ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2013

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1070

Modulation of MolecularProperties

Host–Guest Interactions for Structural Analysisand Chemical Reactions

SARA NORREHED

ISSN 1651-6214ISBN 978-91-554-8747-8urn:nbn:se:uu:diva-207138

Dissertation presented at Uppsala University to be publicly examined in B41, BMC,Husargatan 3, Uppsala, Friday, October 25, 2013 at 09:30 for the degree of Doctor ofPhilosophy. The examination will be conducted in English.

AbstractNorrehed, S. 2013. Modulation of Molecular Properties: Host–Guest Interactions forStructural Analysis and Chemical Reactions. Acta Universitatis Upsaliensis. DigitalComprehensive Summaries of Uppsala Dissertations from the Faculty of Science andTechnology 1070. 61 pp. Uppsala. ISBN 978-91-554-8747-8.

This thesis concerns the construction, use and modulation of various host–guest systems, fromsmall bispidines for binding of inorganic ions to bisporphyrin clips for supramolecular systems.

Small flexible molecules undergo fast conformational movements when in solution.These conformational movements generate time-averaged population-weighted chemical shifts,coupling constants and NOEs when analysed by NMR spectroscopy. A bisporphyrin clip wasdesigned to be used as a host for restriction of conformational movements of small flexiblemolecules by ditopic metal-ligand binding. Based on conformational analysis in combinationwith NMR analysis of molecular flexibility in solution (NAMFIS), the relative stereochemistryof flexible alditol-derived diamines containing three or four consecutive stereocentres could bedetermined.

To further explore the idea of conformational deconvolution via host–guest binding,two flexible molecular tweezers with photoswitchable moieties were developed. Uponphotoswitching cis/trans isomerisation facilitates the opening and closing of these bisporphyrinhosts. A guest molecule could then be exposed to a “catch and stretch” or “catch and release”effect. Preliminary studies have shown that photoisomerisation of the constructed systems ispossible without photodecomposition and preliminary binding studies have been conducted.

Controlled modulation of molecular conformations is of interest especially if theconformational steering activates a unit working as a nucleator in a larger structure orfacilitates a reaction. The protonation-triggered modulation of bispidine conformations has beeninvestigated. In addition to previously reported conformations we have observed that upondiprotonation a bispidine derivative can be driven into the unusual boat-boat conformation.

Finally, the unexpected formation of persistent organic radicals with a cyclophane motif fromthe reaction of N,N´-diphenyl-1-5-diazacyclooctane and AgBF4 is described. Interestingly, thesediradicals exhibit features such as intramolecular π-stacking without lateral displacement andalso intramolecular spin pairing.

Keywords: host-guest systems, conformational analysis, NMR spectroscopy, bispidines,porphyrins

Sara Norrehed, Uppsala University, Department of Chemistry - BMC, Synthetical OrganicChemistry, Box 576, SE-751 23 Uppsala, Sweden.

© Sara Norrehed 2013

ISSN 1651-6214ISBN 978-91-554-8747-8urn:nbn:se:uu:diva-207138 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-207138)

Ett spektrum, flera spektra

List of Papers

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

I Norrehed, S.; Polavarapu, P.; Yang, W.; Gogoll, A.; Grennberg,

H. Conformational restriction of flexible molecules in solu-tion by a semirigid bis-porphyrin molecular tweezer. Tetra-hedron 2013, 69, 7131–7138.

II Norrehed, S.; Johansson, H.; Gogoll, A.; Grennberg, H. Improved stereochemical analysis of conformationally flexi-ble diamines by binding to a bisporphyrin molecular clip. Chem. Eur. J. DOI: 10.1002/chem.201300533

III Norrehed, S.; Blom, M.; Andersson, C.-H.; Huang, H.; Light, M. E.; Grennberg, H.; Gogoll, A. Photomodulable bis-porphyrin molecular tweezers for conformational studies of diamine guests. Preliminary manuscript

IV Norrehed, S.; Erdélyi, M.; Light, M. E.; Gogoll, A. Protonation-triggered Conformational Modulation of an N,N´-Dialkylbispidine: First observation of the elusive boat-boat conformer. Org. Biomol. Chem. 2013, 11, 6292–6299.

V Norrehed, S.; Karlsson, C.; Light, M. E.; Thapper, A.; Huang P.; Gogoll, A. Synthesis of persistent organic dication diradicals incorporating a cyclophane motif with intramolecular π-stacking. Submitted manuscript

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

VI Hamnevik, E.; Blikstad, C.; Norrehed, S.; Widersten, M. Kinet-

ic Characterization of Rhodococcus ruber DSM 44541 Alco-hol Hydrogenase A. Submitted manuscript

Contribution report

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

I Performed and interpreted NMR and UV binding studies, con-

formational analysis, and diffusion studies for 1,ω-diamino-alkanes. Synthesized the 1,20-eicosanediamine guest. Contrib-uted to the writing of the revised manuscript.

II Planned and screened synthetic routes to the diamines and per-formed parts of the final synthetic work. Developed the desalt-ing procedure. Performed all NMR studies, characterisation and the NAMFIS-type analyses. Co-developed the strategy for stereochemical analysis of the target compounds. Contributed to writing the manuscript.

III Performed the majority of the synthetic work, UV and NMR studies. Wrote the first version of the manuscript.

IV Performed all protonation experiments, the majority of the NMR studies and all of the RDC analyses. Took part in selec-tion of suitable NMR pulse sequences. Contributed to writing the manuscript.

V Carried out the synthetic work and performed all analyses ex-cept the X-ray crystallographic analysis and investigations of electrochemical and magnetic properties. Contributed to writing the manuscript.

Contents

1. Introduction .............................................................................................. 11 1.1 What is molecular structure? ........................................................... 11 1.2 Supramolecular chemistry and host–guest systems ........................ 11

1.2.1 Supramolecular tools ............................................................. 12 1.2.2 Bispidines .............................................................................. 13 1.2.3 Bisporphyrins and bispidines as hosts ................................... 14

2. Overview of methods ............................................................................... 15 2.1 NMR spectroscopy for analysis of flexible molecules .................... 15

2.1.1 Chemical shifts ...................................................................... 15 2.1.2 Coupling constants and J-based configurational analysis ...... 17 2.1.3 Nuclear Overhauser effects .................................................... 18 2.1.4 Diffusion coefficients ............................................................ 21

2.2 Strategies for structural analysis of flexible molecules ................... 23 2.2.1 NAMFIS ................................................................................ 23 2.2.2 Residual dipolar couplings ..................................................... 24

2.3 Determination of binding constants and complex stoichiometries . 26

3. Conformational restriction and stereochemical deconvolution of flexible diamines in a bisporphyrin clip (Papers I and II) ........................ 28 3.1 Synthesis of diamines with multiple stereogenic centres ................ 29 3.2 Characterisation of host–guest complexes ...................................... 32 3.3 Assignment deconvolution via J-based arguments and NAMFIS .. 34 3.4 Analysis of relative stereochemistry ............................................... 36 3.5 Conclusions ..................................................................................... 37

4. Photoswitchable bisporphyrin tweezers (Paper III) ................................. 38 4.1 Synthesis of photoswitchable bisporphyrin tweezers...................... 39 4.2 Binding studies with aliphatic diamines ......................................... 41 4.3 Photoisomerisation .......................................................................... 43 4.4 Conclusions ..................................................................................... 43

5. Regulation of bispidine conformations and the observation of a boat-boat conformer (Paper IV) ....................................................................... 44 5.1 Identifying conformations by NOEs and RDCs .............................. 45 5.2 Protonation equilibria ...................................................................... 46 5.3 Conclusions ..................................................................................... 47

6. An unexpected reaction: Formation of persistent dication diradical cyclophanes (Paper V) ............................................................................. 48 6.1 Persistent organic radicals ............................................................... 49 6.2 Formation of cyclophane diradicals ................................................ 49 6.3 Properties of cyclophane diradicals ................................................ 52 6.4 Conclusions ..................................................................................... 53

7. Concluding remarks ............................................................................. 54

8. Summary in Swedish ........................................................................... 55

9. Acknowledgements.............................................................................. 57

10. References ........................................................................................... 59

Abbreviations

CIS Complexation-induced shift

D Self-diffusion coefficient

DCC N,N'-Dicyclohexylcarbodiimide

G Guest

H Host

HG Host−guest complex

HOBt Hydroxybenzotriazole

HPLC High performance liquid chromatography

HSQC Heteronuclear single quantum coherence

J Scalar spin–spin coupling constant

K Binding constant

LEDPGSE Longitudinal encode-decode delay pulsed gradient spin echo

NAMFIS NMR analysis of molecular flexibility in solution

NMR Nuclear magnetic resonance

NOE Nuclear Overhauser effect

NOESY Nuclear Overhauser effect spectroscopy

o.n. Overnight

PBLG Poly(-γ-benzyl-L-glutamate)

Q Cornilescus quality factor

r.t. Room temperature

RDC Residual dipolar coupling

RMS Root mean square

SVD Single value decomposition

TPP Tetraphenylporphyrin

UV/Vis Ultraviolet/visible

δ Chemical shift

11

1. Introduction

1.1 What is molecular structure? The structure of an organic molecule is described by a three-dimensional arrangement of atoms connected by covalent bonds. Often, due to the possi-bility of bond rotations, a series of arrangements called conformations may be adopted. Increasingly, non-covalent interactions between atoms have been recognised to impact the structures and functions of molecules. Over the years, the field of chemistry has changed its primary focus from the mo-lecular structure of individual compounds to the function exhibited due to structure-related properties.1 Traditionally the X-ray crystallographic structure of solid material was con-sidered the definitive representation of a molecule. However, this is actually just the arrangement adopted by the molecule when it is confined to a crystal lattice. A flexible molecule in solution lacks a unique structure. During char-acterisation fast conformational movements will generate a population-weighted time-averaged representation of the molecule. Thus, what we be-lieve we can see may not be what we have, and as most experiments and analyses are performed in solution, these are important considerations.

1.2 Supramolecular chemistry and host–guest systems Supramolecular systems are assemblies of molecules held together by re-versible non-covalent binding interactions. The field of supramolecular chemistry is said to “go beyond the molecule”, striving towards the construc-tion and understanding of molecular assemblies.2 The term ‘host–guest’ de-scribes how (usually) a larger molecule (the host) specifically binds a small-er molecule (the guest).3

The non-covalent interactions that unite the host and guest are weaker than covalent bonds, but when several non-covalent interactions are combined, significant binding strength can be reached. Common non-covalent interac-tions include ion–ion, ion–dipole, dipole–dipole, van der Waals, metal-ligand, and π–π interactions,4 as well as hydrogen bonding5 (Figure 1).

12

Figure 1. Summary of common non-covalent interactions in host–guest systems.

Many parameters affect how a host and guest interact, therefore designing such a system is a difficult task. The main considerations involve size and electronic complementarity, enthalpic and entropic contributions, pre-organisation of the host and solvent effects.6 1.2.1 Supramolecular tools A supramolecular complex that fulfills a function can be called a “supramolecular tool”. Metallated bisporphyrins in which the porphyrins are separated by a spacer are widely used hosts for the ditopic binding of guests7

and are commonly incorporated in supramolecular clips (with rigid spacers) or tweezers (with flexible spacers).8 The properties of the spacer determine the cleft size and flexibility of a host. Hosts can assist in the analysis of guest structure; for example by exhibiting size-selective binding9 or enabling determination of absolute stereochemistry.10 The absolute stereochemistry of a chiral guest has been determined based on the helicity it induced in an achiral molecular clip. The helicity was determined by the absolute stereo-chemistry of the guest, and could be measured by circular dichroism (CD). The process of inducing chirality in an otherwise achiral structure is known as chirogenesis.11 In this type of analysis the guest conformation is usually undetermined or assumed to be dominated by a single conformer. In con-trast, in this thesis, bisporphyrin molecular clips and tweezers have been used to restrict the conformational movements of a flexible guest, allowing its relative stereochemistry to be assigned via ditopic metal−ligand binding (Figure 2).

13

Figure 2. A supramolecular tweezer can assist in restricting the conformational movements of a flexible molecular guest.

1.2.2 Bispidines The bispidine structure (N,N’-3,7-diazabicyclo[3.3.1]nonane) stems from the natural product sparteine (Figure 3) and was originally prepared by Mannich.12 Sparteine is found in several plants and has been investigated extensively due to its pharmacological effects, for example as an anti-arrhythmic drug and an ion-channel blocker.13 Sparteine also has interesting chemical applications as a chiral base or ligand in asymmetric synthesis14. The simpler bispidine backbone has similar pharmacological properties to sparteine but is more easily synthesised and functionalised.15

a b

Figure 3. (a) The natural product sparteine; (b) the bispidine backbone N,N’-3,7-diazabicyclo[3.3.1]nonane.

The bispidine backbone is a bidentate diamine with specific conformational preferences.16 Its high degree of pre-organisation17 makes the bispidine structure a suitable receptor (host) for Lewis acidic guests such as protons or metal ions.

14

1.2.3 Bisporphyrins and bispidines as hosts When a ligand can form two or more bonds to the same guest, a ‘chelate’ is formed. The word chelate is Latin for ‘claws’, and this type of interaction enhances the stability of a complex, which is known as the ‘chelate effect’. For the same reason bispidines are highly basic and act as proton sponges.18 In addition to the chelate effect, the stability of a complex may be further increased by the pre-organisation of the host and the complementarity be-tween host and guest. The chelate effect is particularly evident in systems with multidentate binding,19 such as the binding of a metal cation in the cen-ter of a porphyrin ring. The metallobisporphyrin and bispidine hosts can each bind one or two guests as they each have two binding sites. Alternative-ly, the two binding sites can bind ditopically to a single guest (Figure 4). The bispidine backbone is more pre-organised and will thus be more selec-tive with respect to guest size. A bisporphyrin clip or even more so tweezers, can change its porphyrin-porphyrin sidewall distance to accept a wider range of guest sizes. In the bisporphyrin case, a guest can bind either inside or out-side of the cleft. When ditopic binding is possible a preference for binding inside the cleft can be observed.20

Figure 4. The bisporphyrin and bispidine hosts have the ability to bind to a guest monotopically or ditopically. In the case of monotopic binding, especially to bisporphyrins, several binding scenarios are possible. In the case of weak binding, dynamic exchange between different binding scenarios is likely.

15

2. Overview of methods

2.1 NMR spectroscopy for analysis of flexible molecules

Nuclear magnetic resonance (NMR) spectroscopy is among the most widely used analytical techniques for structural assignment and for elucidating the three-dimensional structures of organic molecules. Also, along with UV/Vis spectroscopy, NMR is one of the most important techniques for studying complexation phenomena in supramolecular chemistry. The main parameters measured and techniques used in this thesis are briefly described below, in the context of their use and limitations in analysis for flexible molecules. In section 2.2 two strategies that combine several parameters to solve flexible structures are described.

2.1.1 Chemical shifts The chemical shift (δ, in ppm) of a nucleus depends partially on the sur-rounding electron density, which is mainly influenced by atoms to which it is directly bonded. Factors that can influence electron density include induc-tion, resonance, and hybridisation.21 In addition to these, there is the anisot-ropy effect which is more closely related to the three-dimensional structure of a molecule. Anisotropy effects arise from the local magnetic fields gener-ated by electrons in chemical bonds. Protons exposed to such local magnetic fields will assume resonance frequencies that depend on their position, and that result in either shielding or deshielding (Figure 5). Due to contributions from the different factors, it is difficult to relate chemical shifts to molecular geometry in a precise manner. Yet, attempts to find empirical correlations between structure and chemical shifts are not uncommon, for example in the conformational characterisation of bispidine derivatives.22,16c

16

δ−

δ+

deshieldinghigher δ

δ+

δ−

δ+

δ− shieldinglower δ

β γ

Figure 5. The anisotropy effect of single bonds impacts the chemical shift of nearby protons depending on their position in the local magnetic field, which is symbolised by the cones.

The ring current effect is a type of anisotropy effect that arises in cyclic ar-omatic systems,23 the classic example of which is the benzene ring (Figure 6) and can be described by the Johnson–Bovey model.24 In a supramolecular clip with bisporphyrin sidewalls, the ring current effect causes particularly large decreases in chemical shift for the protons of a guest molecule that is bound between the porphyrin units. These changes in chemical shift are advantageous as the guest molecule signals are shifted to a non-crowded spectral region and spread apart according to the proximity of the associated protons to the sidewalls.

Figure 6. The ring current effect of benzene. π-Electrons in the ring system circulate under the influence of B0 to produce a current, which in turn creates a local magnetic field that opposes B0. This partially cancels B0 above and below the ring, causing shielding (∆δ < 0), whereas in the plane of the ring, B0 and the local field are addi-tive, causing deshielding (∆δ > 0).

17

2.1.2 Coupling constants and J-based configurational analysis The scalar coupling constant (J) provides geometric information about a molecular structure. In particular the vicinal couplings over three bonds (3JHH) that can be described by the Karplus equation25 show a characteristic relationship to the dihedral angle (φ) between the involved protons. The magnitude of these couplings depends not only on the dihedral angle, but also on bond lengths, electron densities, and substituent effects (Figure 7).

Figure 7. The vicinal coupling constant 3JHH is related to the dihedral angle, as de-scribed by the Karplus equation (▬). Substituent effects result in a considerable variation of these coupling constants, as illustrated by the black curves. The interpretation of J values in rigid molecules can be straightforward. In highly flexible molecules only time-averaged population-weighted J values can be measured. When rotation occurs freely, all rotamers contribute equal-ly to the coupling constant, provided equal energies (Figure 8). When rotamer energies are different the contribution of each rotamer to J is related to its abundance.

gauche trans gauche

AB

A

B

AB

JAB = (Jgauche + Jtrans + Jgauche)/3

Figure 8. The vicinal coupling constant 3J depends on the dihedral angle between the protons and is described by the Karplus equation. When rotation occurs freely and energies are equal all rotamers contribute equally to the coupling constant.

18

As J values are useful in assigning structure and determining important structural features, several methods to interpret J values in the spectra of flexible molecules have been developed; these include J-based configurational analysis26 and Kishi’s Universal NMR Database.27 In a J-based analysis the Newman projection of a molecular segment is considered, and each J is estimated relative to the others, allowing a relative configurational assignment (Figure 9). J-based configurational analyses have been used e.g. to elucidate the structure of several natural products displaying multiple conformer equlibria.28

JlargeJsmall Jsmall Jsmallb

Jsmall

a

Jlarge

A

B

A

B

Figure 9. The principle of a J-based conformational analysis. (a) A dihedral angle of 60o gives a smaller J than a dihedral angle of 180o (b) Via a J-based analysis over a segment of several connected bonds, a relative configuration can be assigned.

2.1.3 Nuclear Overhauser effects The Nuclear Overhauser effect (NOE) is one of the most valuable NMR parameters for elucidating three-dimensional structural features of organic molecules because it relates protons connected through space rather than through bonds. NOEs arise from dipole-dipole relaxation between spins through space and depend strongly on intermolecular distance and molecular motion. NOEs are typically observed between protons up to approximately 5 Å apart.29 In contrast to the case for chemical shift and coupling constants the interconversion of conformations does not generate a NOE that is a simple average, but rather a weighted value (Figure 10 and Figure 11). When con-formational exchange occurs, the NOE is averaged according to Equation 2.1 (η is NOE, <…> denotes the average value), meaning that stronger in-teractions (which occur between protons closer in space) will contribute more to the observed value.30

η ∝< r-6 >-1/6 Equation 2.1

19

4.2 Å 2.8 Å

apparent distance from NOE (not accounting for effect of conformational

averaging)

average distance

2.5 Å 6 Å

Conformer 1 Conformer 2

Figure 10. Fast conformational changes affect the observed NOE, with emphasis being placed on stronger through-space interactions. When measured experimental-ly, two rapidly interconverting conformers in which spins A and B are separated by 2.5 Å and 6.0 Å, respectively, will show an NOE corresponding to a distance of 2.8 Å, as opposed to the actual average distance of 4.2 Å.

Figure 11. In case of conformer equilibria the cross-peak volumes measured by 2D NOESY experiments will be weighted towards the stronger interaction. Often the size of the NOE is interpreted in terms of interatomic distance. However, this size is markedly affected by the molecular correlation time (Figure 12). Therefore, quantitative analysis involving NOEs are usually based on NOE buildup rates. The volume of a cross-peak in a 2D NOESY experiment is related to the interatomic distance between two protons by Equation 2.2. Given a known distance, e.g. the distance between protons in a –CH2– group, other distances can be calculated.

= / 1/6 Equation 2.2

20

Figure 12. The maximum NOE plotted against the logarithmic ωτc scale for steady-state NOE experiments. ω = Larmor frequency, τc = correlation time. The NOE builds up during what is called the “mixing time” in the transient NOE pulse sequence. At the beginning of this period, the two cross-relaxing spins (i.e. nuclei) behave as an isolated spin pair, and NOEs build up linearly with the mixing time. At longer mixing times, spin relaxation starts to com-pete with NOE build-up, causing NOEs to decay and deviate from the linear build-up (Figure 13). In order for equation 2.2 to be valid and interatomic distances to be estimated, the mixing time must be chosen so that NOE build-up remains in the linear phase.30 A build-up curve is normally con-structed out of a series of (at least five) experiments with different mixing times in order to choose an appropriate value. The conformational averaging of the NOEs as described above is the main reason why the analyses of even small conformationally flexible molecules are quite difficult.

Figure 13. The NOEs build up during the mixing time, but start to decay when the simultaneous spin relaxation begins to dominate. Interatomic distances can be esti-mated via Equation 2.2 from NOEs measured during the period of linear build-up. <…> is the average value.

21

2.1.4 Diffusion coefficients The self-diffusion coefficient (D) describes the random translational (Brownian) motion of a molecule or ion in solution, which is thermally driv-en.31 In a given solvent at a given temperature, D depends on the molecular weight, size and shape of the molecule. It is sensitive to aggregation and complexation and to molecular interactions, which makes it to an important parameter for studying interaction phenomena within supramolecular chem-istry. When separate, the components of a host–guest system usually have significantly different self-diffusion coefficients because they are different in size; however, when complexed, they should have the same D. Thus, diffu-sion measurements of host–guest systems can give a direct evidence for binding. There are many pulse sequences available to measure diffusion by NMR, but generally the translational movement of a molecule between two gradient pulses is detected.30,32 The field gradients are applied during the defocusing and refocusing parts of a pulse sequence. Slow-moving molecules will expe-rience almost the same field strength at both times and become refocused, and signal intensity will therefore decay slowly. A fast-moving molecule on the other hand, will experience different magnetic field strengths during the course of the experiment and thus display incomplete refocusing and rapid signal decrease (Figure 14). When equilibria occur between species of dif-ferent shape or size, an average D is obtained.

Figure 14. Left: During the diffusion measurement a slow-moving molecule (a) will experience little change in magnetic field strength and its signal will therefore decay slowly. A fast-moving molecule (b) will experience a larger difference in magnetic field strengths, and its signal will decay rapidly. Right: The collected data is ana-lyzed by plotting the natural logarithm of normalized signal intensity against the square of gradient strength. The slope of the resulting line corresponds to –D, and the actual value is found by relating the measured to a reference with known D, such as H2O in CDCl3.

22

The Stokes–Einstein equation (Equation 2.3 where kb = the Boltzmann con-stant, T = temperature, η = viscosity and rs = Stokes radius) relates the diffu-sion coefficient to molecular size, or rather to the hydrodynamic radius of the diffusing species. This relationship is true for spherical species and may thus be more or less applicable to a system depending on the shape of the molecules involved.30 In practice, most macromolecules are not spherical in solution, and the hydrodynamic radius is only indicative of the apparent size of a solvated system (Figure 15).

= Equation 2.3

Figure 15. The hypothetical radius (rM) of a hard sphere that describes the mass and density of the molecule vs. the hydrodynamic radius (rH) of the same molecule. The hydrodynamic radius represents the spherical volume occupied by a molecule or complex when the associated solvent molecules (hydro) and shape (dynamic) effects are included.

23

2.2 Strategies for structural analysis of flexible molecules

The parameters δ and J as well as NOEs, are frequently used to assign the configuration and conformation of an organic molecule. The obvious draw-back to using these parameters is that they can only be directly interpreted by assuming that a single conformation is dominant. In solution, molecules with flexible structural elements are present as several rapidly interconverting conformers and population-weighted averaged values of δ, J and NOE are therefore obtained. To more accurately describe a flexible molecule one option is to consider it as an ensemble rather than an isolated entity, as im-plemented by the NAMFIS method. Additional parameters such as RDCs in combination with conformational analysis might also render a more realistic description of the molecular structure. A further option may be to minimise the number of ensemble members by binding to a rigid entity, such as a mo-lecular clip.

2.2.1 NAMFIS NMR analysis of molecular flexibility in solution (NAMFIS) is a method that has been developed to analyse conformationally averaged data of flexi-ble molecules in solution by matching experimental J values and NOEs to back-calculated values of computed conformations.33 The degree of match-ing expressed as RMS error is used to quantify the quality of a fit. The NAMFIS method requires high-quality NOE data, measured with ap-propriate mixing time in order to be reliable. Although J values can be used in combination with NOEs, they cannot be related to three-dimensional structures as reliably as NOEs due to the substituent effect on the Karplus equation (Figure 7). Therefore is the NAMFIS analysis usually heavily weighted towards the NOE data. The NAMFIS investigation is used to iden-tify the ensemble that best corresponds to the experimental data (Figure 16), along with predicting the probability of each conformer within that ensem-ble. Ensemble analysis provides a much better representation of molecular struc-ture in solution than the single, virtual conformation resulting from time-averaged NMR parameters. In addition, in contrast to structures derived by X-ray crystallography, an ensemble also accounts for the flexibility (the conformational space available) of a molecule. NAMFIS has been used for the analyses of structures including peptides34 and other small-to-medium-sized molecules.35

24

Figure 16. In a NAMFIS analysis a family of conformers is extracted from a larger set that covers the entire conformational space. Left: Possible conformers of 2,3,4-tri-O-methyl-1,5-diaminopentane according to computations, right: required con-formers to generate the observed NMR data.

2.2.2 Residual dipolar couplings Residual dipolar couplings (RDCs) provide an alternative approach to as-signing the spatial arrangements of atoms in molecules. Dipolar couplings are generated by the through-space interactions of nuclear magnetic dipole moments and depend on the distance between nuclei as well as their orienta-tion with respect to the magnetic field, specified by B0 (Figure 17).

isotropic sample

anisotropic sample

z

x

y

θ

θalignment media

Bo

H

O

O

H

O

O

H

O

O

O

O

H

O

O

HO

O

H

H

O

O

Figure 17. Left: Molecules tumbling isotropically are via an alignment medium ordered into an anisotropic state. Right: The dipolar coupling is measured on an anisotropic sample and is dependent on the orientation of the molecule with respect to B0.

25

In isotropic solution, dipolar couplings are averaged to zero by molecular tumbling, so anisotropy has to be introduced for their detection. Solids are far too ordered, as their dipolar couplings measure up to 20 000 Hz, which is too broad to measure in normal practice. For more practical values, several alignment media that introduce only partial alignment (~1 out of 1000 mole-cules) are available; hence the term “Residual Dipolar Coupling”. The types of alignment media available vary according to the solvent and nature of the molecule, and include liquid crystals, bicells, bacteriophages and various polymer gels.36 The liquid crystal PBLG37 was used for the work described in this thesis (Figure 18). The main advantage of the RDC method is that it can be used in situations where few or no NOEs are available. Because RDCs relate bonds to their orientation in relation to B0, an isolated bond situated anywhere in the mole-cule can in principle be used; without requirement on interaction with other nuclei. RDCs do not rely on parameters that can be difficult to relate to mo-lecular structure, such as chemical shifts, J values or NOEs. However, for a reliable RDC analysis a minimum of five independent RDCs are required. The usual approach to the measurement of RDCs is to perform two HSQC experiments, one in isotropic solution (→ 1JC–H) and one in anisotropic solu-tion (→ 1JC–H + 1DC–H). The difference of the measured couplings for the same C–H bond is the residual dipolar coupling (1DC–H). Because of their magnitude 1JC–H (ca. 120–200 Hz) or 1JN–H (ca. 90 Hz) values are usually measured, but in principle, even smaller long-range C–H or N–H couplings nJX–H, or nJH–H (n=2 or 3, with values of J up to ≈12 Hz), can be used if they can be measured with sufficient accuracy.

Figure 18. Left: Molecular structure of PBLG (poly-γ-benzyl-L-glutamate). Middle: PBLG forms a nematic phase when dissolved in chloroform. Right: B0 enforces alignment of the polymers and the sample becomes anisotropic.

26

The experimental values are matched to back-calculated data from computa-tions and the quality of fit is expressed as Q (Cornilescus quality factor). Also for RDCs it is possible to conduct ensemble analyses.

2.3 Determination of binding constants and complex stoichiometries

The interactions of host–guest complexes are quantified by the binding con-stant K (Equations 2.4 and 2.5). The binding constant is a thermodynamic property that is related to the free energy of the association process via the Gibbs equation (Equation 2.6). The free energy can be further related to en-thalpy and entropy (Equation 2.7), meaning if ∆G is negative, the process is spontaneous whereas if ∆G is positive, the reverse process is spontaneous. ⇋ Equation 2.4

= Equation 2.5

∆G=−RTlnK Equation 2.6

∆G=∆H−T∆S Equation 2.7

Table 1 shows the relationship between the binding constant K and the free energy of binding. A larger binding constant, which indicates stronger bind-ing, corresponds to a larger free energy gain for the binding process. The binding of diamines to zinc bisporphyrins discussed in this thesis are in the range of K = 104–106, which corresponds to a ∆G of approximately 20–30 kJ/mol. In comparison, the free energy of a C–C bond formation is –310 kJ/mol38 and the breaking of a π-bond during a cis/trans isomerization re-quires ≈ 250 kJ/mol.39

Table 1. Binding constants (Ka) and corresponding free energies (∆G).

Ka (M-1) ∆G (kJ/mol) 1 0

Stronger binding 102 –11.4104 –22.8106 –34.2108 –45.6

27

Typically the binding constant is measured in a titration experiment. In supramolecular chemistry, the guest is usually gradually added to a solution of the host while a change in UV absorption or chemical shift (NMR) is monitored. The binding constant is then calculated via a suitable binding model. Many graphical (linearization) methods exist, for example the Benesi–Hildebrand,40 the Schatchard41 and the Scott plot42 are among the most widely used. These methods however have serious limitations. For example they are not applicable to all complex stoichiometries, host:guest concentration ranges or K ranges. A non-linear regression analysis based on iterative fitting that is applicable to all systems is preferrable.43,44 Before a binding constant can be determined, the stoichiometry of the host–guest system must be known. The “method of continuous variation” (Job plot) is a powerful tool frequently used to determine complex stoichiometries.45

Figure 19. Schematic representation of a Job plot indicating different host:guest ratio. 1:1 stoichiometry with curve maximum 0.5, 1:2 stoichiometry at 0.66 and 1:3 stoichiometry at curve maximum at 0.75.

By monitoring an absorbance (UV) or chemical shift (NMR) through a series of solutions containing the host and guest in different ratios, while the total concentration is kept constant, a Job plot can be constructed (Figure 19). The maximum point of the curve indicates the molar ratio of the complex, such as a maximum at 0.5 for a 1:1 stoichiometry, or a maximum at 0.75 for 1:3 stoichiometry. Recently methods for estimating of binding constants and identifying equimolar complexes with stoichiometries other than 1:1 (e.g. 2:2) directly from the Job plot have been developed.46 A number of spectro-scopic techniques can be used for this type of investigation, including UV/Vis, NMR, IR or fluorescence spectroscopy, or any other technique in which a measurable parameter depends upon the molar ratio of the interact-ing species.

28

3. Conformational restriction and stereochemical deconvolution of flexible diamines in a bisporphyrin clip (Papers I and II)

When small flexible molecules in solution are analysed by NMR spectrosco-py, their fast conformational movements result in population-weighted time-averaged chemical shifts, coupling constants and NOEs. If such a molecule has multiple stereocentres these averaged values may not provide enough information to determine their relative stereochemistry. The aim of this work (Papers I and II) was to investigate whether conformational restriction im-posed on a flexible molecule could improve conformational analysis, and thus facilitate the deconvolution of its relative stereochemistry by NAMFIS.

3.1

Figure 20. Bisporphyrin glycoluril clip 3.1

The semi-rigid glycoluril bisporphyrin molecular clip 3.1 (Figure 20) (Paper I) was designed to restrict the conformational movements of small flexible diamine molecules via bidentate metal–ligand coordination of the two Zn centres to the diamine. The CH2 group indicated in Figure 20 (and the three other symmetry related CH2 groups) provides the clip sidewalls with flexibil-ity and allow the distance between the porphyrin units to vary.

29

The variation of potential energy vs. Zn–Zn distance (Figure 21) suggests that the clip should be able to accommodate a variety of guest sizes as the calculated variation in ∆E is small (≈ 50 kJ/mol) for d(Zn–Zn) = 6–14 Å.

Figure 21. Potential energy of closed conformers (aa) vs. Zn–Zn distance for clip 3.1 with the distance indicated below each structure. The lowest-energy open con-former (as) is also shown (Zn–Zn distance ≈ 22.8 Å).

3.1 Synthesis of diamines with multiple stereogenic centres

Alditols, acyclic polyols having several stereogenic centres and known ste-reochemistry, were converted to α,ω−diamino-polyolmethoxy ethers to be used as guest molecules (Figure 22). The secondary hydroxyl groups were capped by methylation to prevent hydrogen bonding at these centres and to improve the guest solubility in non-polar solvents, and the terminal amino groups were chosen to facilitate binding to the zinc metal centres in the porphyrin sidewalls.

3.2Sorbitol Target from Sorbitol

2S,3R,4R,5RXylitol Ribitol

2S,3R,4R,5R 2R,3r,4S 2R,3s,4S

3.4a3.3a3.2a

Figure 22. Parent alditol starting materials used in syntheses aiming for diamines with multiple stereogenic centres.

30

Figure 23 shows the synthetic route for the conversion of sorbitol 3.2a into diamine 3.2. Analogous syntheses converted xylitol 3.3a to 3.3 and ribitol 3.4a was converted to 3.4e. This route has some advantages when compared to previously published alternatives, 47 e.g., it uses cheaper starting materials.

3.2d3.2f

3.2a 3.2b 3.2c

3.2e3.2g

3.2

a b

c

de

f

g

hHCl..HCl

3.2h

Figure 23. Synthetic route to diamines with multiple stereogenic centres. (a) TrCl, Pyr, 100 oC, 69%. (b) MeI, DMF, NaOH, 77%. (c) p-TSA.H2O, MeOH/CH2Cl2, 77%. (d) Et3N, MsCl, CH2Cl2, 99%. (e) NaN3, DMF, 70 oC, 70% (f) Chromato-graphic materials (SiO2, basic Al2O3, neutral Al2O3, or Florisil), 80%. (g) LiAlH4, HCl, 0 oC, 89%. (h) Amberlite IRA-400, MeOH, quant.

During purification of the mesyl intermediates the derivatives from xylitol (3.3e) and sorbitol (3.2e) underwent demethylative cyclization (Figure 24). The impact of stereochemistry on this type of ring formation has previously been discussed for benzylated polyols,48 in that case, final ring products with trans-disposed substituents were preferred. In our work with mesylated polyols we have shown that the sorbitol derivative 3.2e only forms 3.2f and that the ribitol derivate 3.4e does not undergo cyclization, which confirms the preference for trans-disposed methoxy groups from mesylated polyols. More importantly, we found that this unwanted side reaction can be avoided by minimizing contact with chromatographic materials (SiO2, basic and neu-tral Al2O3, and Florisil).

31

a

b

3.2f

Nu:

3.2f'

3.3f'3.3f

3.4f 3.4f'

Nu

:

Figure 24. (a) Possible cyclic products in the demethylative ring closure. Only 3.2f and 3.3f/3.3f′ were formed. (b) Proposed mechanism of the demethylative ring clo-sure. Nu: = nucleophile.

The conversion of the diamine dihydrochlorides to free diamines via the traditional approach, that is, by adding bases such as NaOH or NaHCO3, proved troublesome. Due to the solubility of the protonated species in both the organic and aqueous phases, only mixtures of products were obtained. Instead, Amberlite IRA-400 resin provided a fast and convenient way to obtain clean diamines in quantitative yields. This worked equally well for very basic diamines such as bispidines (pKa ≈20).

32

3.2 Characterisation of host–guest complexes 1H NMR analysis of the complexes formed between clip 3.1 and diamine guests provided evidence for 1:1 binding. The number of signals detected was consistent with a single symmetrical diamine species bound inside the cleft, and only one set of signals was observed (Figure 25).

Figure 25. 1H NMR spectrum of the complex formed between clip 3.1 and guest 3.2, showing CIS (500 MHz, CDCl3 solution). Inset: 1H NMR spectrum of 3.2 in ab-sence of clip.

The CIS (complexation-induced shifts) caused by the porphyrin sidewall ring current effects separated and moved the guest signals to negative chem-ical shifts, thus simplifying analysis. In contrast, the monotopic binding of a diamine guest that occurs in the presence of a monoporphyrin results in loss of symmetry as the binding terminus is more influenced by the ring current effect than the pendant terminus (Figure 26).

Figure 26. Schematic representation of the binding of amine guests to porphyrin hosts and the CIS effect. (a) A free guest (─) experiences no CIS. (b) A guest bound to a monoporphyrin experiences CIS from binding site (─) but the effect diminishes along the chain. (c) The ditopic binding of a guest inside the cleft of a bisporphyrin molecular host causes CIS to occur from both binding sites with reduced effect to-wards the middle of the chain.

ppm-4.0-3.5-3.0-2.5-2.0-1.5-1.0-0.50.00.5

ppm3.5 3.0H-6a

H-1aH-2

H-3

C6-NH�H-6a

H-1bH-5

H-4 C1-NH�

1a1b

6b6a

12

34

56 NH2

NH2

H H

HH

OMe

OMeOMe

OMe

33

An illustrative example of the case in Figure 26b is the 1H NMR spectrum resulting from the binding of pentylamine to Zn(II)TPP, shown in Figure 27. In this case, the protons experience overall less ring current effect, and the signals are not separated enough to enable structural analysis.

Figure 27. 1H NMR spectrum of the monotopic binding of pentylamine to Zn(II)TPP (400 MHz, CDCl3 solution). The spectrum shows how the ring current effect has a greater influence on the chemical shifts of protons closer to the binding terminus (NH2).

No signals that would indicate loss of symmetry were observed for the diamine host–guest complexes with clip 3.1. If mixtures of mono- and ditopically bound diamines were present, several sets of signals should be observed, assuming slow exchange. In the case of fast exchange due to weak binding population-averaged signals for the different species would be ob-served.

The binding constants for the host–guest complexes formed between diamine guests and clip 3.1 were determined by UV/Vis spectroscopy and were in the range of 104–106 M-1. Binding constants for monoamines and aminoalcohols in 3.1 were lower (103 M-1) and for diols, no binding was detected. A red-shift of all bands in the UV/Vis spectra and clear isosbestic points indicated that only one complex was formed for each guest. Further-more, diffusion studies were performed to investigate whether the clip shape changed with guest size.49,50 Complexes with smaller guests (diamines con-nected with up to six carbons) had the same diffusion coefficient as the free clip (D = 7.010-10 m2s-1), whereas larger guests (connected by twelve or more carbons) diffused more slowly than the free clip (D = 4.710-10 m2s-1). The changes in D indicate an alteration of the clip shape, most likely, the guest pushes the porphyrin sidewalls apart.

34

3.3 Assignment deconvolution via J-based arguments and NAMFIS

This section describes an assignment analysis leading to the assignment of peaks in the 1H NMR spectrum of a sorbitol-derived diamine. The deconvolution of its relative stereochemistry is discussed in section 3.4. A corresponding analysis of a xylitol-derived diamine has also been conducted (Paper II).

The first step in a conformational analysis is to assign all 1H NMR signals to the relevant protons in the molecule. In the 1H NMR spectrum of 3.2 it was possible to assign the sequential connectivity of protons by J values and 2D NMR analysis, but not to assign the relative stereochemistry, or to distin-guish the diastereotopic CH2 protons or H–1 from H–6. Using J-based argu-ments in combination with NAMFIS analysis enabled the complete assign-ment. Once the sequential connectivity of 3.2 is known, there are four ways that the 1H NMR signals can be assigned (A–D in Figure 28). Figure 29 shows the Newman projections of 3.2 (leftmost column) and its diastereoisomers along the C–C single bonds, assuming that the all-trans conformation is dominant. The largest R groups are anti-disposed. Following J-based assignment arguments, the proton pair with the largest dihedral an-gle was assigned to the 1H NMR signal with the largest coupling constant, suggesting assignment A to be most likely. If the stereochemistry of the in-vestigated molecules had not been known, merely a J-based analysis would not have been sufficient. As demonstrated by the Newman projections in Figure 29, there is insufficient information to distinguish between all stereo-isomers.

BA

C D

Figure 28. Alternative directional 1H NMR assignments with respect to the C1–C6 connectivity and the diastereotopic protons Ha/Hb of the diamine 3.2.

35

>J1a-2

J1b-2

> J6a-5

J6b-5

>J1a-2

J1b-2

<J6b-5

J6a-5

<J4-5

J2-3 ≈J

4-5J

2-3

> J6a-5

J6b-5 > J

6a-5J

6b-5 <J

6b-5J

6a-5

>J1a-2

J1b-2

J4-5

J2-3

> ≈J4-5

J2-3 ≈J

4-5J

2-3

J1a-2

J1b-2

> J6a-5

J6b-5

J4-5

J2-3

<

C2-C3

2R,3S,4R,5S2R,3R,4S,5S

<J1a-2

J1b-2

Congener of:

C4-C5

C1-C2

C5-C6

GalactitolSorbitol MannitolAllitol Talitol Iditol

<J1a-2

J1b-2

2S,3R,4R,5R 2R,3R,4R,5R 2R,3R,4S,5R 2R,3S,4S,5R

Figure 29. Newman projections along the C–C single bonds in the proposed main rotamers of 3.2 (box) and its diastereomers, identified by their parent alditols. The proton pairs generating the largest J when comparing JH2-H3 vs JH4-H5, JH1a-H2 vs JH1b-

H2 or JH6a-H5 vs JH6b-H5 are marked in red.

NAMFIS supported the J-based analysis above. The experimental J and NOE values of free 3.2 were matched to conformers found by computations with different assignment alternatives (A–D). As expected, matching the experimental data for the free diamine 3.2 against a large population of con-formers did not result in a clearly favored assignment, although assignment A had the highest probability (lowest RMS value, Table 2). NAMFIS sug-gested free 3.2 to exist in several conformations which likely contribute to the poor distinction between A–D. In contrast, the bound diamine 3.2 was suggested to exist mainly in two conformations; 25% of the molecules were all-trans and 75% in a conformation with one gauche kink. When experi-mental data from the bound diamine was matched against a single all-trans conformer, a larger differentiation of RMS values, with preference for as-signment A, was obtained (Table 2).

36

Table 2. Assignment analysis of diamine 3.2 by NAMFIS.

Assignment

RMS, free diaminea

RMS, bound diamine all-trans conformerb

A 0.31 0.43B 0.42 0.71C 0.47 0.59D 0.44 0.66aPopulation from unrestricted conformational search reduced by redundant conformation elimination. NOEs and J values were used in the analysis. bOnly NOEs were used in the analysis.

3.4 Analysis of relative stereochemistry As the binding of a guest to a molecular clip significantly improved assign-ment analysis via NAMFIS, the distinction of stereoisomers by NAMFIS presented an interesting challenge.

The experimental data for free 3.2 was matched against populations from computations of the congeners of the sorbitol diastereomers (excluding enan-tiomers). As expected for the free molecule no clear distinction could be made, although also here the sorbitol congener displayed the best fit, but with an unconvincing ΔRMS (0.07–0.09) over the other congeners (Table 3). When experimental data for the bound diamine was matched against single all-trans conformations of its diastereomers, the distinction between stereoi-somers was clearly improved. The sorbitol congener exhibited the best fit with the lowest RMS value here as well, and furthermore the ΔRMS values were improved to 0.13–0.32.

Table 3. Conformational deconvolution of free and bound 3.2 and its diastereomers.

Isomer derived froma Stereochemistryb RMS, free diaminec

RMS, bound diamined

Sorbitol 2S,3R,4R,5R 0.31 0.37

Allitol 2R,3R,4S,5S 0.38 0.49Galactitol 2R,3S,4R,5S 0.38 0.50Mannitol 2R,3R,4R,5R 0.39 0.69Talitol 2R,3R,4S,5R 0.40 0.57Iditol 2R,3S,4S,5R 0.66 0.49 aStereochemistry indicated by the corresponding alditol. bAssignment in the order used for 3.2. cPopulation from an unrestricted conformational search. dSingle all-trans conformers.

37

3.5 Conclusions Within the scope of Papers I and II we have constructed a semi-rigid bisporphyrin molecular clip that complexes to diamine guests with binding constants in the range of Ka = 104–106 M-1. UV/Vis, diffusion and NMR studies provided evidence for ditopic binding of the diamines inside the cleft. The aim of this work was to use a molecular clip to rigidify small flex-ible molecules with several stereogenic centres to facilitate their NMR anal-ysis. By combining J-based arguments and a NAMFIS analysis, we have shown that assignment deconvolution can be significantly improved and stereochemical analysis made possible by binding a diamine to clip 3.1.

38

4. Photoswitchable bisporphyrin tweezers (Paper III)

In Papers I and II we showed that the use of a bisporphyrin molecular tool can improve the structural analysis of flexible diamine guests. One feature of these systems that was of interest for further exploration was the incorpora-tion of functional units to connect the two porphyrin rings. In Paper III, we describe preliminary results regarding molecular tweezers with photoswitchable backbones that undergo Z/E isomerisation upon irradiation. The aim of this work was to use such molecular switches to create dynamic “catch and release” or “catch and stretch” effects (Figure 30). This addition-al functionality would increase the versatility of such a molecular tweezer. For example, one component of a mixture could be selectively bound, then purified or transported before being released and recovered. Also, conforma-tional analysis of guests of various sizes, especially long guests, could all be made possible with a single tweezer.

Figure 30. Schematic representation of the “catch and stretch” and “catch and re-lease” functions upon photoisomerisation.

39

4.1 Synthesis of photoswitchable bisporphyrin tweezers

A rigid clip appears to be more suitable than a flexible tweezer for the con-formational analysis of a substrate, as the dynamic behaviour of the latter can complicate the investigation. One disadvantage of rigid molecular clips is that their syntheses often are more cumbersome than those involving a flexible backbone. A flexible molecular tweezer can be sufficient for many purposes, and when the host is not restricted to a specific cleft volume, a larger range of guest sizes can be accommodated. Two photoswitchable tweezers were prepared for this investigation: 4.1, with an enediyne spacer and 4.2, with a stiff stilbene spacer (Figure 31).

4.1 4.2

Figure 31. Bisporphyrin tweezers with photoswitchable spacers for cis/trans isomerization.

The porphyrin units were functionalised starting from tetraphenylporphyrin (TPP), as shown in Figure 32. The aminoporphyrin 4.6 was then coupled to the enediyne or stiff stilbene backbone via an amide bond, as shown in Fig-ure 33. A detailed description of the syntheses of backbones and porphyrins can be found in Paper III. The coupling of the porphyrin units to the back-bones was the most difficult aspect of the tweezer synthesis. The amine functionality of the aminoporphyrin was likely deactivated by the large aro-matic system of the porphyrin ring, and also sterically hindered by the phe-nyl groups, and thus behaved as a weaker nucleophile than expected. The peptide coupling reagents DCC, HOBt and HATU, as well as changes in solvent and temperature, were tested to improve the yields of this coupling, however, the maximum yields remained in the range of 20–30 %.

40

e

4.7

4.3 4.4

4.54.6

TPP

a b

c

d

Figure 32. Functionalisation of TPP. (a) Cu(OAc)2·H2O, CH2Cl2, MeOH, reflux, 2.5

h, quant. (b) Cu(NO3)2.3H2O, acetic acid, acetic anhydride, CHCl3, 35 oC, 5 h, 80%.

(c) H2SO4, CH2Cl2, 20 min r.t., 95%. (d) SnCl2.2H2O, HCl, CHCl3, N2 atm, dark, r.t.,

3–4 days. (e) Zn(OAc)2.H2O, CH2Cl2, MeOH, reflux, 2 h, quant.

R 1 = OH R 2 = Cl a

R 1: b or c, dR 2: c, d

Figure 33. General procedures for coupling of aminoporphyrin to backbones. (a) Oxalyl chloride, DMF, THF, 0 oC, 1h. (b) TPPNH2 4.6, DCC, THF, 0 oC, o.n. (c) TPPNH2 4.6, CH2Cl2, r.t., o.n. (d) Zn(OAc)2

.H2O, CH2Cl2, MeOH, reflux, 10 min.

41

4.2 Binding studies with aliphatic diamines Both tweezers 4.1 and 4.2 bound to aliphatic diamine guests ditopically, as was shown by the single set of guest signals in 1H NMR spectra. Clear isosbestic points in the UV/Vis spectra of guest:tweezer mixtures also sug-gested the formation of single complexes. Ka was measured for 1,6-diaminohexane and both Z tweezers to be 104 M-1. Scalar couplings could not be resolved for the guest 1H NMR signals, indicating that dynamic pro-cesses were occurring. As the amide link connecting the porphyrin and backbone permits flexibility in the sidewall orientations, these findings were not surprising. Regardless, the binding was strong enough for conforma-tional analysis, as the signals of the diamine guests were separated and shift-ed to non-crowded regions of the NMR spectra. Variable-temperature NMR experiments further supported the existence of a dynamic system. When the host–guest complexes were analysed by NMR spectroscopy at low temperature (down to –55 oC), several sets of signals appeared, indicating that several conformers were present and did not inter-convert on the NMR time scale. This effect was more pronounced for shorter guests such as 1,6-diaminohexane (Figure 34) than for longer guests such as 1,12-diaminododecane. Apparently, the longer guests are able to adopt a larger number of conformations than the shorter guests.

Figure 34. 1H NMR spectra of a ≈1:1 complex of enediyne tweezer 4.1 and 1,6-diaminohexane (500 MHz, CDCl3 solution) at: (a) 25 oC; (b) 0 oC; (c) –55 oC.

42

There are several possible explanations for the dynamics observed in these host–guest systems. Porphyrins are large aromatic systems that prefer to π-stack face-to-face51 meaning that the tweezers 4.1 and 4.2 are likely not symmetrical, but slightly skewed (Figure 35). At low temperature, the dy-namics of the tweezer are slowed and guests can be observed in different tweezer conformations. From another point of view, conformational move-ments of the guest itself could induce conformational movements in the tweezers.

Figure 35. π−π Stacking dynamics in flexible bisporphyrin tweezers.

In comparison, clip 3.1 (Papers I and II) had a more rigid structure that did not permit intramolecular π–π stacking of the porphyrin units; hence, there were less conformational dynamics in that system. The 1H NMR spectra of a ≈1:1 complex of 1,8-diaminooctane and clip 3.1, when recorded at –55 oC show only one set of signals, as opposed to the spectrum of 1,8-diaminooctane and tweezer 4.1 at the same temperature (Figure 36).

Figure 36. 1H NMR spectrum (500 MHz, CDCl3 solution, –55 oC) of ≈1:1 complex-es (a)1,8-diaminooctane:clip 3.1 (semi-rigid glycoluril spacer) and (b) 1,8-diaminooctane:tweezer 4.1 (enediyne backbone).

43

4.3 Photoisomerisation Preliminary photoisomerisation studies of the enediyne and stiff stilbene backbones and the corresponding bisporphyrin tweezers have been conduct-ed. Ideally a photochemical switch should have a fast response time and interconvert thermally stable molecules whose UV/Vis spectra display well-separated absorptions (∆λmax). The enediyne backbone (as diester) reached a photostationary state at ≈55% Z→E conversion and under similar conditions, the enediyne tweezer 4.1 reached a photostationary state at 10% Z→E conversion. This is in keeping with porphyrin-modified switches being known to reach lower conversion compared to their unmodified switch analogues.52 The stiff stilbene back-bone (as diester) was irradiated semi-selectively at a wavelength with Z and E ∆λmax overlap. Nevertheless, it reached a photostationary state at ≈55% E→Z conversion. The stiff stilbene backbone is expected to reach higher conversion in further studies at irradiation with less ∆λmax overlap. Im-portantly, no degradation was observed during selective irradiation, whereas nonselective irradiation caused the decomposition of all compounds.

4.4 Conclusions Two bisporphyrin molecular tweezers with photoswitchable backbones have been synthesized. 1H NMR and UV/Vis studies of the Z tweezer isomers supported bidentate binding of diamine guest of various sizes. Variable-temperature NMR experiments revealed dynamic systems that exhibit con-formational movement. The preliminary photoisomerisation results showed that the enediyne and stiff stilbene backbones, and tweezer 4.1 could be isomerised without decomposition. To finalise this work, further isomerisa-tion studies of the tweezers will be conducted. The “catch and stretch” and “catch and release” functionalities also deserve further investigation.

44

5. Regulation of bispidine conformations and the observation of a boat-boat conformer (Paper IV)

Bispidines usually favour the chair-chair (CC) conformation, which upon monoprotonation is locked into a rigid bidentate CC complex. Further proto-nation produces a diprotonated boat-chair (BC) conformer.18a Bispidines can therefore be used as pH-dependent molecular switches that may be incorpo-rated into larger structures, such as in peptidomimetics.53

Most structural investigations of bispidines are focused on 1H and 13C chem-ical shift patterns.22 We here saw the opportunity to use RDC NMR, which for small molecules is a relatively new technique, and which does not rely on empirical correlations between NMR parameters and molecular structure. The aim of this work (Paper IV) was to further investigate the modulation of conformational changes in bispidines and to characterise the different con-formations by RDC NMR. We chose 5.1 as our model compound because its free base and monoprotonated form have previously been characterised by X-ray crystallography in our group.18 The species discussed in this chapter are shown in Figure 37. During titrations with acid, it became apparent that an additional, previously unobserved species, the boat-boat (BB) conformer, was formed in solution.

[5.1.2H]2+ BC [5.1.2H]2+ BB5.1 CC [5.1.H]+ CC

endo exo

exo

endo

Figure 37. The bispidine species discussed in this chapter. CC: chair-chair; BC: boat-chair; BB: boat-boat. [5.1.H]+ and [5.1.2H]2+

denote the mono- and diprotonated species, respectively. R = CHPh2. For simplicity the boat ring is shown with C2v rather than D (twist) geometry.

45

5.1 Identifying conformations by NOEs and RDCs Mono- and diprotonated analogues of 5.1 were prepared by adding CH3SO3H to a solution of 5.1 in CDCl3. The monoprotonated [5.1.H]+ was obtained with a 5.1:acid ratio of 1:1 and the diprotonated [5.1.2H]2+ was obtained with a 5.1:acid ratio of ≈1:3. Monitoring the acid addition by 1H NMR spectroscopy revealed the formation of an additional species that had fewer NMR signals, suggesting that it was either a BB or a CC conformation (Figure 38). Surprisingly, the NOEs detected matched only a BB confor-mation (Figure 39).

Figure 38. 1H NMR spectrum of a mixture of [5.1.2H]2+ BC ( boat side, chair side) and [5.1.2H]2+ BB () protonated by an excess of MeSO3H (500 MHz, CDCl3 solu-tion, 25 °C)

Figure 39. Selected NOEs found for the bispidine conformations of mono- or diprotonated 5.1.

The amount of BB conformer was influenced by the rate of acid addition. Slow, stepwise addition generated approximately 15–20% BB conformer, whereas the rapid addition of a large excess of acid generated up to 74% BB conformer (Figure 40).

NN

PhPh

H H

Ph PhH

H

H

HH

H

HN

H H

H

H

N

HPhPh

HH

PhPhH

BC-endo,exo BB-exo,exoCC-exo,exo

N

HN

H

Ph

HPh

Ph Ph

H

H H

H

H

H

46

Preliminary studies were performed to investigate how to control the amount of BB formed, e.g. by varying the acid used, solvent, temperature during addition and concentrations, without further improvements. Attempts to iso-late the BB conformer from a mixture with BC have not yet been successful.

Figure 40. 1H NMR (500 MHz, CDCl3 solution, r.t.) expansion of the Ph2CH pro-tons signals from [5.1.2H]2+ BC and BB mixtures (BB denoted by *). a) Following stepwise addition of excess Me3SO3H upon which 18% BB conformer was formed; b) following rapid addition of excess Me3SO3H, upon which 74% BB conformer was formed.

RDCs were determined by extracting the difference of the C–H coupling constants in clip-HSQC54 spectra that were obtained when comparing iso-tropic (1JCH) and anisotropic (1JCH + 1DCH) samples, using PBLG37 as the alignment medium. The experimentally derived RDCs were matched to val-ues back-calculated from Monte Carlo conformational searches. In all cases, the best-fit structures matched the conformation and R-group positions sug-gested by X-ray crystallography and NOEs.

5.2 Protonation equilibria As both NOEs and RDCs support the observation of a BB conformation, a possible explanation for its formation and the dependence of its prevalence on the rate of acid addition may be found in the protonation equilibria lead-ing up to diprotonation as proposed in Figure 41. At all times separate sets of 1H NMR signals were observed for the various species indicating that exchange between 5.1 and [5.1.H]+, as well as between [5.1.H]+ and [5.1.2H]2+ was slow.

47

To reach a diprotonated BC conformation a nitrogen inversion must follow the first protonation step. If a large excess of acid is added rapidly, the se-cond protonation is likely to occur before nitrogen inversion can occur to a large extent. Also, conversion of the BB to BC conformer would require a deprotonation followed by an inversion, then reprotonation. The lifetime of the monoprotonated species in the presence of a large excess of acid is likely too short to allow this sequence of events to take place.

N inversion

[5.1.H]+ CC exo,exo

[5.1.H]+ CC exo,exo

5.1 CC exo,exo

[5.1.2H]2+ CC exo,exo[5.1.H]+ CC endo,exo

[5.1.2H]2+ BB exo,exo

[5.1.2H]2+ BC endo,exo

+ H+

- H+

+H+-H+

+H+-H+

NH

N

R

H

R

NR

NRH

H

NR NR

H

NR NR

NH

NRR

HNR N

RHR

N NRH

+

+

+

+

+

+

++

+

Figure 41. Proposed protonation equilibria of bispidine 5.1 (R = CHPh2).

5.3 Conclusions In Paper IV, the first reported bispidine boat-boat conformer is described, which was confirmed independently by NOE and RDC NMR. The amount of BB isomer was affected by the rate of acid addition, and could reach over 70%. Further we applied RDC NMR to a small symmetrical molecule in different protonation states. Protonated bispidines can function as nucleators in peptidomimetics, and the finding of the BB conformer and more im-portantly, the possibility of controlling its abundance, is expected to be im-portant for the use of bispidines in this field.

48

6. An unexpected reaction: Formation of persistent dication diradical cyclophanes (Paper V)

The origin of the work described in this chapter was an occasional observa-tion of a green discolouration during the preparation of π-allyl palladium complexes with the N,N´-diphenylbispidine ligand in the presence of silver salts.55 During initial investigations of the green byproduct, useful experi-mental data was hard to come by. No consistent conclusions could be drawn from its mass spectrum or elemental analysis. The 1H NMR spectrum showed significant signal broadening, suggesting the presence of unpaired electrons or perhaps polymerisation, two situations in which an organic mol-ecule essentially becomes “invisible” to NMR spectroscopy. Eventually, after a few crystals of the product were finally isolated, a crystal structure was obtained unveiling the green products identity: A dication diradical with a cyclophane motif (Figure 42).

•

•

2 BF4-

Figure 42. The persistent dication diradical 6.5 found by X-ray crystallography following the reaction of N,N´-diphenyl-3,7-diazacyclooctane 6.1 and AgBF4.

In Paper V, the formation of a series of persistent organic radicals from vari-ations of the N,N´-diphenyl-3,7-diazacyclooctane (6.1) backbone is dis-cussed. The aim of this work was to clarify the preconditions for this reac-tion, to suggest a probable reaction mechanism, and to determine some of the magnetic properties of these new compounds.

49

6.1 Persistent organic radicals Organic radicals are molecules with unpaired electrons. Radicals are gener-ally very reactive and thus short lived. However, a radical can become stabi-lised, consequently prolonging its lifetime and even becoming persistent if structural features such as conjugation, electron-donating heteroatoms or steric effects are present.56 One of the earliest examples is the benzidine (4,4’-diaminobiphenyl) radical.57 Persistent organic radicals have attracted interest within many fields, such as biochemistry,58 chemical synthesis59 and materials chemistry.60 When two unpaired electrons are present in the same molecule and spin pairing can be prevented, the molecule is said to be a diradical.61

6.2 Formation of cyclophane diradicals The formation of the dication diradical cyclophanes was straightforward. Solutions of ligands and AgBF4 were mixed, upon which the solutions im-mediately changed from almost colourless to intense green. At the same time, the formation of Ag(s) was observed. Four ligands were tested under the same reaction conditions and three dication diradical species were ob-tained (Figure 43). The reaction products 6.5 and 6.6 could be isolated in high yields whereas for 6.7, a product mixture was obtained.

6.1 6.2 6.3 6.4

• 2BF4-

6.5 X = 2H, R = H, 87%6.6 X = CH

2, R = H, 73%

6.7 X = CO, R = CO2Me, impure

•

•

Figure 43. Dication diradicals species (6.5–6.7) formed from the reactions of ligand 6.1–6.4 with AgBF4 in acetone at r.t.

50

We propose that the reaction of a diamine ligand with AgBF4 involves the initial chelation of the bidentate ligand to Ag+, followed by a redox reaction generating Ag0 and a radical cation. Dimerisation of the radical diamine ligand via C–C coupling, followed by further redox reactions with Ag+ ions ultimately forms dication diradical cyclophanes (Figure 44). Job plots of the system indicated a ligand:Ag+ ratio of 1:3, with a total six Ag+ being con-sumed in the formation of one dication diradical cyclophane. The mecha-nism proposed here is related to the dimerisation of arylamines to benzidine by copper or iron oxidants.62 There are also examples in the literature of phenols that via para-coupling form 4,4´-dihydroxybiphenyls via radical intermediates.63 The relative rate of cyclophane formation was 6.1 > 6.2 > 6.3. Product stabil-ity followed the same pattern; 6.5 was stable in solution for several months whereas 6.7 deteriorated after a few weeks. As the initial reaction step in-volves the formation of a chelate complex between Ag+ ions and diamine ligands (Figure 44), pre-organisation and the electronic properties of the amine nitrogens could impact the reaction. The degree of pre-organisation is higher in 6.2 and 6.3 than in 6.1, thus the flexibility of 6.1 appears to offer and advantage. This may have meant that the initial cyclic intermediate suffers less strain, as the phenyl groups could be mutually parallel and planar sp2 geometry could be achieved for all nitro-gens. Compounds 6.2 and 6.3 exhibit similar pre-organisation, but bispidinones are weaker electron pair donors than bispidines. Thus, for 6.2, pKa = 13.97, and for 6.3, pKa = 8.13,18 meaning that 6.2 is likely to interact more strongly with Ag+, and thus undergo the reaction more readily. The lack of reactivity of 6.4 towards Ag+ is unlikely to originate in the initial chelation step, as 6.4 would form a five-membered chelate ring and therefore should chelate more strongly than 6.1–6.3 which form six-membered rings.64

Rather, an initial radical 6.4a, analogous to 6.1a (Figure 44), would be una-ble to react further to form stable cyclisation products for steric reasons. However, pending further investigations, this explanation remains specula-tive.

51

•

1

•

1a

1b 1c.2Ag+

8.2Ag+

81e

1c

1d

2 x

- Ag0

•

dimerizationAg+

5

- 2 Ag0

2 Ag+

- 2H+

dimerization

- 2 Ag0

2 Ag+

- 2H+•

•

•

•

N N

Ag

NN NN

NN

N N

N N

NN

Ag

Ag

N N

NN

N N

NN

Ag

Ag

N N

NN

N N

NN

H

H

N N

NN

H

H

N

N N

N N

N N

N

++

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Figure 44. Proposed reaction mechanism for the formation of dication diradicals. BF4

- counterions omitted for clarity.

52

6.3 Properties of cyclophane diradicals The reaction mixtures were investigated by diffusion NMR spectroscopy which supported the formation of dimers. For the samples containing radical species the 19F signal of AgBF4 was used for measurements, as it was not affected by line broadening to any larger extent. The measured diffusion coefficients of the reaction products correlated well with a hydrodynamic radius of a formed dimer co-diffusing with two BF4

- counter ions. The addi-tion of piperazine 6.4 to a solution of AgBF4 did not affect the rate of diffu-sion and essentially the same diffusion constant as for free BF4

- was meas-ured. The cyclophane 6.1 has, in contrast to reported persistent radicals with phe-nyl rings, intramolecularly π-stacked aromatic rings that lack lateral disloca-tion (Figure 45). As a consequence, the two unpaired electrons in dimer 6.5 showed intramolecular spin pairing, as was indicated by weak EPR signals and magnetic moment. Therefore, these should be interesting model com-pounds for further studies of electronic interactions in π-stacked aromatic systems.

Figure 45. Solid state intra- and intermolecular π-stacking of dication diradical 6.5. Left: Crystal packing viewed down crystallographic a axis, with colour coding of symmetry-equivalent units of pairwise π-stacked cyclophanes. Right: Intermolecular π-stacking within a pair of symmetry equivalent molecules, viewed from three dif-ferent directions.

The diradical 6.5 could be chemically reduced to the neutral 6.8 by sodium bisulfite or hydroquinone. The intense green colour of 6.5 disappeared in-stantly upon adding a solution of either reductive species and, when the reac-tion was followed by UV/Vis spectroscopy the characteristic absorption maximum of 6.5 vanished (Figure 46). Neutral species 6.8 could be isolated and analysed by NMR spectroscopy, and its 1H NMR spectrum was con-sistent with the proposed structure.

53

Na2SO

3

acetone/waterr.t., quant.

6.5

•

•

6.8

Figure 46. UV/Vis spectra showing the dication diradical 6.5 (…) being reduced to cyclophane 6.8 (─) upon the addition of sodium bisulfite or hydroquinone solutions. Absorption maxima of compound 6.5 at 403 and 760 nm disappear, leaving a col-ourless solution of the neutral cyclophane 6.8.

6.4 Conclusions Within the scope of Paper V, we have concluded that persistent organic radi-cals form by reaction of Ag+ with compounds having the N,N’-diphenyl-1,5-diazacyclooctane backbone. These diradical products are persistent and can be isolated in high yields. They exhibit magnetic properties but, due to intramolecular spin-paring and delocalisation show weak EPR signals, con-sistent with reported persistent organic radicals. The diradicals exhibit inter-esting features such as intramolecular π-stacking without lateral displace-ment and intramolecular spin-pairing between the two radicals. Therefore, these new compounds should be interesting for further studies of e.g. elec-tronic interactions in π-stacked systems. Studies regarding analogs with ex-tended oligophenyl units, optimisation of the reduction to diamagnetic spe-cies and their properties as hosts for other species are pending.

54

7. Concluding remarks

This thesis concerns the construction, use or modulation of various host–guest systems, from small bispidines for binding of inorganic ions to bisporphyrin clips for supramolecular systems.

We have shown that a semi-rigid bisporphyrin clip can be used to restrict the conformational flexibility of acyclic diamines (Papers I and II). This has been used to develop a strategy for stereochemical analysis of small com-pounds with several stereogenic centers, based on conformational analysis combined with NAMFIS. A new synthetic strategy towards diamines de-rived from alditols has been developed and we have devised a simple meth-od for quantitative conversion of protonated diamines into their free bases (paper II). This strategy might be extended to guests other than diamines by, e. g., replacing the zinc ions in the bisporphyrin by other, such as oxofilic, metals.

Two bisporphyrin tweezers with a photoswitchable backbone have been synthesised (paper III). Preliminary studies have proven their stability during photoswitching as well as their binding to diamines. These systems will be used for investigations into the dynamics of guest binding and conforma-tional modulations. The pH-dependence of bispidine conformational changes has been investi-gated and a previously not observed boat-boat conformer detected (paper IV). We have shown that the amount of boat-boat conformer can be influ-enced by the rate of acid addition. The finding and steering of the boat-boat conformer may be employed for using bispidines as nucleators in larger structures, such as peptidomimetics. Complexation of compounds with the N,N´-diphenyl-1,5-diazacyclooctane backbone to Ag+ ions lead to the formation of persistent dication diradicals with a cyclophane motif. The radicals were stable over several months and could be isolated in high yields. As these compounds exhibit magnetic prop-erties they may be of interest in material chemistry. This study will be broadened by the investigation of extended ligands such as oligophenyls and optimisation of the neutralisation process as well as the use of these cyclophanes as hosts for other species.

55

8. Summary in Swedish

Modulering av molekylära egenskaper Forskningsarbetet som redovisas i avhandlingen rör molekylers struktur, särskilt deras form (konformation), i lösning. Kunskap om molekylers kon-formation bidrar till förståelsen av deras kemiska och biologiska egenskaper t.ex. vid arbete med läkemedelsdesign. En molekyls struktur kan beskrivas genom det tredimentionella nätverk som utgörs av dess atomer och bind-ningarna mellan dessa. För att bestämma molekylstruktur i fast (kristallin) fas är röntgenkristallografi sedan lång tid tillbaka en excellent metod. I kristallen ligger atomerna i stort sätt ”fastfrusna” och molekylen föreligger således i en bestämd konformation. I lösning (och i gasfas) förekommer ofta flera konformationer samtidigt och de står i jämvikt med varandra enligt den statistiska termodynamikens lagar. Detta beror på att rotationsrörelser kring vissa bindningar äger rum mer eller mindre snabbt. Hos en liten rörlig mole-kyl i lösning sker många konformationsändringar alltför snabbt för att kunna detekteras med vanliga analysmetoder. Då observeras endast statistiskt be-stämda medelvärden.

Vidare är forskningen som presenteras i avhandlingen en del av den sup-ramolekylära kemin som introducerades 1969 av Jean-Marie Lehn och be-skrivs som att ”se bortom den enskilda molekylen”, och strävar efter att för-stå molekylär samverkan och interaktion. Inom denna del av kemin talar man ofta om så kallade värd–gäst system, där en, oftast större, värdmolekyl binder en mindre gästmolekyl. Inom en molekyl är bindningarna så kallade kovalenta och relativt starka. Mellan molekyler är de så kallade icke-kovalenta och relativt svaga. När många icke-kovalenta bindingar samverkar kan den totala effekten ändå bli stor.

Ett av de övergripande målen för avhandlingen var att utarbeta strategier för att analysera strukturen av små rörliga molekyler i lösning med hjälp av NMR-spektroskopi. Genom att konstruera en pincettformad värdmolekyl, som genom icke-kovalenta bindningar kunde binda en liten gästmolekyl, kunde flexibiliteten hos de små molekylerna minskas i sådan utsträckning att strukturanalyser kunde genomföras (Figur 1).

56

Figur 1. Till vänster: en schematisk bild av hur en pincett binder en rörlig molekyl. Utanför pincetten är rörligheten hög och molekylerna förekommer i olika konforma-tioner. Till höger visas en konstruerad bisporphyrinpincett som har bundit in en diamin som därmed är i utsträckt konformation och kan analyseras med NMR-spektroskopi.

Om en så kallad switch integreras i pincettens struktur kan den med hjälp av belysning växla mellan öppet och stängt läge. Detta möjliggör till exempel att en liten molekyl kan fångas i pincetten och efter belysning släppas ut (Figur 2).

Figur 2. En pincett med ljuskänslig switch genom belysning variera mellan stängt och öppet läge. En gästmolekyl (röd i figuren) binds till den stängda pincetten men släpps fri när pincetten öppnas.

Den senare delen av denna avhandling berör mindre molekyler som genom icke-kovalenta interaktioner med en enda atom, t.ex. en proton eller metall-jon, ändrar sin konformation. Att kontrollera konformationsändringar är av intresse speciellt då de leder till att en reaktion kan ske eller om ändringen påverkar en större struktur.

Resultaten som presenteras i avhandlingen inspirerar till fortsatt forskning och tillämpning rörande strukturbestämning av flexibla molekyler i lösning med hjälp av molekylära verktyg, samt till användande av kontrollerade konformationsändringar av små molekyler som nukleatorer i större struktu-rer.

57

9. Acknowledgements

I would like to express my gratitude towards all the people that have made my time as a PhD student possible, enjoyable and in many ways unforgetta-ble. First of all, my supervisor Adolf Gogoll – I am absolutely sure I would not have started (or finished) this journey without you. I could not have found a better supervisor and I thank you for your guidance and friendship over the past years. You have encouraged me to stay curious and creative even in my moments of doubt which have been invaluable to me. I will truly miss our daily conversations and I promise you, not everything you said went in one ear and out the other… My co-supervisor Helena Grennberg – you give good advice and bring clari-ty to any situation, and that I have appreciated a lot. Thank you for being a good teacher of chemistry and strong co-supervisor. Máté Erdélyi – I consider you my unofficial co-supervisor and I am tremen-dously grateful for all your help and patience. Your advice on various NMR techniques as well as comments on research life in general has been equally stress-reducing. Also thank you for collaboration on the RDC project. All the past and present members of the GG Group that have made our lab to the best lab. Miranda – I have missed you since you left, never got to talk Idol outfits with anyone else! I am glad we have kept in touch. Classe – the most laid-back teaching partner and lab neighbour ever, thanks for keeping it cool. Magnus – I will miss your shenanigans in the lab and our laughs. PS. maybe you can have my CDCl3 now. Hao – you bring positive energy to the lab every day, keep it up! Anna, Micke, Matt, Xiao – for many jokes, funny notes, cakes and nice conversations. Henrik – for being the best coffee butler and partner in sugars, thank you very much for endeavours in sticky synthe-sis, dangling bonds and sharing your insights about life, even though I may not agree about that many. Your support meant a lot to me. Prasad & Wenzhi for valuable porphyrin knowledge and work on the clip project.

58

Göran Bergsson – for encouragement and wise words along this path. Olle Matsson – it was fun taking part in your book! I have enjoyed our trips and photo missions, as well as our conversations. Thomas Norberg – for your help in the sugar chemistry jungle, photography talks and always asking relevant questions during presentations. Tamara L. Church for linguistic corrections in this thesis, du är en klippa. All the nice people at the department for making everyday life better. You are so many but I would especially like to thank my girls in ‘Fikaklubben’ for all the moments inside and outside BMC. Sofia – for spreading so much positive energy, thank you for being there, bringing cakes and for good taste in tv-shows. Johanna – my partner in all kinds of crime. I can not imagine how much less fun this time would have been without you. Angelica – you’re already a legend with a laugh that can move mountains, I love talking to you about everything. Ann, Diana, Cissi, Åsa, Emil, Micke, Nisse, Rikard, Christian D, Johan V, Johan W, Ola, Christian S, Erika, Peter, Alex, Lotta & Päivi for coffees, lunches, courses, after works and everything in between. Anna – kakboden, you’re a lifesaver. Gunnar & Thomas – for friendly tech-nical support often with a humorous twist. I would also like to thank my non-chemistry related friends that have sup-ported me throughout the years even though they may not have seen me very often lately. Karin, Johanna, Hugo & Anton – my bonus family in Uppsala, for bringing many non-BMC related events to my life and for being the best of friends. Karin & Therese – my photography crew in Uppsala, thank you for cheering for me, photo escapades and unforgettable chat moments. The whole old gang from Gotland – you know who you are and I look forward to spending more new year’s eves, beach days and Munken-nights with you! Tilde – bäst på ön! Vi ses mer snart! Mamma & Pappa – för att ni är ni och alltid tar hand om mig. Finally I would like to thank Danne (och vår pälsmissil) – for looking out for me in every possible way, and saving me from having popcorn dinners every night. I wouldn’t want it any other way – you’re the best!

59

10. References

1. G. Quinkert, E. Egert, C. Griesinger, 1996, Aspects of Organic Chemistry:

Structure, Verlag Helvetica Chimica Acta, Basel, Switzerland, (Eds. M. Volkan Kisakürek, B. Gröne) p. 1-8.

2. J. M. Lehn, Pure & Appl. Chem. 1978, 50, 871. 3. (a) D. J. Cram, J. M. Cram, Science, 1974, 183, 4127. (b) J. Rebek, Jr., Science

1987, 235, 1478. 4. E. A. Meyer, R. K. Castellano, F. Diedrich, Angew. Chem. Int. Ed. 2003, 42,

1210. 5. L. J. Prins, D. N. Reinhoudt, P. Timmerman, Angew. Chem. Int. Ed. 2001, 40,

2382. 6. H.-J. Schneider, A. K. Yatsimirsky, 2000, Principles and Methods in

Supramolecular Chemistry, John Wiley & Sons Ltd., Chichester, England, p. 1-62.

7. I. Beletskaya, V. S. Tyurin, A. Yu. Tsivadze, R. Guilard, C. Stern, Chem. Rev. 2009, 109, 1659.

8. F.-G. Klärner, B. Kahlert, Acc. Chem. Res. 2003, 36, 919. 9. (a) J. T. Groves, Nature 1997, 389, 329. (b) M. C. Feiters, A. E. Rowan, R. J.

M. Nolte, Chem. Soc. Rev. 2000, 29, 375. 10. (a) M. J. Crossley, L. G. Mackay, A. C. Try, J. Chem. Soc. Chem. Commun.

1995, 1925. (b) T. Hayashi, M. Nonoguchi, T. Aya, H. Ogoshi, Tetrahedron Lett. 1997, 38, 1603. (c) N. Berova, G. Pescitelli, A. G. Petrovica, G. Pronic, Chem. Commun. 2009, 5958.

11. M. V. Escárcega-Bobadilla, A. W. Kleij, Chem. Sci. 2012, 3, 2421. 12. C. Mannich, P. Mohs, Chem Ber. 1930, B63, 608. 13. (a) M. K. Pugsley, D. A. Saint, E. Hayes, K. D. Berlin, M. J. A. Walker, Eur. J.

Pharmacol. 1995, 294, 319. (b) P. W. Thies, Pharm. Z. 1986, 15, 172. 14. C. L. Schauf, C. A. Colton, F. A. Davis, J. Pharmacol. Exp. Ther. 1976, 197,

414. 15. P. Comba, M. Kerscher, W. Schiek, 2007, Chaper 9 - Bispidine Coordination

Chemistry - Progress in Inorganic Chemistry Vol 55, John Wiley & Sons, Inc., Ed. K. D. Karlin, p. 613-693.

16. (a) J. E. Douglass, T. B. Ratliff, J. Org. Chem. 1968, 33, 355. (b) M. S. Arias, E. Galvez, J. C. Del Castillo, J. J. Vaquero, J. Chicharro, J. Mol. Struct. 1987, 156, 239. (c) T. Brukwicki, J. Mol. Struct. 1998, 446, 69.

17. G. D. Hosken, R. D. Hancock, Chem. Commun., 1994, 1363. 18. (a) L. Toom, A. Kütt, I. Kaljurand, I. Leito, H. Ottosson, H. Grennberg, A.

Gogoll, J. Org. Chem. 2006, 71, 7155. (b) A. Gogoll, L. Toom, H. Grennberg, Angew. Chem. Int. Ed. 2005, 44, 4729.

19. R. P. Hanzlik, Inorganic Aspects of Biological and Organic Chemistry, 1976, Elsevier, p.119.

60

20. (a) C. A. Hunter, N. Meah, J. K. M. Sanders, J. Am. Chem. Soc. 1990, 112, 5773. (b) H. L. Anderson, C. A. Hunter, N. Meah, J. K. M. Sanders, J. Am. Chem. Soc. 1990, 112, 5780.

21. J. B. Lambert, E. P. Mazzola, 2004, Nuclear Magnetic Resonance Spectroscopy: An Introduction to Principles, Applications, and Experimental Methods, Pearson Education Inc., New Jersey, USA.

22. (a) M. Panda, P.-W. Phuan, M. C. Kozlowski, Chem. Commun. 2002, 1552. (b) M. Breuning, A. Paasche, M. Steiner, S. Dilsky, V. H. Gessner, C. Strohmann and B. Engels, J. Mol. Struct. 2011, 1005, 178.

23. L. Pauling, J. Chem. Phys. 1936, 4, 673. 24. C. E. Johnson, F. A. Bovey, J. Chem. Phys. 1958, 29, 1012. 25. (a) M. Karplus, J. Chem. Phys. 1959, 30, 11. (b) M. Karplus, J. Am. Chem. Soc.

1963, 85, 2870. 26. G. Bifulco, P. Dambruoso, L. Gomez-Paloma, R. Riccio, Chem. Rev. 2007, 107,

3744. 27. (a) C.-H. Tan, Y. Kobayashi, Y. Kishi, Angew. Chem. Int. Ed. 2000, 39, 4282.

(b) S. Higashibayashi, W. Czechtizk, W. Czechtizky, Y. Kishi, J. Am. Chem. Soc. 2003, 125, 14393. (c) H. Seike, I. Ghosh, Y. Kishi, Org. Lett. 2006, 8, 3861.

28. (a) P. Cimino, G. Bifulco, A. Evidente, M. Abouzied, R. Riccio, L. Gomez-Paloma, Org. Lett. 2002, 4, 2779. (b) J. Kobayashi, M. Tsuda, Nat. Prod. Rep. 2004, 21, 77. (c) M. Murata, T. Yasumoto, Nat. Prod. Rep. 2000, 17, 293.

29. D. Neuhaus, M. Williamson, 1989, The Nuclear Overhauser Effect in Structural and Conformational Analysis, VCH Publishers Inc., New York, USA.

30. T. D. W. Claridge, 2009, High-Resolution NMR Techniques in Organic Che-mistry Second Edition: Tetrahedron Organic Chemistry Volume 27, Elsevier Ltd., Amsterdam, The Netherlands.

31. (a) Y. Cohen, L. Avram, L. Frish, Angew. Chem. Int. Ed. 2005, 44, 520. (b) W. S. Price, Concept. Magn. Reson. 1997, 9, 299.

32. J. Harmon, C. Coffman, S. Villarrial, S. Chabolla, K. A. Heisel, V. V. Krishnan, J. Chem. Educ. 2012, 89, 780.

33. (a) D. O. Cicero, G. Barbato, R. Bazzo, J. Am. Chem. Soc. 1995, 117, 1027. (b) N. Nevins, D. Cicero, J. P. Snyder, J. Org. Chem. 1999, 64, 3979. (c) P. Dambruoso, C. Bassarello, G. Bifulco, G. Appendino, A. Battaglia, G. Fontana, L. Gomez-Paloma, Org. Lett. 2005, 7, 983.

34. (a) J. P. Snyder, A. S. Lakdawala, M. J. Kelso, J. Chem. Soc. 2003, 125, 632. (b) J. J. Koivisto, E. T. T. Kumpulainen, A. M. P. Koskinen, Org. Biomol. Chem. 2010, 8, 2103.

35. (a) P. Thepchatri, D. O. Cicero, E. Monteagudo, A. K. Ghosh, B. Cornett, E. R. Weeks, J. P. Snyder, J. Am. Chem. Soc. 2005, 127, 12838. (b) M. Erdélyi, B. Pfeiffer, K. Hauenstein, J. Fohrer, J. Gertsch, K.-H. Altmann, T. Carlomagno, J. Med. Chem. 2008, 51, 1469. (c) A. B. Smith, C. A. Risatti, O. Atasoylu, C. S. Bennett, K. TenDyke, Q. Xu, Org. Lett. 2010, 12, 1792. (d) M. Fridén-Saxin, T. Seifert, L. K. Hansen, M. Grøtli, M. Erdélyi, K. Luthman, Tetrahedron 2012, 68, 7035.

36. C. M. Thiele, Concept. Magn. Reson. A 2007, 30A, 65. 37. (a) M. Sarfati, P. Lesot, D. Merlet, J. Courtieu, Chem. Commun. 2000, 2069.

(b) A. Marx, C. Thiele, Chem. Eur. J. 2008, 15, 254. c) A. Marx, B. Böttcher, C. Thiele, Chem. Eur. J. 2010, 16, 1656.

38. A. Amador, J. Chem. Educ.1979, 56, 4. 39. J. Quenneville, T. J. Martínez, J. Phys. Chem. A 2003, 107, 829. 40. H. A. Benesi and J. H. Hildebrand, J. Am. Chem. Soc. 1949, 71, 2703.

61

41. G. Scatchard, Ann. N. Y. Acad. Sci. 1949, 51, 660. 42. R. L. Scott, Recl. Trav. Chim. Pays-Bas, 1956, 75, 787. 43. L. Fielding, Tetrahedron 2000, 56, 6151. 44. P. Thordarsson, Chem. Soc. Rev. 2011, 40, 1305. 45. (a) L. Ostromisslensky, Ber. Dtsch. Chem. Ges. 1911, 44, 268. (b) R. Denison,

B. Trans. Faraday Soc. 1912, 8, 20. (c) P. Job, Ann. Chim. 1928, 9, 113. 46. (a) E. J. Olson, P. Bühlmann, J. Org. Chem. 2011, 76, 8406. (b) A. Sayago, M.

Boccio, A. G. Asuero, Int. J. Pharm. 2005, 295, 29. 47. J. A. Galbis, M. G. García-Martín, Top. Curr. Chem. 2010, 295, 147. 48. J. Defaye, D. Horton, Carbohyd. Res. 1970, 14, 128. 49. M. Han, R. Michel, B. He, Y.-S. Chen, D. Stalke, M. John, G. H. Clever,

Angew. Chem. Int. Ed. 2013, 52, 1319. 50. (a) T. Branda, E. J. Cabritab, S. Berger, Prog. Nucl. Mag. Res. Sp. 2005, 46,

159. (b) S. P. Gaynor, M. J. Gunter, M. R. Johnston, R. N.Warrener, Org. Biomol. Chem. 2006, 4, 2253. (c) A. Pastor, E. Martínez-Viviente, Coord. Chem. Rev. 2008, 252, 2314. (d) S. Slovak, L. Avram, Y. Cohen, Angew. Chem. Int. Ed. 2010, 49, 428.

51. C.A. Hunter, J. K. M. Sanders, J. Am. Chem. Soc. 1990, 112, 5525. 52. (a) M. Fathalla, J. Jayawickramarajah, Eur. J. Chem. 2009, 6095. (b) M. V.

Peters, R. Goddard, S. Hecht, J. Org. Chem. 2006, 71, 7846. (c) C. A. Hunter, L. D. Sarson, Tetrahedron Lett. 1996, 37, 699. (d) L. J. Esdaile, P. Jensen, J. C. McMurtrie, D. P. Arnold, Angew. Chem. Int. Ed. 2007, 46, 2090.

53. (a) V. Haridas, S. Sadanandan, Y. K. Sharma, S. Chinthalapalli, A. Shandilya, Tetrahedron Lett. 2012, 53, 623. (b) V. Haridas, K. S. Rajgokul, S. Sadanandan, T. Agrawal, V. Sharvani, M. V. S. Gopalakrishna, M. B. Bijesh, K. L. Kumawat, A. Basu, G. R. Medigeshi, PLOS Neglect. Trop. D. 2013, 7, e2005.

54. A. Enthart, J. C. Freudenberger, J. Furrer, H. Kessler, B. Luy, J. Magn. Reson. 2008, 192, 314.

55. A. Gogoll, H. Grennberg, A. Axén, Organometallics 1997, 16, 1167. 56. I. Ratera, J. Veciana, Chem. Soc. Rev. 2012, 41, 303. 57. (a) E. A. Hauser, M. B. Leggett, J. Am. Chem. Soc. 1940, 62, 1811. (b) M.

Hmadeh, H. Traboulsi, M. Elhabiri, P. Braunstein, A. Albrecht-Gary, O. Siri, Tetrahedron 2008, 64, 6522.

58. (a) J. Stubbe and W. A. Van der Donk, Chem. Rev. 1998, 98, 705. (b) G. E. Fanucci and D. S. Cafiso, Curr. Opin. Struct. Biol. 2006, 16, 644.

59. R. Amorati, M. Lucarini, V. Mugnaini, G. F. Pedulli, F. Minisci, F. Recupero, F. Fontana, P. Astolfi, L. Greci, J. Org. Chem. 2003, 68, 1747.

60. C. Train, L. Norel and M. Baumgarten, Coord. Chem. Rev. 2009, 253, 2342. 61. (a) A. Rajca, Chem. Rev. 1994, 94, 871. (b) P. J. Boratyński, M. Pink, S. Rajca,

A. Rajca, Angew. Chem. Int. Ed. 2010, 49, 5459. (c) M. Zalibera, A. S. Jalilov, S. Stoll, I. A. Guzei, G. Gescheidt, S. F. Nelsen, J. Phys. Chem. A 2013, 117, 1439. (d) M. Abe, Chem. Rev. 2013, DOI: 10.1021/cr400056a.

62. (a) M. Kirchgessner, K. Sreenath, K. R. Gopidas, J. Org. Chem. 2006, 71, 9849. (b) K. Sreenath, C. V. Suneesh, V. K. R. Kumar, K. R. Gopidas, J. Org. Chem. 2008, 73, 3245. (c) P. Vitale, L. Di Nunno, A. Scilimati, Arkivoc (iii) 2013, 36.

63. (a) M. J. S. Dewar, T. Nakaya, J. Am. Chem. Soc. 1968, 90, 7134. (b) G. Ghigo, A. Maranzana, G. Tonachini, Tetrahedron 2012, 68, 2161.

64. (a) R. Garner, J. Yperman, J. Mullens, L. C. Van Poucke, Fresen. J. Anal. Chem. 1993, 347, 145. (b) J. E. Douglass, H.-M. Shih, R. E. Fraas, D. E. Craig. Jr., J. Heterocyclic Chem. 1970, 1185.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1070

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.

Distribution: publications.uu.seurn:nbn:se:uu:diva-207138

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2013