and CD signal strength. It is clear that the possibility to form complexes stabilized by guest:guest interactions adds further support to the hypothesis that the ob-served CD signals arise from complexes formed by monotopically binding guests.

Attempts to analyze both the free tweezer and the formed complexes with XRC to further confirm these claims are currently ongoing.

5.1.2 The elusive trans-Tweezer (ii) Another feature of the SS-tweezer requiring further study is the possibility to use the E-SS-tweezer for larger guests. As photo-switching the ZnTPP-tipped tweezer is inefficient due to overlap between absorption bands of the SS-unit and the porphyrins the direct synthesis of the tweezer is a logical starting point to investigate its properties. If the E-SS-tweezer has interesting properties, the synthesis of the potentially photo-switchable MTPFPP analogue might be considered. Numerous attempts have been made to synthesize the E-isomer of the stiff stilbene tweezer (Figure 30).

49

Figure 30. The structure of the E-SS-tweezer, the i. e. trans-tweezer.

With the synthesis of the Z-SS-tweezer being relatively straightforward for a molecule of this size and giving a reasonable yield (20%-30%) no difficulties were initially expected. However, as all attempts to generate the E-tweezer via the same synthetic route failed it became clear that the two tweezers are not as similar as they appear on paper.

In the Z-SS-tweezer synthesis, the amide bond is formed via the reaction of the di-acid chloride stiff stilbene linker and the aminoporphyrin TPPNH2. This reaction has been tried in different solvents and temperatures without results (Figure 31). Attempts to use coupling reagents have also been made. Model compound reactions with terephthalic acid and TPPNH2 gave the best results for HATU. Therefore, both HATU and HBTU have been tried to produce the tweezer from the SS di-carboxylic acid. Another pathway, via the di-acid io-dide was also attempted. Following the procedure reported by Wakeham et. al. the di-acid iodide is generated in situ from the di-carboxylic acid and po-tassium iodide.118 As indicated in Figure 31, all attempts so far have been un-successful.

Figure 31. Various coupling reactions have been tried in order to produce the trans-SS-tweezer. None of the reactions have yielded more than trace amounts of product.

50

The question of why the trans-tweezer is so difficult to synthesize compared to the cis-version is not clear at this time. Solubilities of the E and Z linker compounds are different, with the E linker being less soluble. Since heating, sonication and/or using other solvents or solvent mixtures dissolved the com-pound this cannot be the main problem. One theory is that the porphyrins ten-dency to stack to each other acts as a preorganization factor for the synthesis of the Z-SS-tweezer. Without this aid, the product might be to entropically disfavored to form. More detailed experimental and computational studies into the reaction mechanism and conditions might yield the answer.

5.1.3 The new and exciting “Fleezer” (iii) This group has previously shown that our glycoluril bis-ZnTPP tweezer (Fig-ure 12) can be used in combination with NAMFIS to determine the relative stereochemistry of small flexible diamine guests with several stereocenters.41 Given the few dependable methods to determine the relative stereochemistry of small molecules it is of great interest to increase the range of guests avail-able to this method. In Paper I metalloporphyrins were screened with the pur-pose to find species with strong binding to oxygen containing functional groups. Based on these results a new glycoluril tweezer expected to have bet-ter binding capabilities was synthesized.

Based on the results from Paper I the MgTPFPP was chosen as the new met-alloporphyrin for the tweezer tips. Combined with the semi-rigid glycoluril linker this would result in a new fluorinated bis-porphyrin tweezer dubbed the Fleezer (Figure 32).

Figure 32. The structure of the MgTPFPP-glycoluril tweezer (called the Fleezer).

The plan for the synthesis of the Fleezer was to use the same synthetic route as for the synthesis of the previously studied ZnTPP-glycoluril tweezer.39,41 As shown in Figure 12 the final coupling is a reaction between the TPP dione and the tetra-amino glycoluril linker. To produce the TPP-dione the porphyrin was first nitrated, the product reduced and then Dess-Martin periodinane

51

(DMP) was used to produce the dione product. The synthesis of the Fleezer following this path is shown in Figure 33.

Figure 33. Planned synthesis of the Fleezer based on the previously reported synthe-sis of the ZnTPP-glycoluril tweezer.39,41

The main problem with this approach turned out to be that the stability of the TPFPPNH2 is a lot higher that the stability of the TPPNH2. The TPPNH2 is generally not purified due to its high reactivity whereas the TPFPPNH2 is so stable that it is difficult to make it react at all. The reaction with the DMP gave the desired product, but in a yield of only 4 %. As this is clearly not acceptable, an alternative route was sought. Another aspect to note after observing the stability of the TPFPPNH2 is that the synthesis of the SS-based tweezers uses the aminoporphyrin in the amide coupling reaction. Due to the undesired sta-bility of the TPFPPNH2 another synthetic pathway, or another coupling alto-gether will probably be needed to form this kind of tweezer. Searching the literature for reactions that were reported to work for this exact porphyrin a synthesis of the TPFPP-dione via the dihydroxy-TPFPP was found.119,120 The synthesis starts with the dihydroxylation by OsO4 followed by the Swern oxidation to the dione product (Figure 34).

52

Figure 34. The synthetic route to the MgTPFPP-dione.

The reaction has been carried out both on the free base porphyrin and on var-ious MTPFPP. Even though the Mg(II) was not one of the metals reported the fact that the reactions seemed to produce similar yields independent of metal ion resulted in the decision to carry out the metalation as the first step. The reason for this is that the magnesium metalation reaction of TPFPP rarely gives 100 % yield, 70-80 % being more typical. However, the unreacted start-ing material can be recovered and used later, thus minimizing the loss. The tetra-amino glycoluril linker was received from the reduction of the tetra-nitro glycoluril linker, as previously reported.121 Reaction of the tetra-amino linker with the MgTPFPP-dione in DCM/MeOH in the dark produced the Fleezer in 3 % yield. After purification by column chromatography characterization of the product by NMR, UV/Vis and matrix-assisted laser desorption/ionization mass spectroscopy (MALDI-MS) all confirmed the identity of the Fleezer. These results are very promising and optimization of the reaction conditions are ongoing as well as studies of the Fleezers binding properties.

5.2 Conclusions To summarize the three aims set at the beginning of this chapter: (i) It has been shown that the Z-SS-tweezer can give CD signals for monotopically binding guests. This significantly increase the range of guests for absolute ste-reochemical characterization by ECCD using molecular tweezers. (ii) The synthesis of the E-SS-tweezer was not successful. Another coupling might produce better results. (iii) A new optimal tweezer based on the results from Paper I has been synthesized and characterized. This tweezer could potentially be used for the determination of the relative stereochemistry for guests with only oxygen containing functional groups.

53

6 Conclusion and outlook

The synthesis and application of bis-porphyrin tweezers have been studied from a modular standpoint. The tweezer was divided into the porphyrin tips, metal cations and the connecting linker with the prospect to combine specific porphyrins, metal cations and linkers to form tweezers tailored for certain ap-plications. The targeted applications are the determination of the absolute ste-reochemistry of small chiral molecules by ECCD and the determination of the relative stereochemistry for small flexible molecules with multiple stereocen-ters by NAMFIS. In Paper I the porphyrins and metal cations were studied in order to find can-didates that firstly, strongly bind to oxygen containing functional groups and secondly, also bind amines. Several metalloporphyrins with interesting bind-ing properties emerged from this study (Figure 22). The results can be used to widen the range of guests that can be characterized by using molecular twee-zers. In Paper II the SS linker unit was investigated with the future prospect to use it to synthesize a photo-switchable bis-porphyrin tweezer. The switching properties of the unit turned out to fulfill the requirements but in order to make a functioning tweezer it needs to be combined with porphyrins without UV/Vis absorption overlap. From Paper I we know that the MTPFPPs would fulfill this requirement. If successful, this would allow for larger guests and the stretching of guests.

In Paper III the Z-SS-tweezer is investigated as a host to chiral monoamines for ECCD. We have shown that this tweezer binds guests monotopically with a 1:2 stoichiometry giving a moderate to strong CD signal for eleven out of the nineteen investigated guests (Figure 28). The guests giving CD signals have relatively rigid structures and the potential to form guest:guest -stack-ing and/or hydrogen bonds. This shows that the characterization of many monotopically binding compounds is possible with this tweezer.

In Paper IV a new tweezer is constructed from MTPFPP and the glycoluril linker (Figure 32). A new synthetic path was needed as the stability of the TPFPPNH2 made the one used for the ZnTPP based glycoluril tweezer inac-cessible. Initial characterizations all support the suggested structure of the so-

54

called Fleezer. The hope is that this tweezer will bind eaqually well to amines and oxygen containing functional groups, enabling relative stereochemical characterization of a wide range of flexible guest molecules.

Besides the work presented in the papers this thesis also presents some obvi-ous but failed attempts without which the story would be somewhat frag-mented. Most notable of these is the so far inexplicable difficulties encoun-tered when trying to synthesis the trans-tweezer (Figure 31).

The work of this thesis is the continuation of previous studies and much still remains to be done. From the perspective of the author, the natural continua-tion of this work would be to optimize the synthesis of the Fleezer and then thoroughly investigate its properties. The synthesis of a new SS-tweezer with MTPFPP tips could result in a tweezer usable for stereochemical analysis by ECCD of guests containing different functional groups and binding both mono- and ditopically. Therefore, this should be a priority. The structures of both the new and the existing tweezers, in particular when binding guests could be studied in more detail by XRC. Renewed attempts to form crystals of the tweezers and their complexes should therefore be made.

55

7 Sammanfattning på svenska

Organisk kemi är vetenskapen om kolföreningars egenskaper och reaktioner. Även om dessa föreningar kan vara mycket stora (t.ex. plaster) så är de i regel uppbyggda av små länkade enheter. Dessa kallas små molekyler och de stude-ras flitigt då de används i allt från läkemedel till bränsle och olika material. Eftersom dessa molekyler är mycket små är en av de stora utmaningarna för kemisten att bekräfta att molekylen verkligen ser ut som hen tror. Det finns flera metoder att tillgå men ingen är helt perfekt utan oftast används en kom-bination av flera för att säkerställa molekylens identitet. När vi ser bilder på dessa molekyler så verkar problemet inte så stort, de representeras av några få linjer och bokstäver på ett papper (Figur 1).

Figur 1. Små molekyler representeras ofta på detta sätt, som några linjer och bokstä-ver. Molekylerna ovan (till vänster och i mitten) är även icke-identiska spegelbilder av varandra. I verkligheten är de tredimensionella och väldigt dynamiska (höger).

Ett sätt att börja förstå problemet är att jämföra molekylen med en tvååring. Om man tar en bild av en tvååring, begränsad till ett rum på ca 25 m2 vid 25 oC, så ser man en bild av ett stationärt barn i ett rum. Den som har varit i ett rum med en pigg tvååring vet att säkra metoder för att uppskatta i vilken del av rummet barnet kommer existera vid en viss tidpunkt saknas. Det är även omöjligt att veta om barnet kommer stå, sitta, rulla, skratta eller skrika för full hals. En organisk molekyl i lösning är som en pigg tvååring på en öppen yta. Den rör sig, vrider och sträcker sina bindningar, roterar och viker sig hit och dit. Vad som ser platt och ordnat ut på papper är i själva verket något mycket dynamiskt. För att göra uppgiften av identifikation av små molekyler ännu svårare kan dom existera i nästan identiska par. Dessa kallas enantiomerer och har samma förhållande till varandra som en höger- och vänsterhand, de är spegelbilder

56

utan att vara identiska. För att göra det mer komplicerat kan samma molekyl inehålla flera center med den här typen av symetri. Man kan använda molekylära verktyg som kallas molecular tweezers (mole-kylära pincetter) för att hålla molekylerna stilla under analys. När rörligheten minskar kan vissa befintliga metoder användas för att identifiera molekyler och deras interna symetri. Den här avhandlingen handlar om att förbättra molekylära pincetter så att fler molekyler ska kunna identifieras. En molekylär pincett består av tre delar: två toppar och en länk dem emellan (Figur 2).

Figur 2. Molekylära pincetter består av två toppar och en länk. De kan binda till mo-lekyler på topparna och därmed hålla dem stilla under analys.

De två topprna i våra pincetter utgörs av en typ av molekyl som heter porfyrin. Dessa är starkt färgade föreningar som bla ger färg åt blod och klorofyll. I Paper I undersöktes hur dessa kan fås att binda till flera olika typer av mole-kyler. I Paper II undersöktes hur den länkande delen från en av våra pincetter kan ändra form när den blir belyst av ljus av en viss våglängd. Målet var att se om denna förmåga skulle kunna användas för att ändra form och storlek på pin-cetten och därmed kunna binda större gäster eller sträcka långa kedjor. I den senare delen av avhandlingen undersöktes pincetterna i sig. Det visades hur en, i gruppen, redan framtagen pincett kan användas för att identifiera vissa molekyler som bara kan binda till en av porfyrinerna. Sist och, ur författarens perspektiv, viktigast kommer presentationen av syn-tesen av en ny pincett baserad på de resultat som dom tidigare undersökning-arna givit. Förhoppningen är att denna ska kunna användas på en stor grupp olika molekyler.

57

8 Acknowledgements

I would like to express my deeply felt gratitude towards all the people that have been a part of this adventure! First of all, I would like to thank my supervisor Prof. Adolf Gogoll. The time working for you have been incredibly fun and stimulating. You have been an excellent mentor both when things have been going well and when it has been really tough. You have always given me good advice and I cannot remember ever going from one of our meetings in a bad mood. Thank you for accepting me as a student, I have had a great time and I have grown more, and in more ways, than I could ever have imagined! I wish to thank my co-supervisor Prof. Helena Grennberg for always being ready to listen and give advice on the wide and varied array of problems and concerns I have experienced (and sometimes imagined) over the years. All the past and present members of the AGHG group: Dr. Sara thanks for always answering my panicked porphyrin questions, even years after leaving these tricky rings behind. Dr. Magnus, for the endless amount of energy and enthusiasm you bring to everything you do. Also, thanks for teaching me eve-rything about children, you made having kids sound as fun as it actually is! Dr. Michael, you have been excellent company in all situations, lab, confer-ence, coffee break and badtunna! Dr. Hao, thanks for all the nice talks, team buildings and dinners! Dr. Xiao, thanks for always being positive and support-ive. Dr. Anna, you were a great office and conference mate! I miss our coffee breaks! Dr. Leili, for being great company in and out of the lab. Dr. Rikard, for listening and being a steady support during my thesis-writing-breakdowns. Merve, my current office mate, thanks for always being a happy and positive force in the lab despite whatever challenges are popping up. I also would like to thank some of the students who have worked in the group during my time, Tyran, Romeo, Philipp, Sofie, Johannes, Gustav and Fredrik, it has been fun to get to know you! I want to send a special thanks to my own students: Oscar, thanks for the great job with the SS macrocycles and good luck with the pirating (or was it pilot-ing?). Clara, thanks for the work on the ECCD project, it was great working with you! Julian, thanks for testing all the wild and varied ways we imagined

58

the trans-tweezer might be synthesized, never giving up and always being positive. I want to thank Dr. Christian, for collaborating on the porphyrin project and for teaching me DFT. I also want to thank the other employees at RenFuel for all the nice lunch and fika breaks. Many thanks to the technical and administrative staff: Gunnar, for having so-lutions to all problems there is. Ove, for turning up as fast as sms travels and fix all problems a stubborn fume hood can throw at you. Bosse, for getting the computers to run nicely. Johanna A., for making it a lot easier to work in the course labs. Johanna J., Eva and Hanna for making all the paperwork run smoothly and for nice lunch and fika breaks. I want to thank all the past and present members of the department. There have been too many during my time to mention you all, even though all of you have contributed to making my time unforgettable! Special thanks to Fredrik, Matic, Fabio, Erika, Dr. Emil, Malin, Dr. Hilde, Lina, Charlotte, Stefan, Dr. Johan, Dr. Jie, Dr. Ruisheng, Dr. Rabia, Dr. Kjell, Dr. Aleksandra D., Dr. Maxim, Dr. Jia-Fei, Dr. Carina, Dr. Alexandra B., Dr. Derar, and Dr. Jiajie for making teaching, projects and lunch breaks great times. Also thanks to Dr. Francoise, and Prof. em. Thomas for always providing help and good advice. I need to thank my Girls: Randi, how to thank someone that been around for so long? I don’t know how the Uppsala adventure would have ended without you… It certainly wouldn’t have been this fun and crazy. I can only hope and try to do the same for you as you do for me. Dr. Hanna, you have a great way of putting things in perspective and making them seem less problematic. You have been an awesome support during this time! Jenny, you have a wonderful sense of humour that I have really missed since you moved. I look forward to more, brunches, dinners and spa trips with you all in the future! Finally, I want to thank my family: Mamma, for always being just a phone call away and always listening. Pappa, for always supporting my choices in life and believing in me. Linus, I have always called you when things seem bad, and you always help. Thank you so much! Jonas, you are an inspiration in so many ways! And my favourite hipster in the world! Helen, thank you for cor-recting my English and for all the PhD-to-PhD support talks! Jakob and Sonja, my own little family. Without you two this would have been so difficult. To come home to the two of you have always been the highlight of every day! Jag älskar er.

59

9 References

1. Bryant, R. G. J. Chem. Edu. 1983, 60, 933–935. 2. Inokuma, Y., Yoshioka, S., Ariyoshi, J., Arai, T., Hitora, Y., Takada, K.,

Matsunaga, S., Rissanen, K., & Fujita, M. Nature 2013, 495, 461–466. 3. Rissanen, K. Chem. Soc. Rev. 2017, 46, 2638–2648. 4. Eliel, E. L. & Wilen, S. H., 1994, The Stereochemistry of Organic Compounds,

Wiley Interscience, New York. 5. Cahn, R. S., Ingold, C. & Prelog, V. Angew. Chem. Int. Ed. Engl. 1966, 5, 385–

415. 6. IUPAC, Compendium of Chemical Terminology, 2nd ed. (the "Gold Book")

(1997). Online corrected version: (2006–) "axial chirality". doi:10.1351/goldbook.A00547.

7. IUPAC, Compendium of Chemical Terminology, 2nd ed. (the "Gold Book") (1997). Online corrected version: (2006–) "helicity". doi:10.1351/goldbook.H02763.

8. Blackmond, D. G. Philos. Trans. R. Soc. B Biol. Sci. 2011, 366, 2878–2884. 9. Miller, M. T. Trans. Am. Ophthalmol. Soc., 1991, 89, 623–674. 10. Brooks, W. H., Guida, W. C., & Daniel, K. G. Curr. Top. Med. Chem. 2011,

11, 760–770. 11. IUPAC, Compendium of Chemical Terminology, 2nd ed. (the "Gold Book")

(1997). Online corrected version: (2006–) "relative configuration". doi:10.1351/goldbook.R05260.

12. Harada, N. Chirality 2008, 20, 691–723. 13. Deschamps, J. R. Life Sci. 2010, 86, 585–589. 14. Parsons, S. Tetrahedron Asymmetry 2017, 28, 1304–1313. 15. Yoo, H. D., Nam, S. J., Chin, Y. W., & Kim, M. S. Arch. Pharm. Res. 2016,

39, 143–153. 16. Chhetri, B. K., Lavoie, S., Sweeney-Jones, A. M., & Kubanek, J. Nat. Prod.

Rep. 2018, 35, 514–531. 17. Błażewska, K. M. & Gajda, T. Tetrahedron Asymmetry 2009, 20, 1337–1361. 18. Seco, J. M., Quiñoá, E., & Riguera, R. Chem. Rev. 2012, 112, 4603–4641. 19. McCreary, M. D., Lewis, D. W., Wernick, D. L., Whitesides, G. M., Goering,

H. L., Eikenberry, J. N., Koermer, G. S., Amer, J., M Sanders, J. K., Hanson, S. W., Williams, D. H., & Weissman, S. I. J. Am. Chem. Soc., 1967, 2, 1038-1054.

20. Higashibayashi, S., Czechtizky, W., Kobayashi, Y., & Kishi, Y. J. Am. Chem. Soc. 2003, 125, 14379–14393.

21. Seike, H., Ghosh, I., & Kishi, Y. Org. Lett. 2006, 8, 3861–3864. 22. Bifulco, G., Dambruoso, P., Gomez-Paloma, L., & Riccio, R. Chem. Rev.

2007, 107, 3744–3779. 23. Cicero, D. O., Barbato, G., & Bazzo, R. J. Am. Chem. Soc 1995, 117, 1027–

1033. 24. Grimmer, C. D., & Slabber, C. A. Magn. Reson. Chem. 2015, 53, 590–595.

60

25. Harada, N., Nakanishi K., Circular Dichroic Spectroscopy: Exciton Coupling in Organic Stereochemistry, University Sci-ence Books: Mill Valley, Califor-nia, 1983.

26. Harada, N., & Nakanishi, K. Acc. Chem. Res. 1972, 5, 257–263. 27. Gawronski, J. & Skowronek, P. Curr. Org. Chem. 2004, 8, 65–82. 28. Valderrey, V., Aragay, G., & Ballester, P. Coord. Chem. Rev. 2014, 258–259,

137–156. 29. Berova, N., Pescitelli, G., Petrovic, A. G., & Proni, G. Chem. Commun. 2009,

5958–5980. 30. Palomino-Schätzlein, M., Virgili, A., Gil, S., & Jaime, C. J. Org. Chem. 2006,

71, 8114-8120. 31. Klärner, F. G., & Kahlert, B. Acc. Chem. Res. 2003, 36, 919–932. 32. Leblond, J. & Petitjean, A. ChemPhysChem 2011, 12, 1043–1051. 33. Hardouin-Lerouge, M., Hudhomme, P., & Sallé, M. Chem. Soc. Rev. 2011, 40,

30–43. 34. Williams, R. J. P. Chem. Rev. 1956, 56, 299–328. 35. Pérez, E. M., & Martín, N. Pure Appl. Chem. 2010, 82, 523–533. 36. Schleyer, P. V. R., Maerker, C., Dransfeld, A., Jiao, H., & Van Eikema

Hommes, N. J. R. J. Am. Chem. Soc., 1996, 118, 6317–6318. 37. Gomes, J. A. N. F., & Mallion, R. B. Chem. Rev. 2001, 101, 1349–1383. 38. Peeks, M. D., Claridge, T. D. W., & Anderson, H. L. Nature 2017, 541, 200–

203. 39. Norrehed, S., Polavarapu, P., Yang, W., Gogoll, A., & Grennberg, H. Tetra-

hedron 2013, 69, 7131–7138. 40. Blom, M., Norrehed, S., Andersson, C.-H., Huang, H., Light, M. E., Bergquist,

J., Grennberg, H., & Gogoll, A. Molecules 2015, 21, 16. 41. Norrehed, S., Johansson, H., Grennberg, H., & Gogoll, A. Chem. Eur. J. 2013,

19, 14631–14638. 42. Huang, X., Rickman, B. H., Borhan, B., Berova, N., & Nakanishi, K. J. Am.

Chem. Soc. 1998, 120, 6185–6186. 43. Huang, X., Fujioka, N., Pescitelli, G., Koehn, F. E., Williamson, R. T., Na-

kanishi, K., & Berova, N. J. Am. Chem. Soc. 2002, 124, 10320–10335. 44. Kurtán, T., Nesnas, N., Li, Y. Q., Huang, X., Nakanishi, K., & Berova, N. J.

Am. Chem. Soc. 2001, 123, 5962–5973. 45. Li, X., Tanasova, M., Vasileiou, C., & Borhan, B. J. Am. Chem. Soc. 2008,

130, 1885–1893. 46. Bird, C. W., Cheeseman, G. W. H. & Murray, R. D. H., Naturally occurring oxygen-

ring compounds, Aromatic and Heteroaromatic Chemistry: Volume 5, The Chemi-cal Society, 1977.

47. Livingstone, R. Nature 1963, 200, 473–501. 48. Mishra, B. B., & Tiwari, V. K. Eur. J. Med. Chem. 2011, 46, 4769–4807. 49. Bhyrappa, P., Borovkov, V. V., & Inoue, Y. Org. Lett. 2007, 9, 433–435. 50. Anyika, M., Gholami, H., Ashtekar, K. D., Acho, R., & Borhan, B. J. Am.

Chem. Soc. 2014, 136, 550–553. 51. Borovkov, V. V., Lintuluoto, J. M., & Inoue, Y. J. Am. Chem. Soc. 2001, 123,

2979–2989. 52. Borovkov, V. V., Hembury, G. A., & Inoue, Y. Acc. Chem. Res. 2004, 37, 449–

459. 53. Tanasova, M. & Borhan, B. European J. Org. Chem. 2012, 3261–3269. 54. Dechaine, G. P., Maham, Y., Tan, X., & Gray, M. R. Energ. Fuel. 2011, 25,

737–746.

61

55. Li, X., Burrell, C. E., Staples, R. J., & Borhan, B. J. Am. Chem. Soc. 2012, 134, 9026–9029.

56. Benesi, H. A. & Hildebrand, J. H. J. Am. Chem. Soc. 1949, 71, 2703–2707. 57. Scatchard, G. Ann. N. Y. Acad. Sci. 1949, 51, 660–672. 58. Scott, R. L. Rec. Trav. Chim. Pays-Bas 1956, 75, 787–789. 59. Fielding, L. Tetrahedron 2000, 56, 6151–6170. 60. Thordarson, P. Chem. Soc. Rev. 2011, 40, 1305–1323. 61. Kucharski, T. J., Tian, Y., Akbulatov, S., & Boulatov, R. Energy Environ. Sci.

2011, 4, 4449–4472. 62. Blom, M., 2015, Light-Triggered Conformational Switches for Modulation of

Molecular Recognition: Applications for Peptidomimetics and Supramolecu-lar Systems (PhD dissertation). Uppsala.

63. McMurry, J. E. Chem. Rev. 1989, 89, 1513–1524. 64. Williamson, A. London, Edinburgh, Dublin Philos. Mag. J. Sci. 1850, 37, 350–

356. 65. Rothemund, P. J. Am. Chem. Soc. 1935, 57, 2010–2011. 66. Lindsey, J. S., Hsu, H. C., & Schreiman, I. C. Tetrahedron Lett. 1986, 27,

4969–4970. 67. Kumar, A., Maji, S., Dubey, P., Abhilash, G. J., Pandey, S., & Sarkar, S. Tet-

rahedron Lett. 2007, 48, 7287–7290. 68. Rothemund, P. & Menotti, A. R. J. Am. Chem. Soc. 1948, 70, 1808–1812. 69. Adler, A. D., Longo, F. R., Kampas, F., & Kim, J. J. Inorg. Nucl. Chem. 1970,

32, 2443–2445. 70. Lindsey, J. S., & Woodford, J. N. Inorg. Chem. 1995, 34, 1063–1069. 71. Yao, S. A., Hansen, C. B., & Berry, J. F. Polyhedron 2013, 58, 2–6. 72. Spyroulias, G. A., Despotopoulos, A. P., Raptopoulou, C. P., Terzis, A., de

Montauzon, D., Poilblanc, R., & Coutsolelos, A. G. Inorg. Chem. 2002, 41, 2648–2659.

73. Abdulaeva, I. A., Birin, K. P., Michalak, J., Romieu, A., Stern, C., Bessmertnykh-Lemeune, A., Guilard, R., Gorbunova, Y. G., & Tsivadze, A. Y. New J. Chem 2016, 40, 5758.

74. Beletskaya, I., Tyurin, V. S., Tsivadze, A. Y., Guilard, R., & Stern, C. Chem. Rev. 2009, 109, 1659–1713.

75. Denison, R. B. Trans. Faraday Soc. 1912, 8, 20–34. 76. Likussar, W., & Boltz, D. F. Anal. Chem. 1971, 43, 1265–1272. 77. Lambert, J. B., & Mazzola, E. P., 2004, Nuclear Magnetic Resonance Spec-

troscopy: An Introduction to Principles, Applications, and Experimental Meth-ods, Pearson Education Inc., New Jersey, USA.

78. Staszewska-Krajewska, O., Bocian, W., Maciejko, M., Szcześniak, P., Szymczak, K., Chmielewski, M., & Furman, B. Arkivoc 2014, 3, 143–153.

79. Kleinpeter, E., Krüger, S., & Koch, A. J. Phys. Chem. A 2015, 119, 4268–4276.

80. Karplus, M. J. Chem. Phys. 1959, 30, 11–15. 81. Karplus, M. J. Am. Chem. Soc. 1963, 85, 2870–2871. 82. Neuhaus, D., & Williamson, M. 1989, The Nuclear Overhauser Effect in Struc-

tural and Conformational Analysis, VCH Publishers Inc., New York, USA. 83. Yongye, A. B., Foley, B. L., & Woods, R. J. J. Phys. Chem. A 2008, 112, 2634–

2639. 84. Wu, J., Lorenzo, P., Zhong, S., Ali, M., Butts, C. P., Myers, E. L., & Aggarwal,

V. K. Nature 2017, 547, 436–440. 85. Bame, J., Hoeck, C., Carrington, M. J., Butts, C. P., Jäger, C. M., & Croft, A.

K. Phys. Chem. Chem. Phys. 2018, 20, 7523–7531.

62

86. Leach, A. R. 2001, Molecular modelling, principles and applications, Pearson Education Limited, Essex, UK.

87. Bochicchio, D., & Pavan, G. M. Adv. Phys. X, 2018, 3, 1436408. 88. Nechaeva, L. S., & Butyrskaya, E. V. J. Anal. Chem. 2012, 67, 312–320. 89. dos Passos Menezes, P., dos Santos, P., Dória, G., de Sousa, B., Serafini, M.,

et. al. AAPS PharmSciTech. 2016, 18, 49–57. 90. Harder, E., Damm, W., Maple, J., Wu, C., Reboul, M., Xiang, J. Y., Wang, L.,

Lupyan, D., Dahlgren, M. K., Knight, J. L., Kaus, J. W., Cerutti, D. S., Krilov, G., Jorgensen, W. L., Abel, R., & Friesner, R. A. J. Chem. Theory Comput. 2016, 12, 281–296.

91. Hohenberg, P., & Kohn, W. Phys. Rev. 1964, 136, B864–B871. 92. Sahni, V. in Quantal Density Functional Theory (Springer Berlin Heidelberg,

2013). 99–123. 93. Van Mourik, T., & Gdanitz, R. J. J. Chem. Phys. 2002, 116, 9620–9623. 94. Vondrášek, J., Bendová, L., Klusák, V., & Hobza, P. J. Am. Chem. Soc. 2005,

127, 2615–2619. 95. Becke, A. D. J. Chem. Phys. 1993, 98, 1372–1377. 96. Perdew, J. P., Ernzerhof, M., & Burke, K. J. Chem. Phys. 1996, 105, 9982–

9985. 97. Grimme, S. J. Comput. Chem. 2006, 27, 1787–1799. 98. Kepp, K. P. Inorg. Chem. 2016, 55, 9461–9470. 99. Kadish, K. M., & Bottomley, L. A. J. Am. Chem. Soc. 1977, 99, 2380–2382. 100. Goldammer, E. V. J. Mol. Struct. 1980, 60, 127–130. 101. Andreev, V. P., Sobolev, P. S., & Tafeenko, V. A. Russ. J. Gen. Chem. 2017,

87, 1179–1187. 102. Chumakov, D. E., Khoroshutin, A. V., Anisimov, A. V., & Kobrakov, K. I.

Chem. Heterocycl. Compd. 2009, 45, 259–283. 103. Shanmugathasan, S., Edwards, C., & Boyle, R. W. Tetrahedron 2000, 56,

1025–1046. 104. Bhyrappa, P., & Purushothaman, B. J. Chem. Soc., Perkin Trans. 2 2001, 238–

242. 105. Birnbaum, E. R., Hodge, J. A., Grinstaff, M. W., Schaefer, W. P., Henling, L.,

Labinger, J. A., Bercaw, J. E., & Gray, H. B. Inorg. Chem. 1995, 34, 3625–3632.

106. Bhyrappa, P., & Krishnan, V. Inorg. Chem. 1991, 30, 239–245. 107. Spyroulias, G. A., Despotopoulos, A. P., Raptopoulou, C. P., Terzis, A., de

Montauzon, D., Poilblanc, R., & Coutsolelos, A. G. Inorg. Chem. 2002, 41, 2648–2659.

108. Jaquinod, L., Khoury, R. G., Shea, K. M., & Smith, K. M. Tetrahedron 1999, 55, 13151–13158.

109. Kozlowski, P. M., Piffat, C. A., Rush, T. S., Zgierski, M. Z., Spiro, T. G., & Kumble, R. J. Phys. Chem. B 2002, 104, 5020–5034.

110. Baerends, E. J., Ricciardi, G., Rosa, A., & Van Gisbergen, S. J. A. Coordina-tion Chemistry Reviews 2002, 230, 5–27.

111. Zhang, M. J., Guo, Y. R., Fang, G. Z., & Pan, Q. J. Comput. Theor. Chem. 2013, 1019, 94–100.

112. Hariharan, P. C. & Pople, J. A. Theor. Chim. Acta. 1973, 28, 213–222. 113. Andrae, D., Häußermann, U., Dolg, M., Stoll, H., & Preuß, H. Theor. Chim.

Acta. 1990, 77, 123–141. 114. Zhao, Y., & Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215–241. 115. Miertuš, S., Scrocco, E., & Tomasi, J. Chem. Phys. 1981, 55, 117–129. 116. Miertus,̃ S., & Tomasi, J. Chem. Phys. 1982, 65, 239–245.

63

117. Pascual‐ahuir, J. L., Silla, E., & Tuñon, I. J. Comput. Chem. 1994, 15, 1127–1138.

118. Wakeham, R. J., Taylor, J. E., Bull, S. D., Morris, J. A., & Williams, J. M. J., Org. Lett. 2013, 15, 702–705.

119. Hyland, M. A., Morton, M. D., & Brückner, C. J. Org. Chem. 2012, 77, 3038–3048.

120. Hewage, N., Zeller, M., & Brückner, C. Org. Biomol. Chem 2017, 15, 396. 121. Polavarapu, P., Melander, H., Langer, V., Gogoll, A., & Grennberg, H. New J.

Chem. 2008, 32, 643–651.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1810

Editor: The Dean of the Faculty of Science and Technology

A doctoral dissertation from the Faculty of Science andTechnology, Uppsala University, is usually a summary of anumber of papers. A few copies of the complete dissertationare kept at major Swedish research libraries, while thesummary alone is distributed internationally throughthe series Digital Comprehensive Summaries of UppsalaDissertations from the Faculty of Science and Technology.(Prior to January, 2005, the series was published under thetitle “Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-382219

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2019