Synthesis of Higher Homologues of Cucurbituril Derivatives and Evaluation of Their Supramolecular Properties
Feng Wu
August, 2012
A thesis submitted in fulfilment of the requirement for the degree of Doctor of philosophy
School of Physical, Environmental and Mathematical Sciences (Chemistry) Australian Defence Force Adademy, The University of New South Wales
ORIGINALITY STATEMENT
‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at
UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project’s design and conception or in style, presentation and linguistic expression is acknowledged.’
Signed
Date
ii
COPYRIGHT STATEMENT
‘I hereby grant the University of New South Wales or its agents the right to achieve and to make available my thesis or dissertation in whole or part in the
University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.
I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only).
I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.’
Signed
Date
iii
AUTHENTICITY STATEMENT
‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’
Signed
Date
iv
ACKNOWLEDGEMENTS
I would like to express my deepest gratitude to all who have encouraged, supported and advised me during the work presented in this thesis.
First and foremost, I would like to thank my supervisor Dr Anthony Day for giving me such a great opportunity to work in his group. His valuable ideas, excellent guidance and encouragement throughout this study had made this thesis possible. Thank you very much for your support and understanding throughout this process.
Besides, I would like to thank other members of the ADFA Chemistry discipline.
Thanks to A/Prof. J Grant Collins, A/Prof. Cliff Woodward and Dr Lynne Wallace for advice and points of view during our group meeting. Thanks to Dr Barry
Gray for NMR technical assistances. Special thanks also to Mrs Denise Russell for thesis English editing and correction.
Finally, yet importantly, I would like to express my gratitude to my family and friends for the support throughout my studies.
v
CONTENTS
ORIGINALITY STATEMENT...... ii COPYRIGHT STATEMENT ...... iii AUTHENTICITY STATEMENT ...... iv ACKNOWLEDGEMENTS ...... v ABBREVIATIONS ...... ix ABSTRACT ...... xvi CITATIONS ...... xviii
CHAPTER 1 Introduction ...... - 2 - 1.1 General introduction - motivation and thesis outline ...... ‐ 2 ‐ 1.2 Physical properties and chemical behaviour of cucurbit[n]uril family .. ‐ 4 ‐ 1.2.1 Synthesis of Q[n]...... ‐ 5 ‐ 1.2.2 Physical properties of Q[n] ...... ‐ 6 ‐ 1.2.3 Chemical behaviour of Q[n]s and their applications ...... ‐ 9 ‐ 1.3 New cucurbit[n]uril derivatives, analogues and related compounds .. ‐ 16 ‐ 1.3.1 Synthesis of new Q[n] derivatives ...... ‐ 16 ‐ 1.3.2 Synthesis of Q[n] analogues and related compounds ...... ‐ 23 ‐ 1.4 Aims of study ...... ‐ 25 ‐ 1.5 References ...... ‐ 26 ‐
CHAPTER 2 CyclopentanoQ[n] family ...... - 33 - 2.1 Introduction ...... ‐ 33 ‐ 2.2 Aims of study ...... ‐ 38 ‐ 2.3 Experimental ...... ‐ 39 ‐ 2.3.1 Materials ...... ‐ 39 ‐ 2.3.2 Instrumental methods ...... ‐ 39 ‐ 2.3.3 Experimental procedures ...... ‐ 40 ‐ 2.4 Results and discussion ...... ‐ 46 ‐ 2.4.1 Structural analyses of glycoluril diethers relative to changes in substitution ...... ‐ 46 ‐
vi
2.4.2 Synthesis of CyPnQ[n] from cyclopentanoglycoluril ...... ‐ 48 ‐
2.4.3 Modification of reaction conditions for the synthesis of CyPnQ[n] ...... ‐ 57 ‐
2.4.4 Guest selectivity between CyPnQ[n] and Q[n] ...... ‐ 59 ‐
2.4.4.1 Dioxane selectivity between CyP6Q[6] and Q[6] ...... ‐ 59 ‐ + 2.4.4.2 Comparative affinity of CyP7Q[7] and Q[7] for adamantylNH3 (Ada) ...... ‐ 63 ‐ 2.5 Conclusions ...... ‐ 65 ‐ 2.6 References ...... ‐ 67 ‐
CHAPTER 3 Five-membered carbon ring glycoluril derivatives with functionality ...... - 71 - 3.1 Introduction ...... ‐ 71 ‐ 3.2 Aims of study ...... ‐ 74 ‐ 3.3 Experimental ...... ‐ 75 ‐ 3.3.1 Materials ...... ‐ 75 ‐ 3.3.2 Instrumental methods ...... ‐ 75 ‐ 3.3.3 Synthetic procedures ...... ‐ 76 ‐ 3.4 Results and discussion ...... ‐ 106 ‐ 3.4.1 Synthesis of functional five-membered carbon ring glycolurils . ‐ 106 ‐ 3.4.2 Synthesis of glycolurils with norbornane ring substituent ...... ‐ 117 ‐ 3.5 Conclusions ...... ‐ 128 ‐ 3.6 References ...... ‐ 129 ‐
CHAPTER 4 Guest exchange moderated by substituents on the host ...... - 134 - 4.1 Introduction ...... ‐ 134 ‐ 4.2 Aims of study ...... ‐ 137 ‐ 4.3 Experimental ...... ‐ 138 ‐ 4.3.1 Materials ...... ‐ 138 ‐ 4.3.2 1H NMR experiments ...... ‐ 139 ‐ 4.3.3 Experimental procedures ...... ‐ 139 ‐ 4.4 Results and discussion ...... ‐ 143 ‐ 4.4.1 BmaFc/Ada guest exchange in substituted Q[7]s ...... ‐ 145 ‐ 4.4.2 Guest exchange mechanism in Q[7] ...... ‐ 147 ‐
vii
4.4.3 Exchange kinetics and guest size ...... ‐ 156 ‐ 4.5 Conclusions ...... ‐ 163 ‐ 4.6 References ...... ‐ 165 ‐
CHAPTER 5 Conclusions ...... - 169 - 5.1 Synthesis of higher homologues of SQ[n] ...... ‐ 169 ‐ 5.2 Synthesis of substituted glycolurils with functionalised cyclopentane rings ...... ‐ 171 ‐ 5.3 Modification of electronic properties of substituted Q[n] by variations of substituents ...... ‐ 173 ‐ 5.4 Recommendations for future research ...... ‐ 174 ‐ 5.5 References ...... ‐ 176 ‐
APPENDIX ...... - 177 -
viii
ABBREVIATIONS
The following abbreviations are used in this work:
Ac Acetate
Bz Benzoyl
CD Cyclodextrin
CDCl3 Deuterated chloroform
CyH Cyclohexano
CyP Cyclopentano
CyPnQ[n] Cyclopentanocucurbit[n]uril d Doublet
DCM Dichloromethane
Chemical shift (NMR)
DMF Dimethylformamide
DMSO Dimethyl sulfoxide
D2O Deuterium oxide
ES-MS Electrospray ionization mass spectrometry
Et3N Triethylamine
EtOAc Ethyl acetate
EtOH Ethanol
HR-MS High resolution mass spectrometry
Hz Hertz, [s-1]
IR Infrared spectroscopy
J Coupling constant m Multiplet
Me Methyl
ix
MeOH Methanol
Mp Melting point
NMNO N-methylmorpholine N-oixde
NMR Nuclear magnetic resonance
NOSEY Nuclear overhauser enhancement spectroscopy
PhH Benzene
ppm Parts per million
PTSA p-Toluenesulfonic acid
Py Pyridine
Q[n] Cucurbit[n]uril rt Room temperature s Singlet
SQ[n] Fully substituted cucurbit[n]uril t Triplet
TEA Triethylamine
TFA Trifluoroacetic acid
TFAA Trifluoroacetic anhydride
THF Tetrahydrofuran
TLC Thin layer chromatography
Ts Tosyl
x
xi
xii
xiii
xiv
xv
ABSTRACT
The cucurbit[n]uril (Q[n]) family provides real potential for a wide range of applications, including drug delivery, complex self-sorting systems, chemical sensors, environmental protection and supramolecular catalysis. However, compared with other host molecules, such as crown ethers, cyclodextrins and calixarenes, its further development has been limited by the challenge of achieving its functionalisation. The introduction of substituents, especially those carrying reactive functional groups, into Q[n] has met with significant problems, such as their unavailability or the fact that they are restricted to the smaller homologues, where n = 5–6 and, in some cases, only n = 5.
In this thesis, we describe the synthesis of a new family of cyclopentanoQ[n] (n
= 5–7) which opens the way to obtaining significant proportions of the substituted higher homologues. In describing our achievement, we also identify reasons why certain publications report the synthetic difficulties in trying to obtain other substituted higher homologues.
We have shown that alkyl substituents on Q[6,7] modify the host–guest binding behaviour with regard to increased binding constants and slower rates of exchange. The reasons for these changes have been discussed with reference to their substitutent effects upon the electronic characters of their portals and cavities. In this context, the shape and size of the guest, as well as its charge, are considered. In particular, the neutral guest dioxane and the ionic guest adamantyl ammonium ion have been studied. The rates of guest exchange for
xvi substituted Q[7] have been found to be much slower than those for normal Q[7] when studied in detail using the adamantyl ammonium ion and substituted ferrocene ammonium ion derivatives. Two guest exchange mechanisms are possible: dissociation and replacement; and displacement or ‘shunting’. Both have been evaluated in relation to the kinetics of the exchange reaction.
Our study of the synthesis of the new family of cyclopentanoQ[n] derived from the cyclopentane substituent provides opportunities for achieving similar substituents which bear reactive functional groups. Some of these are explored, and at least two potentially applicable starting glycolurils, in which the cyclopentane substituent ring carries reactive functional groups as esters or acetate-protected alcohols, were successfully synthesised. Our cyclopentanoglycoluril example, which introduces substitution into the Q[n] family by virtue of the positive effect of a five-membered carbon ring on the Q[n] synthesis, and, in particular, achieves the higher homologues, led us to also examine the strained rings of the norbornane structure. Glycolurils derived from norbornane derivatives could potentially offer greater opportunities to the synthesis of higher homologues of substituted Q[n]. We report our findings in relation to the ease or difficulty of the norbornane derivatives to form glycolurils.
We discovered intermediates that would normally not have been isolated, however, they help to explain the general reaction process in the synthesis of aliphatic glycolurils, which are less common than aromatic glycolurils. These intermediates provide insight that may prove useful for future aliphatic glycoluril syntheses.
xvii
CITATIONS
Material from this thesis is preparation for publication or has been published or presented in the following forms:
1. Wu, F.; Wu, L.-H.; Xiao, X.; Zhang, Y.-Q.; Xue, S.-F.; Tao, Z.; Day, A. I.
“Locating the cyclopentano cousins of the cucurbit[n]uril family” J.
Org. Chem. 2012, 77, 606–611.
2. Wu, F.; Mandadapu, V.; Day, A. I. “Why are alkyl substituted
glycolurils from dicarbonyls less common? – an insight into the
intermediates and divergent paths of a multi-step process”
manuscript in preparation, August, 2012.
3. Wu, F.; Woodward, C. E.; Day, A. I. “Guest exchange moderated by
the substituents of the molecular host: exchange where equilibrium
is reached in days” manuscript in preparation, August, 2012.
4. Wu, F.; Xue, S.-F.; Tao, Z.; Day, A. I. “Cyclopentanocucurbit[n]uril
family” 2nd International Conference on Cucurbiturils, Cambridge, UK,
June, 2011.
5. Wu, F.; Xue, S.-F.; Tao, Z.; Day, A. I. “The synthesis of
cyclopentanocucurbit[n]uril family” 6th International Symposium on
Macrocyclic and Supramolecular Chemistry, Brighton, UK, July, 2011.
xviii
6. Day, A. I.; Wu, F. “Guest exchange at a glacial pace in spite of the
double door access to the cavity of cucurbit[n]uril” 2nd
Supramolecular Chemistry in New Zealand and Australia meeting,
Sydney, Australia, October, 2011.
xix
CHAPTER 1
INTRODUCTION
- 1 -
CHAPTER 1
Introduction
1.1 General introduction - motivation and thesis outline
Cucurbit[n]uril (Q[n]) is a relatively new family of cavitands which has interesting molecular host–guest properties. A cucurbituril is a macrocyclic molecule shaped like an open-ended barrel, the inner part of which is capable of encapsulating small molecules. The driving forces for encapsulation are hydrophobic effects, van der Waals contact forces, and ion–dipole and dipole– dipole interactions, including hydrogen bonding (Figure 1.1) [1]. The openings to cucurbit[n]uril cavities, known as portals, are rimmed by electron-rich carbonyl oxygens and it is these that are important in the ion–dipole interactions.
Exploitation of these interactions has been demonstrated by many examples using cations generated from protonated amines, quaternary ammonium ions and metal ions [1]. Hydrophobic interactions have been reported to form neutral molecule complexes, such as gases, various hydrocarbons, carborane, ferrocenes and molecules possessing both hydrophobic and positively charged parts [1].
- 2 -
Figure 1.1 Q[6] with its supramolecular host–guest interaction regions highlighted.
Cucurbit[n]urils have potential application in a number of areas, including drug delivery systems [2], complex self-sorting systems [3, 4], chemical sensors [5-9], environmental protection and supramolecular catalysis [10-18]. To date, reported examples of catalysis have usually taken advantage of the host properties of Q[n] to orient cationic reactants into conformations suitable for reactions or to stabilise a reaction intermediate. However, these reactions have only been carried out in the cavities of the simplest members of the cucurbit[n]uril family [1]. Although positive results showing increased reaction rates, improved stability of the catalysed products and control over regiochemistry have been found, other cavity-based reactions have yet to be realised.
Supramolecular assemblies based on Q[n] also play an important role in some biologically relevant reactions, such as the energy cycle of the cell and, metabolic pathways for function and detoxification (e.g., pharmaceutical processing), which are of great interest in a biomimetic context. Many of these biological reactions are catalytic processes of some enzymes involving, for example, redox reactions, carbon–carbon bond formation and hydrolysis.
- 3 -
Common active components are thiols, thioesters and disulfides [19]. Kaifer et al. observed the properties of Q[6] inclusion complexes of disulfides or thiols and demonstrated that the inclusion of Q[6] leads to a pronounced stabilisation of these disulfides or thiols in the solution [20].
The development of substituted cucurbit[n]urils and, in particular, their substitution with reactive functional groups is relatively recent with there being only a few examples [1]. Substitution of Q in its development has been used to define a functional group, which is attached to the methane carbon of the glycoluril moiety. In the past, a number of fully and partially substituted Q[n]s, such as the fully substituted CyH5,6Q[5,6] (cyclohexano, CyH) [6], Me10Q[5] [21],
Me12Q[6] [22] and (HO)2nQ[n] [23], and the partially substituted Ph2Q[6] [24],
Me4Q[6] [25] and (Me2CyP)nQ[6] (cyclopentano, CyP) [22], have been synthesised. This is an area that provides the possibilities for more complex structures and, thus, new opportunities for chemical reactivity [26].
1.2 Physical properties and chemical behaviour of cucurbit[n]uril family
A cucurbituril composed of C36H36N24O12 (cucurbit[6]uril), the first compound of the Q[n] family, was first synthesised by Behrend et al. in 1905 who used the condensation reaction of glycoluril with an excess of formaldehyde in concentrated sulfuric acid at a fairly high temperature [27]. However, at that time, there were few analytical techniques, such as X-ray crystallography, that are routinely used today to establish structures of molecules. The chemical structure of Q[6] remained unknown for almost 8 decades until 1981 when
- 4 -
Freeman, Mock and Shih reproduced Behrend et al.’s 1905 synthesis and obtained a colourless crystalline compound, the structure of which was characterised by X-ray diffraction analysis [28]. Cucurbit[6]uril is a symmetrical, cage-like structure consisting of 6 glycoluril units linked by methylene bridges
(Figure 1.2) and was called “cucurbituril” by Mock due to the similarity of its molecular visual image to Cucurbitacee (the pumpkin family). About two decades later, cucurbit[n = 5, 7, 8]urils (Q[5], Q[7] and Q[8]) were prepared and isolated by two research groups in 2000 [29, 30], and the largest homologue, cucurbit[10]uril (Q[10]), was reported in 2002 [31]. In this thesis, cucurbiturils consisting of n repeat glycoluril units are abbreviated to Q[n]s or referred to as cucurbit[n]urils.
Figure 1.2 Side and top views of Q[6] shown as ball and stick models.
1.2.1 Synthesis of Q[n]
In the condensation reaction of glycoluril and formaldehyde, Behrend et al. and
Mock et al. initially obtained yields in their syntheses of Q[6] at around 20% and did not detect any other macrocyclic homologues composed of a different number of glycoluril fragments [27, 28]. About two decades later, the synthesis
- 5 - used to produce Q[n] was optimised by the two research groups of Day and Kim who detected and isolated more Q homologues (Q[5–8] and Q[5]@Q[10])
(Scheme 1.1) [29-31]. Although there is NMR spectroscopic evidence indicating the existence of Q[9] and other Q[n], where n is larger than 10, these homologues have not yet been isolated. Cucurbit[10]uril is the largest homologue that has been isolated and there is no evidence that there is a Q[n] smaller than Q[5] [32]. Day et al. found that, to obtain a number of homologues of Q[n], the condensation reaction could be carried out in one pot using milder conditions (Scheme 1.1) [32] instead of those reported by Mock et al.,
Buschmann et al. and Behrend et al. [28].
Scheme 1.1 Synthesis of Q[n] from glycoluril under milder conditions.
Conditions: a) CH2O, HCl, 100 °C, 18 h.
1.2.2 Physical properties of Q[n]
As a new family of host molecules, Q[n]s have often been compared to the cyclodextrins (CDs) as they have similar cavity sizes and both have been studied in aqueous solutions [1]. Similar to CDs, Q[n]s can accommodate small organic molecules within their hydrophobic cavities. However, unlike CDs, the two carbonyl group rich openings (portals) allow Q[n]s to bind ions and
- 6 - molecules through ion–dipole, as well as hydrogen-bonding, interactions.
Another important difference in structural features between the two host molecules is the Q[n]s’ structural rigidity. Therefore, Q[n]s virtually retain their shapes upon the binding of different guests and, consequently, may be expected to be more shape-selective than CDs in their formations of host–guest complexes.
Sizes
Cavity size is the primary determinant as to which cucurbituril could be used in the formation of inclusion complexes. Size and shape dependencies, as well as ion–dipole interactions, are important factors for achieving good incorporation of guests into the cavities of cucurbiturils. The diameter of the cavity or portal of a cucurbituril is a function of the number of glycoluril fragments forming the cucurbituril. The cavity volumes of Q[6], Q[7] and Q[8] parallel those of the -,
- and -CDs (Table 1.1). Free Q[10] has a cavity volume of 870 Å3 and forms a water-soluble inclusion complex with sizeable guests [33].
- 7 -
Table 1.1 Dimensions of cucurbit[n]urils (Q[n]s) and cyclodextrins (CDs) [1, 33].
Cucurbit[n]urils (Q[n]s) Cyclodextrins (CDs)
Q[5] Q[6] Q[7] Q[8] Q[10] -CD -CD -CD
Cavity 11.3– Dia. 4.4 5.8 7.3 8.8 4.7–5.3 6.0–6.3 7.5–8.3 12.4 (Å) Cavity Vol. 82 164 279 479 870 174 262 427 (Å3)
Stability
The Q[n] family shows significantly stable thermal/physical properties (> 370 °C) which are superior to those of CDs [34, 35]. Day et al. studied the relative stability of Q[n]s in the presence of concentrated HCl at 100 °C and found that
Q[5], Q[6] and Q[7] were stable with no detectable decomposition. However,
Q[8] in concentrated HCl at 100 °C after 24 h showed that it was converted to the smaller homologues. The very low proportion of Q[9] (< 1%) in the synthetic reaction of Q[n]s also indicated that the larger Q[n] homologues were thermally unstable compared with the smaller ones [32]. In addition, Q[n]s exhibit relatively high degrees of chemical stability as they do not react chemically with acids, bases or other chemicals. The reactions of regular Q[n]s with K2S2O8, which can be attributed to the high chemical stability of Q[n]s, have been the only direct functionalisation method reported to date.
- 8 -
Solubility
Both Q[6] and Q[8] have relatively low solubility in pure water but this can be substantially increased in the presence of ammonium ions, alkali and alkaline metals, due to their formation of complexes with cations [1]. Q[5] and Q[7] have a substantially higher solubility in aqueous media than Q[6] and Q[8] which has allowed the investigation of their supramolecular behaviour in neutral water [1].
Acidity
One of the structural features of Q[n]s that makes them unique is that their two portals, which are lined by ureido carbonyl groups, can function as proton
(hydrogen ion) acceptors. The pKa value of the conjugate acid of Q[7] has been measured by titration with dilute HCl to be 2.20 [36] and that of the conjugate acid of Q[6] to be 3.02 [37]. It is predictable that there would be small variations in acidity for other Q[n]s because of their difference in portal size and the number of electronegative carbonyl oxygens.
1.2.3 Chemical behaviour of Q[n]s and their applications
Their hydrophobic inner cavities with different sizes and the presence of carbonyl oxygens that line their ends make the homologous series of Q[n]s behave efficiently as host molecules in supramolecular chemistry. They have shown their remarkable binding affinities and selectivities in aqueous solutions toward a variety of molecules which have allowed them to play an important role in drug delivery systems [2], chemical sensors [5-9], separation technologies, environmental protection and catalysts [10-18].
- 9 -
Q[5]
Q[5], the smallest member or homologue of the Q family, contains 5 glycoluril units. It should be noted that, although its binding cavity is similar to that of the
-CD, it is rather rigid and slightly larger in diameter than its openings (portals).
The supramolecular chemistry of Q[5] is comparatively limited due to its low volume cavity (82 Å3), and a great deal of the study of it has focused on its formation of metal or ammonium ion complexes using its two ureidyl carbonyl- lined portals [1]. However, Q[5] is capable of accommodating small molecules, such as gases (e.g., certain noble gases, N2, O2, NO, CO, CO, etc.) and some solvents (e.g., CH3OH and CH3CN), within its cavity. These results suggest that
Q[5] is of potential interest for gas separation and purification by the selective adsorption of one or more components from gas mixtures [1].
Q[6]
Q[6], the first member of the Q family introduced earlier in this section, is the major product in the Q[n]-forming reaction using sulfuric or hydrochloric acids.
Due to its larger-sized cavity, it is capable of binding a variety of small organic molecules by ion–dipole interactions, hydrogen bonds and the hydrophobic effect [1]. In their early work, Mock et al. studied the binding affinities of Q[6] toward a series of alkylammonium ions through 1H NMR competition
1 experiments in 40% aqueous HCO2H and measured binding constants of 10 –
108 M-1. They also described the mechanism of guest exchange in the cavity of
Q[6] which indicated that the exchange process involved two steps, the vacation of the cavity by dissociation and its filling by an alternate guest [38]. Q[6] is very
- 10 - slightly soluble in aqueous or organic solvents. The majority of binding studies of
Q[6] have required that alkali metal ions or protons be added to achieve appreciable solubility in water. However, early work reported by Werner et al. on the effect of salts or acid concentration on the supramolecular behaviour of Q[6] indicated that the thermodynamic and kinetic data for Q[6] did not truly reflect the supramolecular affinity between the host and guest because it involved the presence of mainly metal ions or protons [39]. Therefore, considerable caution should be exercised when comparing the reported binding constants.
Q[7]
Relative to Q[5] and Q[6], Q[7] is capable of incorporating a wider-range of guests inside its more spacious cavity. It is known to bind a wide range of positively charged guests, such as adamantanes, bicyclooctanes, naphthalene, ferrocene, cobaltocene and their derivatives [1]. It also binds metal complexes, such as multinuclear platinum complexes, which is one example of its utility as an efficient drug delivery vehicle [40]. The biocompatible and relatively low- or non-toxic properties of Q[7] are essential to its successful application in drug delivery [2]. Depending on the situation, the complexation between Q[7] and a drug may increase the aqueous solubility of the drug [40], reduce its undesired toxicity [41], improve its stability or activate it [42].
Inspired by the high affinity of the avidin-biotin complex (ca. 1015 M-1) through non-covalent interactions, Inoue et al. established the ferrocene@Q[7] complexation system which yielded high levels of binding affinity in aqueous solutions similar to those of the avidin-biotin complex; for example, the dicationic
- 11 - ferrocene derivative 1,1'-bis(trimethylammoniomethyl)ferrocene exhibited an
15 -1 extremely high equilibrium association constant with Q[7] (Ka = 3 × 10 M ) and the monocationic guest, (ferrocenylmethyl)trimethylammonium, also exhibited a
12 -1 very high Ka value (3 × 10 M ) with the same host in pure water [43]. The relatively high affinity between Q[7] and such cationic guests is achieved because of the hydrophobic effect of the ferrocenes’ core in the cavity of the host as well as the positively charged ammonium ions closely associated with each of the portals. In 2011, Inoue et al. proposed a new series of guest molecules, including bicycle[2.2.2]octane and adamantane derivatives, which could bind Q[7] with affinities similar to those of analogous ferrocene derivatives [44].
Q[8]
Despite its poor intrinsic solubility in water and organic solvents, Q[8] has received considerable attention, primarily due to its relatively large cavity which sometimes can accommodate two guests simultaneously [30, 45, 46]. Kim et al. demonstrated that an electron-deficient guest and an electron-rich guest could form a stable 1:1:1 ternary complex with Q[8]. However, by themselves, electron-rich guests do not bind to Q[8], thus suggesting that the major driving force for the ternary complex formation is a strong interaction between those two guests within the host molecule Q[8]. Importantly, the presence of Q[8] leads to the enhancement and stabilisation of the charge-transfer interaction between the guest pair which is probably due to their close contact inside the cavity of the host [46].
- 12 -
Two research groups have taken advantage of the stability of this type of charge-transfer complex. Kaifer et al. demonstrated that the preparation of a charge-transfer complex between two dentrimers with π-donor and π-acceptor units can be mediated by the presence of host molecule Q[8] to construct a dentrimer assembly. This finding offers a versatile method for the building of macromolecular assemblies of desirable types and sizes [47]. Scherman et al. also used the charge-transfer complex to advantage in polymer synthesis in which they described the complex as a ‘cuff-link’ [48].
Scheme 1.2 (a) Schematic formation process of 1:2 host–guest complex based on cucurbit[8]uril in water. (b) Thermodynamic and kinetic parameters of second-guest binding. The figure was taken from [48].
Q[10]
Q[10] was initially isolated as the Q[5]@Q[10] inclusion complex shown in
Figure 1.3 [31]. In 2005, Isaacs’s group reported the isolation of free Q[10] and disclosed its binding properties, including forming termolecular complexes,
- 13 - which indicated similar behaviour to that of Q[8] [49]. The portal dimensions and cavity volume of Q[10] are significantly larger than those of Q[8]. Its vast volume can offer more opportunities for the binding of larger molecules or the deeper encapsulation of smaller molecules with flexible moieties that can fold inside its cavity.
Figure 1.3 Molecular gyroscope made from Q[5]@Q[10] complex.
In 2008, Isaacs et al. studied the supramolecular behaviour of porphyrin@Q[10] complexes and found that host Q[10] was able to protect and preserve the photophysical and electrochemical properties of encapsulated porphyrins. Such complexed porphyrins may be important in a variety of applications, such as enzyme-mimetic catalysts for oxidation reactions and agents for targeted phototherapeutic systems [50].
Dinuclear ruthenium (Rubbn) complexes (Figure 1.4) are cytotoxic and exhibit distinguishing antibiotic activity against bacteria [51]. In 2010, Day et al. reported the formation and characterisation of Rubbn@Q[10] complexes and found that ruthenium complexes can be encapsulated and released slowly over
- 14 - several hours from the cavity of Q[10]. These results indicate that Q[10] has considerable potential as a delivery mechanism for the controlled release of large dinuclear ruthenium complexes [52].
Figure 1.4 Structures of dinuclear ruthenium drugs (Rubbn).
In 2006, Isaacs et al. reported the isolation and recognition properties of nor- seco-cucurbit[10]uril (ns-Q[10]) with two offset cavities, each of which was comparable in size to those of Q[6] and Q[7]. Interestingly, ns-Q[10] exhibited homotropic allostery based upon a guest size-induced pre-organisation mechanism [53].
From a practical point of view, the application of the Q[n] family is still limited, in large part due to its poor solubility in aqueous and organic media. We, and others, have been studying the synthesis and recognition properties of Q[n] derivatives as a remedy to the limited solubility. Preliminary results from substitution on cucurbiturils have indicated that substituents can modify and manipulate the electronic character at the portals and within the cavity, which play a crucial role in molecular recognition events in both aqueous and organic solvents [26, 55-57].
- 15 -
1.3 New cucurbit[n]uril derivatives, analogues and related compounds
1.3.1 Synthesis of new Q[n] derivatives
Despite the wide range of applications in which the regular Q[n] compounds may be used, there is a real need for the preparation of Q[n] derivatives with substituents at their peripheries. The potential gains from producing Q[n] derivatives are: (1) to modify their solubility to a range of solvents as well as increase their aqueous solubility [1]; (2) to modify and manipulate the electronic character of both their portals and cavities [26, 55, 56]; and (3) to open a route to more complex structures.
There are three potential pathways available for the synthesis of Q[n] derivatives. The first is the condensation reaction between glycolurils carrying substituents and formaldehyde under acid-catalysed conditions which produces
Q[n] derivatives bearing the substituents at the ‘‘equatorial’’ position. This pathway can be modified to produce partially or fully substituted Q[n]s (Scheme
1.3 (a), (b)). The second pathway is the direct (per)hydroxylation of Q[n] developed by Kim et al. in 2003 (Scheme 1.3 (c)) [23]. The third pathway, which also involves a direct derivatisation, produces a substituent linked between a pair of glycoluril N or something similar (see Page 19).
We firstly discuss the use of glycoluril carrying substituents to give fully or partially substituted Q[n]s.
- 16 -
Scheme 1.3 Syntheses of Q[n] derivatives. (a) Fully substituted Q[n], CH2O,
HCl, heat; (b) Partially substituted Q[n], n = s + u, CH2O, HCl, heat; (c) Direct oxidation of Q[n], K2S2O8, H2O, 85 °C.
Fully substituted Q[n]
The method outlined in Scheme 1.3 (a), which gave the first fully substituted
Q[5] (R = Me, Me10Q[5]) reported by Stoddart et al. in 1992, was demonstrated to be possible but limited [21]. Almost a decade later, Kim et al. isolated the fully substituted cyclohexanocucurbit[n]uril, CyHnQ[n], where n = 5 and 6, the distinguishing feature of which was its enhanced solubility in both aqueous and organic solutions [6]. The drawback to the synthesis of a fully substituted Q[n] is that the substituted glycolurils so far reported favour the formation of smaller cucurbituril homologues [1, 33]. Fully substituted Q[n], including Me10Q[5] and
+ CyH5,6Q[5,6] (cyclohexano, CyH), have found limited applications, such as in K
- 17 - sensor or gas binding, because of their relatively low cavity volumes [1, 6]. To date, no fully substituted Q[7] has been reported prior to this thesis where the method of Scheme 1.3 (a) has been applied.
A variation of this method, using a mixture of substituted and unsubstituted glycolurils, gives partially substituted Q[n]s (Scheme 1.3 (b)).
Partially substituted Q[n]
The reaction of a mixture of glycoluril and its derivatives with formaldehyde randomly gives a complicated mixture of partially substituted and regular Q[n]s which are difficult to isolate in a pure form [24, 58]. The benefits of introducing glycoluril moieties into a substituted Q[n] is that the presence of unsubstituted glycoluril can reduce the angle influence imposed by the adjacent substituted glycoluril upon the methylene linking groups between them, thus readily forming higher homologues of partially substituted Q[n] [25]. The question has arisen as to whether there is a strategy that can be used to induce a degree of control over the distribution of partially substituted cucurbituril homologues. Day et al. first demonstrated the controlled synthesis of the symmetrical Me4Q[6] utilising the dimer of glycoluril which delivered Me4Q[6] in a 30% yield on a multi-gram scale (Scheme 1.4) [25]. Interestingly, the introduction of substituted glycoluril moieties can affect the shape of a partially substituted Q[n]’s cavity which also affects the accommodation of some organic guests. As an example, the symmetrical Me4Q[6] has an ellipsoidal cavity and binds 2,2'-bipyridine with its aromatic rings parallel to the long axis of its cavity [25]. A variety of partially substituted Q[n]s have been synthesised and isolated utilising a similar strategy.
- 18 -
Scheme 1.4 Controlled synthesis of Me4Q[6].
Other paths for producing functionalised Q[n]
Another way of synthesising a fully substituted Q[n] is the direct functionalisation of Q[5–8] developed by Kim et al. which is shown in Scheme
1.3 (c) [23]. While the yields for the oxidation reactions of Q[5] and Q[6] were reasonable (42% and 45%, respectively), those for the higher homologues (Q[7] and Q[8]) were low (< 5%). The reason for this is unknown.
Scheme 1.5 Synthesis of monofunctionalised Q[6] derivatives. Conditions: (a) 9
M H2SO4 or conc. HCl, rt. Adapted from Isaacs et al. [59].
Recently, Isaacs et al. demonstrated one novel approach toward monofunctionalised Q[6]-type compounds based upon Q[n] compounds lacking one or more bridging CH2-groups. For example, the reaction of readily available methylene-bridged glycoluril hexamer with substituted phthalaldehydes yielded
- 19 - a Q[6] derivative containing one reactive substituent sitting at the methylene bridge of the Q[6] skeleton (Scheme 1.5). The host–guest recognition properties of this type of monofunctionalised Q[6] derivative are similar to those known for unsubstituted Q[6] [59]. This group of researchers also reported the different template effects (positive or negative) of p-xylylenediammonium ions upon the formation of methylene-bridged glycoluril hexamer and regular Q[6] [59]. This result suggests that a similar strategy utilising larger templates could be used for the synthesis of unknown glycoluril heptamers or octamers which could also act as building blocks for the construction of higher homologues of monofunctionalised Q[n] derivatives.
Application of substituted Q[n]
An important factor governing the bioavailability of a drug candidate is its aqueous solubility. The host–guest nature of substituted Q[n]s and their enhanced solubility in an aqueous medium can facilitate their potential as a drug delivery vehicle. For example, Day’s group showed that the symmetrical
Me4Q[6] facilitated the solubility of the poorly soluble class of cytotoxic drugs, the benzimidazoles – albendazole (ABZ) and (2-methoxyethyl) 5-propylthio-1H- benzimidazole-2-yl carbamate (MEABZ) (Figure 1.5), for which encapsulation in
Me4Q[6] increased their water solubility by 2400-fold and 300-fold, respectively
[54, 60].
- 20 -
Figure 1.5 Albendazole (ABZ), (2-methoxyethyl) 5-propylthio-1H- benzimidazole-2-yl carbamate (MEABZ) and host Me4Q[6].
Another important benefit of substituents is their ability to modify and manipulate the electronic characters of both the portals and cavity which play a significant role in supramolecular chemistry [26, 55, 56].
Substituted Q[n]s can be used as fundamental building blocks for the construction of solid-state frameworks. In their recent work, Day et al. indicated that the increased electronegativity of the carbonyl oxygens of the substituted glycoluril moiety could favour the formation and stability of supramolecular rings constructed from substituted Q[5] and metal ions [55]. Also, their subsequent work in 2010 revealed that a general feature of substituted Q[5] was that it coordinated directly with metal ions and formed various substituted Q[5]-based metal-organic frameworks with unusual structures which could act as absorption materials for gases and volatiles (Figure 1.6) [56].
- 21 -
Figure 1.6 Three substituted Q[5]s and their formations of supramolecular rings. The figure was taken from [56].
Kim et al. achieved a major breakthrough for substitution in the synthesis of perhydroxylated Q[n] derivatives which could undergo further covalent derivatisation reactions to yield a variety of Q derivatives [23]. In 2007, a model of a targeted drug delivery vehicle was reported by the same research group using Q[6]-based carbohydrate clusters which were synthesised by photochemical reaction on the hydroxylated Q[6] compound (Scheme 1.6) [61].
The potential for extending this type of targeting is, in principle, available to the higher homologues of cucurbiturils which could be loaded with a wider range of drugs for delivery to a specific set of diseased cells and unloaded.
- 22 -
Scheme 1.6 Synthesis of carbohydrate clusters based on Q[6] (average number of sugar groups attached approximately 11). The figure was taken from [2].
Kim’s group also developed Q[6] derivatives with a number of different substituted groups attached as (per)ethers or (per)thiols. These derivatised Q[6] can be incorporated into a polymer nanocapsule with a drug encapsulated in the polymer nanocapsule’s core. Some molecules bound within the cavity of derivatised Q[6] can function as targeting groups employed to control the destination of drugs. This is an alternative example of a targeted drug delivery system based upon supramolecular chemistry and the special properties of Q[6] derivatives [62].
1.3.2 Synthesis of Q[n] analogues and related compounds
A variation of the synthesis of Q[n]s, which involved replacing some of the glycolurils with alternative molecular units, was developed by Isaacs et al. who utilised the condensation reaction of bis(phthalhydrazide) and glycoluril cyclic ether to produce Q[n] analogues with several novel properties (Scheme 1.7), such as: (1) electrochemically and fluorescently active barrel, UV/Vis; (2)
- 23 - elongated shapes of their cavities; and (3) good solubility in both organic and aqueous media depending on their substituents [63]. Further research into Q[n] analogues could involve syntheses of higher homologues of the Q[n] analogues with more spacious cavities which could be realised through the use of longer and functionalised bis(phthalhydrazide) in their synthesis reactions.
Scheme 1.7 Synthesis of Q[n] analogues. Conditions: (a) MeSO3H, 80 °C.
Miyahara et al. expanded the Q[n] family through synthesising hemicucurbiturils that resembled cucurbiturils cut in half along the equator but exhibited different supramolecular properties (Scheme 1.8) [64].
Scheme 1.8 Synthesis of hemicucurbiturils. Conditions: (a) CH2O, HCl.
- 24 -
In contrast to the well-known macrocyclic hosts cucurbituril and hemicucurbituril, bambusuril is a relatively new family of macrocyclic compounds (Figure 1.7) which exhibits a pronounced preference to interact with various anions with high affinity and selectivity in both organic solvent mixtures and aqueous solutions [65].
Figure 1.7 Dodecamethylbambus[6]uril (Me12BU[6]).
1.4 Aims of study
To date, considerable effort has been expended in deriving functionalised Q[n] derivatives in order to extend their potential. While some progress has been made, a major limitation is the availability of higher homologues with substituted functional groups, such as SQ[7], SQ[8] and SQ[10], where S is the substituent.
The predominant motivation for this current work is to understand this limitation in order to overcome it and to derive substituted and functionalised Q[n]s using suitably substituted glycolurils as starting materials.
- 25 -
1.5 References
1. Lagona, J.; Mukhopadhyay, P.; Chakrabatri, S.; Isaacs, L. Angew. Chem.,
Int. Ed. 2005, 44, 4844–4870.
2. Day, A. I.; Collins, J. G. Supramol. Chem. Mol. Nanomater. 2012, 3, 983–
1000.
3. Mukhopadhyay, P.; Wu, A.; Isaacs, L. J. Org. Chem. 2004, 69, 6157–
6164.
4. Liu, S.; Ruspic, C.; Mukhopadhyay, P.; Chakrabarti, S.; Zavalij, P. Y.;
Isaacs, L. J. Am. Chem. Soc. 2005, 127, 15959–15967.
5. Mock, W. L.; Pierpont, J. J. Chem. Soc., Chem. Commun. 1990, 1509–
1511.
6. Zhao, J.; Kim, H.-J.; Oh, J.; Kim, S.-Y.; Lee, J. W.; Sakamoto, S.;
Yamaguchi, K.; Kim, K. Angew. Chem., Int. Ed. 2001, 40, 4233–4235.
7. Sindelar, V.; Cejas, M. A.; Raymo, F. M.; Chen, W.; Parker, S. E.; Kaifer,
A. E. Chem.–Eur. J. 2005, 11, 7054–7059.
8. Bush, M. E.; Bouley, N. D.; Urbach, A. R. J. Am. Chem. Soc. 2005, 127,
14511–14517.
9. Wagner, B. D.; Boland, P. G.; Lagona, J.; Isaacs, L. J. Phys. Chem. B
2005, 109, 7686–7691.
10. Mock, W. L.; Irra, T. A.; Wepsiec, J. P.; Adhya, M. J. Org. Chem. 1989,
54, 5302–5308.
11. Jon, S. Y.; Ko, Y. H.; Park, S. H.; Kim, H.-J.; Kim, K. Chem. Commun.
2001, 1938–1939.
- 26 -
12. Pattabiraman, M.; Natarajan, A.; Kaanumalle, L. S.; Ramamurthy, V. Org.
Lett. 2005, 7, 529–532.
13. Wang, R.; Yuan, L.; Macartney, D. H. J. Org. Chem. 2006, 71, 1237–
1239.
14. Kloeck, C.; Dsouza, R. N.; Nau, W. M. Org. Lett. 2009, 11, 2595–2598.
15. Basilio, N.; Garcia-Rio, L.; Moreira, J. A.; Pessego, M. J. Org. Chem.
2010, 75, 848–855.
16. Lu, X.; Masson, E. Org. Lett. 2010, 12, 2310–2313.
17. Pemberton, B. C.; Singh, R. K.; Johnson, A. C.; Jockusch, S.; Da Silva, J.
P.; Ugrinov, A.; Turro, N. J.; Srivastava, D. K.; Sivaguru, J. Chem.
Commun. 2011, 47, 6323–6325.
18. Tuncel, D.; Uenal, O.; Artar, M. Isr. J. Chem. 2011, 51, 525–532.
19. Nelson, D. L.; Cox, M. M. Principles of Biochemistry 4th ed. (Freeman
Publishers: New York, 2004).
20. Berbeci, L. S.; Wang, W.; Kaifer, A. E. Org. Lett. 2008, 10, 3721–3724.
21. Flinn, A.; Hough, G. C.; Stoddart, F. J.; Williams, D. J. Angew. Chem., Int.
Ed. 1992, 31, 1475–1477.
22. Sasmal, S.; Sinha, M. K.; Keinan, E. Org. Lett. 2004, 6, 1225–1228.
23. Jon, S. Y.; Selvapalam, N.; Oh, D. H.; Kang, J.-K.; Kim, S.-Y.; Jeon, Y. J.;
Lee, J. W.; Kim, K. J. Am. Chem. Soc. 2003, 125, 10186–10187.
24. Isobe, H.; Sato, S.; Nakamura, E. Org. Lett. 2002, 4, 1287–1289.
25. Zhao, Y.-J.; Xue, S.-F.; Zhu, Q.-J.; Tao, Z.; Zhang, J.-X.; Wei, Z.-B.; Long,
L.-S.; Hu, M.-L.; Xiao, H.-P.; Day, A. I. Chin. Sci. Bull. 2004, 49, 1111–
1116.
- 27 -
26. Day, A. I.; Xiao, X.; Zhang, Y.-Q.; Zhu, Q.-J.; Xue, S.-F.; Tao, Z. J. Incl.
Phenom. Macrocycl. Chem. 2011, 71, 281–286.
27. Behrend, R.; Meyer, E.; Rusche, F. Justus Liebigs Ann. Chem. 1905, 339,
1–37.
28. Freeman, W. A.; Mock, W. L.; Shih, N. Y. J. Am. Chem. Soc. 1981, 103,
7367–7368.
29. Day, A. I.; Arnold, A. P.; Blanch, R. J. (Unisearch Limited, Australia), PCT
Int. Appl. WO 2000068232, 2000 [Chem. Abstr. 2000, 133, 362775].
30. Kim, J.; Jung, I.-S.; Kim, S.-Y.; Lee, E.; Kang, J.-K.; Sakamoto, S.;
Yamaguchi, K.; Kim, K. J. Am. Chem. Soc. 2000, 122, 540–541.
31. Day, A. I.; Blanch, R. J.; Arnold, A. P.; Lorenzo, S.; Lewis, G. R.; Dance, I.
Angew. Chem., Int. Ed. 2002, 41, 275–277.
32. Day, A. I.; Arnold, A. P.; Blanch, R. J.; Snushall, B. J. Org. Chem. 2001,
66, 8094–8100.
33. Isaacs, L. Chem. Commun. 2009, 619–629.
34. Germain, P.; Letoffe, J. M.; Merlin, M. P.; Buschmann, H. J. Thermochim.
Acta 1998, 315, 87–92.
35. Lee, J. W.; Samal, S.; Selvapalam, N.; Kim, H. J.; Kim, K. Acc. Chem. Res.
2003, 36, 621–630.
36. Hwang, I.; Jeon, W. S.; Kim, H.-J.; Kim, D.; Kim, H.; Selvapalam, N.;
Fujita, N.; Shinkai, S.; Kim, K. Angew. Chem., Int. Ed. 2007, 46, 210–213.
37. Buschmann, H. J.; Cleve, E.; Schollmeyer, E. Inorg. Chim. Acta. 1992,
193, 93–97.
38. Mock, W. L.; Shih, N. Y. J. Org. Chem. 1986, 51, 4440–4446.
- 28 -
39. Marquez, C.; Hudgins, R. R.; Nau, W. M. J. Am. Chem. Soc. 2004, 126,
5806–5816.
40. Zhao, Y.-J.; Buck, D. P.; Morris, D. L.; Pourgholami, M. H.; Day, A. I.;
Collins, J. G. Org. Biomol. Chem. 2008, 6, 4509–4515.
41. Wheate, N. J.; Day, A. I.; Blanch, R. J.; Arnold, A. P.; Cullinane, C.;
Collins, G. Chem. Commun. 2004, 12, 1424–1425.
42. Saleh, N.; Koner, A. L.; Nau, W. M. Angew. Chem., Int. Ed. 2008, 47,
5398–5401.
43. Rekharsky, M. V.; Mori, T.; Yang, C.; Ko, Y. H.; Selvapalam, N.; Kim, H.;
Sobransingh, D.; Kaifer, A. E.; Liu, S.; Isaacs, L.; Chen, W.; Moghaddam,
S.; Gilson, M. K.; Kim, K.; Inoue, Y. Proc. Natl. Acad. Sci. U.S.A. 2007,
104, 20737–20742.
44. Moghaddam, S.; Yang, C.; Rekharsky, M. V.; Ko, Y. H.; Kim, K.; Inoue, Y.;
Gilson, M. K. J. Am. Chem. Soc. 2011, 133, 3570–3581.
45. Kim, H.-J.; Heo, J.; Jeon, W. S.; Lee, E.; Kim, J.; Sakamoto, S.;
Yamaguchi, K.; Kim, K. Angew. Chem., Int. Ed. 2001, 40, 1526–1529.
46. Ko, Y. H.; Kim, E.; Hwang, I.; Kim, K. Chem. Commun. 2007, 1305–1315.
47. Wang, W.; Kaifer, A. E. Angew. Chem., Int. Ed. 2006, 45, 7042–7046.
48. Appel, E. A.; Biedermann, F.; Rauwald, U.; Jones, S. T.; Zayed, J. M.;
Scherman. O. A. J. Am. Chem. Soc. 2010, 132, 14251–14260.
49. Liu, S.; Zavalij, P. Y.; Isaacs, L. J. Am. Chem. Soc. 2005, 127, 16798–
16799.
50. Liu, S.; Shukla, A. D.; Gadde, S.; Wagner, B. D.; Kaifer, A. E.; Isaacs, L.
Angew. Chem., Int. Ed. 2008, 47, 2657–2660.
- 29 -
51. Li, F.; Mulyana, Y.; Feterl, M.; Warner, J. M.; Collins, J. G.; Keene, F. R.
Dalton Trans. 2011, 40, 5032–5038.
52. Pisani, M. J.; Zhao, Y.; Wallace, L.; Woodward, C. E.; Keene, F. R.; Day,
A. I.; Collins, J. G. Dalton Trans. 2010, 39, 2078–2086.
53. Huang, W.-H.; Liu, S.; Zavalij, P. Y.; Isaacs, L. J. Am. Chem. Soc. 2006,
128, 14744–14745.
54. Li, G.-P.; Yu, F.-Q.; Lin, J.-H.; Chen, Y.-Z. Yingyong Huaxue 2010, 27,
563–566.
55. Ni, X.-L.; Lin, J.-X.; Zheng, Y.-Y.; Wu, W.-S.; Zhang, Y.-Q.; Xue, S.-F.;
Zhu, Q.-J.; Tao, Z.; Day, A. I. Cryst. Growth Des. 2008, 8 (9), 3446–3450.
56. Li, Z.-F.; Wu, F.; Zhou, F.-G.; Ni, X.-L.; Feng, X.; Xiao, X.; Zhang, Y.-Q.;
Xue, S.-F.; Zhu, Q.-J.; Lindoy, L. F.; Clegg, J. K.; Tao, Z.; Wei, G. Cryst.
Growth Des. 2010, 10, 5113–5116.
57. Wu, F.; Wu, L.-H.; Xiao, X.; Zhang, Y.-Q.; Xue, S.-F.; Tao, Z.; Day, A. I. J.
Org. Chem. 2012, 77, 606–611.
58. Day, A. I.; Arnold, A. P.; Blanch, R. J. Molecules 2003, 8, 74–84.
59. Lucas, D.; Minami, T.; Iannuzzi, G.; Cal, L.; Wittenberg, J. B.;
Anzenbacher, P. Jr.; Isaacs, L. J. Am. Chem. Soc. 2011, 133, 17966–
17976.
60. Zhao, Y. PhD thesis, Australian Defence Force Academy, The University
of New South Wales, 2009.
61. Kim, J.; Ahn, Y.; Park, K. M.; Kim, Y.; Ko, Y. H.; Oh, D. H.; Kim, K. Angew.
Chem., Int. Ed. 2007, 46, 7393–7395.
62. Park, K. M.; Lee, D.-W.; Sarkar, B.; Jung, H.; Kim, J.; Ko, Y. H.; Lee, K.
E.; Jeon, H.; Kim, K. Small 2010, 6, 1430–1441.
- 30 -
63. Lagona, J.; Fettinger, J. C.; Isaacs, L. Org. Lett. 2003, 5, 3745–3747.
64. Miyahara, Y.; Goto, K.; Oka, M.; Inazu, T. Angew. Chem., Int. Ed. 2004,
43, 5019–5022.
65. Svec, J.; Necas, M.; Sindelar, V. Angew. Chem., Int. Ed. 2010, 49, 2378–
2381.
- 31 -
CHAPTER 2
CYCLOPENTANOQ[N] FAMILY
- 32 -
CHAPTER 2
CyclopentanoQ[n] family
2.1 Introduction
Cucurbit[n]urils (Q[n]s) is a new family of supramolecular hosts which has potential for use in such diverse areas as drug delivery [1], complex self-sorting systems [2, 3], gas purification, waste-water remediation and supramolecular catalysis [4]. However, compared with the areas of other hosts, such as crown ethers, cyclodextrins and calixarenes, Q[n]’s applications are relatively limited because of their low solubility in aqueous and organic media, and the difficulty involved in functionalising them.
Figure 2.1 Fully substituted Q[n] (SQ[n]) (substituted glycoluril moiety highlighted in bold where R = ally or oxy radicals).
Substantial effort has been made to introduce substituents into a cucurbituril in order to allow it to have good solubility in common solvents and develop new supramolecular structures [4, 5]. Substituents on substituted Q[n]s during their
- 33 - development have been defined as alkyl or oxy radical functional groups which are attached to the symmetrical carbons of the junction of the cis-fused imidazolone rings of the glycoluril moieties, as shown in Figure 2.1 [6-9]. There are two types of substituted Q[n]s: fully substituted which are discussed in this chapter; and partially substituted which were discussed in the introductory chapter (Section 1.3.1, Chapter 1). Here, fully substituted Q[n]s, where n = the number of repeated substituted glycoluril moieties, are abbreviated to SQ[n].
In principle, a straightforward approach toward a SQ[n], as previously highlighted in Chapter 1, involves a condensation reaction of the substituted glycolurils with formaldehyde under acidic conditions. However, only a limited number of SQ[n]s have been synthesised in the past by this method due to various problems [6-8]. The greatest problem or limitation of this approach is that the smaller homologues consistently form as the major products when substituted glycolurils are used, as found with methylQ[n] and cyclohexanoQ[n]
(Table 2.1). An alternative method which satisfactorily introduces hydroxyl groups into the smaller homologues Q[5,6], but gives poor results for the higher ones (yield 5%) has been reported by Kim et al. (Table 2.1) [9].
- 34 -
Table 2.1 Product distributions of SQ[n] in their synthetic reactions.
MethylQ[n] CyclohexanoQ[n] [b] (HO)2nQ[n] [9] (Weight%[a]) [6, 7] (Weight%[a]) [8]
SQ[5] 98% 80% 45%
SQ[6] 2% 10% 42%
SQ[7] 0 0 5%
[a] Mol percentages were determined by 1H NMR. [b] Defined as yields.
A long-standing issue in the cucurbituril area is the preparation and isolation of the higher homologues of SQ[n] (n = 7, 8 and 10). The potential benefits from producing higher homologues of SQ[n] are the capability to: (1) encapsulate a wider range of guest molecules with a selection of cavity sizes; and (2) introduce new intrinsic properties delivered by the substituents carried, as discussed in Section 1.3.1 (Chapter 1).
We have been intrigued by the factors responsible for the strikingly high yields observed for the smaller homologues of SQ[n] (n = 5, 6). This synthetic phenomenon is especially pronounced for the preparation of fully substituted methylQ[n], as shown in Table 2.1, in which it can be seen that methylQ[5] is almost the only product formed [6, 7].
- 35 -
Figure 2.2 1,5-diaxial interactions between two adjacent substituents (R and R') on a SQ[n].
A possible explanation for this synthetic phenomenon could be the 1,5-diaxial interactions between substituent groups situated at neighbouring glycoluils in the SQ[n], as suggested by Isaacs et al. (Figure 2.2) who proposed that this type of interactions was primarily responsible for the favourable formation of the smaller homologues of SQ[n] [4].
Day et al. studied the formation of an ellipsoid cavity in both ,-(CyH)2Q[6]
(cyclohexano, CyH) and ,',,'-Me4Q[6] (methyl, Me), and suggested that the substituted glycoluril moieties favoured more acute angles of the methylene- linked ureide N than unsubstituted glycoluril [10]. These results indicate that the substituent groups can impose a subtle angle influence upon the methylene linking groups to the adjacent glycoluril and that if the adjacent glycoluril is also substituted that this will result in a sharper curve as the endo-oligomer forms.
Such a cumulative effect of substituted glycolurils on the SQ[n]-forming reaction will lead more readily to the small homologues. This raises the question as to whether there is a substituent group that will reduce the extent of the influence of substituted glycolurils upon the angles of the adjacent methylene-linked
- 36 - ureide N. One feature of the data in Table 2.1 is worthy of further comment. It should be noted that there is a small percentage increase in the proportion of cyclohexanoQ[6] compared with that of methylQ[6]. This synthetic phenomenon leads to a question – does the substituent dominate and limit the formation of higher homologues?
Table 2.2 Selected details derived from X-ray structures of substituted glycoluril diethers 1 and 2.
Angle Angle Compound (°)[a] (°)[b] 1 [11, 12] 117.09 108.88 2 [13] 113.19 109.31
[a] Defined as angle (R1–C6b–C6a). [b] Defined as dihedral angle of concave face.
Glycoluril diethers 1 and 2 in Table 2.2 are the precursors to the syntheses of the SQ[n], methylQ[n] and cyclohexanoQ[n]. Studies of the angle in 1 and 2 by X-ray crystallographic methods have found a 3.9º difference between the two glycoluril diethers while the angle in 2 is slightly larger than that in 1 (Table
2.2). It seems likely that the precursor glycolurils featuring the small angle in the substituent and the large dihedral angle of the concave face lead to the
- 37 - formation of the higher homologues. This assumption and the small percentage increase in the proportion of cyclohexanoQ[6] compared with that of methylQ[6] led us to contemplate the possibility that a tighter ring substituent, such as a cyclopentane, might improve the outcome achieved by a cyclohexano or methyl substituent. It was expected that the smaller five-membered carbon ring would subtly change the bond angles in the glycoluril structure to allow the formation of the higher homologues of Q derivatives. In this chapter, cyclopentanoQ[n] is abbreviated as CyPnQ[n] where n = the number of repeating cyclopentanoglycoluril units.
2.2 Aims of study
The aims of this section of the thesis are as follows.
1. To evaluate the hypothesis that the glycoluril bond angles influence the
synthesis of a SQ[n] where n ≥ 7, and synthesise and isolate a new
family of SQ[n], the cyclopentanoQ[n] (n = 5–7).
2. To study the characteristic binding difference between CyP6,7Q[6,7] and
Q[6,7] toward guest molecules in order to evaluate the possible
difference between them in terms of the electronic character of either
their portals or cavities being potentially influenced by the substituents.
- 38 -
2.3 Experimental
2.3.1 Materials
The starting materials were purchased from commercial sources and used without purification. Dowex 50WX2 cation-exchange resin was obtained from
Bio-Rad Laboratories.
2.3.2 Instrumental methods
Analytical thin-layer chromatography (TLC) was performed on glass plates (5 ×
1.5 cm) pre-coated (0.25 mm) with silica gel (DC-Fertigplatten Kieselgel 60 F254).
The staining of a TLC plate with iodine vapour is the method used for the visualisation of SQ[n] compounds. The column chromatography was performed with Dowex 50WX2 cation-exchange resin. NMR spectra were recorded at 400
MHz for the 1H nuclei and 100 MHz for the 13C nuclei. Chemical shifts for 1H are
13 reported with the solvent as the internal standard D2O 4.78 ppm and C with external dioxane 67.4 ppm. All NMR experiments were conducted at 25 °C unless otherwise stated. Melting points (Mp) were measured on a Digimelt apparatus in open capillary tubes and uncorrected. Infrared (IR) spectra were recorded as KBr pellets and are reported in cm-1. Elemental analyses (C, H, N) were performed by the Australian National University Microanalytical Service using a Carlo Erba 1106 Automatic analyser. Single crystal X-ray data was collected on a CCD Smart diffractometer using graphite-monochromated Mo KR radiation ( 0.71073 Å) with the scan mode. A structural solution and full
- 39 - matrix least-squares refinement based on F2 were performed using the
SHELXS-97 and SHELXL-97 program packages, respectively.
2.3.3 Experimental procedures
1,6:3,4-Bis(2-oxapropylene)tetrahydro-3a,6a-propanoimidazo[4,5-d]imidazol- 2,5(1H,3H)-dione
A mixture of paraformaldehyde (1.5 g, 50 mmol), water (3.6 mL) and HCl (37%,
2.9 mL) was heated to 50 °C. When the paraformaldehyde had completely dissolved, the solution was cooled to rt. To this clear solution was added cyclopentanoglycoluril 5 (0.91 g, 5 mmol) in one portion and the reaction mixture was stirred at rt for 18 h. The heavy white precipitate was collected by filtration, washed with ethanol and air-dried, giving a mixture (1.39 g) of 3 and
1,6:3,4-bis(2,4-dioxapentylene)tetrahydro-3a,6a-propanoimidazo[4,5-d]imidazol-
2,5(1H,3H)-dione (tetracyclic tetraether) [14]. The ratio of 3 to the tetracyclic tetraether was ~ 1:19, as determined by 1H NMR. The slow vapour diffusion of acetone into a formic acid and acetic acid solution of the product mixture gave colourless crystals of 3 (0.04 g, 3%). Alternatively, the tetracyclic tetraether
(1.27 g) was heated in a mixture of formic and acetic acids at 100 °C followed by slow cooling to rt over 1 d which gave 3 in a higher yield (0.48 g, 46%). Mp >
260 °C. IR (KBr, cm-1): 3441 w, 2986 w, 2936 m, 1748 s, 1728s, 1466 s, 1412 s,
- 40 -
1381 s, 1323 s, 1254 s, 1018 s, 937 s, 895 s, 768 s, 575 m, 424 m. 1H NMR
(DMSO-d6): 5.22 (4H, d, J = 11.6 Hz), 4.89 (4H, d, J = 11.2 Hz), 2.23 (6H, t, J
13 = 6.8 Hz), 1.87–1.80 (2H, m). C NMR (DMSO-d6): 159.1, 80.1, 71.1, 35.0,
+ 24.5, MS (ESI): m/z 289.3 (100, [M + Na] ). Anal. calcd for C11H14N4O4 (266.25):
C, 49.62; H, 5.30; N, 21.04. Found: C, 49.48; H, 5.11; N, 20.78. X-Ray crystal structure (see Appendix 1).
Typical Reaction Conditions for the Synthesis of CyPnQ[n] HCl (37%, 4 mL) was added to a mixture of cyclopentanoglycoluril 5 (1.28 g, 7 mmol), paraformaldehyde (0.448 g, 14.9 mmol) and Li2CO3 (0.08 g, 1.1 mmol). After being stirred at rt for 30 min, the mixture was heated at 50 °C for 80 min, and then heated at 90–95 °C for 3 h. The resultant light amber solution was cooled to rt and the HCl removed in vacuo. Water (10 mL) and acetone (5 mL) were added to the residue and the precipitate collected by filtration, washed with acetone and then dried to give a nearly colourless solid (1.42 g). The 1H NMR spectra of the crude reaction mixture revealed that the white solid contained
CyP5Q[5] (45%), CyP6Q[6] (38%) and CyP7Q[7] (16%), together with small amounts of higher homologues. The crude mixture (250 mg) was dissolved in a minimum volume of 50% formic acid and purified by Dowex cation-exchange resin, eluting with 1:1 1 M HCl/formic acid. The homologues eluted in the order of CyP6Q[6], CyP5Q[5] and CyP7Q[7], respectively. After the removal of the solvent in vacuo, the fractions were dried at 100 °C in vacuo (0.3 mmHg) for 12 h. Fractions containing predominantly CyP6Q[6] (99 mg) were crystallised from water with the addition of dioxane, and CyP5Q[5] (53 mg) crystallised from
- 41 - water, and CyP7Q[7] (23 mg) was crystallised from water after the addition of
+ adamantylNH3 and NH4PF6.
Cyclopentanocucurbit[5]uril
Mp > 260 °C. IR (KBr, cm-1): 3412 s, 3385 s, 1734 s, 1471 s, 1461 s, 1382 m,
1 1333 m, 1294 w, 1170 w, 933 m, 908 m, 766 w. H NMR (D2O): 5.62 (10H, d,
J = 16.0 Hz), 4.38 (10H, b s), 2.26 (20H, narrow m), 1.75 (10H, narrow m);
(alternative 0.1 M KCl D2O): 5.67 (10H, d, J = 16.0 Hz), 4.45 (10H, d, J = 16.0
13 Hz), 2.34 (20H, narrow m), 1.85 (10H, narrow m). C NMR (D2O): 156.2, 84.8,
46.5, 37.0, 23.5. MS (ESI): m/z 1069.4 (100, [M + K]+). Anal. calcd for
C45H50N20O10·8H2O (1175.14): C, 45.99; H, 5.66; N, 23.84. Found: C, 45.59; H,
5.64; N, 24.02. X-Ray crystal structure (see Appendix 1).
- 42 -
Dioxane@cyclopentanocucurbit[6]uril
Mp > 260 °C. IR (KBr, cm-1): 3468 m, 3412 m, 1750 s, 1734 s, 1469 m, 1461 m,
1 1325 m, 1292 w, 929 m, 908 w. H NMR (D2O): 5.66 (12H, d, J = 16.0 Hz),
4.32 (12H, d, J = 16.0 Hz), 2.67 (8H, s, dioxane@CyP6Q[6]), 2.30 (24H, narrow
13 m), 1.75 (12H, narrow m). C NMR (D2O): 156.6, 85.2, 67.2
2+ (dioxane@CyP6Q[6]), 46.8, 35.0, 23.6. MS (ESI): m/z 641 (100, [M + 2Na] /2)
2+ 685 (95, [M + dioxane + 2Na] /2). Anal. calcd for C54H60N24O12·C4H8O2·16H2O
(1175.14): C, 43.17; H, 6.25; N, 20.83. Found: C, 43.20; H, 6.23; N 20.65. X-
Ray crystal structure (see Appendix 1).
Adamantylammonium@cyclopentanocucurbit[7]uril Hexafluorophosphate
Mp > 260 °C. IR (KBr, cm-1): 3462 s, 3408 s, 1734 s, 1718 s, 1456 m, 1448 m,
1 1323 m, 1313 m, 1292 m, 1163 w, 932 m, 910 w, 849 m. H NMR (D2O): 5.72
(14H, d, J = 15.6 Hz), 4.37 (14H, t, J = 15.6 Hz), 2.37 (28H, narrow m), 1.81
- 43 -
(14H, narrow m); [1.36 (3H, s), 1.17 (3H, d, J = 12.0 Hz), 1.10 (6H, s), 0.79 (3H,
13 d, J = 12.0 Hz) adamantyl resonances]. C NMR (D2O): 156.9, 156.3, 85.5,
52.8, 47.14, 47.07, 41.0, 35.0, 34.2, 29.1, 23.4. MS (ESI): m/z 809 (70, [M +
+ 2+ + 2+ (adamantylNH3 ) + Na] /2), 817 (100, [M + (admantylNH3 ) + K] /2). HR-MS
+ + for C73H88N29O14 [M + (adamantylNH3 )] : calcd 1594.7066, found 1594.7072.
Modification of the Reaction Conditions for the Synthesis of CyPnQ[n] These alternative reaction conditions were the same as those in the typical procedure described above except that the reaction was carried out with the addition of 2.2 mmol of CsCl or without an alkali metal salt present. The reaction at high reactant concentrations was performed by adding HCl (37%, 0.1 mL) to a mixture of cyclopentanoglycoluril 5 (91 mg, 0.5 mmol) and paraformaldehyde
(31.5 mg, 1.05 mmol) at rt with stirring for 30 min. The reaction mixture was heated to 50 °C for 80 min and then maintained at 95 °C for 20 h. The workup procedure was the same as that described above for the typical synthesis.
Measurement of Binding Constant for Q[6] toward Dioxane The study of the absolute binding constant for Q[6] toward dioxane was performed in a 10 mM of
-4 K(C6H4CO2DCO2)/DCl (pD 4.0) buffer system. Solutions of Q[6] (1.700 × 10 M)
-3 and dioxane (2.5988 × 10 M) in D2O were prepared for this study. In each step, a small amount of dioxane (2–20 µl) was added and 1H NMR measurements were carried out at a constant temperature of 25 °C.
Job’s Method for Determination of Stoichiometry of Dioxane@Q[6] Complex A series of samples containing Q[6] and dioxane in a 10 mM of
K(C6H4CO2DCO2)/DCl (pD 4.0) buffer system were prepared and allowed to
- 44 - reach equilibrium. The molar ratios of the two reactants, dioxane and Q[6], were varied while their total concentrations were kept at 0.15 mM. The concentrations of the dioxane@Q[6] complexes were calculated through integrations of the appropriate resonances in the 1H NMR spectra. The stoichiometry was obtained from the x-coordinate at the maximum point of the Job’s curve.
Comparative Binding Studies Comparative binding studies were performed in
D2O at the specified temperatures and pD conditions. Comparative binding studies for CyP6Q[6] (0.25 mM) relative to Q[6] and a limited quantity of guest dioxane (relative initial ratio 1:2.3:0.5) were performed in a 10 mM of
K(C6H4CO2DCO2)/DCl (pD 4.0) buffer system and the whole system was kept at rt for 24 h to reach equilibrium. The comparative binding of CyP7Q[7] and Q[7]
+ - was performed by adding a Q[7] solution to (adamantylNH3 @CyP7Q[7])PF6
0.25 mM in a ratio of 1:1 (no buffer solution pD 3.5). Equilibrium was forced to conclusion at 85 °C over 12 h, as determined by 1H NMR.
Preparation of 2, 3, CyP5Q[5]·2KCl·19H2O and dioxane@CyP6Q[6]·18.5H2O
The single crystal of the tetracyclic diether 2 and 3 was obtained by the slow diffusion of acetone vapour into an aqueous solution of glycoluril 2 or 3. The solution was allowed to stand at rt and colourless crystals formed after several days. Single crystals of CyP5Q[5] and dioxane@CyP6Q[6] were prepared by dissolving CyP5Q[5] and CyP6Q[6] in an aqueous solution of KCl or a water solution with a slow infusion of dioxane. The final solution was mixed thoroughly and allowed to stand at rt. Colourless crystals were formed after several days.
- 45 -
2.4 Results and discussion
2.4.1 Structural analyses of glycoluril diethers relative to changes in substitution
We re-examined the crystal structures of glycoluril diethers 1–3 and, from this data, found that the choice of the glycoluril with a cyclopentane substituent for the formation of the higher homologues could be better supported by an alternative argument to that previously proposed by Isaacs et al. (see Section
2.1).
The collected data presented in Table 2.3 shows that a trend was found in the
N–C–N bond angles ( ) of the concave faces of the glycoluril diethers 1, 2 and
3. The structural data was specifically collected for these glycoluril diethers in preference to their parent compounds for three reasons: (1) they serve as primary synthetic intermediates for the formation of SQ[n]; (2) they incorporate the essential N alkyl group inherent in the SQ[n] structure, albeit in a six- membered, not an eight-membered, ring; and (3) because the presence of 4 unprotected N atoms on their parent glycolurils can introduce additional factors, especially H-bonded structures which, potentially, could lead to fluctuations in bond angles.
- 46 -
Table 2.3 Selected details derived from X-ray structures of substituted glycoluril diethers 1, 2 and 3.
Mean Angle plane Compound angle O7−O8 (Å) (°)[a] (°)[b] 1 [11, 12] 117.09 108.88 4.991 2 [13] 113.19 109.31 5.074 3 [15] 106.40 110.08 5.083
[a] Defined as angle (R1–C6b–C6a). [b] Defined as dihedral angle of concave face.
X-ray crystallographic structure determinations were made for glycoluril diethers
2 and 3 (see Appendix 1), while the crystal data for 1 was obtained from the literature [11, 12]. It can be seen that the dihedral angle ( ) of the concave face in each crystal exhibits a widening trend with the change in substitutent (R1,
R2 = Me 1; cyclohexano 2; to the smallest ring cyclopentano 3 (Table 2.3)).
Equally striking is the observation that the distance between the two carbonyl O atoms (O7–O8) increases from 4.991 to 5.083 Å in the following order (1 < 2 <
3). It can also be noted that there is a correlation between bond angle and dihedral angle (Table 2.3): the smaller bond angle , the larger dihedral angle
- 47 -
. This is an interesting correlation which could be attributed to the steric effect of the substituents.
Figure 2.3 Glycoluril dimer 4 showing two angles ( 1 = 113.86º and 2 = 115.16º) of its concave face.
The diether of unsubstituted glycoluril (R1, R2 = H) versus compounds 1–3 was not possible as this compound is unknown. However, in our collaborative laboratories, we previously analysed the structure of the glycoluril dimer 4, as shown in Figure 2.3. It had two angles ( 1 = 113.86º and 2 = 115.16º concave face angles, Figure 2.3) [16] and a rudimentary comparison of it with 4 indicated that the trend was achieved in the right direction. The hypothesis that there is benefit in increasing angle can be further tested by the synthesis of a
SQ[n] where n ≥ 7 can be obtained or the increased proportion of SQ[6] occurs in the reaction mixtures.
2.4.2 Synthesis of CyPnQ[n] from cyclopentanoglycoluril
Cyclopentanoglycoluril 5 was synthesised from 1,2-cyclopentadione and urea according to the known procedure reported by Gompper et al. in which the acid catalyst was anhydrous HBF4. This method provides the highest yield of 60%
- 48 -
[17] whereas an alternative method using TFA as an acid catalyst produces much lower yields [14].
+ + Scheme 2.1 Synthesis of CyPnQ[n] (n = 5–7). Conditions: (a) H , Li /H2O, rt, 30 min, 50 °C, 80 min, then 90–95 °C for 3 h.
The synthesis of CyPnQ[n] (Scheme 2.1) was evaluated using a strategy similar to that developed by our group for the synthesis of normal Q[n] [18]. The controlling factors were expected to include the acid type, acid concentration, temperature, reactant concentrations and templating ions. It was found that the condensation reaction of cyclopentanoglycoluril 5 with paraformaldehyde
(Scheme 2.1) was best performed in a 37% HCl solution in the presence of the templating metal ion Li+. Also, the process involved a three-step sequence of temperature conditions of: rt for ~ 30 min; 50 °C for ~ 80 min; and then 90–
95 °C for 3 h. The reaction mixture was always a light amber colour which indicated that there was little or no decomposition during the course of the reaction.
This version of the condensation reaction gave the best results, with CyP7Q[7] in a relatively high proportion in the reaction mixtures, as indicated in the 1H
- 49 -
NMR spectrum shown in Figure 2.4. The proportion of CyP7Q[7] was lower in the absence of the metal ion but still significant.
Figure 2.4 1H NMR spectrum of crude reaction mixtures from synthesis of + CyPnQ[n] with adamantylNH3 added to demonstrate presence of CyP7Q[7] (The addition of the guest had the added advantage of causing some resonance shifts of CyP7Q[7], as indicated by the arrows; otherwise all resonances are superimposed upon each other).
1 The H NMR spectrum of the crude reaction mixtures was obtained using D2O salt solutions as the solvent. The typical chemical shifts for the methylene proton resonances are at 4.1–4.4 and 5.4–5.7 which are in a similar region to those of normal Q[n]. Unlike mixtures of normal Q[n], the doublet sets for each homologue are not conveniently distinguishable. The host Q[n] constitutes a 1H
NMR shielding region, resulting in significant upfield chemical shifts of the 1H resonances of a cavity-encapsulated guest molecule. This feature can be used to identify the existence and proportion of each homologue within a product mixture by the addition of a molecular probe (guest) that can selectively form a complex with a particular homologue. As dioxane, adamantylammonium salt and dimethylphenanthroline salt each have the property of being encapsulated
- 50 - in the cavities of Q[6], Q[7] and Q[8], respectively, they have been considered suitable as probes for identifying Q[6]-, Q[7]- and Q[8]-sized cavities in the product mixture of CyPnQ[n] [3, 19, 20]. In employing these molecular probes to determine the proportion of a particular homologue in the product mixtures, an excess of the probe molecule was added to some individual samples to ensure full representation. By comparing the integrals of the bound guest resonances with those of the total CyPnQ[n] methylene proton resonances, and assuming that the ratio of the binding guest to a CyPnQ[n] was comparable to a normal
Q[n], we could roughly calculate the proportions of the higher homologues in the
CyPnQ[n] mixture. Repeating the reaction several times under typical conditions for the synthesis of CyPnQ[n] (Section 2.3.3) gave the proportions of CyP6Q[6] and CyP7Q[7] as > 30 and > 15 weight% of the crude reaction mixture each time. After each CyPnQ[n] was isolated and purified, the binding ratio assumptions were either verified or corrected.
Recrystallisation from HCl is generally performed to separate normal Q[5] and
Q[6] with very high purities from the crude reaction mixtures. A combination of chromatography for the enrichment of one homologue, followed by crystallisation for the final purification can be used to isolate CyP5Q[5] and
CyP6Q[6] in a crude reaction mixture. The enrichment of CyP5Q[5] and CyP6Q[6] was obtained by Dowex cation-exchange resin chromatography, eluting with 1:1
1 M HCl/formic acid. These fractions were generally two homologues one of which was enriched to ~ 80%. Each homologue was then purified by crystallisation. The addition of dioxane to the CyP6Q[6] as an aqueous solution produced crystals of the dioxane@CyP6Q[6] complex. Normally, when Q[n] is
- 51 - purified by the same chromatography method, the order of elution is Q[5], Q[6] and Q[7]. However, the elutions for the CyPnQ[n] mixture were CyP6Q[6],
CyP5Q[5] and CyP7Q[7], respectively.
A cucurbituril (and its derivatives) has a spheroidal hydrophobic cavity and an electronegative rim at each of its two portals. It has been found that, for an alkyl-substituted cucurbit[5]uril, the affinity of the carbonyl oxygens of the substituted glycoluril moieties for potassium ion is enhanced primarily due to the electron-donating effect of the alkyl substituents [21]. We suggest that the cause of a higher affinity of CyP5Q[5] for the cation is probably due to its substituent effects as well as its suitable portal size.
Both CyP5Q[5] and dioxane@CyP6Q[6] have been thoroughly characterised by
1 13 H and C NMR and ESM spectroscopy. The significant feature of CyP5,6Q[5,6] is their symmetries, which result in relatively simply 1H NMR spectra (see
Appendix 2). Their methylene proton resonances appear as doublets in the regions of 5.6–5.7 and 4.3–4.4, respectively. The proton resonances in the regions of 2.2–2.3 and 1.7–1.9 with a ratio of 2:1 are consistent with the methylene protons of the cyclopentane ring. Analyses by ESMS have further supported these structures with molecular ions, such as m/z 1069 [CyP5Q[5] +
+ 2+ K] ; m/z 641.3 and 685.3 consistent with ([CyP6Q[6] + 2Na] /2) and ([CyP6Q[6]
+ dioxane + 2Na]2+/2), respectively.
For the dioxane@CyP6Q[6] complex, we observed that the proton resonance of a bound dioxane was shifted upfield by 1.03 ppm relative to that of a free dioxane. In addition, the binding ratio of 1:0.88 for CyP6Q[6]/dioxane was
- 52 - determined when an excess of dioxane (10 mol equiv.) was added to a pure
CyP6Q[6] sample in D2O.
We also obtained structural information for CyP5Q[5]·KCl and dioxane@CyP6Q[6] by single crystal X-ray diffraction studies. Both presented the classical spheroid cavity of a Q[5] or Q[6]. Table 2.4 presents data for the dimensions of CyP5,6Q[5,6] and Q[5,6] for comparison from which it can be seen that there is no, or only a slight, difference between CyP5,6Q[5,6] and normal Q[5,6] in terms of the dimensions of both their portals and cavities.
[a] Table 2.4 Comparison of dimensions of CyP5,6Q[5,6] and normal Q[5,6] derived from X-ray crystal structures.
Portal O−O Cavity C−C Depth Q[n] Av. Dia. Å Av. Dia. Å Å
CyP5Q[5]·KCl 2.03 5.23 9.07
Q[5]·KI [22] 2.06 5.30 9.10
Dioxane@CyP6Q[6] 3.94 6.89 9.11
THF@Q[6]·NaCl [23] 3.90 6.80 9.13
[a] Dimensions include van der Waals radii.
- 53 -
Figure 2.5 X-ray crystal structures of potassium ion-coordinated ring structure of CyP5Q[5]·KCl (left) and dioxane@CyP6Q[6] (right) (water molecules omitted for clarity).
One interesting aspect of the packing of CyP5Q[5]·KCl in the crystal is the observation of a six-membered beaded-ring structure (Figure 2.5 left) which is similar to previously reported structures of partially substituted Q[5] with K+ or
Sr2+ and solid-state templated Q[5]·KI [10, 21, 22, 24]. However, the difference in the case of CyP5Q[5]·KCl is that there is virtually no channel formed at the core of the six-membered ring due to the cyclopentane ring substituents of
CyP5Q[5] which occupy the channel space that previously existed in other Q[5] structures.
CyP7Q[7] was purified by a method which took advantage of the selective binding of its cavity, as indicated in Scheme 2.2, and has been proven to be
+ most effective in isolating CyP7Q[7]. AdamantylNH3 was added to crude reaction mixtures which were then purified by chromatography. Dowex cation- exchange resin (with 1:1 1 M HCl/formic acid as the eluent) was used for the
+ separation of the adamantylNH3 @CyP7Q[7] complex from most of the other
- 54 -
+ components to achieve > 75% purity. AdamantylNH3 is a very effective guest for Q[7]-sized cavities with relatively high binding constants [3]. The
+ adamantylNH3 @CyP7Q[7] cation has a higher affinity toward the cation-
+ exchange resin. The separation of adamantylNH3 @CyP7Q[7] and uncharged
CyP5,6Q[5,6] can be accomplished by eluting the cation-exchange resin with 50%
+ formic acid. Then, the adamantylNH3 @CyP7Q[7] complex releases from the resin can be initiated by an influx of a competing cation (0.5 M HCl) which
+ displaces the bound adamantylNH3 @CyP7Q[7] from the cationic binding sites.
Experimentally, the major difficulty in the purification of
+ adamantylNH3 @CyP7Q[7] was its separation from CyP5Q[5] as their solubilities in water are very similar and, usually, they co-crystallise or drop out of the
+ solution together. Following chromatography, the adamantylNH3 @CyP7Q[7] complex can be further purified by the addition of NH4PF6 salt to give crystals of
+ the adamantylNH3 @CyP7Q[7]·PF6 complex which has low solubility in an
- aqueous solution. PF4 salts often have organic solvent solubility. This
+ adamantylNH3 @CyP7Q[7]·PF6 salt was found to be soluble in CH3CN.
- 55 -
+ Scheme 2.2 Purification procedure for adamantylNH3 @CyP7Q[7].
Following chromatography, CyP7Q[7] was isolated as an
+ adamantylNH3 @CyP7Q[7]·PF6 complex, which was confirmed by ESMS, giving three characteristic ions at m/z 1595, 809 and 817
+ + + 2+ ([adamantylNH3 @CyP7Q[7]] , [adamantylNH3 @CyP7Q[7] + Na] /2 and
+ 2+ 1 [adamantylNH3 @CyP7Q[7] + K] /2, respectively). One aspect of the H and
13 + C NMR spectra of adamantylNH3 @CyP7Q[7] is noteworthy: the portal methylene proton resonances in 1H NMR were found at 5.72 as a doublet and at 4.37 as a pseudotriplet (two overlapping doublets) (see Appendix 2). This upfield doublet pair arose from the magnetic non-equivalence between the two
+ 13 portals due to the presence of the NH3 ion in only one of them. The C NMR spectrum which had a small difference in the C=O (156.86, 156.32) and the portal CH2 (47.14, 47.07) resonances also indicated the existence of two magnetically non-equivalent portals.
- 56 -
2.4.3 Modification of reaction conditions for the synthesis
of CyPnQ[n]
In 2001, Day et al. identified some important factors that determine the relative proportions of different homologues in the synthetic reaction of cucurbituril [18].
The study presented here took advantage of these optimising factors in the mediation of the synthesis of CyPnQ[n] in order to obtain increased proportions of the higher homologues in product mixtures. The reaction conditions shown in
Table 2.5 are the same as those described in Section 2.3.3. It was found that the reaction at relatively high reactant concentrations could significantly change the product distribution, discriminating against CyP5Q[5] and producing higher
1 proportions of CyP6,7Q[6,7]. H NMR analyses of the reaction at the reactant concentration of 910 mg/mL showed the proportion of each homologue (Table
2.5) plus oligomer (~ 40%) indicating incomplete reaction. This result is similar to that of the synthesis of normal Q[10] with the same reaction conditions except that the Q[10] reaction was complete [18]. Extending the reaction time did not improve the outcome as the oligomer still remained.
+ Attempts to increase the proportion of CyP7Q[7] using Li salt as a template at a relatively high concentration (320 mg/mL) were successful. This synthetic condition increases the proportion of CyP7Q[7] at the expense of CyP5Q[5], as shown in Table 2.5. The mechanism for this Li+-templated reaction remains unknown. It has been suggested that a possible correlation exists between the radii of Li+ and the preferential formation of particular intermediate oligomers, such as endo-trimers and tetramers [25], which may condense with each other to form higher proportions of CyP7Q[7].
- 57 -
Table 2.5 Homologue distribution of synthesis of CyPnQ[n] with two templating alkali metal ions at different reactant concentrations.
[Glycoluril] Template Weight%[a]
+ mg/mL M [d] CyP5Q[5] CyP6Q[6] CyP7Q[7] CyP8Q[8] HCl[b] 160 None 80 20 0 0
320 None 60 32 8 0
320 Li+ 45 38 16 1[f]
320 Cs+ 49 37 12 -
910[c] none -[e] 23 10 -
1 [a] Weight% were determined by H NMR in D2O for total CyPnQ[n] mixture by comparing relative ratios of integrals of bound probe molecule, dioxane for
CyP6Q[6] and adamantylamine salt for CyP7Q[7], and CyPnQ[n] methylene proton resonances. [b] HCl was 37% in all reactions. [c] Reaction times, rt for 30 min; 50 °C for 80 min; then 90–95 °C for 20 h. Reaction was incomplete. [d]
Proportion of CyP5Q[5] was calculated as remaining material. [e] Proportion of
CyP5Q[5] cannot be determined due to incomplete reaction. [f] Indicated but not conclusive.
The protonated 2,9-dimethylphenanthroline (neocuprineH+) is an effective probe molecule for normal Q[8] [20]. Using it, we could potentially determine the presence or absence of CyP8Q[8] within the reaction product mixtures (Table
2.5). Unfortunately, although there were weak resonances observed after the addition of neocuprineH+ that were consistent with this guest being encapsulated inside the cavity of CyP8Q[8], we were unable to be confident of the presence of CyP8Q[8] in the product mixtures.
- 58 -
2.4.4 Guest selectivity between CyPnQ[n] and Q[n]
In their pioneering work, Mock et al. indicated the importance of both ion–dipole attraction and size and shape complementarity in the formation of Q[6] complexes [26]. In this section, we discussed how we used 1H NMR competitive binding experiments to compare the affinity difference between CyP6,7Q[6,7]
+ and normal Q[6,7] toward dioxane or adamantylNH3 guests in aqueous solutions. It was found that the CyP6,7Q[6,7] exhibited a slightly higher binding affinity toward those guest molecules than did the normal Q[6,7].
2.4.4.1 Dioxane selectivity between CyP6Q[6] and Q[6]
We studied the binding constant of normal Q[6] toward dioxane by 1H NMR titration experiments, and then employed 1H NMR competition experiments to compare the complexation behaviours of CyP6Q[6] and Q[6] toward dioxane in aqueous solutions.
As normal Q[6] experiences low solubility in neutral water and requires alkali metal ions for moderate solubility, 10 mM of a K(C6H4CO2DCO2)/DCl (pD 4.0) buffer system was used in our experiments. A constant pH is important for obtaining reliable comparative data but we are also aware of the potential competitive effects of metal ions or the binding of cavity-bound guests. In their early work, Werner et al. demonstrated that the presence of alkali metal ions had an effect on the binding behaviour of the host Q[6] toward guest molecules in an aqueous solution [27]. Therefore, we also evaluated this effect in relation
- 59 - to CyP6Q[6], the binding affinity of which toward dioxane also decreased with increasing Na+ concentrations, as shown in Figure 2.6.
1 Figure 2.6 H NMR spectra in D2O at 25 °C of dioxane@CyP6Q[6] with increasing amounts of Na2SO4 (initial concentrations of CyP6Q[6] and dioxane,
0.44 mM). (a) [Na2SO4] = 0 mM, [dioxane@CyP6Q[6]] = 0.36 mM; (b) [Na2SO4]
= 3 mM, [dioxane@CyP6Q[6]] = 0.27 mM; (c) [Na2SO4] = 5 mM,
[dioxane@CyP6Q[6]] = 0.23 mM.
It was found that the binding ratios for dioxane/CyP6Q[6] also decreased, as shown in Figure 2.6, from 0.8 (no Na2SO4) to 0.6 (3 mM of Na2SO4) to 0.5 (5 mM of Na2SO4). In a 10 mM of K(C6H4CO2DCO2)/DCl (pD 4.0) buffer system, the binding ratio for dioxane/CyP6Q[6] is 0.48.
Dioxane forms a relatively weak inclusion complex with Q[6] in a 10 mM of
K(C6H4CO2DCO2)/DCl (pD 4.0) buffer system and exhibits slow exchange kinetics between free and bound dioxane on the 1H NMR time scale. Therefore, the absolute Ka of the dioxane@Q[6] complex can be determined accurately by
1H NMR titration measurements.
- 60 -
The Job’s method of continuous variation was employed to determine the stoichiometry of the dioxane@Q[6] complex formed. The experimental conditions are described in Section 2.3.3. The maximum of the complex product appears at x = 0.5 in Figure 2.7 which indicates that a 1:1 complexation is predominant at the equilibrium.
0.03 (mM)
0.02
0.01
0 [Dioxane@Q[6]] 0 0.2 0.4 0.6 0.8 1
[Q[6]]0/([Q[6]]0+[Dioxane]0)
Figure 2.7 Job’s plot of dioxane–Q[6] interaction.
The binding constant Ka was obtained by plotting the [Q[6]] × [dioxane] versus the [dioxane@Q[6]] (Figure 2.8), where [Q[6]], [dioxane] and [dioxane@Q[6]] are the concentrations of free Q[6], free dioxane and the dioxane@Q[6] complex, respectively. The slope of the solid line in Figure 2.8 equals the
-1 binding constant of dioxane@Q[6] (Ka = 2783 M ).
- 61 -
0.1 0.08 y = 2.783x + 0.010 R² = 0.973
M) 0.06
3 ‐
10 0.04 × ( 0.02 [Dioxane@Q[6]] 0 0 0.005 0.01 0.015 0.02 0.025 0.03 [Q[6]]×[Dioxane] (× 10‐6 M2)
Figure 2.8 Plot of [dioxane@Q[6]] as function of [Q[6]] x [dioxane] in which [dioxane@Q[6]], [Q[6]] and [dioxane] are concentrations of dioxane@Q[6], free -1 Q[6] and free dioxane, respectively (Ka = 2783 M ).
We then resorted to the competition experiments based on 1H NMR spectroscopy introduced by Mock et al. in their pioneering work on Q[6] [26].
The experiments were performed in a 10 mM buffer system of
K(C6H4CO2DCO2)/DCl (pD 4.0) and their detailed procedures are described in
Section 2.3.3.
- 62 -
The thermodynamics of the host–guest system and competition experiments are defined by Equations 1–3. Substitutions of the various concentrations determined by a NMR integration into Equation 4 produced a value of Krel. The
Krel of CyP6Q[6] for dioxane between CyP6Q[6] and Q[6] was found to be Krel =
1.8 (KCyP6Q[6]/KQ[6]). The Krel value was then referenced to the absolute KQ[6] measured by the 1H NMR titration experiment using Equation 5.
2.4.4.2 Comparative affinity of CyP7Q[7] and Q[7] for + adamantylNH3 (Ada)
+ The binding affinity of CyP7Q[7] toward the adamantylNH3 (Ada) guest was also found to be higher than that of Q[7]. A competitive binding experiment for
- Ada was conducted in which a Q[7] solution was added to (Ada@CyP7Q[7])PF6
0.25 mM in a ratio of 1:1 (no buffer solution pD 3.5). The exchange reaction was brought to equilibrium at 85 °C over 12 h, as determined by 1H NMR (Scheme
2.3). The same experiment at rt progressed to < 12% toward equilibrium after
45 days! From the completed equilibrium experiment, the relative binding constant for CyP7Q[7] was determined to be KCyP7Q[7] = 1.9 KQ[7].
Scheme 2.3 Competitive binding reaction between CyP7Q[7] and Q[7] for + adamantylNH3 .
A possible reason for the higher affinity of CyP6,7Q[6,7] toward dioxane or
+ adamantylNH3 could come from the difference in the dimensions of the portal and cavity between the two host families but from X-ray crystal structure data
- 63 - for CyP5Q[5] and CyP6Q[6], this is not substantiated and therefore should not be relevant to the difference in binding constant. An alternative consideration could come from the substitution, which increases the electron density on the carbonyl oxygens of the substituted glycoluril moiety, thereby apparently increasing the ion–dipole interactions. The slightly higher affinity observed for
+ the CyP7Q[7] binding to the adamantylNH3 guest in this study could reflect these increased cation–dipole interactions. The suggested increased electronegativity at the portal was also reinforced by the reported Mulliken atomic charge calculations and solid-state metal coordinated structures of partially substituted Q[5] [21, 24]. Establishing the reason for the higher binding constant of CyP6Q[6] toward dioxane is not that straightforward as there are no cation–dipole interactions in this complexation process. In the case of this uncharged guest (dioxane), the H2O molecule associated with the portal after complexation through hydrogen bonds could be responsible for the increased stability. We suggest that the increased electronegativity at the portal described above also increases the strengths of hydrogen bonds of this type, thereby promoting binding. In addition, incidental hydrogen bonding to the dioxane O may also increase binding stability. To investigate this, particular molecular guests without electronegative atoms would need to be further investigated.
However, this was beyond the scope of the current study.
- 64 -
2.5 Conclusions
In this chapter, the synthesis of the first family of fully substituted cucurbit[n]uril, cyclopentanoQ[n] has been discussed. The members of the family synthesised, where n = 5–7, have been fully characterised and their higher binding affinities
+ identified for dioxane in CyP6Q[6] and adamantylNH3 in CyP7Q[7].
It has been demonstrated that the dihedral angle of the concave face of the starting glycoluril is important for determining whether higher homologues will form. In comparison to the ethers of cyclohexanoglycoluril and cyclopentanoglycoluril, the dimethylglycoluril ether has the sharpest angle on the concave face as a result of the intrinsic steric repulsion between the two methyl substituents. Introducting a six-membered carbon ring, as in cyclohexanoglycoluril, relieves such steric repulsion and the concave face angle is widened while introducting a five-membered carbon ring, as in cyclopentanoglycoluril, widens it further. The consequence is that the increased curvature of the growing intermediate oligomers formed under acid-catalysed conditions increases the preference for the higher SQ[n] homologues at the expense of SQ[5] [18]. The synthesis of CyPnQ[n] can also be manipulated to a small degree by the addition of templating alkali metal ions to reaction mixtures.
+ The addition of Li produced a relatively high proportion of CyP7Q[7] up to a 16 weight%.
The cucurbit[n]urils (n ≥ 6) with much more spacious cavities are of prime interest as these homologues offer greater opportunities for supramolecular chemistry. In particular, the new cyclopentanoQ[n] developed in this research
- 65 - suggests it has important potential to increase binding constants, as demonstrated for the neutral guest dioxane and the ionic guest adamantyl ammonium ion.
- 66 -
2.6 References
1. Day A. I.; Collins J. G. Supramol. Chem. Mol. Nanomater. 2012, 3, 983–
1000.
2. Mukhopadhyay, P.; Wu, A.; Isaacs, L. J. Org. Chem. 2004, 69, 6157–
6164.
3. Liu, S.; Ruspic, C.; Mukhopadhyay, P.; Chakrabarti, S.; Zavalij, P. Y.;
Isaacs, L. J. Am. Chem. Soc. 2005, 127, 15959–15967.
4. Lagona, J.; Mukhopadhyay, P.; Chakrabatri, S.; Isaacs, L. Angew. Chem.,
Int. Ed. 2005, 44, 4844–4870.
5. Masson, E.; Ling, X.; Joseph, R.; Kyeremeh-Mensah, L.; Lu, X. RSC Adv.
2012, 2, 1213–1247.
6. Flinn, A.; Hough, G. C.; Stoddart, F. J.; Williams, D. J. Angew. Chem., Int.
Ed. 1992, 31, 1475–1477.
7. Sasmal, S.; Sinha, M. K.; Keinan, E. Org. Lett. 2004, 6, 1225–1228.
8. Zhao, J.; Kim, H.-J.; Oh, J.; Kim, S.-Y.; Lee, J. W.; Sakamoto, S.;
Yamaguchi, K.; Kim, K. Angew. Chem., Int. Ed. 2001, 40, 4233–4235.
9. Jon, S. Y.; Selvapalam, N.; Oh, D. H.; Kang, J.-K.; Kim, S.-Y.; Jeon, Y. J.;
Lee, J. W.; Kim, K. J. Am. Chem. Soc. 2003, 125, 10186–10187.
10. Zheng, L.-M.; Zhu, J.-N.; Zhang, Y.-Q.; Zhu, Q.-J.; Xue, S.-F.; Tao, Z.;
Zhang, J.-X.; Zhou, X.; Wei, Z.-B.; Long, L.-S.; Day, A. I. Supramol.
Chem. 2008, 20 (8), 709–716.
11. Schouten, A.; Kanters, J. A. Acta Crystallogr., Sect. C 1990, C46, 2484–
2486.
- 67 -
12. Wagner, B. D.; Boland, P. G.; Lagona, J.; Isaacs, L. J. Phys. Chem. B
2005, 109, 7686–7691.
13. Wu, F. Master Thesis, Guizhou University, 2005.
14. Wu, L.-H.; Ni, X.-H.; Wu, F.; Zhang, Y.-Q.; Zhu, Q.-J.; Xue, S.-F.; Tao, Z.
J. Mol. Struct. 2009, 920, 183–188.
15. Wu, F.; Wu, L.-H.; Xiao, X.; Zhang, Y.-Q.; Xue, S.-F.; Tao, Z.; Day, A. I. J.
Org. Chem. 2012, 77, 606–611.
16. Ma, P.-H.; Xiao, X.; Zhang, Y.-Q.; Xue, S.-F.; Tao, Z. Acta Crystallogr.,
Sect. E 2008, E64, o1795.
17. Gompper, R.; Noth, H.; Rattay, W.; Schwarzensteiner, M.-L.; Spes, P.;
Wagne, H.-U. Angew. Chem., Int. Ed. 1987, 26, 1039–1041.
18. Day, A. I.; Arnold, A. P.; Blanch, R. J.; Snushall, B. J. Org. Chem. 2001,
66, 8094–8100.
19. Day, A. I.; Arnold, A. P.; Blanch, R. J. Molecules 2003, 8, 74–84.
20. Fu, H.; Xue, S.-F.; Mu, L.; Du, Y.; Zhu, Q.-J.; Tao, Z.; Zhang, J.-X.; Day,
A. I. Sci. China, Ser. B 2005, 48, 305–314.
21. Li, Z.-F.; Wu, F.; Zhou, F.-G.; Ni, X.-L.; Feng, X.; Xiao, X.; Zhang, Y.-Q.;
Xue, S.-F.; Zhu, Q.-J.; Lindoy, L. F.; Clegg, J. K.; Tao, Z.; Wei, G. Cryst.
Growth Des. 2010, 10, 5113–5116.
22. Feng, X.; Chen, K.; Zhang, Y.-Q.; Xue, S.-F.; Zhu, Q.-J.; Tao, Z.; Day A. I.
Cryst. Eng. Comm. 2011, 13, 5049–5051.
23. Jeon, Y.-M.; Kim, J.; Whang, D.; Kim, K. J. Am. Chem. Soc. 1996, 118,
9790–9791.
24. Ni, X.-L.; Lin, J.-X.; Zheng, Y.-Y.; Wu, W.-S.; Zhang, Y.-Q.; Xue, S.-F.;
Zhu, Q.-J.; Tao, Z.; Day A. I. Cryst. Crowth Des. 2008, 8 (9), 3446–3450.
- 68 -
25. Day, A. I.; Blanch, R. J.; Coe, A.; Arnold, A. P. J. Incl. Phenom.
Macrocycl. Chem. 2002, 43, 247–250.
26. Mock, W. L.; Shih, N. Y. J. Org. Chem. 1986, 51, 4440–4446.
27. Marquez, C.; Hudgins, R. R.; Nau, W. M. J. Am. Chem. Soc. 2004, 126,
5806–5816.
- 69 -
CHAPTER 3
FIVE-MEMBERED CARBON RING GLYCOLURIL DERIVATIVES WITH FUNCTIONALITY
- 70 -
CHAPTER 3
Five-membered carbon ring glycoluril derivatives with functionality
3.1 Introduction
Due to cucurbit[n]urils’ (Q[n]s) high affinity toward guest molecules and highly selective molecular binding in aqueous media, supramolecular chemists’ interest in them has increased dramatically [1-6]. Applications of Q[n]s based on these properties have been described in a recent review article entitled
“Cucurbituril chemistry: a tale of supramolecular success” by Masson el al. [7].
As discussed in Chapter 1 of this thesis, other areas of application for Q[n]s include drug delivery systems [8], gas purification [9-11] and supramolecular catalysis [12-20]. However, adopting a long-term view, their applications could potentially be extended by taking advantage of Q[n] derivatives with functionality which could be further modified and functionalised for use in syntheses of more complex supramolecular systems.
As a preliminary example of functionalised Q, Kim et al. were the first to achieve the synthesis of perhydroxyQ[n] by the direct oxidation of Q[5–8] (Scheme 3.1)
[21]. However, this method has some limitations. It favours introducing hydroxyl groups into the smaller homologues (Q[5,6], yielding 42% and 45%, respectively) but reports obtaining only very low yields for the higher homologues (ca. 5%) [21]. The introduction of hydroxyl groups provides the
- 71 - opportunity to further modify functionality to achieve a number of Q[5,6] derivatives (Scheme 3.1) [21]. However, only with partial hydroxylation have derivatives of Q[7] been possible [22].
Scheme 3.1 Direct functionalisation of Q[n] (n = 5–8) and examples of further functionalisation by some conventional methods. The figure was taken from [21].
Our research group has been interested in the syntheses of the higher homologues of substituted Q[n] which, until the recent successful synthesis of cyclopentanoQ[5–7] (Chapter 2), had mostly only been possible for partially substituted Q[n] (Section 1.3.1, Chapter 1). However, our greater goal has been to achieve higher homologues with substituents that are also functionalised. Our approach to cyclopentanoQ could potentially be extended to the preparation of higher homologues of functionalised Q by using a glycoluril derivative carrying a functional group on a five-membered carbon ring substituent. The synthetic
- 72 - conditions for functionalised Q were expected to be similar to those developed for the synthesis of cyclopentanoQ[n] but be able to yield higher homologues of functionalised Q. This line of inquiry led us to prepare a number of glycoluril derivatives with a functionalised five-membered carbon ring substituent.
In this chapter, the substantial effort that has been devoted to the synthesis of glycolurils carrying a functionalised five-membered carbon ring substitutent is discussed. This type of functionalised glycoluril is potentially a credible alternative to the perhydroxylation method developed by Kim et al. [21], especially as syntheses of the higher homologues are possible with five- membered carbon ring substituents, as previously highlighted in Chapter 2.
Scheme 3.2 Generalised scheme for synthesis of glycolurils.
Glycolurils were first synthesised more than a century ago and the method used since has remained essentially the same [23, 24]. Generally, it involves acid- catalysed urea condensation on -diones or -keto-aldehydes with different solvents at different temperatures. It is generally accepted that the synthesis of glycoluril B involves the reaction intermediate diol A as the first condensation product in a two-step reaction sequence, as presented in Scheme 3.2 [23, 25,
26]. In the presence of 2 mole equivalents of urea, the reaction intermediate diol
- 73 -
A can proceed to glycoluril B (step b) through the dehydration and condensation of a second molecule of urea under conditions in which an acid catalyst is present. Our interest is in synthesising B where R1 = R2 = (CH2)3
(five-membered carbon ring) carries functionality (Scheme 3.2). Through this approach, we take advantage of the five-membered carbon ring substituent but also include functionality in order to achieve our goal.
In this chapter, we report the successful syntheses of a number of suitably functionalised glycolurils. We also examine in detail examples of some original intermediates found on the pathway to glycoluril formation, and identify the products that occur in either step a or b of the glycoluril synthesis (Scheme 3.2) which can prevent the efficient formation of B. These results are discussed in more detail in Section 3.4.
3.2 Aims of study
The aims of this section of the thesis are as follows.
1. To synthesise a number of glycoluril derivatives based on a five-
membered carbon ring substituent carrying a functional group. It was
anticipated that these glycolurils would provide useful building blocks for
the preparation of Q[n] derivatives containing reactive functional groups.
2. To gain an insight into possible intermediates of the synthesis of
glycolurils and their derivatives in order to better understand the reasons
for the unavailability of many alkyl glycolurils or to determine why their
yields are sometimes poor.
- 74 -
3.3 Experimental
3.3.1 Materials
The starting materials, such as triethyl propane-1,2,3-tricarboxylate, diethyl oxalate, cis-2-butene-1,4-diol, camphorquinone, epimeric mixture of endo (~
90%) 5-norbornen-2-yl acetate, norbornylene, 3-methyl-1,2-cyclopentadione and diethyl diallylmalonate were purchased from Sigma-Aldrich company and used without purification. An OsO4 solution (4 wt.% in H2O), N- methylmorpholine N-oxide, trifluoroacetic anhydride, benzoyl chloride, borane tetrahydrofuran complex solution (1 M in THF), hydrogen peroxide (30 wt.% in
® H2O), TFA, trifluoroacetic anhydride (TFAA), urea, Oxone , 3-phenyl-1H-inden-
1-ylidene[bis(1-butyl-phoban)] ruthenium (Ⅱ) dichloride (Ru-catalyst) and
Grubbs’ Catalyst 2nd Generation were purchased from Johnson Matthey
Company. DMSO, CH2Cl2, Et3N, THF and pyridine were dried and freshly distilled before use. The hexane, toluene and EtOAc used in chromatography were also distilled prior to use. Anhydrous MgSO4 was used as a common drying agent.
3.3.2 Instrumental methods
Analytical thin-layer chromatography (TLC) was performed on glass plates (5 ×
1.5 cm) pre-coated (0.25 mm) with silica gel (DC-Fertigplatten Kieselgel 60 F254).
Compounds were visualised by exposure to UV light or by staining with H2SO4-
EtOAc Vanillin followed by heating on a hot plate. Flash column chromatography was performed using silica gel (60, 230–400 mesh) with
- 75 - eluents in the indicated v:v ratio. NMR spectra were recorded at 400 MHz for the 1H nuclei and 100 MHz for the 13C nuclei. Chemical shifts were reported with the solvent as the internal standard (CDCl3 7.24 and 77.2 ppm, DMSO
2.50 and 39.5 ppm for 1H and 13C, respectively). Phase-sensitive NOSEY spectra were acquired using 2048 data points in t2 for 256 t1 values with a pulse repetition delay of 1.7 s for mixing times of 350 ms. COSY experiments were conducted using 2048 data points in t2 (with a spectral width of 4200 Hz) for 256 t1 values with a pulse repetition delay of 1.7 s with 16–128 scans per FID. All
NMR experiments were conducted at 25 °C unless otherwise stated. Melting points (Mp) were measured on a Digimelt apparatus in open capillary tubes and left uncorrected. Infrared (IR) spectra were recorded as KBr pellets or thin films on NaCl plates and were reported in cm-1. Elemental analyses (C, H, N) were performed by the Australian National University’s Microanalytical Service using a Carlo Erba 1106 automatic analyser.
3.3.3 Synthetic procedures
General Procedure for OsO4 Oxidation A solution of alkene (2 mmol) in 4:1 acetone/water (5 mL) was heated to 40 °C. N-methylmorpholine N-oxide (2.2 mmol) was added to the above solution and the reaction mixture was stirred for
5 min. OsO4 (0.06 mL) was added and the reaction continued at 40 °C for over
18 h, as monitored by TLC. After the removal of acetone in vacuo, the remaining aqueous solution was extracted with EtOAc and the combined organic extracts were dried over MgSO4. The EtOAc was removed in vacuo and the crude residue was purified by chromatography to yield a pure diol product.
- 76 -
General Procedure for Swern Oxidation Dry DMSO (3.5 mL) was added very slowly to a cooled (-78 °C) solution of trifluoroacetic anhydride (44 mmol) in anh.
CH2Cl2 (25 mL), and the reaction mixture was stirred for 10 min. Then, diol (15 mmol) in anh. CH2Cl2 (8 mL) was added and the reaction continued at -78 °C for 5.5 h. After cooling to -100 °C, anh. Et3N (12 mL) was added and the solution was stirred at -78 °C for a further 3 h. The reaction was allowed to reach 0 °C in an ice-water bath, was acidified with 3 M HCl and extracted with
CH2Cl2. The combined organic extracts were washed with brine and dried over
MgSO4. The solvent was removed in vacuo and the crude residue was purified by chromatography to yield a pure dione product.
General Procedure for Urea Condensation Reactions In a 50-mL flask, equipped with a condenser and a Dean-Stark trap for water removal, a slurry of urea (15 mmol), dicarbonyl compound (5 mmol) and TFA (0.2 mL) in 20 mL of benzene (toluene or xylenes where specified) was placed. After being stirred at rt for 40 min, the mixture was heated at reflux for 4 h. The resultant light amber solution was cooled to rt and the solvent thoroughly decanted. A minimum volume of EtOH was added to the residue and the solid suspension was collected by filtration, washed with CH3COCH3 and then dried under high vacuum to give a light yellow solid product.
- 77 -
Triethyl tricarballylate
This was a modification of the known procedure reported by Anthonsen et al.
[27]. In a 100-mL round-bottomed flask, a mixture of 1,2,3-propane-tricarboxylic acid (5 g, 28 mmol), absolute EtOH (40 g, 0.87 mol) and conc. H2SO4 (1.5 mL) was placed. The reaction mixture was heated at reflux for 8 h. The excess EtOH was removed in vacuo, the residue dissolved in CH2Cl2 (10 mL) and then washed with sat. aq. NaHCO3 until bubbling ceased. The organic solution was dried over MgSO4 and the solvent was evaporated in vacuo, affording crude compound 7 (6.18 g, 85%). The physical data of 7 were consistent with those previously reported [27].
Triethyl 4,5-dioxocyclopentane-1,2,3-tricarboxylate
25 mL of absolute EtOH was placed in a 100-mL round-bottomed flask equipped with a reflux condenser, continuously stirred under an inert atmosphere and then 1.10 g (48 mmol) of Na pieces weighing 50–100 mg were added to it. The reaction was initially carried out at rt and then the temperature was gradually raised to near the boiling point of the ethanol to dissolve all the
- 78 - solid sodium. When the solution was saturated, 7 (6.18 g, 23.8 mmol) was added dropwise to the resultant sodium alcoholate solution which was stirred at rt for 10 min and then had diethyl oxalate (3.48 g, 23.8 mmol) added. The reaction was continued at rt for 10 min and heated to reflux for an additional 18 h. The reaction mixture was then cooled to rt, a few drops of sat. aq. NH4Cl were added and the ethanol was removed in vacuo. A yellow solid was collected, washed with anh. ether and dried under high vacuum to afford disodium salt 8 (5.11 g, 60%). To disodium salt 8 (4.37 g, 12.2 mmol), 10%
H2SO4 (9.3 mL) was added in portions at 0 °C. The resulting solution was extracted with EtOAc (8 × 40 mL) and the combined organic extracts were dried over MgSO4. After the removal of EtOAc in vacuo, the residue contained a mixture of tautomers 8a and 8b (2.87 g, 75%).
3,4-Dioxocyclopentane-1,2-diethyl ester
To the disodium salt 8 (8.8 g, 24.6 mmol), a mixture of glacial acetic acid (30 mL), conc. H2SO4 (3 mL) and water (20 mL) was added and the resulting mixture heated at reflux for 3 h. After the removal of the solvents in vacuo, the addition of acetone (80 mL) provided a precipitate that was filtered, and the filtrate was concentrated under reduced pressure to give a brown liquid product.
Further purification was not necessary for the preparation of dione 10 in the following step. An analytical sample of 9 was obtained by recrystallisation from
- 79 -
1 water. H NMR (CDCl3): 4.40–4.26 (2H, m), 4.21–4.13 (2H, m), 3.80 (1H, dd,
J = 4.0, 8.0 Hz), 2.77 (1H, dd, J = 8.0, 20.0 Hz), 2.49 (1H, dd, J = 4.0, 20.0 Hz),
1.32 (3H, t, J = 8.0 Hz), 1.26 (3H, t, J = 8.0 Hz). MS (ESI): m/z 265.2 (100, [M +
+ Na] ). Anal. calcd for C11H14O6 (242.23): C, 54.54; H, 5.83. Found: C, 54.71; H,
5.85.
3,4-Dioxocyclopentane-1-carboxylic acid
Compound 9 (9.27 g) was added into a mixture of glacial acetic acid (32 mL), conc. H2SO4 (6 mL) and water (22 mL), and the solution was heated at reflux for
5 h. The excess acids were neutralised by the addition of sat. aq. NaOAc. After the removal of solvents in vacuo, the addition of acetone (100 mL) provided a precipitate that was removed by filtration, and the filtrate was then concentrated under reduced pressure to give a liquid product. Purification of the resulting residue by chromatography (hexane/EtOAc, 1:1) gave dione 10 as a colourless
1 solid (4.41 g, 81%). H NMR (D2O): 6.65 (1H, d, J = 4.0 Hz), 3.79–3.75 (1H, m), 2.77–2.60 (2H, m).
- 80 -
(Z)-Butene-1,4-diyl-bis[toluenesulfonate]
This modification of the known procedure [28] delivered 14 in a reasonably high yield. Cis-2-butene-1,4-diol (8.8 g, 0.1 mol) and methyl trioctyl ammonium chloride (Aliquat 336, 2 g, 5 mmol) were dissolved in a solution of 150 mL of
KOH (aq. 50%) and 150 mL of dioxane. TsCl (38.1 g, 0.1 mol) dissolved in 150 mL of dioxane was added dropwise over a period of 45 min at -4 °C, stirred at -
4 °C for 40 min and at rt for a further 3 h. The reaction mixture was filtered to remove salts and the organic layer was concentrated under reduced pressure, leaving crude 14 as a light amber viscous oil. To this crude 14 was added 150 mL of CH2Cl2. The resulting solution was filtered again and the filtrate was washed with sat. brine and dried over MgSO4. After the removal of solvent in vacuo, the residue was 31.9 g of 14 (80%). All physical data of 14 were consistent with those previously reported [28].
Diethyl 3-cyclopentene-1,1-dicarboxylate
Cyclopentene 15 was prepared by two different methods. Method 1: This was a modification of the known procedure reported by Angier et al. [29]. Diethyl malonate (6.13 g, 38.3 mmol) in anh. THF (45 mL) was added dropwise to a slurry of NaI (1.43 g, 9.5 mmol) and NaH (3.06 g, 76.5 mmol) in anh. THF (460
- 81 - mL) at 0 °C. 14 (15.16 g, 38.3 mmol) in anh. THF (150 mL) was added dropwise to the above solution with stirring at 0 °C. After 3 h at 0 °C, the mixture was stirred at rt for 40 h and then maintained at reflux for 6 h. The reaction mixture was allowed to stand at rt overnight and then the finely divided inorganic salts were removed by filtration. After the removal of THF in vacuo, the residual colourless oil was distilled at 90–110 °C/75 Pa to give cyclopentene 15 (4.62 g,
57%) as a colourless oil. The physical data of 15 were consistent with those previously reported by an alternative method [30].
Method 2: This was a modification of a known procedure [30] which utilised much smaller amounts of Ru-catalyst with increased reaction times. 2,2-diallyl- malonic acid diethyl ester 16 (5 g, 20.8 mmol), anh. CH2Cl2 (105 mL) and Ru- catalyst (0.05 g, 0.067 mmol) were used in this transformation. The reaction continued at rt for 24 h while being monitored by TLC. The purification process was the same as that described above and gave cyclopentene 15 (3.85 g, 87%) as a colourless oil. The physical data of 15 were consistent with those previously reported [30].
(3S,4R)-Diethyl 3,4-dihydroxycyclopentane-1,1-dicarboxylate
Prepared as described in the general procedure for OsO4 oxidation from cyclopentene 15 (0.424 g, 2 mmol). The crude product was purified by chromatography (hexane/EtOAc, 3:1) to give diol 17 (0.44 g) in a yield of 89%.
- 82 -
The physical data of 17 were consistent with those reported by Bowden et al. utilising an alternative method [31].
1,1-Diethyl 3,4-dioxocyclopentane-1,1-dicarboxylate
Prepared as described in the general procedure for Swern oxidation from diol
17 (3.71 g, 15.1 mmol). The crude product was purified by chromatography
(hexane/EtOAc, 1:1) to give dione 19 (3.07 g, 84%) as a light brown oil. 1H
NMR (CDCl3): 6.55 (1H, s), 4.22–4.15 (4H, m), 2.95 (2H, s), 1.99 (1H, s), 1.22
(6H, t, J = 8.0 Hz).
Dione 19 could also be prepared from 18 by the Swern oxidation general procedure. DMSO (0.4 mL), trifluoroacetic anhydride (0.7 mL), 18 (0.41 g, 1.68 mmol) and Et3N (1.5 mL) were used in this transformation and, after purification, gave 0.26 g of 19 in a yield of 65%.
3-Cyclopentene-1,1-dimethanol
The preparation of 20 was a modification of a known procedure [32]. A stirred suspension of LiAlH4 (6.22 g, 0.164 mol) in anh. THF (130 mL) was cooled to
0 °C in an ice-water bath. A solution of cyclopentene 15 (8.62 g, 40.6 mmol) in
- 83 - anh. THF (100 mL) was added dropwise. The reaction mixture was stirred at reflux for 20 h, cooled to 0 °C and quenched by the careful addition of EtOAc
(20 mL) and then water (13 mL). The reaction mixture was brought to pH 6–7 with H3PO4, filtered through celite to remove the precipitated solid and the filtrate was evaporated to dryness in vacuo to give 4.52 g of the diol 20 in a yield of 87%. An analytical sample of diol 20 was obtained by recrystallisation from water. The physical data of 20 were consistent with those reported in the literature [32].
Cyclopent-3-ene-1,1-diylbis(methylene) diacetate
Acetyl chloride (2.71 g, 34.5 mmol) was slowly added to a solution of diol 20
(1.74 g, 13.6 mmol) in anh. pyridine (20 mL) at 0 °C. The reaction mixture was stirred at 0 °C for 1 h and then at rt for 24 h. The reaction mixture was cooled to
0 °C in an ice-water bath, brought to pH 7–8 with 2 N NaOH and then extracted with EtOAc (2 × 100 mL). The combined organic extracts were washed several times with sat. brine and dried over MgSO4. After the removal of EtOAc in vacuo, the residue was purified by chromatography (toluene/EtOAc, 4:1) to give diacetate 21 as a light amber liquid (2.07 g, 72%). The physical data of 21 were consistent with those previously reported by an alternative method [32].
- 84 -
2,2-Diallylpropane-1,3-diol
This was a modification of the known procedure reported by Hii et al. [33]. To a stirred suspension of LiAlH4 (3.10 g, 0.082 mmol) in anh. THF (65 mL) at 0 °C, a solution of diester 16 (4.81 g, 20 mmol) in anh. THF (50 mL) was added dropwise. The mixture was refluxed for 20 h, cooled to 0 °C and quenched by careful addition of EtOAc (10 mL) and then water (7 mL). The reaction mixture was brought to pH 6–7 with H3PO4, filtered through celite to remove the precipitated solid and the filtrate was evaporated to dryness in vacuo. The residue was purified by chromatography (hexane/ EtOAc, 1:1) to give diol 22 as a light amber liquid (2.69 g, 86%). The physical data of 22 were consistent with those previously reported [33].
2,2-Diallylpropane-1,3-diacetate
Acetyl chloride (3.41 g, 43.4 mmol) was slowly added to a solution of diol 22
(2.66 g, 17.1 mmol) in anh. pyridine (25 mL) at 0 °C. The reaction mixture was stirred at 0 °C for 1 h and then at rt for 24 h. The reaction mixture was cooled to
0 °C in an ice-water bath, brought to pH 7–8 with 2 N NaOH and then extracted with EtOAc (2 × 150 mL). The combined organic extracts were washed several times with sat. brine and dried over MgSO4. After the removal of EtOAc in
- 85 - vacuo, the residue was purified by chromatography (EtOAc) to give diacetate 23 as a light amber liquid (3.24 g, 79%). The physical data of 23 were consistent with those previously reported by an alternative method [34].
(3S,4R)-3,4-Dihydroxycyclopentane-1,1-di(acetyloxy)methyl
Prepared as described in the general procedure for OsO4 oxidation from cyclopentene 21 (3.11 g, 14.7 mmol). The crude product was purified by chromatography (toluene/EtOAc, 4:1) to give diol 24 as a light amber liquid
1 (3.07 g, 85%). H NMR (CDCl3): 4.13–4.08 (2H, m), 4.07 (2H, s), 3.89 (2H, s),
2.38 (2H, s), 2.04 (3H, s), 2.03 (3H, s), 1.77 (2H, dd, J = 8.0, 16.0 Hz), 1.66 (2H,
13 dd, J = 8.0, 12.0 Hz). C NMR (CDCl3): 171.3, 171.2, 73.8, 68.9, 68.2, 42.7,
+ 37.4, 21.1. MS (ESI): m/z 269.3 (100, [M + Na] ). Anal. calcd for C11H18O6
(246.26): C, 53.65; H, 7.37. Found: C, 53.60; H, 7.32.
3,4-Dioxocyclopentane-1,1-di(acetyloxy)methyl
Prepared as described in the general procedure for Swern oxidation from diol
24 (3.11 g, 12.6 mmol). The crude product was purified by chromatography
(hexane/EtOAc, 1:1) to give dione 26 (2.44 g, 80%) as a light amber liquid. 1H
NMR (CDCl3): 6.28 (1H, s), 4.10 (2H, d, J = 12.0 Hz), 4.00 (2H, d, J = 8.0 Hz),
- 86 -
13 2.31 (2H, s), 2.00 (6H, s). C NMR (CDCl3): 201.6, 171.0, 154.2, 129.4, 66.6,
+ 42.6, 39.8, 20.8. MS (ESI): m/z 265.2 (100, [M + Na] ). Anal. calcd for C11H14O6
(242.23): C, 54.54; H, 5.83. Found: C, 54.41; H, 5.95.
(±)-5,6-Dioxobicyclo[2.2.1]hept-2-yl-acetate (mixture of endo and exo)
Prepared as described in the general procedure for Swern oxidation from diols
30 (2.4 g, 12.8 mmol). The crude product was purified by chromatography
(hexane/EtOAc, 7:1) to give diones 31 as a viscous oil (1.64 g, 70%). The physical data of 31 were consistent with those previously reported [35].
(±)-(6-exo)-6-Hydroxybicyclo[2.2.1]-hept-2-yl acetate (mixture of 2-endo and 2- exo); (±)-(5-exo)-5-Hydroxybicyclo[2.2.1]-hept-2-yl acetate (mixture of 2-endo and 2-exo)
This was a modification of the known procedure reported by Holy´ et al. [36]. 33 mL of borane (1 M) was added dropwise to a stirred solution of (±)-5-norbornen-
2-yl acetate 29 (9.14 g, 60 mmol) in anh. THF (30 mL) at 0 °C under an inert atmosphere. The reaction mixture was stirred at 0 °C for 4.5 h. Excess borane was decomposed by the addition of EtOH (30 mL). A solution of 30% H2O2
(46.2 mL) in sat. aq. NaHCO3 (110 mL) was added dropwise at 0 °C over 20
- 87 - min. The reaction mixture was maintained at 0 °C for a further 30 min, neutralised with 1 N HCl and extracted with diethyl ether (3 × 100 mL). The combined organic extracts were washed with sat. brine and dried over MgSO4,
After the removal of the solvent in vacuo, the residue was purified by chromatography (toluene/EtOAc, 4:1) to give a mixture of regioisomeric alcohols 32 and 33 (7.79 g, 77%) as a light amber oil. Further purification was not necessary for the preparation of 34 and 35 in the following step.
(±)-(6-exo)-6-Benzoyloxybicyclo[2.2.1]-heptan-2-ol (mixture of 2-endo and 2- exo); (±)-(5-exo)-5-Benzoyloxybicyclo[2.2.1]-heptan-2-ol (mixture of 2-endo and 2-exo)
A mixture of 34 and 35 (1.03 g, 3.76 mmol) was added to a solution of 32% HCl
(1.2 mL) in MeOH (55 mL) and the reaction mixture was continually stirred at rt for 3 d. The resulting solution was diluted with EtOAc (90 mL), treated with sat. aq. NaHCO3 until bubbling ceased and washed with sat. brine. After the removal of EtOAc in vacuo, the residue was diluted with CH2Cl2, filtered and the filtrate was dried over MgSO4. The solvent (CH2Cl2) was removed in vacuo and the residue was purified by chromatography (toluene/EtOAc, 4:1) to give the alcohol mixture of 36 and 37 as a waxy light amber liquid (0.66 g, 76%).
- 88 - exo 2-Oxobicyclo[2.2.1]heptan-6-yl benzoate; exo 2-Oxobicyclo[2.2.1]heptan-5- yl benzoate
This was a modification of the known procedure reported by Holy´ et al. [36]. A mixture of anh. pyridine (37 mL), anh. CH2Cl2 (370 mL) and CrO3 (22.5 g, 0.225 mol) was cooled to 0 °C in an ice-water bath. Then, a solution of 36 and 37
(8.63 g, 37.2 mmol) in anh. CH2Cl2 (340 mL) was added for 15 min and the resulting solution was stirred at 0 °C for a further 8 h. After filtration and rotary evaporation, the residue was purified by chromatography (toluene/EtOAc, 4:1) to give a mixture of ketones 38 and 39 as a colourless solid (5.85 g, 68%). exo-2,3-Dioxobicyclo[2.2.1]heptan-5-yl-benzoate
A mixture of ketones 38 and 39 (0.55 g, 2.39 mmol), SeO2 (0.27 g, 2.43 mmol) and xylenes (2.6 mL) was heated at reflux for 24 h. The reaction mixture was filtered, the xylenes was removed in vacuo, and the residue was purified by chromatography (hexane/EtOAc, 1:1) to give dione 40 as a white solid (0.41 g,
70%). An analytical sample of dione 40 was obtained by recrystallisation from
EtOAc. Mp 183 °C. IR (KBr, cm-1): 1714 s, 1695 m, 1315 m, 1273 s, 1242 s,
1 1192 m, 1107 s, 1045 m, 966 w, 700 s. H NMR (DMSO-d6): 7.97 (2H, d, J =
- 89 -
8.0 Hz), 7.67 (1H, t, J = 8.0 Hz), 7.53 (2H, t, J = 8.0 Hz), 5.50–5.45 (1H, m),
3.07–2.97 (2H, m), 2.42–2.32 (1H, m), 2.24–2.15 (1H, m), 2.08–2.00 (1H, m),
13 1.99–1.90 (1H, m). C NMR (DMSO-d6): 175.6, 174.3, 165.2, 133.5, 129.6,
129.3, 128.8, 77.8, 49.7, 41.5, 35.4, 31.4. MS (ESI): m/z 283 (100, [M + K]+),
+ 245 (45, [M + H] ). Anal. calcd for C14H12O4 (244.24): C, 68.85; H, 4.95. Found:
C, 68.69; H, 4.86.
(±)-(5-exo,6-exo)-5,6-Dibenzoyloxybicyclo[2.2.1]-heptan-2-ol (mixture of 2-endo and 2-exo)
Compound 41 (9.35 g, 23.7 mmol) was added to a solution of 32% HCl (7.5 mL) in MeOH (330 mL) and the reaction mixture was continually stirred at rt for 3 d.
The solution was diluted with EtOAc (600 mL), treated with sat. aq. NaHCO3 until bubbling ceased and extracted with sat. brine. The combined organic extracts were dried over MgSO4. After the removal of the solvent in vacuo, the residue was purified by chromatography (hexane/EtOAc, 1:1) to give the alcohol mixture 42 (7.4 g, 89%) as a waxy solid in an approximate endo/exo-2-hydroxy ratio of 4:1, as determined by 1H NMR. The physical data of 42 were consistent with those previously reported by an alternative method [35].
- 90 -
Bicyclo[2.2.1]heptane-2,3-diol
Prepared as described in the general procedure for OsO4 oxidation from norbornene 45 (5 g, 53.1 mmol). The crude product was a light green solid of diol 46 (6.09 g, 89%). An analytical sample of the diol 46 was obtained by
1 recrystallisation from EtOAc. H NMR (CDCl3): 3.67 (2H, s), 2.23 (2H, b s),
2.12 (2H, s), 1.75 (1H, d, J = 8.0 Hz), 1.48–1.41 (2H, m), 1.09 (1H, d, J = 8.0
13 Hz), 1.04 (2H, d, J = 8.0 Hz). C NMR (CDCl3): 75.0, 43.2, 31.8, 24.7. MS
+ (ESI): m/z 167 (100, [M + K] ). Anal. calcd for C7H12O2 (128.17): C, 65.60; H,
9.44. Found: C, 65.74; H, 9.31.
Bicyclo[2.2.1]heptane-2,3-dione
Prepared as described in the general procedure for Swern oxidation from diol
46 (6.09 g, 47.5 mmol). The crude product was purified by chromatography
1 (EtOAc/Toluene, 1:1) to give dione 47 (4.36 g, 74%). H NMR (CDCl3): 3.05
13 (2H, s), 2.15–1.99 (4H, m), 1.81–1.72 (2H, m). C NMR (CDCl3): 202.3, 48.6,
31.8, 23.9.
- 91 -
5-Norbornene-2-exo,3-exo-dimethanol,diacetate
This was a modification of the known procedure reported by Yan et al. [37].
Acetyl chloride (2.59 g, 32.9 mmol) was slowly added to a solution of diol 48 (2 g, 13 mmol) in anh. pyridine (18 mL) at 0 °C, and the reaction mixture was stirred at 0 °C for 1 h and at rt for 24 h. The solution was cooled to 0 °C in an ice-water bath, brought to pH 7–8 with 2 N NaOH and extracted with EtOAc (2 ×
75 mL). The combined organic extracts were washed several times with sat. brine and dried over MgSO4. After the removal of the solvent in vacuo, the residue was purified by distillation at reduced pressure (112–114 °C/0.4 mmHg) to give acetate 49 as a white solid (2.73 g, 88%). The physical data of 49 were consistent with those previously reported [37].
(5-exo,6-exo)-5,6-Dimethanol,diacetatebicyclo[2.2.1]-heptan-2,3-diol
Prepared as described in the general procedure for OsO4 oxidation from acetate 49 (3.64 g, 15.3 mmol). The crude product was a white solid of diol 50
(3.95 g, 95%). An analytical sample of the diol 50 was obtained by recrystallisation from EtOAc. Mp 102 °C. IR (KBr, cm-1): 3350 s, 3235 s, 2972 s,
2909 m, 1736 s, 1368 s, 1240 s, 1111 w, 1028 s, 980 m, 899 w, 795 w, 606 w,
- 92 -
1 492 w. H NMR (CDCl3): 4.08 (2H, dd, J = 4.0, 12.0 Hz), 3.98–3.90 (2H, m),
3.77 (2H, s), 2.11 (2H, s), 2.04 (6H, s), 1.90–1.80 (2H, m), 1.70 (1H, d, J = 12.0
13 Hz), 1.40 (1H, d, J = 12.0 Hz). C NMR (CDCl3): 171.1, 74.2, 63.5, 46.9, 39.5,
+ 27.4, 21.2. MS (ESI): m/z 295 (100, [M + Na] ). Anal. calcd for C13H20O6
(272.29): C, 57.34; H, 7.40. Found: C, 57.50; H, 7.52.
(exo,exo)-5,6-Dimethanol,diacetatebicyclo[2.2.1]heptane-2,3-dione
Prepared as described in the general procedure for Swern oxidation from diol
50 (3.23 g, 11.9 mmol). The crude product was purified by chromatography
(EtOAc/Toluene, 1:2) to give dione 51 (2.23 g, 70%). IR (KBr, cm-1): 3447 w,
2967 w, 1776 s, 1759 s, 1738 s, 1393 w, 1369 s, 1238 s, 1036 s, 997 m. 1H
NMR (CDCl3): 4.26 (2H, dd, J = 8.0, 12.0 Hz), 4.16–4.10 (2H, m), 3.04 (2H, s),
2.58–2.48 (2H, m), 2.42 (1H, d, J = 12.0 Hz), 2.08 (6H, s), 2.02 (1H, d, J = 12.0
13 Hz). C NMR (CDCl3): 200.4, 170.7, 62.2, 52.4, 38.9, 27.3, 21.1. MS (ESI):
+ + m/z 291 (100, [M + Na] ). HR-MS for C14H21N2O7Na [M + Na] : calcd 291.0845, found 291.0845.
- 93 -
Diethyl 1H,4H-3a,6a-propanoimidazo[4,5-d]imidazole-2,5(3H,6H)-dioxo-8,8- dicarboxylate
Prepared as described in the general procedure for the urea condensation reaction from urea (91.8 mg, 1.53 mmol) and dione 19 (0.123 g, 0.51 mmol) in 2 mL of toluene. Yield: 34.3 mg (21%) of 19a as a white solid. Mp > 260 °C. IR
(KBr, cm-1): 3370 w, 3240 w, 1717 s, 1682 s, 1651 s, 1456 m, 1298 m, 1219 m,
1 1123 m, 862 w. H NMR (DMSO-d6): 7.41 (4H, s), 4.13–4.06 (4H, m), 2.61
13 (4H, s), 1.17 (6H, t, J = 8.0 Hz). C NMR (DMSO-d6): 169.4, 159.7, 80.8, 61.7,
59.8, 46.7, 13.8. MS (ESI): m/z 327 (47, [M + H]+), 349 (100, [M + Na]+). Anal. calcd for C13H18N4O6 (326.31): C, 47.85; H, 5.56; N, 17.17. Found: C, 47.81; H,
5.69; N, 17.09.
- 94 -
1H,4H-3a,6a-Propanoimidazo[4,5-d]imidazole-2,5(3H,6H)-dioxo-8,8- dicarboxylate disodium salt
A mixture of glycoluril 19a (0.326 g, 1 mmol), NaOH (0.08 g, 2 mmol) and deionised H2O (2 mL) was heated at 80 °C for 3 h. After cooling to rt, the solvent was removed under reduced pressure and the residue was dried under high vacuum to give 19b (0.29 g, 92%) as a light yellow solid. Mp > 260 °C. IR
(KBr, cm-1): 3379 s, 3294 s, 1709 s, 1674 s, 1582 s, 1474 m, 1420 m, 1335 m,
1 13 1165 w, 748 w, 652 w. H NMR (D2O): 2.73 (4H, s). C NMR (D2O): 178.4,
162.2, 82.8, 65.1, 47.0.
1H,4H-3a,6a-Propanoimidazo[4,5-d]imidazole-2,5(3H,6H)-dioxo-8,8- dicarboxylic acid
. A solution of PTSA H2O (0.19 g, 1 mmol) in H2O (2 mL) was added to glycoluril
19b (0.157 g, 0.5 mmol). The excess H2O was removed by rotary evaporation and the resulting residue was dried under high vacuum. The residue was
- 95 - washed with MeOH (5 × 1 mL) and dried under high vacuum to give 19c (0.054 g, 40%) as a light yellow solid. Mp 245 °C. IR (KBr, cm-1): 3387 s, 1686 s, 1477 w, 1385 m, 1285 w, 1246 w, 1157 m, 1034 w, 764 w, 617 w. 1H NMR (DMSO-
13 d6): 7.38 (4H, s), 4.03 (2H, b s), 2.58 (4H, s). C NMR (DMSO-d6): 171.5,
159.8, 81.2, 59.9, 46.7.
1H,4H-3a,6a-Propanoimidazo[4,5-d]imidazole-2,5(3H,6H)-dioxo-8-carboxylic acid
Glycoluril 19a (0. 10 g, 0.30 mmol) dissolved in 6 mL of 37% HCl was heated at
90 °C for 24 h. After the removal of the excess HCl by rotary evaporation, the residue was dried under high vacuum to give 10a (0.044 g, 65%) as a brown
1 solid. H NMR (DMSO-d6): 7.42 (2H, s), 7.30 (1H, s), 7.17 (2H, s), 2.84–2.73
13 (1H, m), 2.40–2.32 (2H, m), 1.91 (2H, t, J = 12.0 Hz). C NMR (DMSO-d6):
174.2, 160.6, 159.8, 80.8, 43.8, 41.7.
- 96 -
Diethyl 1,6:3,4-bis(2-oxapropylene)tetrahydro-3a,6a-propanoimidazo[4,5- d]imidazol-2,5(1H,3H)-dioxo-8,8-dicarboxylate
Glycoluril 19a (0.143 g, 0.44 mmol) was dissolved in TFA (0.5 mL) and then paraformaldehyde (0.132 g, 4.4 mmol) was added in one portion. The mixture was heated at reflux for 16 h, concentrated and dried under high vacuum. A mixture of the crude material, PTSA (0.49 g, 2.58 mmol) and ClCH2CH2Cl (13 mL) was heated under an inert atmosphere at reflux under an addition funnel filled with molecular sieves (4 Å) for 2 h. The reaction mixture was diluted with
EtOAc (35 mL), washed with sat. aq. Na2CO3 and brine, concentrated and then the residue was purified by chromatography (toluene/EtOAc, 1:1) to give diether of glycoluril 19d as a light yellow solid (0.163 g, 90%). Mp 151 °C. IR (KBr, cm-1):
1736 s, 1466 m, 1408 s, 1381 s, 1319 m, 1269 m, 1250 s, 1207 m, 1173 s,
1 1092 s, 1026 s, 941 m, 903 m, 806 w, 679 w, 579 w. H NMR (CDCl3): 5.46
(4H, d, J = 12.0 Hz), 4.69 (4H, d, J = 12.0 Hz), 4.22–4.13 (4H, m), 3.00 (4H, s),
13 1.20 (6H, t, J = 8.0 Hz). C NMR (CDCl3): 168.8, 158.2, 79.4, 71.8, 62.8, 61.3,
42.1, 14.1.
- 97 -
8,8-Di(acetyloxy)methyl 1H,4H-3a,6a-propanoimidazo[4,5-d]imidazole-2,5(3H, 6H)-dione
Prepared as described in the general procedure for the urea condensation reaction from urea (30.6 mg, 0.51 mmol) and dione 26 (41 mg, 0.17 mmol) in 3 mL of benzene with efficient stirring for 40 min at rt, and then was heated at reflux for 17 h. Yield: 5.9 mg (11%) of 26a as a white solid. Mp > 260 °C. IR
(KBr, cm-1): 3292 m, 3258 w, 1740 m, 1659 s, 1487 m, 1364 m, 1223 m, 1136
1 m, 1034 s. H NMR (DMSO-d6): 7.46 (4H, s), 3.93 (4H, s), 2.07 (4H, s), 2.01
13 (6H, s). C NMR (DMSO-d6): 170.3, 159.8, 81.9, 65.7, 46.7, 45.2, 20.6. MS
+ + (ESI): m/z 327 (47, [M + H] ), 349 (100, [M + Na] ). Anal. calcd for C13H18N4O6
(326.31): C, 47.85; H, 5.56; N, 17.17. Found: C, 47.68; H, 5.77; N, 16.93.
4-Methyl-3-oxo-N-cyclopenten-2-yl urea
Prepared as described in the general procedure for the urea condensation reaction from urea (1.80 g, 30 mmol) and 3-methyl-1,2-cyclopentadione 11
- 98 -
(1.12 g, 10 mmol). The crude product was purified by chromatography using
EtOAc as an eluent to give 11a as a pale yellow solid (0.71 g, 45%). Mp 175 °C.
IR (KBr, cm-1): 3412 m, 3331 m, 3192 w, 1682 s, 1614 s, 1537 s, 1371 m, 1287
1 m, 1132 s, 770 m, 638 s, 613 s. H NMR (CDCl3): 8.10 (1H, s), 7.58 (1H, t, J
= 4.0 Hz), 5.50 (2H, b s), 2.95–2.84 (1H, m), 2.47–2.36 (1H, m), 2.23 (1H, dt, J
13 = 4.0, 20.0 Hz), 1.18 (3H, d, J = 8.0 Hz). C NMR (CDCl3): 208.5, 156.4,
138.1, 136.9, 38.4, 34.3, 16.4. MS (ESI): m/z 177 (100, [M + Na]+). Anal. calcd for C7H10N2O2 (154.17): C, 54.54; H, 6.54; N, 18.17. Found: C, 54.68; H, 6.74;
N, 18.01.
(3aS,4R,7S,7aS)-Octahydro-7a-hydroxy-4,7-methano-2-oxo-7,8,8-trimethyl-2H- benzimidazol-3a-yl urea
Prepared as described in the general procedure for the urea condensation reaction from urea (1.80 g, 30 mmol) and camphorquinone 28 (1.66 g, 10 mmol).
Yield: 0.56 g (21%) of 28a as a white solid. An analytical sample of 28a was obtained by recrystallisation from water. Mp 237 °C. IR (KBr, cm-1): 3325 w,
3300 w, 1688 s, 1657 s, 1530 m, 1414 w, 1364 w, 1103 s, 824 w. 1H NMR
(DMSO-d6): 7.03 (1H, s), 6.64 (1H, s), 6.57 (1H, s), 6.24 (1H, s), 5.88 (2H, s),
1.74 (1H, d, J = 4.0 Hz), 1.52–1.40 (3H, m), 1.24–1.17 (1H, m), 1.17 (3H, s),
13 0.86 (3H, s), 0.79 (3H, s). C NMR (DMSO-d6): 159.5, 158.5, 95.5, 78.7, 55.2,
+ 53.3, 47.4, 31.0, 22.7, 22.4, 22.1, 10.9. MS (ESI): m/z 269 (100, [M + H] ). Anal.
- 99 - calcd for C12H20N4O3·0.5H2O (277.32): C, 51.97; H, 7.63; N, 20.20. Found: C,
52.14; H, 7.59; N, 20.14.
(7S,10R) 1H,4H-7,11,11-Trimethyl-7,10-methano-3a,6a-butanoimidazo[4,5- d]imidazole-2,5(3H,6H)-dione
Two methods were used for the preparation of this compound. Method 1: It was prepared as described in the general procedure for the urea condensation reaction from urea (0.90 g, 15 mmol) and camphorquinone 28 (0.83 g, 5 mmol) in 20 mL of xylenes. Yield: 0.19 g (15%) of 28c as a white powder. An analytical sample of 28c was obtained by recrystallisation from water. Mp > 260 °C. IR
(KBr, cm-1): 3219 m, 3067 w, 1730 m, 1680 s, 1661 s, 1487 m, 1395 m, 1136 m,
1 1109 s, 1047 m, 779 m. H NMR (DMSO-d6): 7.29 (1H, s), 7.21 (1H, s), 7.15
(2H, s), 1.90 (1H, d, J = 4.0 Hz), 1.62–1.48 (3H, m), 1.34–1.24 (1H, m), 1.08
13 (3H, s), 0.94 (3H, s), 0.85 (3H, s). C NMR (DMSO-d6): 162.2, 160.9, 83.0,
79.6, 52.1, 52.0, 49.3, 30.7, 25.0, 23.1, 20.3, 10.4. MS (ESI): m/z 251 (100, [M
+ + H] ). Anal. calcd for C12H18N4O2 (250.30): C, 57.58; H, 7.25; N, 22.38. Found:
C, 57.40; H, 7.41; N, 22.37.
Method 2: Compound 28a (0.029 g, 0.108 mmol) and TFA (0.02 mL) were dissolved in 3 mL of formamide and stirred at rt for 30 min. The solution was heated at 140 °C for 6 h and the reaction mixture was cooled to rt. After the
- 100 - removal of the formamide by rotary evaporation, the residue was 15 mg of a solid mixture of 28a and 28c (28c is 37% in the product mixture, determined by
1H NMR, yield 20%).
(3aR,4S,5R,7R,7aS)-Octahydro-5-acetoyloxy-3a,7a-dihydroxy-4,7-methano-2H- benzimidazol-2-one; (3aS,4R,6S,7R,7aR)-Octahydro-6-acetoyloxy-7a-hydroxy- 4,7-methano-2-oxo-2H-benzimidazol-3a-yl urea
Prepared as described in the general procedure for the urea condensation reaction from urea (0.405 g, 6.75 mmol) and dione 31 (0.41 g, 2.25 mmol). Yield:
0.23 g (36%) of a mixture of 31a and 31b as a light yellow solid in an approximate ratio of 1:9, as determined by 1H NMR. Analytical samples of 31a
1 and 31b were obtained by recrystallisation from water. 31a: H NMR (DMSO-d6):
6.99 (1H, s), 6.96 (1H, s), 5.76 (1H, s), 5.41 (1H, s), 4.84 (1H, b s), 4.67 (1H, s), 2.13 (1H, s), 1.95 (3H, s), 1.93–1.83 (1H, m), 1.78 (1H, d, J = 12.0 Hz), 1.49
13 (1H, d, J = 12.0 Hz), 1.33 (1H, d, J = 12.0 Hz). C NMR (DMSO-d6): 170.6,
159.3, 89.4, 88.7, 74.1, 48.1, 45.9, 30.8, 28.0, 21.1. 31b: Mp 247 °C. IR (KBr, cm-1): 3439 w, 3181 w, 1728 m, 1680 m, 1639 s, 1526 m, 1333 w, 1240 s, 1159
1 m, 1036 m. H NMR (DMSO-d6): 7.12 (1H, s), 6.55 (1H, s), 6.37 (1H, s), 6.23
(1H, s), 6.01 (2H, s), 4.92–4.85 (1H, m), 2.72 (1H, d, J = 4.0 Hz), 2.14 (1H, d, J
= 4.0 Hz), 2.01(3H, s), 1.97–1.87 (1H, m), 1.66–1.55 (2H, m), 1.38 (1H, d, J =
13 12.0 Hz). C NMR (DMSO-d6): 170.5, 159.9, 158.8, 90.4, 74.2, 73.4, 48.5,
- 101 -
46.3, 31.1, 28.0, 21.1. MS (ESI): m/z 307 (100, [M + Na]+). Anal. calcd for
C11H16N4O5 (284.27): C, 46.48; H, 5.67; N, 19.71. Found: C, 46.14; H, 5.65; N,
19.46.
(3aR,4S,5R,7R,7aS)-Octahydro-5-benzoyloxy-3a,7a-dihydroxy-4,7-methano- 2H-benzimidazol-2-one
Prepared as described in the general procedure for the urea condensation reaction from urea (0.063 g, 1.05 mmol) and dione 40 (0.085 g, 0.35 mmol).
Yield: 0.03 g (28%) of 40a as a white solid. An analytical sample of 40a was
1 obtained by recrystallisation from water. Mp 225 °C. H NMR (DMSO-d6): 7.94
(2H, d, J = 8.0 Hz), 7.65 (1H, t, J = 8.0 Hz), 7.51 (2H, t, J = 8.0 Hz), 7.10 (1H, s),
7.05 (1H, s), 5.35 (1H, s), 5.26 (1H, d, J = 8.0 Hz), 5.25 (1H, s), 2.48–2.42 (1H, m), 2.41 (1H, s), 2.25 (1H, d, J = 4.0 Hz), 1.60 (1H, d, J = 12.0 Hz), 1.53 (1H, d,
13 J = 8.0 Hz), 1.41 (1H, d, J = 8.0 Hz). C NMR (DMSO-d6): 165.2, 159.2,
133.3, 130.1, 128.7, 128.8, 88.3, 88.0, 72.8, 52.2, 45.9, 32.9, 29.4. MS (ESI):
+ m/z 305 (100, [M + H] ). Anal. calcd for C15H16N2O5 (304.3): C, 59.21; H, 5.30;
N, 9.21. Found: C, 59.19; H, 5.33; N, 9.28.
- 102 -
(3aS,4S,5R,6S,7R,7aS)-Octahydro-5,6-dibenzoyloxy-3a,7a-dihydroxy-4,7- methano-2H-benzimidazol-2-one
Prepared as described in the general procedure for the urea condensation reaction from urea (0.031 g, 0.51 mmol) and dione 44 (0.062 g, 0.17 mmol).
Yield: 0.017 g (24%) of 44a as a light yellow solid. An analytical sample of 44a was obtained by recrystallisation from water. Mp 248 °C. IR (KBr, cm-1): 3393 s,
3358 s, 3069 w, 1721 s, 1705 s, 1686 s, 1601 w, 1450 m, 1404 s, 1317 s, 1287
1 s, 1128 s, 1051 w, 951 w. H NMR (DMSO-d6): 7.76 (4H, d, J = 8.0 Hz), 7.55
(2H, t, J = 4.0 Hz), 7.32 (4H, t, J = 8.0 Hz), 7.28 (2H, s), 5.61 (2H, s), 5.57 (2H, s), 2.46 (2H, s), 2.05 (1H, d, J = 12.0 Hz), 1.64 (1H, d, J = 12.0 Hz). 13C NMR
(DMSO-d6): 164.8, 159.1, 133.4, 129.2, 129.1, 128.5, 87.8, 72.1, 51.9, 28.0.
+ MS (ESI): m/z 447 (100, [M + Na] ). Anal. calcd for C22H20N2O7 (424.41): C,
62.26; H, 4.75; N, 6.60. Found: C, 62.50; H, 4.76; N, 6.48.
(3aR,4R,7S,7aS)-Octahydro-3a,7a-dihydroxy-4,7-methano-2H-benzimidazol-2- one
Prepared as described in the general procedure for the urea condensation reaction from urea (0.18 g, 3 mmol) and dione 47 (0.124 g, 1 mmol). Yield: 0.04
- 103 - g (16%) of 47a as a light yellow solid. An analytical sample of 47a was obtained by recrystallisation from water. Mp 216 °C. IR (KBr, cm-1): 3424 w, 3296 m,
1659 s, 1628 m, 1470 m, 1366 m, 1290 m, 1136 s, 1099 s, 773 s. 1H NMR
(DMSO-d6): 6.83 (2H, s), 5.05 (2H, s), 2.10 (2H, s), 1.74 (2H, d, J = 8.0 Hz),
1.51 (1H, d, J = 12.0 Hz), 1.26 (2H, d, J = 8.0 Hz), 1.14 (1H, d, J = 12.0 Hz). 13C
NMR (DMSO-d6): 159.6, 89.0, 46.6, 32.3, 22.0. MS (ESI): m/z 185 (100, [M +
+ H] ). Anal. calcd for C8H12N2O3 (184.19): C, 52.17; H, 6.57; N, 15.21. Found: C,
51.97; H, 6.66; N, 15.09.
(3aS,4R,5R,6S,7S,7aR)-Octahydro-5,6-di(acetyloxy)methyl-3a,7a-dihydroxy- 4,7-methano-2H-benzimidazol-2-one
Prepared as described in the general procedure for the urea condensation reaction from urea (0.18 g, 3 mmol) and dione 51 (0.268 g, 1 mmol) in 4 mL of toluene. Yield: 0.065 g (20%) of 51a as a white solid. An analytical sample of
51a was obtained by recrystallisation from DMF. Mp 238 °C. IR (KBr, cm-1):
3431 s, 3329 s, 1717 s, 1383 m, 1369 m, 1263 s, 1248 s, 1140 m, 1034 m, 976
1 w, 905 w, 748 w, 610 w. H NMR (DMSO-d6): 6.93 (2H, s), 5.19 (2H, s), 4.04
(2H, dd, J = 4.0, 12.0 Hz), 3.94–3.84 (2H, m), 2.61–2.52 (2H, m), 2.04 (2H, s),
1.99 (6H, s), 1.53 (1H, d, J = 12.0 Hz), 1.35 (1H, d, J = 12.0 Hz). 13C NMR
(DMSO-d6): 170.2, 159.3, 88.7, 63.2, 49.9, 35.9, 27.0, 20.8. MS (ESI): m/z
- 104 -
+ + + 329 (50, [M + H] ), 351 (100, [M + Na] ). HR-MS for C14H21N2O7 [M + H] : calcd
329.1349, found 329.1349.
- 105 -
3.4 Results and discussion
3.4.1 Synthesis of functional five-membered carbon ring glycolurils
Recently, our group has been engaged in the syntheses of the types of glycoluril derivatives that constitute the fundamental substructure of the substituted cucurbit[n]uril. Our basic approach has been through the condensation reaction of an urea with an -dicarbonyl compound under acid- catalysed conditions. It is the choice of an -dicarbonyl compound that can potentially generate a desired substituent on the glycoluril. In practice, many glycoluril derivatives with alkyl or functionalised substituents are either currently unavailable or limited by poor yields. The condensation of a glycoluril derivative and formaldehyde in concentrated HCl yields substituted Q[n] composed of different numbers of glycoluril rings. The studies of cyclopentanoQ[n] discussed in Chapter 2 have unambiguously established that cyclopentanoglycoluril with a five-membered carbon ring substituent enables the formation of higher homologues of substituted Q[n]. An extension of this methodology could realise the introduction of functionality into Q[n]s through a functionalised five- membered carbon ring glycoluril moiety. Our focus has been on synthesising glycoluril derivatives that carry a functionalised group on a five-membered alkyl ring, as shown in Figure 3.1.
- 106 -
Figure 3.1 Structure of target glycoluril.
The functional substituent groups at C-4 in the glycoluril in Figure 3.1 are preferable because these glycolurils have the advantage of symmetrical structures and these have simplicity. The synthesis of Q using glycolurils with unsymmetrical structures would produce many Q isomers with less control over the locations of substituents and product isolation would be either, practically impossible or quite laborious.
Figure 3.2 Known glycolurils each with a substituent based on the five- membered carbon ring.
Only two glycolurils are known to carry a five-membered carbon ring substituent cyclopentanoglycoluril 5 and dimethylcyclopentanoglycoluril 6 (Figure 3.2). Until the work undertaken for this thesis, preparation of a glycoluril with a functional group on a five-membered carbon ring substituent was unknown.
- 107 -
The simplest example of a glycoluril with a five-membered carbon ring substituent is cyclopentanoglycoluril 5 (Figure 3.2) which was synthesised from
1,2-cyclopentadione and urea with the assistance of an acid catalyst to give the best reported yield of 60% [38, 39].
Scheme 3.3 Synthesis of dione 10. Conditions: (a) EtONa, diethyl oxalate, reflux, 18 h; (b) 10% H2SO4; (c) CH3CO2H, H2SO4, H2O, reflux, 3 h; (d)
CH3CO2H, H2SO4, H2O, reflux, 5 h.
The introduction of functionality into glycoluril 5 (Figure 3.2) was potentially accomplishable through a functionalised five-membered cyclic -dione. The dione 10 shown in Scheme 3.3 was expected to provide an example of the structural feature that we required. Although it was first synthesised by Gault et al. in 1910, only limited experimental detail and no spectroscopic data were provided [40]. We repeated this series of reactions and isolated dione 10 in a
- 108 - yield of 81% as a colourless solid (Scheme 3.3). Surprisingly, attempts to synthesise glycoluril 10a from dione 10 and urea in the presence of an acid catalyst failed. The reason for this was not absolutely clear but one possible explanation is that the urea condensation intermediate involved an endocyclic double bond formation, such as 10b in the substituent ring (Scheme 3.4).
Another possible explanation is that the carboxylic acid group in the intermediate 10c may have reacted with the OH group to form the intermediate lactone 10d (Scheme 3.4).
Scheme 3.4 Possible intermediates (10b, 10d) for condensation reaction between dione 10 and urea.
While neither intermediates 10b nor 10d were isolated, evidence for this type of side reaction could be found in the condensation reaction of 3-methyl-1,2- cyclopentadione 11 or keto-aldehyde 13 with urea, as shown in Scheme 3.5 and Scheme 3.6.
3-Methly-1,2-cyclopentadione (11) was commercially available and it was found that the acid-catalysed condensation reaction of dione 11 with urea gave a product mixture from which neither diol nor glycoluril could be detected.
- 109 -
However, the alternative urea condensation product 11a was isolated in a 45% yield by chromatography.
Scheme 3.5 Synthesis of 11a and glycoluril 12a. Conditions: (a) PhH,
H2NCONH2, TFA, rt, 40 min, then reflux, 4 h.
It seemed that the formation of 11a, which cannot easily lead back to glycoluril formation, prevailed. Previous work in our laboratories had also clearly demonstrated that, when an alkyl group was present to block at least one path to an endocyclic by-product (Scheme 3.5), the synthesis of glycoluril 12a was possible, albeit in a very low yield of 1%.
A divergent path to an intermediate, such as cyclic ether 10d (Scheme 3.4), was also demonstrated by the example of 13d being isolated from the condensation reaction between dione 13 and urea, as shown in Scheme 3.6.
- 110 -
Scheme 3.6 Synthesis of glycoluril 13c.
Dione 13 (Scheme 3.6), with a branched alkyl substituent attached to an - carbon atom, blocked the divergent path to a double bond during the synthetic reaction to a glycoluril but also produced the alternative condensation product
13d. We observed in Scheme 3.6 that the first step, a, appeared relatively facile but the subsequent reaction to the second step, b, was less efficient and produced significant amounts of imadazolone 13d (36%), a product resulting from the hydrolysis of the intermediate product 13a prior to ring closure. This result suggested that a difficulty occured in the ring closure at the neopentyl C- centre to furnish the glycoluril 13c (23%). This was possibly a consequence of the steric bulk near the position of dione 13.
The type of substituted glycolurils (Figure 3.1) we wanted appeared to be possible given that dimethylcyclopentanoglycoluril 6, in which two identical substituent groups were present at C-4 in its substituent ring, had previously
- 111 - been reported (Figure 3.2) [41]. This structural feature led us to contemplate the possibility that 4,4-difunctionalised 1,2-cyclopentadiones, such as 19 and 26 in
Scheme 3.7, might have produced a better outcome than the monofunctional group of dione 10 (Scheme 3.4).
Scheme 3.7 Synthesis of diones 19 and 26. Conditions: (a) CH2(CO2Et)2, NaI, NaH, THF, 0 °C, 3 h, rt, 40 h, reflux, 6 h, rt, overnight; (b) Grubbs’ catalyst
(metathesis), DCM, 24 h; (c) OsO4/NMNO/Me2CO-H2O, 40 °C, 18 h; (d) Grubbs’ ® II catalyst, NaHCO3, Oxone , rt, 2 h; (e) DMSO-TFAA/DCM/TEA, -78 °C, 5.5 h;
(f) DMSO-TFAA/DCM/TEA, -78 °C, 5.5 h; (g) LiAlH4, THF, reflux, 20 h; (h)
LiAlH4, THF, reflux, 20 h; (i) AcCl/Py, rt, 24 h; (j) AcCl/Py, rt, 24 h; (k) ® OsO4/NMNO/Me2CO-H2O, 40 °C, 18 h; (l) Grubbs’ II catalyst, NaHCO3, Oxone ; (m) DMSO-TFAA/DCM/TEA, -78 °C, 5.5 h; (n) DMSO-TFAA/DCM/TEA, -78 °C.
- 112 -
To this end, we set out to synthesise diones 19 and 26 using the synthetic scheme summarised in Scheme 3.7. Our first approach was the cycloalkylation of diethyl malonate with 14 in anh. THF in the presence of sodium hydride to give cyclopentene 15 in a 57% yield, with the remaining by-product being a vinylcyclopropane isomer [42]. Due to the formation of this by-product and the difficulty of purifying cyclopentene 15, a reported alternative synthesis of 15 was investigated [30] in which it was readily synthesised by the Ru-catalysed ring- closing metathesis reaction of the diethyl diallylmalonate 16 in excellent yields
(up to 87%). We investigated two paths to diones 19 and 26 utilising cyclopentene 15 and diethyl diallylmalonate 16. There is no doubt that OsO4 was a very effective catalyst for the formation of 1,2-diols with good yields from alkenes as was also found to be true in steps c and k of our Scheme 3.7 for producing diols 17 and 24, both of which were oxidised under Swern conditions to produce diones 19 and 26 in good yields.
However, as OsO4 is expensive and highly toxic, an alternative route that avoided it was the ring-closing metathesis reaction of 16 or 23 (Scheme 3.7) followed by oxidation with Oxone® to give hydroxyketone 18 or 25, and then the synthesis of dione 19 (26) from the hydroxyketone 18 (25) by a Swern oxidation.
This path was also subject to some restrictions, such as relatively low yields and laborious workup procedures.
- 113 -
Scheme 3.8 Synthesis of glycolurils 19a, 19b, 19c, 19d, and 10a. Conditions:
(a) Toluene, H2NCONH2, TFA, rt, 40 min, then reflux, 4 h; (b) HCl, 90 °C, 24 h;
(c) TFA, (CH2O)n, reflux, then PTSA, ClCH2CH2Cl; (d) H2O, NaOH, 80 °C, 3 h; (e) PTSA.
Having successfully synthesised dione 19, we condensed it with urea using TFA as the catalyst to afford the required glycoluril 19a, but it produced a relatively low yield (21%). The possible reasons for this could be the same as those discussed earlier in the example of dione 10 (Scheme 3.4). We tested hydrolysis under basic conditions to achieve 19b and its conversion to the carboxylic acid 19c was accomplished by treatment with an excess of PTSA.
Base hydrolysis and acid-catalysed decarboxylation afforded glycoluril 10a
- 114 - which demonstrated that 10a was stable in an acidic solution with relatively higher temperatures. This result also suggested that the stability of 10a was not responsible for its failed preparation directly from dione 10 shown in Scheme
3.3. This series of glycolurils carrying ester or carboxylic acid functional groups had a high degree of solubility in both aqueous and organic media, indicating their potential for the improved solubility of a substituted Q[n] under similar solution conditions.
Glycoluril diether had been one of the important starting materials for the synthesis of particular partially substituted Q by heteromeric cyclisation, as discussed in Section 1.3.1 (Chapter 1) [43, 44]. For the formation of glycoluril diether 19d, we chose anhydrous acidic conditions to avoid hydrolysis of the
CO2Et groups of 19a (Scheme 3.8 c). The purification of 19d in pure form in high yields (90%) was easily accomplished by chromatography.
Scheme 3.9 Synthesis of glycoluril 26a. Conditions: (a) Toluene, H2NCONH2, TFA, rt, 40 min, then reflux, 17 h.
The potential use of glycoluril 19a in the synthesis of substituted Q[n] included its propensity to form glycoluril moiety with a monofunctional group (Scheme 3.8,
10a) which was attributed to its cyclic germinal diesters which underwent
- 115 - hydrolysis followed by decaboxylation. Compared with 19a, heating glycoluril
26a, which was prepared from dione 26 (Scheme 3.9) under similar conditions, did not result in the formation of a glycoluril moiety with a monofunctional group.
We believe that this estate-disubstituted glycoluril derivative holds excellent promise for the synthesis of particular Q[n] derivatives.
We have demonstrated that the preparation of a glycoluril with a functionalised five-membered carbon ring substituent was possible, albeit with a relatively low yield. A study of the synthesis of this type of glycoluril showed that, for those with a five-membered carbon ring substituent, their synthetic reactions could be accompanied by the formation of an endocyclic double bond in their substituent rings. Could this divergent pathway be responsible for the low yields of acquired glycolurils?
As an extension of the idea of avoiding the endocyclic double bond formation, we next considered diones with norbornane-based structures. Diones of this type appeared potentially valuable as they had the advantage of being unable to form double bonds as a by-product during their synthetic reactions to glycolurils and had a strained five-membered ring system with angles slightly smaller than those of cyclopentane, a desirable structural feature, as discussed in Chapter 2.
- 116 -
3.4.2 Synthesis of glycolurils with norbornane ring substituent
Scheme 3.10 Synthesis of glycoluril 47b. Five-membered rings in structures highlighted in bold. defined as angle (C2–C1–C6).
In this section, we discuss our synthetic exploration toward glycolurils with the norbornane structure as a substituent. -Dione 47 is the simplest norbornane structure that is potentially suitable for condensation with urea to synthesise a glycoluril such as 47b (Scheme 3.10). Within the norbornane structure shown in
Scheme 3.10, a five-membered carbon ring is highlighted to indicate its structural similarity to the cyclopentane-substituted glycolurils discussed in
Section 2.4.1 (Chapter 2). In this regard, a rudimentary comparison was made between the relevant internal ring angles, , of glycoluril 3 and an orthocarbonate-substituted norbornane 27 (Figure 3.3). As both of these structures had angles derivable from crystal structure data, we could determine that the appropriate average angles in 27 were ~ 2º smaller than those in glycoluril 3. Given that angle in the substituent of a glycoluril has been determined to have a direct correlation to the dihedral angle of the concave face of the glycoluril, this appeared to be significant and promising (Section 2.4.1,
Chapter 2). Potentially, if a glycoluril with a norbornane-based structure such as
47b also had a smaller angle, this could translate into a larger dihedral angle for
- 117 - the concave face of the glycoluril and, hence, improve the prospects for the synthesis of even greater proportions of SQ[n] as higher homologues (Chapter
2).
Figure 3.3 Structures of cyclopentanoglycoluril diether 3 and norbornane-exo- cis-2,3-dily 1',2'-phenylene orthocarbonate 27 ( (angle C6a–C6b–C6c) in 3
106.40° [6] and (angle C6–C1–C2) in 27 104.48° [45]).
Scheme 3.11 Synthesis of 28a and glycoluril 28c. Conditions: (a) Xylenes,
H2NCONH2, TFA, rt, 40 min, then reflux, 4 h; (b) PhH, H2NCONH2, TFA, rt, 40 min, then reflux,4 h; (c) Formamide, TFA, 140 °C, 6 h.
Although the synthesis of camphorquinoneglycoluril (28c) has been the only norbornane-type glycoluril reported, experimental details and spectroscopic
- 118 - data are scant (Scheme 3.11) [46]. We repeated the reported synthesis and found that the reaction was not as straightforward as implied. Camphorquinone
(28) condensed with urea in the presence of TFA to give glycoluril 28c (yield
15%) as the primary product but only if the reaction was carried out at high temperatures (> 110 °C). When camphorquinone (28) was reacted with urea in refluxing benzene with TFA, there was a preferential formation of the intermediate imadazolone urea 28a (21%). In this case, the steric constraint of the 1-methyl group of the camphor structure came into play and the temperature of 80 °C was insufficient to overcome such a steric barrier.
However, urea 28a in a solution of acidic (TFA) formamide heated to a temperature of 140 °C resulted in an incomplete cyclocondensation reaction even after extended periods of 6 h (yield 37%). Compared with other solvents, formamide was an excellent solvent for 28a as it had both polar and protic properties. These intrinsic properties were expected to stabilise the carbocation intermediate 28b and drive the condensation reaction forward to glycoluril 28c.
- 119 -
Scheme 3.12 Synthesis of diones 31, 40, 44 and 47. Conditions: (a)
OsO4/NMNO/Me2CO-H2O, 40 °C, 21 h; (b) DMSO-TFAA/DCM/TEA, -78 °C, 5.5 h; (c) BH3-THF, 0 °C; then sat. aq. NaHCO3, aq. H2O2, 30 min; (d) BzCl/Py, rt,
24 h; (e) HCl/MeOH, rt, 3 d; (f) CrO3, Py/DCM, rt, 6 h; (g) SeO2/xylenes, 140 °C,
24 h; (h) BzCl/Py, rt, 24 h; (i) HCl/MeOH, rt, 3 d; (j) CrO3, Py/DCM, rt, 6 h; (k)
SeO2/xylenes, 140 °C, 24 h; (l) OsO4/NMNO/Me2CO-H2O, 40 °C, 21 h; (m) DMSO-TFAA/DCM/TEA, -78 °C, 5.5 h.
Having identified the difficulties of synthesising camphorquinoneglycoluril 28c and then overcoming them, we turned our attention to the norbornane derivatives carrying functional groups. Not only did we want glycolurils that could be used for the synthesis of substituted Q[n] as higher homologues but also those which carried functionality.
- 120 -
The endo 5-acetate dione (31) was readily achievable in two steps from commercially available acetate 29 via diol 30 followed by Swern oxidation
(Scheme 3.12 b). Dione 31, under the condensation reaction conditions with urea and acid at 80 °C, afforded the partial condensation products of the imadazolone urea mixture (31a and 31b) (Scheme 3.13). Both 31a and 31b were isolated by crystallisation and fully characterised by NMR spectra. None of the anticipated glycoluril 31c was detected in the reaction mixture under these conditions.
Scheme 3.13 Synthesis of 31a and 31b. Conditions: (a) PhH, H2NCONH2, TFA, rt, 40 min, then reflux, 4 h.
From a steric point of view, the endo acetate substituent in 31b could explain the failure of the ring closure. In light of the effect of increasing the reaction temperature to overcome the thermal barrier found in the camphorquinoneglycoluril example, it was expected that it may also be possible to achieve 31d (Scheme 3.13). However, for dione 31, a complete
- 121 - decomposition of all materials was observed when increasing the reaction temperatures to 120–140 °C. Furthermore, heating urea 31a or 31b under similar conditions as urea 28a (140 °C in the acidic formamide solution) resulted in the decomposition of all starting materials. These results could indicate that the acetate group attached to the norbornane ring may have had a negative effect on the stability of the non-classical carbocation ion intermediate 31c, as shown in Scheme 3.13.
Concluding that the steric effect arising from an endo substituent was most likely responsible for the failure of the reaction to go to completion to produce
31d led us to think that this steric hindrance would be minimised by using diones such as 44 or 40 with exo substituents (Scheme 3.14).
Scheme 3.14 Synthesis of 44a, 40a and 47a. Conditions: (a) PhH, H2NCONH2, TFA, rt, 40 min, then reflux, 4 h.
Dibenzoatedione 44 was readily synthesised from norbornene 29 according to the procedures reported by Fernández et al. [35], with some variations to improve yields (see Experimental detail, Section 3.3). Curiously, only the imidazolidinonediol 44a was formed when dione 44 reacted with urea at 80 °C,
- 122 - with no glycoluril detected in the crude reaction mixtures (Scheme 3.14). The two exo benzoate groups at C5 and C6 in 44a (Scheme 3.14) were not expected to sterically hinder the approach of the second molecule of urea to form glycoluril 44b. Therefore, to validate this assumption, mono exo benzoate
40 was prepared (Scheme 3.12) [36]. Under the same condensation reaction conditions as dibenzoatedione 44, 40 exclusively gave the imidazolidiol product
40a (Scheme 3.14). This result was even more surprising as exo-benzoate 40 should at least have reacted in a similar way to the previously described endo- acetate 31 (Scheme 3.13) to produce a urea product analogous to 31b.
Reactions carried out in refluxing xylenes (140 °C) for both 44 and 40 only led to a complete decomposition. To add to this puzzle, the preferential formation of imidazolidinonediol 47a was also found for the reaction of norbornadione 47 which carried no substituents. These results were not consistent with our experience of synthesising camphorquinoneglycoluril (28c) which was prepared directly from dione 28 at higher temperatures. Attempts to achieve the same by using higher reaction temperatures in solvents of xylenes or formamide failed and resulted in only decomposition.
- 123 -
Scheme 3.15 Non-classical carbocation ion intermediate involved in synthesis of norbornane-type glycoluril.
The synthetic reactions of norbornane-type glycolurils involved formations of imidazolidinonediols (step a in Scheme 3.15) which appeared relatively facile, as seen in examples 28a, 31a, 40a, 44a and 47a. Under conditions of acid catalysis, the overall reaction may have proceeded via a non-classical carbocation ion intermediate (Scheme 3.15) which may have been destabilised by the ester group or may have facilitated decomposition. However, this was unclear. It seemed likely that a key factor in this successful transformation was the presence of alkyl substituents on the norbornane ring at C2 or perhaps C3
(Scheme 3.15) which was able to stabilise the non-classical carbocation ion intermediate.
- 124 -
Scheme 3.16 Synthesis of dione 51. Conditions: (a) AcCl/Py, rt, 24 h; (b)
OsO4/NMNO/Me2CO-H2O, 40 °C, 21 h; (c) DMSO-TFAA/DCM/TEA, -78 °C, 5.5 h.
Scheme 3.17 Synthesis of 51a. Conditions: (a) PhH, H2NCONH2, TFA, rt, 40 min, then reflux, 4 h.
As a final attempt to solve this puzzle, we investigated the urea condensation of dione 51 which carried alkyl substituents at C3 and C4 as CH2OAc (Scheme
3.17). Dione 51 was readily synthesised from diol 48, as shown in Scheme 3.16.
Surprisingly, again only a partial condensation product, 51a, was formed in the reaction mixture at the temperatures of refluxing benzene, toluene or xylenes. A complete decomposition of 51a in acidic formamide was observed when the reaction mixture was heated to 140 °C or above.
- 125 -
We had examined six diones with norbornane structures (28, 31, 40, 44, 47 and
51) and, in each case, the first step to condensation with urea toward the synthesis of a glycoluril appeared facile to give imadazolidinonediols (31a, 40a,
44a, 47a and 51a). The exception to this was the reaction of the camphorquinone 28 which gave the product of the first stage of the second urea condensation, urea 28a. This reaction proceeded without stopping at the diol.
Therefore, the objective of synthesising new glycolurils with norbornane substituents each of which also carried a functional group was not achieved.
This is difficult to explain, except to suggest that the substituents may not have been sufficiently stable to the reaction conditions required to achieve them. It was found that the reaction of the second urea condensation could be progressed to some degree (e.g., ureas 28a and 31b) with higher reaction temperatures but only if their intermediate products were sufficiently stable. The ring closure of ureas 28a and 31b to achieve a glycoluril, could only be effected for urea 28a. Also, contrary to expectations, urea 31b could be obtained even though the benzoate exo analogue could not. It seems that the solutions to these synthetic problems may lie in finding methods for glycoluril synthesis that can operate under milder conditions.
Instead of the regular acid catalyst methods, a recently published procedure used polyoxometalate catalysts for synthesising some relatively simple glycolurils [47] which may suggest that alternative catalysts may be the future for improving glycoluril yields. Recently, our group investigated an alternative to acid catalyst methods and found that the Lewis acid-type catalyst Bi(OTf)3 could overcome some of the difficulties discussed here. Related studies involving
- 126 - synthesising glycolurils with a norbornane ring utilising this catalyst are being conducted by our research group [48].
- 127 -
3.5 Conclusions
In this chapter, the synthesis of a number of glycoluril derivatives with a functionalised five-membered carbon ring as a substituent has been described.
It was found that syntheses of glycolurils of this type were usually accompanied by the formation of endocyclic double bonds in their substituent rings, resulting in low yields of the acquired glycolurils. We examined the norbornane structure as a substituent for glycolurils as it had the advantage of being unable to form a double bond as a by-product and had strained five-membered ring skeletons with angles comparable to those of a cyclopentane ring substituent. Some original intermediates were found on the path to the formation of norbornane- type glycolurils. It was also discovered that the first step in diketone-urea condensation was a facile process while the second was a more difficult reaction, with the potential for an alternative path, such as double bond formation or the prevention of reaction due to steric constraints. The formation of stable urea intermediates that could not be closed to form the second imidazolidinone ring in order to realise a glycoluril also pointed to the difficulty of the reaction. Collectively, these results help to explain why some aliphatic glycolurils are difficult to synthesise, especially when alternative reaction paths may be possible.
Furthermore, subsequent to this study, it has been found that the glycoluril- formation reaction appears to be more efficiently carried out using the Lewis acid-type catalyst Bi(OTf)3. We are currently engaged in further research to exploit this new catalyst for the preparation of new glycolurils.
- 128 -
3.6 References
1. Mock, W. L.; Shih, N. Y. J. Org. Chem. 1986, 51, 4440–4446.
2. Liu, S.; Ruspic, C.; Mukhopadhyay, P.; Chakrabarti, S.; Zavalij, P. Y.;
Isaacs, L. J. Am. Chem. Soc. 2005, 127, 15959–15967.
3. Rekharsky, M. V.; Mori, T.; Yang, C.; Ko, Y. H.; Selvapalam, N.; Kim, H.;
Sobransingh, D.; Kaifer, A. E.; Liu, S.; Isaacs, L.; Chen, W.; Moghaddam,
S.; Gilson, M. K.; Kim, K.; Inoue, Y. Proc. Natl. Acad. Sci. U.S.A. 2007,
104, 20737–20742.
4. Kaifer, A. E.; Li, W.; Yi, S. Isr. J. Chem. 2011, 51, 496–505.
5. Nau, W. M.; Florea, M.; Assaf, K. I. Isr. J. Chem. 2011, 51, 559–577.
6. Wu, F.; Wu, L.-H.; Xiao, X.; Zhang, Y.-Q.; Xue, S.-F.; Tao, Z.; Day, A. I. J.
Org. Chem. 2012, 77, 606–611.
7. Masson, E.; Ling, X.; Joseph, R.; Kyeremeh-Mensah, L.; Lu, X. RSC Adv.
2012, 2, 1213–1247.
8. Day, A. I.; Collins, J. G. Supramol. Chem. Mol. Nanomater. 2012, 3, 983–
1000.
9. Lim, S.; Kim, H.; Selvapalam, N.; Kim, K.-J.; Cho, S. J.; Seo, G.; Kim, K.
Angew. Chem., Int. Ed. 2008, 47, 3352–3355.
10. Huber, G.; Legrand, F.-X.; Lewin, V.; Baumann, D.; Heck, M.-P.;
Berthault, P. ChemPhysChem. 2011, 12, 1053–1055.
11. Kim, H.; Kim, Y.; Yoon, M.; Lim, S.; Park, S. M.; Seo, G.; Kim, K. J. Am.
Chem. Soc. 2010, 132, 12200–12202.
12. Mock, W. L.; Irra, T. A.; Wepsiec, J. P.; Adhya, M. J. Org. Chem. 1989,
54, 5302–5308.
- 129 -
13. Jon, S. Y.; Ko, Y. H.; Park, S. H.; Kim, H.-J.; Kim, K. Chem. Commun.
2001, 1938–1939.
14. Pattabiraman, M.; Natarajan, A.; Kaanumalle, L. S.; Ramamurthy, V. Org.
Lett. 2005, 7, 529–532.
15. Wang, R.; Yuan, L.; Macartney, D. H. J. Org. Chem. 2006, 71, 1237–
1239.
16. Kloeck, C.; Dsouza, R. N.; Nau, W. M. Org. Lett. 2009, 11, 2595–2598.
17. Basilio, N.; Garcia-Rio, L.; Moreira, J. A.; Pessego, M. J. Org. Chem.
2010, 75, 848–855.
18. Lu, X.; Masson, E. Org. Lett. 2010, 12, 2310–2313.
19. Pemberton, B. C.; Singh, R. K.; Johnson, A. C.; Jockusch, S.; Da Silva, J.
P.; Ugrinov, A.; Turro, N. J.; Srivastava, D. K.; Sivaguru, J. Chem.
Commun. 2011, 47, 6323–6325.
20. Tuncel, D.; Uenal, O.; Artar, M. Isr. J. Chem. 2011, 51, 525–532.
21. Jon, S. Y.; Selvapalam, N.; Oh, D. H.; Kang, J.-K.; Kim, S.-Y.; Jeon, Y. J.;
Lee, J. W.; Kim, K. J. Am. Chem. Soc. 2003, 125, 10186–10187.
22. Lee, D.-W.; Park, K. M.; Banerjee, M.; Ha, S. H.; Lee, T.; Suh, K.; Paul,
S.; Jung, H.; Kim, J.; Selvapalam, N.; Ryu, S. H.; Kim, K. Nat. Chem.
2011, 3, 154–159.
23. Petersen, H. Synthesis 1973, 243–292.
24. Lagona, J.; Fetinger, J. C.; Isaacs, L. Org. Lett. 2003, 5, 3745–3747.
25. Grillon, E.; Gallo, R.; Pierrot, M.; Boileau, J.; Wimmer, E. Tetrahedron
Lett. 1988, 29, 1015–1016.
26. Chegaev, K. Y.; Kravchenko, A. N.; Lebedev, O. V.; Strelenko, Y. A.
Mendeleev Comm. 2001, 11, 32–33.
- 130 -
27. Kvittingen, L.; Partali, V.; Braenden, J. U.; Anthonsen, T. Biotech. Lett.
1991, 13, 13–18.
28. Malanga, C.; Pagliai, L.; Menicagli, R; Synth. Commun. 1990, 20, 2821–
2826.
29. Murdock, K. C.; Angier, R. B. J. Org. Chem. 1962, 27, 2395–2398.
30. Boeda, F.; Clavier, H.; Jordaan, M.; Meyer, W. H.; Nolan, S. P. J. Org.
Chem. 2008, 73, 259–263.
31. Mwangi, M. T.; Schulz, M. D.; Bowden, N. B. Org. Lett. 2009, 11, 33–36.
32. Julie, B.; Nicolas, J.; Silvia, D. G.; Sabine, B. R.; Thomas, Z.; Steven, P.
N.; Luigi, A. A. Tetrahedron 2009, 65, 1162–1170.
33. Adrio, L. A.; Quek, L. S.; Taylor, J. G.; Hii, K. K. M. Tetrahedron 2009, 65,
10334–10338.
34. Perch, N. S.; Widenhoefer, R. A. J. Am. Chem. Soc. 1999, 121, 6960–
6961.
35. Hergueta, A. R.; López, C.; García-Mera, X.; Fernández, F. Tetrahedron
2004, 60, 10343–10352.
36. Šala, M.; De Palma, A. M.; Hrˇebabecky´, H.; Nencka, R.; Dracˇinsky´, M.;
Masojidkova, M.; Leyssen, P.; Neyts, J.; Holy´, A. Bioorg. Med. Chem.
2010, 18, 4374–4384.
37. Liu, B.; Li, Y.; Shin, B.-G.; Yoon, D. Y.; Kim, I.; Zhang, L.; Yan, W. J.
Polym. Sci., A: Polym. Chem. 2007, 45, 3391–3399.
38. Wu, L.-H.; Ni, X.-L.; Wu, F.; Zhang, Y.-Q.; Zhu, Q.-J.; Xue, S.-F.; Tao, Z.
J. Mol. Struct. 2009, 920, 183–188.
39. Gompper, R.; Noth. H.; Rattay, W.; Schwarzensteiner, M.-L.; Spes, P.;
Wagne, H.-U. Angew. Chem., Int. Ed. 1987, 26, 1039–1041.
- 131 -
40. Gault, H. Compt. Rend. 1910, 150, 1241–1242.
41. Sasmal, S.; Sinha, M. K.; Keinan, E. Org. Lett. 2004, 6, 1225–1228.
42. Depre´s, J.-P.; Greene, A. E. J. Org. Chem. 1984, 49, 928–931.
43. Zhao, Y.-J.; Xue, S.-F.; Zhu, Q.-J.; Tao, Z.; Zhang, J.-X.; Wei, Z.-B.; Long,
L.-S.; Hu, M.-L.; Xiao, H.-P.; Day, A. I. Chin. Sci. Bull. 2004, 49, 1111–
1116.
44. Zheng, L.-M; Zhu, J.-N.; Zhang, Y.-Q.; Zhu, Q.-J.; Xue, S.-F.; Tao, Z.;
Zhang, J.-X.; Zhou, X.; Wei, Z.; Long, L.; Day, A. I. Supramol. Chem.
2008, 20 (8), 709–716.
45. Richard, B.; Peter, K. Acta Crystallogr., Sect. E 2008, E64, o399.
46. Tudolf, G.; Heinrich, N.; Wilfried, R.; Marie-Luise, S.; Peter, S.; Hans-
Ulrich, W. Angew. Chem., Int. Ed. 1987, 26, 1039–1041.
47. Esmail, R-S.; Reza, Palma, T. J. Chem. Pharm. Res. 2011, 3, 103–107.
48. Day et al. unpublished results.
- 132 -
CHAPTER 4
GUEST EXCHANGE MODERATED BY SUBSTITUENTS ON THE HOST
- 133 -
CHAPTER 4
Guest exchange moderated by substituents on the host
4.1 Introduction
The molecular recognition that is of principal supramolecular interest, is a nanoscopic process through which molecular components are assembled into large supramolecular assemblies, which often have prominent features strikingly different from those of their precursors [1]. The driving forces for the formation of assemblies are noncovalent bonding, such as hydrogen bonding, hydrophobic forces, van der Walls forces and ion–dipole and dipole–dipole interactions [1]. Molecular recognition in supramolecular chemistry sometimes displays exchanges of guest molecules within assemblies of host–guest structures, that depend upon the chemical natures of the host or guest molecules.
Several studies of molecular recognition that proceed through guest exchanges within assemblies have been reported [1, 2]. As this project focuses on the study of the supramolecular chemistry of Q[n], an illustrative example of the kinetic study of self-sorting systems based on Q[n] complexes, that reported by
Isaacs et al. is highlighted (Scheme 4.1). They found that, at the initial stage of the reaction, the molecular host pair Q[6]/Q[7] accommodate the adamantanebutylammonium (Scheme 4.1, 2a) at its butyl moiety and
- 134 - cyclohexanediammonium (Scheme 4.1, 1) under kinetic control, respectively.
After 56 days, a thermodynamically more stable system, with the adamantyl unit bound inside Q[7] and the cyclohexyl unit in Q[6], was observed (Scheme 4.1)
[3]. This reaction was driven by the 4.3 × 103-fold gain in binding affinity when the adamantane moiety displaced the cyclohexane unit within Q[7] [4]. This self- sorting process occured in several steps (Scheme 4.1) which constituted the reaction mechanism. A mechanistic understanding of guest exchange, including the mechanisms of guest molecules entering or exiting the host cavity, is significant for controlling self-sorting systems of this type.
Scheme 4.1 Example of kinetic and thermodynamic self-sorting. The figure was taken from [3].
The cucurbit[n]uril (Q[n]) family is one of several supramolecular hosts that has become important as a noncovalently bonding agent. The structural features that give it this property have been discussed in detail in Section 1.2, Chapter 1,
- 135 - with two of the most important features being the ureidyl-carbonyl lined portals and the hydrophobic interior of a Q[n]. The portals allow a guest molecule access to this hydrophobic interior which is slightly larger in diameter (~ 1.9 Å larger) than its portals [5, 6]. In general, many host–guest exchange processes are measured on a time scale of fractions of seconds to minutes which is particularly true for certain molecular hosts, such as cyclodextrins and calixarenes [7-11], and some reports also involve Q[n] [5, 12]. In contrast, in their pioneering work, Werner et al. demonstrated that the complexation kinetics of Q[6] is much slower than those observed for host systems featuring unobstructed openings [7-11, 13]. This is attributed mainly to the relatively tight portals of Q[6] which can cause restrictions relating to a physical barrier against complexation [13]. Examples of slow guest exchanges within Q[n] can also be found in Isaacs’s self-sorting system described above [3]. Since the host Q[n] does not require any part of its structure to be dismantled to effect exchange, the mechanism for such a slow exchange involves restrictions relating to the tightness of binding and relative selectivity [3].
We have extended this exploration of exchange by focusing on tightly bound guests, using very slow exchange kinetics as a tool, to analyse the effect of subtle derivative variations at the equatorial skirt of the cavity of the host, substituted Q[7]. Our recent success with the synthesis of fully substituted cyclopentanoQ[7] (CyP7Q[7]) provided us with the opportunity to evaluate the possible exchange kinetic differences between the simplest Q[7] and a substituted counterpart, CyP7Q[7]. This appeared to be especially poignant in
- 136 - view of an observed increase in the relative binding constant of the adamantyl ammonium ion when comparing the hosts Q[7] and CyP7Q[7] [14].
4.2 Aims of study
The aims of this section of the thesis are as follows.
1. To examine the influence of substituents on a substituted Q[7] upon the
exchange rate of molecular guests to determine the effects of
substituents on the electronic character of either the Q’s portals or cavity.
This arose as an extension of the discovery that the binding constants
increased, as discussed in Chapter 2.
2. To explore the mechanism of molecular guest exchange within Q[7]. It
was expected that the relatively high affinity of a pre-existing complexed
guest toward Q[7] would have a significant effect upon the mechanism of
guest exchange.
- 137 -
4.3 Experimental
4.3.1 Materials
Figure 4.1 Cucurbituril hosts and guests examined.
Figure 4.1 shows the cucurbituril hosts and guests studied in this chapter. Q[7] and CyP7Q[7] were prepared according to the methods developed in our labs
[14, 15]. A mixture of Ada@Me2Q[7]·PF6 and Ada@,-Me4Q[7]·PF6 in a ratio of
7:3 was synthesised and isolated, as described in this chapter. All guest@substituted Q[7] complexes were recrystallised solids which showed typical pD 3.5 when prepared as reaction kinetics solutions. The known ferrocene-based guest molecules (BmaFc, MaFc and MaFcH) were synthesised and purified according to procedures in the literature [16, 17] while the BmaFcH and AdaG guests were synthesised as described in this chapter. The adamantyl ammonium ion (Ada) was purchased from Aldrich and D2O obtained from
- 138 -
Cambridge Isotope Laboratories. Unless otherwise indicated, commercial materials were used as received without purification.
4.3.2 1H NMR experiments
All spectra were recorded at 400 MHz on a Varian Unityplus-400 spectrometer and analysed using Varian’s VNMR software.
4.3.3 Experimental procedures
3-(Adamantan-1-ylamino)propane-1,2-diol
A mixture of amantadine (1.40 g, 9.27 mmol), glycidol (0.23 g, 3.10 mmol) and
MeOH (40 mL) was heated at reflux for 24 h. The reaction mixture was concentrated by rotary evaporation and dried under high vacuum. The resulting product was purified by chromatography (Al2O3, EtOH) to give a white solid, which was treated with HCl to yield the hydrochloride salt of AdaG. The product was washed with CH3COCH3 (10 mL) and then dissolved in hot CH3CN (30 mL).
After a few hours the salt was obtained as a microcrystalline solid by filtration
(0.49 g) with a yield of 70% based upon glycidol. Mp 179 °C. IR (KBr, cm-1):
3213 s, 3090 s, 3024 s, 2914 s, 2856 s, 1610 w, 1384 m, 1364 m, 1129 w, 1077
1 s, 1042 w, 930 w. H NMR (D2O): 3.98–3.90 (1H, m), 3.69–3.57 (2H, m),
- 139 -
3.25–3.19 (1H, m), 3.06–2.99 (1H, m), 3.21 (3H, s), 1.92 (6H, t, J = 16.0 Hz),
13 1.80–1.64 (6H, m). C NMR (D2O): 67.7, 63.3, 57.9, 41.8, 37.8, 34.8, 28.8.
+ + MS (ESI): m/z 226 (100, [M] ). HR-MS for C13H24NO2 [M] : calcd 226.1807, found 226.1807.
Ferrocene-1,1'-diylbis(methyldimethylammonium bromide)
1.35 g of HBr (30 wt.% in acetic acid) was slowly added to a solution of 1,1'- bis(N,N-dimethylaminomethyl)ferrocene [18] (0.75 g, 2.50 mmol) in MeOH (15 mL) and the resulting solution heated under reflux for 10 min before being allowed to cool. Dry ether (80 mL) was added and BmaFcH precipitated as a yellow solid. This was collected by filtration, washed with CH2Cl2 (10 mL) and then ether (10 mL), before being dried (0.62 g, 54%). The salt BmaFcH decomposed above 196 °C but did not melt below 260 °C. IR (KBr, cm-1): 3417
1 s, 2953 m, 2700 s, 2656 s, 2033 w, 1384 s, 1243 w, 1043 w, 1011 w, 921 m. H
13 NMR (D2O): 4.53 (4H, b s), 4.46 (4H, b s), 4.20 (4H, s), 2.77 (12H, s). C
+ NMR (D2O): 75.0, 71.9, 71.3, 57.0, 41.5. MS (ESI): m/z 301 (100, [M – H] ).
+ HR-MS for C16H25N2Fe [M – H] : calcd 301.1367, found 301.1367.
- 140 -
Ada@Me2Q[7]·PF6 and Ada@,-Me4Q[7]·PF6 (7:3)
The methylQ[7]s were prepared according to our previously reported synthesis with the following modifications [19]. To the crude reaction mixture from the reported procedure, adamantyl amine HCl (Ada) was added and this mixture subjected to chromatography on a Dowex cation-exchange resin eluting with a mixture of 0.5 M HCl in 50% formic acid. The fractions containing the mixture of
Ada@Me2Q[7] and Ada@Me4Q[7] salts were treated with NH4PF6 and crystallised from water to give a mixture of Ada@Me2Q[7]·PF6 and
Ada@Me4Q[7]·PF6. As repeated crystallisation from water could not separate the two methylQ[7]s, they were used as a mixture in a ratio of 7:3, respectively.
As each product was indistinguishable from the other by 1H NMR, the ratio was calculated from the integral of the relative integrals of methine and methyl
1 resonances. H NMR (D2O): 5.83 (14H, m), 5.61 (11.66H, narrow m), 4.35
(14H, m), 1.89 (7.9H, s); [1.51 (3H, s), 1.23 (3H, d, J = 10.4 Hz), 1.21 (6H, s),
0.94 (3H, d, J = 10.4 Hz) adamantyl resonances]. MS (ESI): m/z 1371 (60,
+ + [adamantylNH3 @Me4Q[7]]), 1343 (60, [adamantylNH3 @Me2Q[7]]), 697 (100,
+ 2+ + [adamantylNH3 @Me4Q[7] + Na] /2, 683 (70, [adamantylNH3 @Me2Q[7] +
2+ Na] /2. HR-MS for C56H68N29O14: calcd 1370.5505, found 1370.5501; and for
C54H64N29O14: calcd 1342.5188, found 1342.5250.
- 141 -
Guest Exchange Studies in Substituted Q[7]s All the exchange reactions were carried out in D2O at the specified temperatures. The sample NMR tube was inverted several times to ensure proper mixing of the reacting solutions. The times for solution mixing were not included in data analyses. The 1H NMR spectra were recorded at set intervals using an automated routine and acquired with delay times of 5 times the longest T1 to ensure that systematic errors due to differences in relaxation times were eliminated. Free and encapsulated guest species in the host molecules (Q[7], Me2Q[7]/Me4Q[7] and CyP7Q[7]) were easily identified by 1H NMR as the interior of host Q constituted of a 1H NMR shielding region, thus resulting in significant upfield shifts of the 1H resonances of the encapsulated species while resonances of the free guest appeared further downfield. The interchange reaction of the Ada guest between hosts
1 (CyP7Q[7] and Q[7]) was also indentified by H NMR.
Initial Rates Method for Determining Rate Law of Exchange Reaction All stock solutions were prepared in D2O and equilibrated to 37 °C. Each reaction solution was prepared in a NMR tube from the same stock solutions of BmaFc or Ada@Q[7]. Table 4.1 shows the initial concentrations of BmaFc and
Ada@Q[7] required by the experiments. The NMR tubes containing Ada@Q[7] and BmaFc were inverted several times to ensure complete mixing of the reacting solutions and 1H NMR spectra were recorded at 15 min intervals at
37 °C using an automated routine.
- 142 -
Table 4.1 Initial concentrations of BmaFc and Ada@Q[7] for initial rates method.
[BmaFc][a] [Ada@Q[7]][a] Experiment (mM) (mM) 1 0.5 0.5 2 1 0.5 3 0.5 1
[a] Initial concentration.
Comparative Binding Studies Comparative binding studies for guest 1 (0.250 mM) relative to guest 2 and a limited quantity of host Q[7] (relative initial ratio
1:1:1) were performed in D2O and the whole system kept at 85 °C over 2 h, as determined by 1H NMR.
4.4 Results and discussion
The guest@substituted Q[7] complexes involved in this study all exhibited slow exchange between the free and bound guests on the 1H NMR time scale. Guest exchange in guest@substituted Q[7] complexes with a variety of guests underwent sufficiently slow exchange that the kinetic process could be easily followed by 1H NMR. In order to obtain clearly defined changes over time, guests with at least one clearly resolved resonance were chosen. This condition was necessary for either free or encapsulated guests in order to more accurately determine mole proportions of each species over time. Also, a deliberate choice was made to exclude the use of buffers in order to avoid the participatory effects of their ions in the exchange process [3, 20].
- 143 -
Section 2.4.4.2 (Chapter 2) described the thermodynamic difference in the
+ complexation behaviours of CyP7Q[7] and Q[7] toward the adamantylNH3 ion
(Ada) and established their relative binding constant as KCyP7Q[7] = 1.9 KQ[7] using a 1H NMR competition experiment. The rate of the guest Ada interchange between the two hosts (CyP7Q[7] and Q[7]) was found to be extremely slow and the experiment at rt only proceeded to < 12% toward equilibrium after 45 days!
We considered that the factors which might have retarded the rate of interchange could be the bulkiness of the guest (Ada) and the substituents of the molecular host. The substitution, as discussed in Chapter 2, could potentially increase the electron density on the carbonyl O of the glycoluril moiety, thereby increasing ion–dipole interactions and, hence, binding constants. This specific example led us to hypothesise that the study of hosts with substituents in guest exchange supramolecular systems, might result in interesting dynamic behaviour and provide an understanding that could prove useful in future chemical manipulations.
- 144 -
4.4.1 BmaFc/Ada guest exchange in substituted Q[7]s
Scheme 4.2 General scheme for displacement of Ada (0.250 mM) inside substituted Q[7]s (R = H, methyl, and cyclopentano, 0.250 mM) by BmaFc (1.125 mM) at 37 °C.
The encapsulated guest Ada within the cavity of Q[7] (Scheme 4.2, Q[7] R = H)
14 -1 12 -1 (Ka = 1.7 × 10 M neutral conditions, 4 × 10 M buffered [21, 22]) had the potential to be replaced or displaced by another molecule providing the second molecule either had a higher affinity or was present in a large excess. This was efficiently achieved by the bismethylammoniummethyl ferrocene ion (BmaFc) which showed a slow exchange process (Table 4.2, t1/2 = 8 h, 37 °C). The higher binding affinity for BmaFc (3 × 1015 M-1) [21, 23] drove the equilibrium strongly to the RHS of the exchange process (Scheme 4.2, R = H) but should have had little effect on the rate of exchange.
The 1H NMR spectra for this exchange reaction are shown in Figure 4.2 in which upfield-shifted resonances of the encapsulated Ada are clearly identified.
A relatively higher concentration (3 mM) was employed in this case in order to
- 145 - obtain clear spectra within a reasonable time period following the exchange
(Figure 4.2).
Figure 4.2 Representative 1H NMR data following exchange of Ada for BmaFc within Q[7] ([Ada@Q[7]], [BmaFc] = 3 mM, T = 37 °C).
Interestingly, the exchange for the same guest pair as above, but for the fully substituted cyclopentanoQ[7] (CyP7Q[7] where R = ring (CH2)3 in Scheme 4.2), underwent a very slow exchange of t1/2 = 38 h (Table 4.2) and an intermediate effect was found for the partially substituted methylQ[7], Me2Q[7]/Me4Q[7] (7/3), which gave an exchange rate intermediate between Q[7] and CyP7Q[7] of t1/2 =
18 h (Table 4.2). Due to their inseparable natures, a mixture of Me2Q[7]/Me4Q[7]
(7/3) was assessed in our experiment (see Section 4.3.3). The rates of exchange within CyP7Q[7], Me2Q[7]/Me4Q[7] and Q[7] produced an anticipated trend which related the exchange rates to the number of alkyl substituents on a
- 146 - substituted Q[7]. A slow guest exchange was expected when this number increased.
Table 4.2 Comparison of rates of guest exchange relative to different substituents on host substituted Q[7] at 37 °C.
Host Substituted Q[7] Guest Rate of Exchange
Ingress/Egress t1/2 (h) Q[7] 8 BmaFc/Ada[a] Me2Q[7]/Me4Q[7] (7/3) 18
CyP7Q[7] 38
[a] BmaFc was added at a 4.5 mole equiv. relative to Ada@substituted Q[7] (0.250 mM).
In their pioneering work, Mock et al. studied the exchange of various guests bearing ammonium ions in the cavity of Q[6] and indicated that the mechanism of a guest exchange of this type involves a two-step process, the vacation of the cavity by dissociation and its filling by an alternative guest [12]. The examples presented here illustrate guest exchanges of another type within substituted Q[7] which are much slower in their rates of exchange than those reported by Mock et al. [12]. In the following section, the exchange reactions within normal Q[7] involving another three pairs of guests are employed to explore the mechanism of guest exchange to better understand the exchange process.
4.4.2 Guest exchange mechanism in Q[7]
One of the significant structural features of the Q[n] family is their two openings
(portals), the diameters of which are approximately 2 Å narrower than the cavity of the macrocycle [5]. For the purposes of distinction, we refer to the Q[n] as an
- 147 -
‘open-door’ type of host in a supramolecular guest exchange system due to the presence of two opposite and permanent openings to its cavity. The freedom for exchange is explicit for ‘open-door’ hosts, such as cyclodextrin [24, 25] and cucurbituril [5], but less so for alternative hosts, such as capsules derived from calixarenes (including pyrogallolarenes and resorcinarenes) [26, 27], glycoluril derivatives [28, 29] and metal-coordinated catachol cages [1, 29].
As described above, as an ‘open-door’ host, Q[7] has two opposite and equal- sized openings to the cavity which, importantly, are permanently open. Given this condition, there would appear to be two possible mechanisms for a guest exchange within Q[7] (Scheme 4.3): dissociation and replacement; or displacement e.g., ‘shunting’. A displacement or ‘shunting’ mechanism proceeds when a guest can simultaneously enter the cavity as the bound guest exits, which is only possible due to the existence of two opposite and equal- sized openings. A replacement mechanism involves the cavity firstly being vacated by dissociation and then being filled by a free guest.
Scheme 4.3 BmaFc/Ada exchange within Q[7] at 37 °C.
Based on a general knowledge of guest exchange reactions within a molecular host, the reactions of this type are generally reversible so that the potential reversibility of a reaction must be taken into account, as indicated in Scheme
4.3. In the initial stage, only the reactants (BmaFc and Ada@Q[7]) were present but, as the reaction proceeded, the concentration of reactants decreased and
- 148 - that of the product increased. Finally, an equilibrium was reached at which point no further change in concentrations of the reactants and products occurs. We chose to explore the initial stage of exchange to probe the potential mechanism as this gave the advantage of simplicity. The reverse reaction could be largely ignored when considering low concentrations of free Ada and BmaFc@Q[7]
(products in Scheme 4.3).
Of the two potential mechanisms, (a) displacement and (b) replacement, one took the dominant role in the initial stage of guest exchange (Scheme 4.3), as illustrated by the example of exchange between the BmaFc/Ada guest pair within Q[7] in Scheme 4.4. The depiction of a particular arrangement of intermediate states in mechanism (a) is discussed in detail later in this chapter.
- 149 -
Scheme 4.4 Proposed mechanisms for BmaFc/Ada initial exchange within Q[7]: (a) guest displacement; and (b) guest replacement.
Increasing the concentration of BmaFc or Ada@Q[7] (Scheme 4.3) at a temperature of 37 °C resulted in an increase in the rate of exchange. The following method of initial rates was used to determine the order of reaction:
the rate law of which has the form:
where [BmaFc] and [Ada@Q[7]] express the concentrations of the BmaFc guest and Ada@Q[7] complex, respectively. Experimental conditions for the method of initial rates are described in Section 4.3.3.
- 150 -
Table 4.3 Data by initial rates method.
[b] [b] Experiment[a] [BmaFc] [Ada@Q[7]] Initial Rate (mM) (mM) (mM·s-1) 1 0.5 0.5 7.2 × 10-6
2 1 0.5 1.4 × 10-5
3 0.5 1 1.4 × 10-5
[a] Exchange temperature is 37 °C. [b] Initial concentration.
The order of the exchange reaction for BmaFc and Ada@Q[7] was calculated from the initial rates data in Table 4.3, and it was found that the experiments gave a value of 1 for both x and y in the rate equation. Therefore, the exchange reaction of BmaFc/Ada within Q[7] exhibited second-order kinetic behaviour, with a rate constant of k = 2.78 × 10-5 (mM·s)-1.
From the point of view of the rate law, it is apparent that the BmaFc/Ada exchange within Q[7] (Scheme 4.3) was substantially different from the examples of exchanges of substituted ammonium ion guest pairs within Q[6] reported by Mock et al. The rates of exchange in their study were little influenced by the concentration of the incoming guest [12]. Based on their observations, Mock et al. concluded that the exchange rate is independent of the concentration of the incoming guest; that is, the initial guest dissociates from the cavity in a slow step, followed by a competition between the initial and incoming guests for the empty cavity [12]. This replacement mechanism could also potentially operate for our exchange system of BmaFc/Ada within Q[7] (Scheme
4.4 (b)). To illustrate this point, an experiment on the interchange of the Ada guest between hosts Q[7] (0.250 mM) and CyP7Q[7] (ratio 1:1), with a relative
- 151 - binding affinity of KCyP7Q[7] = 1.9 KQ[7] (see Chapter 2 for determination of relative binding affinity), was conducted (Scheme 4.5) and can provide some insight into the above assumption.
Scheme 4.5 Interchange of Ada between CyP7Q[7] and Q[7].
The interchange reaction in Scheme 4.5 was chosen as a reference reaction because it proceeded by rate-determining dissociation of the Ada@CyP7Q[7] complex to an empty CyP7Q[7] and a free guest Ada. This dissociation was followed by Ada’s rapid association with an empty Q[7] or re-association with a free CyP7Q[7]. Independent experiments have shown that the Ada’s association with a free Q[7] is so rapid that thermodynamic equilibrium is immediate upon the mixing of the two components. At the initial stage of the exchange process, the reverse reaction is considered negligible because of the lower concentration of free CyP7Q[7] than Q[7] in the solution, and the mechanism of the interchange of Ada between CyP7Q[7] and Q[7] is illustrated in Scheme 4.6.
- 152 -
Scheme 4.6 Proposed mechanism for initial stage of interchange of Ada between CyP7Q[7] and Q[7].
The process of the interchange of Ada between CyP7Q[7] and Q[7] is shown in
Figure 4.3 for which changes in the concentrations of Ada@Q[7] and
Ada@CyP7Q[7] from their initially specified concentrations were determined over time at 37 °C.
0.25 Ada@Q[7] 0.2 Ada@CyP7Q[7]
0.15
0.1 Conc (mM) 0.05
0 0 100 200 300 400
Time (Hours)
Figure 4.3 Interchange of Ada (0.250 mM) between CyP7Q[7] (0.250 mM) and Q[7] (0.250 mM) at 37 °C.
The whole interchange process proceeded at an initial rate that was approximately linear (Figure 4.3). The initial rate of the Ada guest interchange
- 153 - between Q[7] and CyP7Q[7] (Scheme 4.6) was compared with the exchange reaction of Ada@CyP7Q[7] with BmaFc (Scheme 4.7) and it was found that the
-8 -1 former occurred 11 times more slowly (ki = 8.3 × 10 mM·s ) than the latter (ki
= 8.9 × 10-7 mM·s-1). This result clearly demonstrated that the dissociation or replacement mechanism (Scheme 4.4 (b)) could not be ruled out, but that its contribution to the overall exchange rates in Schemes 4.3 and 4.7 did not appear to be significant.
Scheme 4.7 BmaFc/Ada exchange within CyP7Q[7] at 37 °C.
The fact that the rate of exchange was much faster than the dissociation step supports an alternative mechanism which we describe here as ‘shunting’.
Accordingly, the initial Ada@Q[7] complex was attacked by an incoming BmaFc guest from the side opposite the leaving Ada guest, with the former entering simultaneously with the latter leaving. This ‘shunting’ is pictorially illustrative for describing the unique feature of this mechanism in which the incoming guest squeezed through one portal and knocked the resident guest out off the opposite equal-sized portal. This process of attack by an incoming BmaFc guest initially generated an association with the portal of Q[7] prior to displacement.
One piece of evidence in favour of the existence of this intermediate association complex was the observation that a titration of BmaFc into a solution of
Ada@Q[7] gave small upfield chemical shifts of proton c (-Hz2–8 Hz) over a time period too short for displacement to occur (Figure 4.4).
- 154 -
Here, Hz has the form:
where Hz < 0 refers to an upfield relative shift.
2 [BmaFc]/[Ada@Q[7]]
0 0 0.25 0.5 0.75 1 1.25 1.5 ‐2
‐4 Hz Proton c ‐6 Proton d ‐8 y = 2.379x3 ‐ 8.077x2 + 11.99x ‐ 9.628 R² = 0.998 ‐10
‐12
Figure 4.4 Plot of chemical shifts of methylene proton resonances c (blue diamonds) with increasing concentrations of BmaFc relative to Ada@Q[7] before replacement of Ada guest. Ada protons at d (red squares) of complex Ada@Q[7] showed zero shift relative to reference. The same titration in the absence of Ada@Q[7] gave no chemical shift changes.
Verifications of these chemical shifts were performed by 3 repetitive titrations and the resonances of the cavity-bound Ada (proton d) were found not to shift at all (Figure 4.4). Similar plots were also obtained for the BmaFc aromatic protons a and b (see Appendix 3). While this evidence was weak, the repetitions
- 155 - provided reproducible shifts consistent with a portal association of BmaFc most likely occurring at the ion-free portal of the Ada@Q[7] complex. In doing so, the two positive ionic groups of the exchanging guests (BmaFc and Ada) involved were as far apart as possible and electrostatic repulsion was minimised.
Another aspect of the 1H NMR spectra of BmaFc during titration is also noteworthy: no obvious shift was observed for the BmaFc methyl hydrogens.
The resonances that shifted were only the aromatic and methylene protons of the ferrocene which implies that a hydrophobic driving force existed with the closest association being between the aromatic rings and portal (see exchange intermediate in Figure 4.4). Alignment in this way could assist in displacing the resident guest through the ferrocene aromatic rings ingressing into the cavity.
4.4.3 Exchange kinetics and guest size
This section attempts to show the interrelationships between the exchange kinetics and guest sizes to further examine the displacement or ‘shunting’ guest exchange mechanism. As described above, it is customary to measure the initial rate (ki) of the exchange to make any reversible reaction less significant.
This discussion focuses on the exchange reactions within Q[7] using three different guest pairs: Ada and its glycerol derivative AdaG; the protonated ferrocene bis-derivative BmaFcH and the quaternary ammonium ion ferrocene bis-derivative BmaFc; and the mono-derivatives of ferrocene, as protonated and quaternised salts MaFcH and MaFc, respectively (Figure 4.1). Their relative binding affinities toward Q[7] are shown in Table 4.4 in which it can be seen that the Ka differences within a guest pair are relatively small but suitable for the
- 156 - study of exchange kinetics. The experimental conditions for comparative binding studies are described in Section 4.3.3.
Table 4.4 Relative binding affinities of six guests encapsulated in Q[7].
[a] Guest pair Ka ratio
AdaG/Ada 1.03 BmaFc/BmaFcH 4 MaFc/MaFcH 1.7
[a] Competitive experiments were brought to equilibrium at 85 °C over 2 h, as determined by 1H NMR.
Scheme 4.8 MaFc/MaFcH exchange within Q[7] at 25 °C.
The facility for one guest to be ‘shunted’ from the host by another was evaluated and it was found that the structural features of the guest moderated the exchange process. The mono-ferrocene derivatives, MaFcH and MaFc, exchanged particularly readily from the Q[7] host (Scheme 4.8). This exchange rate was obtained using a saturation transfer experiment and found to be exceptionally fast relative to other exchange reactions in this study (k = 1.6 s-1 at 25 °C, not substantially slowed at 5 °C) (Figure 4.5). From the time of the addition of the exchange guest (e.g., MaFc added to MaFcH@Q[7]) until the immediate measurement, thermal equilibrium was already complete at rt as it was for the reverse order of addition.
- 157 -
0 0 0.5 1 1.5 2 2.5 3 ‐1
‐2 y = ‐1.629x ‐ 1.228 ‐3 R² = 0.996
I(infinity)]/I(infinity)} ‐4 ‐
‐5 ln{[I(t) ‐6 Duration of the saturating pulse to peak t (s)
Figure 4.5 Plot of saturation transfer in reaction of MaFcH@Q[7] (0.25 mM) with MaFc (0.25 mM) in D2O. Thermodynamic equilibrium at 25 °C was reached quickly and methyl resonance of free MaFcH (at shift clear of other resonances) was irradiated to determine rate of exchange from slope of line in plot.
Scheme 4.9 Ada/MaFcH exchange within Q[7] at 5 °C.
In contrast, for the complex MaFcH@Q[7] subjected to exchange conditions of guest replacement by Ada (Scheme 4.9), the result was a dramatically slower
-6 -1 exchange (ki = 9.8 × 10 mM·s , or t1/2 = 3 h, 5 °C) which suggests that the rate-determining step was possibly ‘shunting’. Given that the kingress of both
MaFcH and Ada were excessively fast for the empty Q[7] compared with the initial rates of the exchange reactions in Schemes 4.8 and 4.9, there would appear to have been an inhibition to the ingress of Ada. This could best be explained by a ‘shunting’ mechanism related to the dimensions of the host and incoming guest. The portal opening to Q[7] was ~ 5.4 Å and the cavity dia. 7.3 Å while the cross-sectional dimensions of the adamantyl frame were 6.4 and 7.1 Å
- 158 -
(Figure 4.6) [5, 30]. Compared with MaFc, the ‘shunting’ ability of Ada decreased which is what would be expected from its tight-fit relative to that of the Q[7] portal (Figure 4.6).
Figure 4.6 Structures of BmaFc, MaFcH and Ada, and models showing dimensions in relation to the Q[7] portal. Dimensions of BmaFc and Ada were extracted from reported X-ray crystal structures (distances from atom centre to atom centre) [30, 31].
Scheme 4.10 AdaG/Ada exchange within Q[7] at 37 °C.
Manoeuvring the Ada guest for narrower regions of the frame and flexing the portal opening allowed the guest to enter easily where the cavity was empty but retarded its ingress when it already accommodated a guest (MaFcH). Further to
- 159 - this was the finding that the exchange rates were much more seriously affected when the incoming and leaving guests were both tight-fitting molecules, as demonstrated by the exchange between the guest pair of AdaG/Ada within Q[7]
-7 (Scheme 4.10) which resulted in an initial rate of exchange of ki = 9.4 × 10
-1 mM·s or t1/2 > 15 h, 37 °C. Here, the difficulties emanated from a combination of the molecular sizes for the ingress and egress as all other factors, such as ion–dipole interactions, hydrophobicity and stability constants, were relatively equal. The AdaG exchange reaction with Ada@Q[7] was 5 times slower than when BmaFc was the exchanging guest (ingress) (Scheme 4.3). When the size of a guest molecule was less important for either ingress or egress, the rate of exchange was controlled primarily by the ability of an incoming guest to ‘shunt’ and the resistance of an outgoing guest to depart, the latter being reflected in the stability constant of the guest@host. For the guest pair BmaFc/BmaFcH, where BmaFc displaced BmaFcH from the complex BmaFcH@Q[7] (Scheme
-6 -1 4.11), the initial rate of exchange was 3.1 × 10 mM·s , 37 °C (t1/2 = 10 h) and the size of the guest was not as important a feature (Figure 4.6). This reaction rate was comparable to the ‘shunting’ of Ada as a guest in Ada@Q[7] by BmaFc
(Scheme 4.3, Table 4.5) even though the binding constant of BmaFcH was ~ 5 times higher than Ada in Q[7] (Table 4.4).
Scheme 4.11 BmaFc/BmaFcH exchange within Q[7] at 37 °C.
- 160 -
Table 4.5 Exchange reactions from initial state of complex guest@Q[7]; egress of encapsulated guest to ingress of added guest.
Guest@Q[7][a] Initial Rate of Exchange at Ingress/Egress 37 °C (mM·s-1) BmaFc/Ada 4.8 x 10-6
BmaFc/BmaFcH 3.1 x 10-6
AdaG/Ada 9.4 x 10-7
Ada/MaFcH[b] 9.8 x 10-6
[a] Ingress guest was added at 4.5 mole equiv. relative to guest@Q[7] (0.250 mM). [b] Exchange reaction was conducted at lower temperature of 5 °C.
The rates of exchange of guests for the ‘open-door’ host Q[7] were shown to be very slow when: the ingress of the guest was affected by the tightness of the fit which limited its ability to ‘shunt’ the resident guest; and egress of a tight-fitting guest limited the ingress of a new guest and an increased stability of the guest@Q[7] complex through ion–dipole and van der Waals interactions and hydrophobicity contributed to the resistance of the resident guest to be displaced by the incoming guest. This latter point was succinctly demonstrated when we examined the substituted Q[7] as the host in Scheme 4.2. Another example of substituent effects on the exchange rate can be seen in a comparison of the rate for the ‘shunting’ of BmaFcH as a guest in Q[7] by BmaFc with the same reaction for the host CyP7Q[7] (Scheme 4.12). The initial rate of exchange for
-6 -1 the above guest pair in CyP7Q[7] was 1.5 x 10 mM·s , t1/2 = 26 h, which was
-6 -1 2.1 times slower than that in normal Q[7] (3.1 x 10 mM·s , t1/2 = 10 h)
(Scheme 4.11). In this case, the sizes of the guests involved in the exchange
- 161 - were less important and the reduced rates of exchange were achieved by subtle derivative changes on the molecular host substituted Q[7].
Scheme 4.12 BmaFc/BmaFcH exchange within CyP7Q[7] at 37 °C.
The guest exchange process for Q[n] could be affected by ion–dipole interactions which was particularly true for our guest pair bearing ammonium ions. The substituents on a substituted Q[7] may have increased the electron density on the carbonyl O of the substituted glycoluril moiety, thereby increasing the strength of the ion–dipole interactions. The progressive decrease in the rate of exchange from Q[7] with no substitution to partially substituted
Me2Q[7]/Me4Q[7] to fully substituted CyP7Q[7] could reflect that the difference in the electronegative rims of the three Q portals was potentially influenced by the substituents. Although an alternative or additional consideration could come from a restricted portal to ingress or egress (e.g., smaller portal), X-ray crystal structural data for CyP6Q[6] or CyP5Q[5] (Section 2.4.2, Chapter 2), does not substantiate this, therefore, it should not be relevant for CyP7Q[7] either.
- 162 -
4.5 Conclusions
In this chapter, we explored the mechanism of guest exchange within the host
Q[7] in which the guests involved exhibited high-affinity binding toward the host molecule. For this study, we selected several guest pairs, including adamantylammonium ion Ada and its glycerol derivative AdaG, ferrocene bis- and mono- derivatives BmaFcH, BmaFc and MaFcH, MaFc, because the guests within these pairs had virtually identical shapes, sizes, charges and complimentary stability constants. Observations of the guest exchange dynamic behaviour by 1H NMR spectroscopy indicated that a displacement or ‘shunting’ mechanism was the most likely used for guest exchanges within Q[7]s with or without substituents and at least with the guests studied here.
Q[7] demonstrated a surprisingly slow exchange (in hours) for guests such as adamantyl ammonium ion derivatives and ferrocene ammonium ion derivatives.
The rates of exchange were further slowed when Q[7] was substituted. The contributing factors of the size of the guest and the electronics of the host, as discussed, moderated the relative rates. We found that, at least for a guest with a high binding constant, the dominant exchange involved a ‘shunting’ mechanism. However, in the earlier work by Mock et al., a guest exchange occurred where the alternative mechanism of dissociation and replacement was reported to operate [12]. This may mean that both mechanisms operate but that they are perhaps dependent upon the binding constant of the guest. The work
8 -1 of Mock et al. involved guests with Ka < 10 M [12] where as our guests had
12 -1 Ka > 10 M .
- 163 -
The Q[n] family has a significant potential for the delivery of drugs, as carriers, protectors, slow release and targeted delivery, as both relatively simple structures and complex supramolecular structures [32]. The tuning of the rates of exchange in this study could push the drug delivery area of Q[n] application to higher levels, with greater control over delivery times and, perhaps, the sites of release, by the carrying of appropriately functionalised substituents.
- 164 -
4.6 References
1. Pluth, M. D.; Raymond, K. N. Chem. Soc. Rev. 2007, 36, 161–171.
2. Ariga, K.; Ito, H.; Hill, J. P.; Tsukube, H. Chem. Soc. Rev. 2012, 41,
5800–5835.
3. Mukhopadhyay, P.; Zavalij, P. Y.; Isaacs, L. J. Am. Chem. Soc. 2006,
128, 14093–14102.
4. Masson, E.; Ling, X.; Joseph, R.; Kyeremeh-Mensah, L.; Lu, X. RSC Adv.
2012, 2, 1213–1247.
5. Lagona, J.; Mukhopadhyay, P.; Chakrabatri, S.; Isaacs, L. Angew. Chem.,
Int. Ed. 2005, 44, 4844–4870.
6. Yu, J.-S.; Wu, F.-G.; Tao, L.-F.; Luo, J.-J.; Yu, Z.-W. Phys. Chem. Chem.
Phys. 2011, 13, 3638–3641.
7. Lucarini, M.; Luppi, B.; Pedulli, G. F.; Roberts, B. P. Chem.–Eur. J. 1999,
5, 2048–2054.
8. Yoshida, N. J. Chem. Soc., Perkin Trans. 2 1995, 12, 2249–2256.
9. Zhang, X.; Gramlich, G.; Wang, X.; Nau, W. M. J. Am. Chem. Soc. 2002,
124, 254–263.
10. Nau, W. M.; Zhang, X. J. Am. Chem. Soc. 1999, 121, 8022–8032.
11. Izatt, R. M.; Pawlak, K.; Bradshaw, J. S. Chem. Rev. 1995, 95, 2529–
2586.
12. Mock, W. L.; Shih, N. Y. J. Org. Chem. 1986, 51, 4440–4446.
13. Marquez, C.; Hudgins, R. R.; Nau, W. M. J. Am. Chem. Soc. 2004, 126,
5806–5816.
- 165 -
14. Wu, F.; Wu, L.-H.; Xiao, X.; Zhang, Y.-Q.; Xue, S.-F.; Tao, Z.; Day, A. I. J.
Org. Chem. 2012, 77, 606–611.
15. Day, A. I.; Arnold, A. P.; Blanch, R. J.; Snushall, B. J. Org. Chem. 2001,
66, 8094–8100.
16. Glidewell, C.; Royles, B. J. L.; Smith, D. M. J. Organomet. chem. 1997,
527, 259–261.
17. Bo, W. Acta Crystallogr., Sect. E 2010, E66, m686.
18. Pauson, P. L.; Sandhu, M. A.; Watts, W. E. J. Chem. Soc. C 1966, 251–
255.
19. Zhao, Y.-J.; Xue, S.-F.; Zhu, Q.-J.; Tao, Z.; Zhang, J.-X.; Wei, Z.-B.; Long,
L.-S.; Hu, M.-L.; Xiao, H.-P.; Day, A. I. Chin. Sci. Bull. 2004, 49, 1111–
1116.
20. Tang, H.; Fuentealba, D.; Ko, Y. H.; Selvapalam, N.; Kim, K.; Bohne, C. J.
Am. Chem. Soc. 2011, 133, 20623–20633.
21. Moghaddam, S.; Yang, C.; Rekharsky, M.; Ko, Y. H.; Kim, K.; Inoue, Y.;
Gilson, M. K. J. Am. Chem. Soc. 2011, 133, 3570–3581.
22. Liu, S.; Ruspic, C.; Mukhopadhyay, P.; Chakrabarti, S.; Zavalij, P. Y.;
Isaacs, L. J. Am. Chem. Soc. 2005, 127, 15959–15967.
23. Rekharsky, M. V.; Mori, T.; Yang, C.; Ko, Y. H.; Selvapalam, N.; Kim, H.;
Sobransingh, D.; Kaifer, A. E.; Liu, S.; Isaacs, L.; Chen, W.; Moghaddam,
S.; Gilson, M. K.; Kim, K.; Inoue, Y. Proc. Natl. Acad. Sci. U.S.A. 2007,
104, 20737–20742.
24. Saudan, C.; Dunand, F. A.; Abou-Hamdan, A.; Bugnon, P.; Lye, P. G.;
Lindoln, S. F.; Merbach, A. E. J. Am. Chem. Soc. 2001, 123, 10290–
10298.
- 166 -
25. Saito, H.; Yonemura, H.; Nakamura, H.; Matsuo, T. Chem. Lett. 1990,
535–538.
26. Kvasnica, M.; Chapin, J. C.; Purse, B. W. Angew. Chem., Int. Ed. 2011,
50, 2244–2248.
27. Dsouza, R. N.; Nau, W. M. J. Org. Chem. 2008, 73, 5305–4310.
28. Santamaria, J.; Martin, T.; Hilmersson, G.; Craig, S. L.; Rebek, J., Jr.
Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 8344–8347.
29. Davis, A. V.; Raymond, K. N. J. Am. Chem. Soc. 2005, 127, 7912–7919.
30. Mestrovic, E.; Bucar, D.-K. Crystallogr., Sect. E 2005, E61, m522.
31. Cambridge data base CCDC 645542.
32. Day A. I.; Collins J. G. Supramol. Chem. Mol. Nanomater. 2012, 3, 983–
1000.
- 167 -
CHAPTER 5
CONCLUSIONS
- 168 -
CHAPTER 5
Conclusions
This project focused on the introduction of the cyclopentane substituent into the equator of a Q[n]. Its success has demonstrated that the structural features of precursor glycolurils in substituted Q[n]-forming reactions are of great importance as they have a significant effect upon the potential to form higher homologues. An extension of this methodology has the potential to permit the introduction of functionality into the higher homologues through glycolurils carrying functionalised cyclopentane rings. To this point, several synthetic routes to glycolurils of this type were explored, and at least two potentially applicable glycolurils were synthesised.
The new cyclopentanoQ[n] developed in this thesis also indicated opportunities relating to increases in binding constants. As an extension of this discovery, a comparison of exchange rates of tightly bound guest pairs within alkyl- substituted Q[7] and Q[7] provided an insight into the effects of substituents upon the electronic character of either the portals or cavity.
5.1 Synthesis of higher homologues of SQ[n]
The synthesis of the first family of percyclopentanoQ[n = 5–7] (CyPnQ[n]) confirmed our hypothesis that the dihedral angle (Figure 5.1) of the concave
- 169 - face of the precursor glycoluril was important in determining whether higher homologues could be formed.
Figure 5.1 Substituted glycoluril diether showing two perpendicular views highlighting angle (R1–C6b–C6a) and dihedral angle of concave face.
An investigation into the structural features of dimethyl-, cyclohexano- and cyclopentano-substituted glycoluril diethers showed that dimethylglycoluril ether had the sharpest angle on the concave face compared with those of cyclohexanoglycoluril and cyclopentanoglycoluril, probably because of steric repulsion between the two methyl groups. This steric repulsion was reduced when the two methyl substituents were replaced by a 6-membered ring, as in cyclohexanoglycoluril, and the concave face angle widened. In accord with this trend was the fact that the introduction of a 5-membered ring, as in cyclopentanoglycoluril, further widened the angle. The consequence was that the natural curvature of the growing precursor oligomers formed in the acid- catalysed synthesis of cucurbituril was larger and, therefore, the potential for the formation of higher homologues was greater. This could also be manipulated to a small degree by the addition of templating Li+ ions into the reaction.
- 170 -
Given that angle in the substituent was determined to have a direct correlation to the dihedral angle of the concave face of the glycoluril (Figure 5.1), a more strained five-membered ring system, such as a norbornane or currently unknown four-membered carbon ring substituted glycoluril, could provide the possibility for further improvements in the synthesis of higher homologues with substitution.
The study described above also indicated that the development of new fully, or even partially, substituted higher homologues of Q was greatly limited by the synthetic availability of suitable glycolurils.
5.2 Synthesis of substituted glycolurils with functionalised cyclopentane rings
The synthesis of glycolurils with a functionalised cyclopentane ring discussed above is a worthy research pursuit since they could potentially be used to achieve the formation of higher homologues of functionalised Q[n] derivatives.
The symmetry of functionality on a cyclopentano substitutent had the advantage of controlling the locations of substituents when synthesising Q. By looking at the single example of the known germinal dimethylcyclopentanoglycoluril
(Section 3.4.1, Chapter 3), it was apparent that germinally placed substituents were potential entries to alternative analogue glycolurils. This germinal structural feature was investigated through several trials, from which two potentially suitable glycolurils with di-functional groups in the substituent ring were successfully synthesised. Although their yields were not ideal, given more
- 171 - time to optimise the reaction conditions, a better result may have been achieved.
A possible explanation for the lower yields was reasoned to be that they were a result of the formation of endocyclic double bonds in their substituent rings. As part of our response to this, we examined a different approach of utilising the norbornane structure as a substituent for glycolurils. The structure of this bicyclic system had the advantage of avoiding the double bond formation because of a strained ring system that could also provide opportunities for the synthesis of higher homologues of substituted Q[n].
Six diones with norbornane structures were considered but, unfortunately, only the reaction of the camphorquinone gave the product of acquired glycoluril under acidic conditions. It was found that the first step in the condensation of urea with -diones or -keto-aldehydes was a facile process, whereas the second was a more difficult reaction. The latter was elucidated through the isolation of some half- or three-quarter glycoluril intermediates (Section 3.4.2,
Chapter 3). The addition of the second ‘urea’ was either difficult or incomplete.
These intermediates provided an insight into the ease or difficulty of formation in each step which helped to explain why some aliphatic glycolurils were difficult to synthesise.
Future efforts to overcome this difficulty could be directed toward the use of alternative catalysts under milder conditions.
- 172 -
5.3 Modification of electronic properties of substituted Q[n] by variations of substituents
As we successfully obtained a new family of substituted Q[n], the CyP5–7Q[5–7], it was necessary to evaluate their host–guest properties compared with those of the simple Q[5–7]. The comparative binding studies of CyP6,7Q[6,7] and Q[6,7] using neutral or ionic probes revealed that the probes were bound more tightly in CyP6,7Q[6,7] than in Q[6,7]. We suggest that the reason for this was the enhanced electron density on the carbonyl O of the substituted glycoluril moieties caused by the alkyl-substituents.
The Q[n]-guest systems with exchanging guest molecules suggested that increased electronegativity at the host portal played an important role that allowed the rates of guest exchange for substituted Q[7] to be much slower than those for normal Q[7]. Detailed analysis of the exchange reaction within Q[7] revealed that several factors could contribute to the rate of exchange, such as the size of a guest molecule and the stability constant of guest@Q[7]. A great deal of evidence, as discussed in detail in Section 4.4.2 (Chapter 4), indicated that a displacement or ‘shunting’ mechanism was the best explanation for the guest exchange reaction we studied.
In the main, this study demonstrated an increase in the portal binding ability of substituted Q influenced by the alkyl-substituents it carried. We expect that the effect of substituents upon the electronic character of the Q’s cavity will be similar to that of substituents upon portals but, to establish this, particular neutral guest molecules need to be investigated in the future.
- 173 -
5.4 Recommendations for future research
Due to the importance of the functionalised Q[n] derivatives regarding their potential for building more complex systems, a number of synthetic routes to Q of this type have been developed by several research groups [1-4]. It is imperative to note that, as the introduction of functionalised substituents to the higher homologues still remains a challenge in this area, further research is the next step that we, and others, aim to take. Recently, we reported the Li+- templated synthesis of the CyPnQ[n] family with a higher proportion of CyP7Q[7], up to a 16 weight% [5]. The results suggested to us that the synthesis of higher homologues with functionality using a precursor glycoluril with a functionalised cyclopentane ring is possible. Two glycolurils of this type discussed in Chapter
3 are promising candidates for this purposes and it is expected that the synthetic conditions developed for the synthesis of CyPnQ[n] could be easily applied to the synthesis of a series of a fully functionalised Q[n] family.
Another possible area of pursuit is the synthesis of higher homologues with a partial substitution that carries functionality. The partial substitution of Q[7] in a controlled manner has been achieved using methyl substituents on the glycoluril moiety [6] which indicates that a five-membered carbon ring substituent could allow successful extrapolation to partial substitution using the same strategy.
The introduction of functionalised substituents with controlled positions at the
‘equator’ of Q would provide the potential to synthesise supramolecular structures from Q, such as dimers of substituted Q or even 3D polymers in the form of molecular tubes made from Q.
- 174 -
Host–guest binding studies have clearly shown that members of the Q[n] family have great potential as drug delivery vehicles [7, 8]. The results obtained from the research in this thesis have demonstrated that alkyl-substituents on a Q modify the host–guest binding behaviour with regard to increased binding constants and slower rates of exchange. The potential to ‘tune’ the rates of exchange for drug delivery systems is highlighted by our discoveries. The control of rates of release could provide a novel way of delivering drugs over variable time periods, and substitution with functionality could provide opportunities for attaching site-targeting groups [8].
To obtain a variety of substituted and functionalised Q[n], the appropriate precursor glycolurils must first be available. To this point, we have provided some insight into the synthetic problems and reaction intermediates that occur in glycoluril-formation reactions. From this study arises the need for improved reaction conditions and, perhaps, alternative catalysts. The Lewis acid-type catalyst Bi(OTf)3 has received considerable attention from synthetic chemists during the past decade. Recently, our group found that, for our glycoluril- formation reactions, this catalyst was very effective under milder conditions and with improved yields of glycolurils. Therefore, it may be worthwhile proceeding with future syntheses of glycolurils using this new catalyst.
- 175 -
5.5 References
1. Jon, S. Y.; Selvapalam, N.; Oh, D. H.; Kang, J.-K.; Kim, S.-Y.; Jeon, Y. J.;
Lee, J. W.; Kim, K. J. Am. Chem. Soc. 2003, 125, 10186–10187.
2. Lucas, D.; Minami, T.; Iannuzzi, G.; Cal, L.; Wittenberg, J. B.;
Anzenbacher, P. Jr.; Isaacs, L. J. Am. Chem. Soc. 2011, 133, 17966–
17976.
3. Zhao, N.; Lloyd, G. O.; Scherman, O. A. Chem. Commun. 2012, 48,
3070–3072.
4. Vinciguerra, B.; Cao, L.; Cannon, J.-R.; Zavalij, P.-Y.; Fenselau, C.;
Isaacs, L. J. Am. Chem. Soc. 2012, 134, 13133–13140.
5. Wu, F.; Wu, L.-H.; Xiao, X.; Zhang, Y.-Q.; Xue, S.-F.; Tao, Z.; Day, A.-I. J.
Org. Chem. 2012, 77, 606–611.
6. Day et al. unpublished results.
7. Masson, E.; Ling, X.; Joseph, R.; Kyeremeh-Mensah, L.; Lu, X. RSC Adv.
2012, 2, 1213–1247.
8. Day A. I.; Collins J. G. Supramol. Chem. Mol. Nanomater. 2012, 3, 983–
1000.
- 176 -
APPENDIX
1. Crystal structure information
Crystal data for 2 (CCDC-848923): [C12H16N4O4] (280); monoclinic, space group
P2(1)/n; colourless block, a = 8.402(3) Å, b = 11.908(4) Å, c = 12.710(4) Å; =
90.00º, = 101.856(13)º, = 90.00º, V = 1244.5(7) Å3; Z = 4; T = 273(2) K; R(F)
2 = 0.0392; GOF on F = 1.051. Crystal data for 3 (CCDC-670588): [C11H14N4O4]
(268); orthorhombic, space group Pnma; colourless block, a = 12.4218(7) Å, b =
11.3954(7) Å, c = 8.1359(5) Å; = 90.00º, = 90.00º, = 90.00º, V = 1151.65(7)
Å3; Z = 4; T = 293(2) K; R(F) = 0.0374; GOF on F2 = 1.058. Crystal data for
CyP5Q[5]2KCl·19H2O (CCDC-748653): [C135H258Cl6N60O87K6]; monoclinic, space group P21/c; colourless block, a = 20.2298(8) Å, b = 33.2253(11) Å, c =
29.4252(10) Å; = 90.00º, = 95.186(2)º, = 90.00º, V = 19696.9(12) Å3; Z = 4;
T = 273(2) K; R(F) = 0.0405; GOF on F2 = 1.052. Crystal data for
Dioxane@CyP6Q[6]·18.5H2O (CCDC-848924): [C116H208N48O61]; monoclinic, space group P 21/c; colourless block, a = 12.6912(5) Å, b = 12.5739(5) Å, c =
23.0762(10) Å; = 90.00º, = 98.670(2)º, = 90.00º, V = 3640.4(3) Å3; Z = 1; T
= 293(2) K; R(F) = 0.0405; GOF on F2 = 1.090.
- 177 -
1 13 2. H and C NMR spectra of CyP5Q[5]·KCl, + dioxane@CyP6Q[6] and adamantylNH3 @CyP7Q[7]
- 178 -
3. The plot of the chemical shifts of the BmaFc aromatic protons a and b with increasing concentration of BmaFc relative to Ada@Q[7] before replacement of the Ada
- 179 -
1 [BmaFc]/[Ada@Q[7]] 0 0 0.25 0.5 0.75 1 1.25 1.5 ‐1
‐2
Hz Proton a ‐3 Proton d y = 2.555x3 ‐ 8.226x2 + 10.04x ‐ 5.693 ‐4 R² = 0.982
‐5
‐6
1 [BmaFc]/[Ada@Q[7]]
0 0 0.25 0.5 0.75 1 1.25 1.5 ‐1
Proton b Hz ‐2 Proton d ‐3 y = 2.202x3 ‐ 7.652x2 + 9.630x ‐ 5.240 ‐4 R² = 0.988
‐5
- 180 -