Supramolecular Chemistry - 2009 - Steed - Cation‐Binding Hosts.pdf

Cation-Binding Hosts 3 ‘Now spurs the lated traveller apace To gain the timely inn. Ourself will mingle with society And play the humble Host.’ William Shak...

Cation-Binding Hosts 3 ‘Now spurs the lated traveller apace To gain the timely inn. Ourself will mingle with society And play the humble Host.’ William Shakespeare (1564–1616), Macbeth  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 106 Cation-Binding Hosts 3.1 Introduction to Coordination Chemistry 3.1.1 Supramolecular Cation Coordination Chemistry Lehn, J.-M., ‘Supramolecular chemistry – scope and perspectives molecules, supermolecules, and molecular devices’, Angew. Chem. Int. Ed. Engl., 1988, 27, 89–112; Cram, D. J., ‘The design of molecular hosts, guests, and their complexes (Nobel Lecture)’, Angew. Chem. Int. Ed. Engl., 1988, 27, 1009–1020; Pedersen, C. J., ‘The discovery of crown ethers (Nobel Lecture)’, Angew. Chem. Int. Ed. Engl., 1988, 27, 1021–1027. The examples of natural ionophores, such as valinomycin, nonactin and the enniatins discussed in Chapter 2, have provided an enormous impetus to supramolecular chemists to synthesise artificial ionophore mimics and model compounds capable of exhibiting selective complexation and transport, not only of alkali metal cations but also the majority of the s, p d and f block metals and nonmetallic cations such as NH4
and organic ammonium salts. A particular challenge has been the enantiospecific binding of chiral species such as protonated amino acids. Cation complexation chemistry has effec- tively given birth to the whole field of molecular recognition (the area of chemical research involving the selective binding by hosts that are structurally and electronically preprogrammed for complexation of particular guests). An enormous and diverse range of ligands (hosts for metal ions) has been pre- pared, exhibiting both remarkable selectivities and useful reactivity properties. In this chapter, we will survey some of the most common classes of ligand and use these as examples in the discussion of the basis for their molecular recognition properties. We will then look at the application of these concepts to more complex or specialised systems. 3.1.2 Useful Concepts in Coordination Chemistry Housecroft, C.E. and Sharpe, A.G., Inorganic Chemistry, 3rd edn., Prentice Hall: Upper Saddle River, 2007. We saw in Section 1.3 how Alfred Werner formulated the modern concept of coordination chemistry, which supramolecular chemistry generalises to a ‘complete coordination chemistry’. Prior to Werner’s time the ‘chain’ theory of coordination compounds was popular. The chain and Werner formulations of ‘Co(NH3)4Cl3’ and ‘Co(NH3)3Cl3’ are shown in Figure 3.1. While both theories predict that ‘Co(NH3)4Cl3’ will exhibit one labile chloride ion per molecule, the chain theory also predicts that ‘Co(NH3)3Cl3’ will have one labile chloride, while Werner’s theory ultimately correctly predicted that the chloride is not labile in this case. The study of supramolecular complexes of metal cations is really nothing more than the coordination chemistry of relatively labile (i.e. ligand substitution is relatively rapid under ambient conditions) metal ions and relatively elaborate, usually chelating or multidentate ligands (see Section 1.5).1 It is therefore worth spending a little time reviewing some basics of coordination chemistry before looking at specific supramolecular systems. Experts read no further! Complex Compound or Coordination Compound A complex compound is compound formed between reactants of which the valencies are already formally saturated, e.g. CoCl2
2 Et4N
Cl (Et4N) 2 [CoCl4]. All three compounds are stable individually. We generally use square brackets to denote the coordination complex. Supramolecular Chemistry, 2nd edition J. W. Steed and J. L. Atwood © 2009 John Wiley & Sons, Ltd ISBN: 978-0-470-51233-3  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Introduction to Coordination Chemistry 107 + NH3 Cl H3 N NH3 Cl Co H3 N Co NH3NH3NH3NH3 Cl Cl Cl Cl NH3 H3 N Cl Cl Co Co NH3NH3NH3 Cl H3 N Cl Cl Cl Werner formulation chain formulation Figure 3.1 Werner and historical chain formulations of ‘Co(NH3) 4Cl3’ and ‘Co(NH3)3Cl3’. Ligand A ligand* is an ion or a molecule which is bonded to the central atom(s) in a complex compound. In the example above chloride is the ligand in (Et4N)2 [CoCl4]. Generally ions or molecules which act as ligands are Lewis bases (electron donors; often lone pair donors) and are capable of independent existence. Ligands can often bind through more than one donor atom or group and the number of such donors is referred to as the ligand’s denticity ( the number of ‘teeth’ it has that can ‘bite’ onto the metal). Chloride is a monodentate ligand, 1,2-diaminoethane is bidentate etc. In cation supramolecular chemistry we call relatively elaborate, multidentate ligands ‘hosts’. Oxidation State The oxidation state of an element in a compound is the resultant charge on the central atom(s) when the attached groups (ligands) are removed in their closed electron shell configurations (e.g. in [CoCl4] 2 the chloride ligands would be removed as Cl because this has a closed valence electron shell con- taining 8 electrons). For example: [CoCl4] 2 can be dismantled to give four chloride ions Cl and a cobalt(II) ion, Co2
. The oxidation state of the chlorine is thus –1, or (-I) and that of cobalt is
2, or (II). These are usually written as Cl(-I) and Co(II). Bonding Bonding in coordination complexes ranges from entirely ionic ion-dipole type interactions in which a ligand lone pair of electrons forms a dative bond to a positively charged metal cation to entirely covalent in which there is significant orbital overlap between metal and ligand valence orbitals. Most compounds are somewhere in between and hence have at least some non-covalent component to the interaction. The degree of covalency will be determined by the metals charge and size and the nature of the ligand. Small, highly charged metal ions, and those with closed valence shells such as Na
or Al3
, tend to form more ionic compounds. The coordination number and geometry of the complex is then dependent on the number of ligands that can pack around the small metal centre. As a result coordina- tion geometries can be highly irregular and dependent on the geometry of the ligand, particularly if it is a macrocycle or constrained in some way. As a result matching the metals’ coordination geometry * From the Latin ligare, to bind. For an interesting discussion of the use and spread of this term see: Brock, W. H., Jensen, K. A., Jørgensen, C. K., Kauffman, G. B., The origin and dissemination of the term ‘ligand’ in chemistry. Polyhedron 1983, 2, 1–7.  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 108 Cation-Binding Hosts tendencies is less important in host design for these kinds of metals. In contrast metals in lower oxida- tion states or those with unfilled sub-shells such as the transition metals tend to form more covalent complexes with well-defined coordination geometries and a preference for more polarisable ligands. A detailed discussion of the bonding in transition metal complexes can be found in the key reference. Coordination Number In solid state chemistry where there are often a number of equal short interatomic contacts, the simplest definition of coordination number is the number of nearest neighbours. In coordination chemistry (solid, solution and gas phases) a more practical definition of coordination number is the number of atoms or ligating groups (such as a C= C double bond) bound to a metal. Thus in [CoCl4] 2 the coordination number is four. The majority of coordination complexes are either 4- or 6-coordinate. For 6-coordinate the most important geometrical arrangement is the octahedron (or pseudooctahedron if the six ligands are different from one another). This term implies that the six ligands lie on the carte- sian axes at 90o to one another. Drawing imaginary lines from each ligand to those adjacent gives an 8-sided 3D polyhedron; an octahedron. Arbitrarily we refer to the top and bottom sites as axial and the middle four as equatorial. In reality these sites are indistinguishable in a complex in which all the ligands are the same. The other possible 6-coordinate geometry is the trigonal prism, as in [ZrMe6] 2 and [Re(S2C2Ph2)3]. For 4-coordinate complexes there are two geometries with approximately equal occurrence: square planar and tetrahedral. Square planar complexes are not common outside transition metal chemistry (ICl4 is one example) and are mostly found in complexes of Rh(I), Ir(I), Ni(II), Pd(II), Pt(II) and possibly Cu(II). [NiCl4] 2 is tetrahedral, whereas [Ni(CN) 4] 2 is square planar. Other coor- dination numbers are also encountered such as linear 2-coordinate Ag(I) in [Ag(NH3)2]
. The reaction of a bidentate ligand such as 1,2-diaminoethane with Ag
does not lead to a chelate complex, but instead to a linear two coordinate coordination polymer. One reason for this is that bidentate ligands can not exist in trans arrangements unless they are extremely long and flexible, i.e. they cannot span 180o. Coordination numbers for the large lanthanoid metal ions (La3
– Lu3
) are commonly 8–12. Thus the ability of multidentate macrocycles to bind lanthanoid cations can be combined with additional water coordination in their use as Magnetic Resonance Imaging (MRI) contrast agents (Section 11.3.2). Nomenclature Connelly, N. G. (ed.) Nomenclature of Inorganic Chemistry IUPAC Recommendations 2005, Royal Society of Chemistry: Cambridge, 2005. In both names and formulae of salts (i.e. ionic compounds) cations precede anions. When writing the formulae of binary compounds the usual rule is to write the more electropositive element first. So we have SiF4 not F4Si. An ‘element seniority sequence’, which differs slightly from the electronegativities has been drawn up by IUPAC, which explains why we have H2O but NH3. Coordination entities are named by writing the ligands in alphabetical order with appropriate numerical prefixes, followed by the central atom and its oxidation state. The numerical prefixes are ignored when establishing the alphabetical order. Thus, tribromo precedes iodo. Although the alphabetical order is used for names, when writing formulae the sequence is: 1. Central atom, 2. Anionic ligands, 3. Neutral ligands, and the formula of the coordination entity is enclosed in square brackets. Some deviation is allowed if it is desired to convey specific structural information. The names of anions usually end in –ide if there are monatomic or homopolyatomic, or –ate for heteropolyatomic anions and coordination compounds. Anionic ligands usually end in –o. Thus Cl− is chloride, I3− is triiodide and SO42− is sulfate but become, chloro, triiodo and sulfato if they are coordinated as ligands. The names of complex anions are not always derived from the familiar name  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Introduction to Coordination Chemistry 109 of the element. Thus iron anions are ferrates, silver anions are argentates, gold anions are aurates, lead anions are plumbates, and tin anions are stannates. Cations or neutral compounds are not given any special ending, but coordinated water is referred to as aqua (previously aquo) and coordinated ammonia is referred to as ammine. The IUPAC rules generally use numerical prefixes derived from Greek. Latin prefixes (e.g. uni instead of mono) were used and will be encountered in older literature. Some of the higher number prefixes revert to Latin, so that nona is used in preference to ennea. The multiplicative prefixes are necessary to avoid ambiguity, for example, bis(methylamine) is (CH3NH2)2 but dimethylamine is (CH3)2NH. Italics are used for geometrical and structural prefixes e.g. cis-, trans-, fac-, mer- but not for other prefixes (bis, tris, tetrakis etc.). Note that the geometrical and structural prefixes are used with a hyphen, but numerical prefixes are not. Bridging atoms are designated with Greek µ. The Greek letter κ (kappa) is used to designate which atom of a polyatomic ligand is coordinated to the metal atom. So [(NH3)5CoONO] 2
is the pentaamminenitrito–κO-cobalt(III) ion and [(NH3)5CoNO2] 2
is the pentaamminenitrito–κN-cobalt(III) ion. The old names for these ions were pentaamminenitritoco balt(III), and pentaamminenitrocobalt(III) respectively. Greek derived prefix multiplicative Latin prefix 1. mono prefixes uni 2. di bis bi 3. tri tris tri (ter) 4. tetra tetrakis quadri 5. penta pentakis quinque 6. hexa hexakis sexa Synthesis Generally transition metal coordination compounds are prepared simply by mixing a solution of the central (metal) ion in a suitably receptive state with the ligand. Not all ions undergo rapid reactions (e.g. complexes of the inert Cr(III) and Co(III) which are often prepared from Cr(II) and Co(II)). In many cases of first row transition metals the hydrated metal ions are readily available and water may be replaced by more strongly complexing ligands. In some cases a solid product can only be isolated if a large counter ion (e.g. BPh4 or NEt4
) is used. This is because the lattice energy is proportional to 1/rA
rC (where rA and rC are the anion and cation radii, respectively) but the solvation energy is proportional to 1/rA
1/rC so if either rA or rC are small the solvation energy will tend to be large and the salt will be soluble. e.g. (NEt4)2 [CoCl4] is insoluble. Complexes of macrocyclic and Schiff base ligands are often made by template synthesis methods where the host ligand is prepared in the presence of the metal guest, see Section 3.9. Hard and Soft Acid and Base Theory The idea of hard and soft acids and bases (HSAB) is a development of the concept of Lewis acids and bases, so that acids are electron acceptors and bases are electron donors. Hard acids: the acceptor atom is of high positive charge, small size, and does not have easily excited outer electrons, i.e. non-polarisable. Examples: H
, Na
, Ca2
, high oxidation states of the transition metals.  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 110 Cation-Binding Hosts Soft acids: the acceptor atom is of low positive charge, large size, and has several easily excited outer electrons, i.e. it is polarisable. Examples: Pt2
, Rh
, Tl
, Hg2
, low oxidation states of the transition metals. Hard bases: the donor atom is of low polarisability, high electronegativity, hard to reduce, and associated with empty orbitals of high energy and hence inaccessible. Examples: F, ligands with oxygen or to some extent nitrogen donor atoms. Soft bases: the donor atom is of high polarisability, low electronegativity, easily oxidised and associated with empty, low-lying orbitals. Examples: I, R3P, R2S, H. The Principle of Hard and Soft Acids and Bases states that hard acids form more stable complexes with hard bases and soft bases form more stable complexes with soft acids. In orbital terms hard molecules have a large gap between the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO), and soft molecules have a small HOMO-LUMO gap. In recent years it has been possible to correlate the hardness with the electronic properties of the atoms involved. Thus, if the enthalpy of ionisation (I) and the electron affinity (A) are known the so-called absolute hardness (η) and absolute electronegativity (χ) can be found from: η  (I – A) / 2 and χ  (I
A) / 2. For example, the first and second ionisation enthalpies of sodium are 5.14 and 47.29 eV. Thus for Na
, I 
47.29 and A 
5.14, so η  (47.29 – 5.14) / 2  21.1. Similarly for silver the first and second ionisation enthalpies are 7.58 and 21.49eV, so for Ag
we have, η  (21.49 – 7.58) / 2  6.9. There is no sharp dividing line between the classes and the class depends on the oxidation state. High oxidation states will have more hard character than low oxidation states; Cu(II) is hard whereas Cu(I) is soft. The Leden-Chatt triangle (Figure 3.2) is a useful way to remember the soft metal ions. The division is far from perfect, however! An interesting example of the HSAB principle is given by the reaction: LiI
CsF → LiF
CsI. This proceeds from left to right as predicted by HSAB but contrary to Pauling’s electronegativity which gives more energy for large electronegativity differences. It is worth noting that while the HSAB theory deals predominantly with electronic (ionic and covalent) effects it has been suggested that there is also a significant steric component based on donor atoms’ size.2 Optical isomerism in Coordination Chemistry Optical isomerism occurs when a molecule contains neither a centre of symmetry (inversion centre, i) nor a mirror plane (m) or other reflection operation such as improper rotation (Sn). The molecule is thus chiral and will rotate the plane of polarised light if present as a pure enantiomer. Chiral molecules exist in left and right handed forms (isomers) usually labelled with a relative configuration d or l (from the Latin dextro- and laevo- meaning right and left relative to the configuration of glyceraldehyde). Chiral molecules can also be denoted
or – according to their direction of rotation of plane polarised light. Note there is not relationship betweel the d or l relative configuration and the rotation of plane polarised light. In organic compounds we can usually discover whether a molecule is chiral by look- ing for a carbon atom with four different substituents and we can assign its absolute handedness (R or S according to the Cahn-Ingold-Prelog rules; it takes a combination of structural and optical rotation measurements to tie together the (
/) optical rotation properties with the absolute configuration). Co Ni Cu Zn Ga Rh Pd Ag Cd In Ir Pt Au Hg Tl Figure 3.2 The Leden-Chatt triangle with copper at the apex. A useful way to remember some of the soft metal ions.  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Introduction to Coordination Chemistry 111 H2N NH2 NH2 NH2 H2N NH2 Co Co NH2 NH2 H2N NH2 H2N NH2 ∆-[Co(en) 3]3+ Λ-[Co(en)3]3+ en = ethylenediamine, NH2CH2CH2NH2 Figure 3.3 Optical isomers of [Co(en)3]3
. The Cahn-Ingold-Prelog rules work for inorganic compounds too but coordination complexes often have coordination numbers greater then four and may exhibit helical chirality, for example, denoted ∆ and Λ (or P and M in the Cahn-Ingold-Prelog system). The formal condition for chirality is that the molecule should not have an improper axis of rotation (i.e. a rotation
reflection axis, Sn; S1  mirror, S2  i) of any order (including 1). In coordination chemistry optical activity is common when chelate rings are present and hence is common in supermolecules. Hence, [M(L-L)3] and cis- [M(L-L)2X2] complexes (where L-L  didentate chelate) are chiral, although trans-[M(L-L)2X2] is not, e.g. [Co(H2NCH2CH2NH2)3]3
, Figure 3.3. Enantiomers have identical physical and chemical properties to one another except the direction in which they rotate plane polarised light (clockwise or anticlockwise). They may be separated by inter- action with a second chiral species. This gives two diastereoisomers (if the two chiral centres are the same we can describe the diastereoisomers as optically pure meso ∆Λ and the racemic or rac form which itself occurs as two pairs of enantiomers, ∆∆ and ΛΛ) which do differ in their physical proper- ties (e.g. have different NMR spectra, can be separated by achiral chromatography etc.). For example, Scheme 3.1 shows the experimental resolution of [Co(en)3]3
using tartrate. CO2- H - OH tartrate2 = racemic (+)-Co(en)33+ and (-)-Co(en)33+ HO H CO2- (+)-tartrate (chiral resolving agent) 50% 50% (+)Co(en)3[(+)tartrate]Cl (-)Co(en)3[(+)tartrate]Cl precipitate soluble add Co(en)32+ to catalyse racemisation 25% 25% (+)Co(en)3[(+)tartrate]Cl (-)Co(en)3[(+)tartrate]Cl precipitate soluble continue to near quanitative yield Scheme 3.1 Experimental resolution of [Co(en)3]3
using optically pure
-tartrate.  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 112 Cation-Binding Hosts Me NH2 Ph Me NH2 NH2 Ph NH2 Me H Me H Pt Pt Ph Ph NH2 NH2 NH2 H NH2 H Tetrahedral - mirror plane in plane of paper Square planar - no mirror plane or inversion centre so molecule is chiral Figure 3.4 Use of chirality to establish the square planar geometry of Pt(II). One of the most elegant uses of optical activity was the experiment by Mills and Quibell which demonstrated the square planar geometry of Pt(II). They prepared the Pt(II) complex of isobuty- lenediamine and meso-stilbenediamine and found that it was chiral. If the geometry at platinum is tetrahedral the complex has a mirror plane and so is not chiral, Figure 3.4. Enantiomers may be distinguished by their influence on the rotation of plane polarised light. This is measured by a type of electronic spectroscopy called circular dichroism. The ability of the complex to rotate plane polarised light is given in terms of an optical rotation given in degrees at a particular wavelength (see Box 6.2). X-ray crystallography may also distinguish enantiomers as long as there are ‘heavy’ atoms present (usually heavier than Na, although specialist experiments can distinguish enantiomers of organic compounds as well). This is a special type of crystallographic experiment termed absolute structure determination and makes use of a property of crystals termed anomalous dispersion.3 3.1.3 EDTA – a Classical Supramolecular Host Ethylenediamine tetraacetic acid (H4EDTA) in its deprotonated form, EDTA4, is an extremely com- mon, chelating ligand that is able to bind to most metal centres. It represents an excellent illustration of how a supramolecular host is nothing more than a classical multidentate ligand. It may be deprotonated up to four times to give EDTA4 which acts as a multiply chelating ligand. EDTA titrations are used for the analytical determination of metal ion concentrations (e.g. Mg2
and Ca2
in a 24 hour urine sam- ple). This is possible because EDTA4 binds so strongly to most metal ions that essentially 100% of the ion is bound to the ligand. A dye such as the murexide ion is used as an indicator and hence the reaction shown in Equation 3.1 takes place. Decomplexation of the metal-bound indicator causes a change in the energy of the indicator HOMO and hence a change in the electronic absorption energy and hence colour. This kind of indicator displacement assay is a useful sensing method (Section 11.3.3). [M·indicator] n

EDTA4 → [M(EDTA)](4-n)
free indicator (3.1) Colour 1 Colour 2 H O O H HO2C CO2H N N N N O N O HO2C N N CO2H H H O O H4EDTA the Murexide anion used as an indicator in EDTA titrations  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Introduction to Coordination Chemistry 113 Table 3.1 Stability constants (Log K) in aqueous solution for metal complexes of EDTA4. Mg2
8.7 Zn2
16.7 La3
15.7 2
2
3
Ca 10.7 Cd 16.6 Lu 20.0 2
2
3
Sr 8.6 Hg 21.9 Sc 23.1 2
2
3
Ba 7.8 Pb 18.0 Ga 20.5 2
3
3
Mn 13.8 Al 16.3 In 24.9 2
3
4
Fe 14.3 Fe 25.1 Th 23.2 2
3

Co 16.3 Y 18.2 Ag 7.3 2
3

Ni 18.6 Cr 24.0 Li 2.8 2
3

Cu 18.8 Ce 15.9 Na 1.7 Binding constants as Log K values for the equilibrium: Mn

EDTA4  [M(EDTA)](n-4)
are given in Table 3.1. They cover over 23 orders of magnitude from 1.7 for the relatively large, low charge and labile Na
to 25.1 for the highly charged, complementary Fe3
. Significant factors are the cations’ charge and ionic radius. For example for Na
, Ca2
and Lu3
the ionic radius is very similar but the charge increases, Table 3.2. If we look at the data for the alkaline earths we see that Mg2
is anomalous because the value is lower than we would expect from charge and radius considerations, Table 3.3. This reflects the high degree of solvation of the small, polarising Mg2
and that fact that it is too small to form an optimum chelate geometry. Moreover the hard Mg2
is not particularly complementary to the softer tertiary amine nitrogen donors. As a result the complex is 7-coordinate with an open face that interacts with an additional aquo ligand (Figure 3.5). In contrast the smaller Co(III) ion is able to coordinate optimally to six EDTA4 donor atoms, while the larger Ni(II) interacts with an aquo ligand instead of an EDTA donor atom. Note that the divalent Ni2
and Mg2
are insufficiently polarising to form stable complexes with the fully deprotonated EDTA4. Table 3.2 Correlation of ionic charge with affinity for EDTA4. Na
Ca2
Lu3
radius (Å) 0.95 0.99 0.93 log K (EDTA) 1.7 10.7 20.0 Table 3.3 correlation of ionic radius with affinity for EDTA4; Mg2
is anomalous. Mg2
Ca2
Sr2
Ba2
radius (Å) 0.65 0.99 1.13 1.35 charge/radius 3.07 2.02 1.77 1.48 log K (EDTA4–) 8.7 (7 coord.) 10.7 8.6 7.8  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 114 Cation-Binding Hosts Figure 3.5 X-ray molecular structures of (a) [Co(EDTA)] ⴚ, (b) [Ni(HEDTA)(H 2O)] ⴚ and (c) [Mg(H2O)(HEDTA)] ⴚ. 3.2 The Crown Ethers G. W. Gokel, Crown Ethers, Royal Society of Chemistry: Cambridge, 1990. 3.2.1 Discovery and Scope The crown ethers are among the simplest and most appealing macrocyclic (large ring) ligands, and are ubiquitous in supramolecular chemistry as hosts for both metallic and organic cations. They consist simply of a cyclic array of ether oxygen atoms linked by organic spacers, typically —CH2CH2— groups. While the metal binding ability of unidentate ethers such as the common solvent diethyl ether is very poor, the crown ethers are much more effective by virtue of the chelate effect and the partial preorganisation arising from their macrocyclic structure (Section 1.5). The discovery of the crown ethers in 1967 gained a share of the 1987 Nobel Prize for Chemistry for Charles Pedersen, a chemist working at the American du Pont de Nemours company. Oddly, however, Pedersen’s initial synthesis of the fi rst crown ether, dibenzocrown-6 (3.4) was acciden- tal. In trying to carry out the synthesis of the linear di-ol (3.3), which he hoped might act as a ligand for the catalytic vanadyl ion, Pedersen carried out the reaction shown in Scheme 3.2. The starting material was the catechol (1,2-dihydroxybenzene) derivative (3.1), in which one of the hydroxyl groups is protected by a tetrahydropyran ring to prevent its reaction. Unknown to Pedersen, his starting material was slightly contaminated by some free catechol (3.2). The resulting product was a mixture of the desired compound (3.3) along with a small amount of dibenzocrown-6, formed in only 0.4 % yield. It is a tribute to Pedersen’s skills that he was able to isolate and characterise this small amount of by-product.4 Pedersen’s interest was aroused by the interesting solubility properties of (3.4), and by its high degree of crystallinity (suggesting a molecular compound rather than a polymer). The compound dissolved sparingly in methanol, but its solubility was enhanced significantly on addition of alkali metal salts. Pedersen soon synthesised the compound in much better yield. He found that it dissolved inorganic salts like KMnO4 in organic solvents such as benzene to give it a purple colouration (this was termed ‘purple benzene’). Pedersen also became aware of the crown ether’s ability to dissolve the alkali metals themselves to give interesting blue solutions of what are now known to be alkalide and electride salts  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License The Crown Ethers 115 OH 1. NaOH / BunOH OH HO Cl Cl + O 2. H+, H2O O O O O O 3.1 3.3 + + OH O OH O O 3.2 O O O 3.4 Scheme 3.2 Accidental synthesis of the first crown ether, dibenzocrown-6.4 (Section 3.13). He eventually concluded that ‘the potassium ion had fallen into the hole at the centre of the molecule’ (Figure 3.6), at the time a bold and highly imaginative statement. Events were quickly to prove him to be entirely correct. This initial result led rapidly to the synthesis of a related family of species (Figure 3.2) to which Pedersen gave the name ‘crown ethers’ because of the crown-like shape of the capsular complex between 3.4 and K
. Since the early work of Pedersen, the crown ethers have become very popular in cation complexation chemistry. Their versatile binding ability and their almost infinite synthetic malleability have led to an enormous variety of functionalised and derivatised crowns and crown analogues with varying selectiv- ity for an enormous variety of guest species, Figure 3.7. The analogy of the K
-binding ability of 3.4 Figure 3.6 Space-filling diagram of the complex between crown-6 (3.4) and K
.  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 116 Cation-Binding Hosts O O O crown-5 O 3.5 Complementary to Na+ O O O O 3.8 O O O Dicyclohexylcrown-6 O More conformationally rigid O O crown-6 3.6 O O Complementary to K+ O O O O O O O O 3.10 O O 3.9 O O O O O Dibenzocrown-10 Dibenzocrown-4 O O O Binds two Na+ ions an unusual example 3.7 crown-7 O O Complementary to Cs+ O O Figure 3.7 Some common crown ethers. and the metal complexes of the ionophores such as valinomycin and nonactin discussed in Section 2.1 is immediately apparent and extensive studies have been made of the crown ethers as models for ionophore behaviour. 3.2.2 Synthesis D. Parker, Macrocycle Synthesis: A Practical Approach, OUP, Oxford, 1996. Pedersen described a total of six different methods of crown ether synthesis in his original work. These still form much of the basis for modern crown ether preparations (Scheme 3.3). Most new crown ethers are prepared by either method (a) or (b). All of the methods shown in Scheme 3.3 are examples of the Williamson ether synthesis, although it was an example of method (c) that led to the original preparation of dibenzocrown-6. In fact, dibenzocrown-6 may be prepared in much better yield (about 80 per cent yield, versus 45 per cent) via method (b). Furthermore, the unreacted catechol may cause problems in the separation of the [1
1] and [2
2] cycloaddition products in (c). The method shown in scheme (d) was used by Pedersen only for the preparation of macro- cycle 3.10. Method (e) (intramolecular cyclisation) is not a particularly viable one in general terms because of the nonavailability of the starting materials and usually poor yields. For example, crown-6 (3.6) may be prepared by this method in 1.8 % yield, although excellent yields have been obtained for oligoethyleneglycol [HO(CH2CH2O) nH] cyclisations by slow addition of p-toluenesulfo- nyl chloride to a suspension of alkali metal hydroxide, which acts as the base. In general, increasing use of the tosylate (p-toluenesulfonate, MeC6H4SO2, Ts) function as a leaving group instead of chloride has often led to increasingly better yields. Method (f) is a catalytic reduction to produce saturated cyclohexyl rings from the correspond- ing arenes. In the case of dibenzocrown-6 this can, in principle, result in up to five isomers of dicyclohexylcrown-6 (3.8), of which only the first four have been isolated (Figure 3.8). The different configurations of these materials (which are noninterconvertable) result in a wide variation of their binding ability. Their affinity for Na
in methanol, for example, ranges over about 1.5 orders of magnitude, the cis-syn-cis isomer being the most effective.  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License The Crown Ethers 117 OH O 2 NaOH + Cl R Cl R OH O (a) OH OH HO O T O NaOH Cl T Cl 2 + Cl S Cl OH S O 2 NaOH O O S O (b) OH O U O 4 NaOH 2 + 2 Cl U Cl OH O U O (c) OH 2 NaOH O V O O V Cl O V O (d) OH O KOBut V O V Cl O (e) O O H2 / RuO2 O O n n O O O O (f) Scheme 3.3 Methods for synthesising the crown ethers (R – V are organic linker groups). O O O O O O O O O O O O O O O O O O cis-syn-cis cis-anti-cis cis-t rans O O O O O O O O O O O O trans-syn-trans trans-anti-trans Figure 3.8 Isomers of dicyclohexylcrown-6.  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 118 Cation-Binding Hosts 3.3 The Lariat Ethers and Podands 3.3.1 Podands Gokel, G. W. and Murillo, O. ‘Podands’, in Comprehensive Supramolecular Chemistry, J. L. Atwood, J. E. D. Davies, D. D. MacNicol and F. Vögtle (eds), Pergamon: Oxford, 1996, vol. 1, 1–33. Acyclic hosts with pendant binding sites are termed podands. The simplest podands are simply acyclic analogues of the crown ethers such as pentaethyleneglycol dimethylether (3.11), analogous to crown-6, or its di-ol analogue (3.12). O O O O O O O O O O OH HO Me Me 3.11 3.12 Podand hosts generally exhibit less cation affinity than their cyclic analogues, as a result of their lack of preorganisation (Section 1.6), but they may adopt similar wrapping conformations to the crown ethers in the presence of suitable metal cations, such as the highly charged lanthanoids (Figure 3.9). The extra flexibility of podand hosts, however, also allows them to engage in multiple bridging and helical binding modes unknown for the crown ethers (Figure 3.10). As early as 1972, podand Ca2
carriers with transmembrane ionophore-type behaviour were reported by the Eidgenössische Technische Hochschule (ETH) group of Simon in Zürich.5 These hosts (e.g. 3.13) rely on the polar amide groups for their binding ability, and the fact that the long ester arms are flexible and able to bend back on themselves to simultaneously coordinate the metal cation. Figure 3.9 X-ray molecular structure of the europium(III) podand complex [Eu(H2O)3 (3.12)]3
.  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License The Lariat Ethers and Podands 119 Figure 3.10 Alternative coordination modes of the simple podand trioxyethylene glycol. Also, the aliphatic chains contribute greatly to the lipophilicity of the compound, hence its membrane transport ability. O Me N OEt O O O O O O O N N N O O O O O O O N OEt Me O MeO O O O 3.13 3.14 OMe O MeO O O O N O O N O O 2+ O Ca O HO OH 3.17 + + Me N O O N Me Me Me 3.15 3.16 The term ‘podand’ was coined by Vögtle and Weber in 1979,6 whose early work involved quinoline- terminated podands such as 3.14, which are able to form stable crystalline compounds with a wide variety of alkali metal salts as well as transition metals and even uranyl nitrate, UO2 (NO3) 2.6H2O. As part of their work on podands, these researchers also formulated the important end group con- cept. One of the problems with podands is their high degree of flexibility, allowing them to adopt non-binding open conformations. If the podand is terminated by a rigid functionality (e.g. aryl, ester, amide), however, binding is enhanced by the extra degree of organisation given to the podand host by the rigidifying endgroup. A good example of this concept is 3.15, in which the conjugated benzoic acid moieties provide a rigid, planar and partially preorganised binding cleft, especially in conjunction with the short diethyleneglycol bridge. The rigid-end-group host 3.16, which is terminated by quinone monoimine groups, was studied as a possible means of sensing the presence of cations by changes in the UV–visible absorption spectrum of the host (a chromoionophore). This zwitterionic podand, binding via two anionic oxygen donors, gave large spectral changes in the presence of divalent cations. Extending the podand concept to three dimensions gives tripodal  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 120 Cation-Binding Hosts 3.6 3.11 3.18 3.19 crown-6 O O O O O O O O Me O O NH HN Me N N O O O O Me O O O O O O O O log K 6.08 2.3 2.04 4.8 Me Figure 3.11 Binding constants (log K) for K
binding by simple podand, crown and lariat ether compounds in methanol at 25 °C. molecules, such as 3.17, which are able to encapsulate their guests more fully. Again, the rigid- end-group philosophy is applicable since tripodal hosts are often highly flexible as a consequence of inversion at the bridgehead nitrogen atoms. 3.3.2 Lariat Ethers Gokel, G. W., ‘Lariat ethers - from simple sidearms to supramolecular systems,’ Chem. Soc. Rev. 1992, 21, 39–47. The term ‘lariat ether’ refers to a crown ether or similar macrocyclic derivative with one or more accompanying appendages designed to enhance metal cation complexation ability by giving some three-dimensionality to the binding (cf. 3.17). The compounds may be regarded as a crown type macrocycle with a podand side-arm. The name comes from the Spanish la reata, meaning ‘the rope’, and suggests the concept of the macrocycle representing the loop of a lasso and the podand side-arm representing the rope by which it is held. They thus combine the higher rigidity and preorganisation of macrocyclic compounds with the additional stability and flexibility (which leads to fast cation- binding kinetics) of podand complexation. A comparison between the binding ability of some simple podand, crown and lariat ether compounds is shown in Figure 3.11. Clearly crown-6 is a much more effective ligand for K
than the podand pentaethyleneglycol dimethylether (3.11) (the K
complex is almost four orders of magnitude more stable). At fi rst sight, the lariat ether 3.19 seems to lie somewhat in between the binding ability of the cyclic and acyclic compounds. However, the affinity for K
is also regulated by the character of the donor atoms. A more direct comparison between the azacrowns 3.18 and 3.19 suggests that it is the three-dimensional binding by the lariat ether (Figure 3.12) that is the more effective of the two, although the greater amine basicity in 3.19 also enhances the binding constant (cf. N,N′-di-n-butyldiazacrown-6, log K  3.8). The number of donor atoms (eight as opposed to six) is also a key consideration. Me O O O O O O Me N N Me N N O O N N O O O O O O Me O O O O Me Me Figure 3.12 Metal complexation by a lariat ether.  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License The Lariat Ethers and Podands 121 TsCl OMe OMe O TsO Ts = O2SC6H4Me pyridine 0 oC, 10 min OMe OH MeO Na2CO3 NaH / dmf N OH O O H N + N TsO(CH2CH2O)4Ts OH OTs O O O Me OH O (a) BrNHCOMe O Br ROH RO 3.20 R = o-C6H4OMe O O HO(CH2CH2O)4H 3.21 R = p-C6H4OMe O O (b) OR Scheme 3.4 Syntheses of (a) N-pivot and (b) C-pivot lariat ethers 3.20 and 3.21.7 Compound 3.19 makes use of a tertiary amine nitrogen atom as a pivot where the podand side chain is attached to the crown ether ring. However, effective lariat ethers may also be synthesised by use of the carbon backbone as the point of attachment, as in 3.20. Example syntheses of both types of com- pound are shown in Scheme 3.4.7 Interestingly, while 3.20 may be readily prepared in excess of 70 per cent yield, the analogous para isomer 3.21, which is sterically incapable of coordinating both side-arm donors to the same metal cation, is produced in only 30 per cent yield, highlighting the mechanism of the reaction, which involves binding of the reactants to the metal cation during the transition state. The synthesis proceeds under kinetic control, using the metal ion to organise the reactants. This is referred to as the kinetic template effect and is discussed in detail in Section 3.9. 3.3.3 Bibracchial Lariat Ethers The lariat ether concept may be readily extended to produce so-called bibracchial lariat ether (BiBLE) hosts, which have two podand side-arms, thus displaying even more three-dimensional binding coverage of the metal ion guest’s surface. A wide range of N-pivot BiBLE ligands has been prepared by Gokel et al. according to the convenient, although relatively low-yielding, method outlined in Scheme 3.5.8 NH2 Na2CO3 O O O O R R + N N R MeCN I I O O * * * OMe O N * * R= (* = point of attachment) Scheme 3.5 Simultaneous four-bond coupling to produce BiBLE ligands.8  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 122 Cation-Binding Hosts 3.4 The Cryptands Lehn, J. -M. Supramolecular Chemistry: Concepts and Perspectives, VCH: Weinheim, 1995. Shortly after Pedersen’s work, Jean-Marie Lehn, then a young researcher at the University of Strasbourg, decided to design three-dimensional analogues of the crown ethers. In this way it was anticipated that metal ions could be encapsulated entirely within a crown-like host with consequent gains in cation selectivity and enhancements in ionophore-like transport properties. Accordingly, the bicyclic cryptands (named because of their ability to spherically surround or ‘entomb’ a metal cation, as in a crypt; from the Greek kruptos meaning ‘hidden’) and an ever-increasing range of related compounds were synthesised using the high-dilution technique (Section 3.9.2), the major- ity in remarkably high yields (Scheme 3.6). The fi rst and most important member of the series is [2.2.2]cryptand (3.22), which is sold commercially under the trade name Kryptofi x ®. Because it is based on a similar sized ring to crown-6, this host also exhibits selectivity for K
over the other alkali metals. The binding of K
by [2.2.2]cryptand in methanol is, however, some 10 4 times stronger than its crown analogue. Similarly, [2.2.1]cryptand (3.23) is selective for Na
. The key to the dramatically enhanced metal cation binding ability of cryptands over crown ethers is the pre- organised, three-dimensional nature of their cavity, which enables spherical recognition of the M
ion to take place. High-dilution syntheses, while highly versatile, do not readily enable large quantities of material to be prepared, and often involve many steps, particularly the final reductive decarbonylation with diborane. Since the early work of Lehn, a wide variety of synthetic procedures has been developed to prepare an enormous range of cryptands of varying degrees of sophistication, including chiral species and those containing three different bridges. Many of these follow a stepwise approach, which may be summarised as follows: 1. building up two linear chains possessing suitable reactive groups at each chain end; 2. cyclisation reaction of these two chains leading to a corand (crown ether-like macrocycle); 3. addition of a third chain to the corand to give a macrobicyclic compound. Because of the length and tedium of such approaches, much simpler routes, often taking advantage of the template effect (Section 3.9.1), have been devised, which, in many simple cases, can be remarkably effective. Such alternative syntheses of 3.22 and related compounds are shown in Scheme 3.7. The synthetic versatility of azacrowns (crown ethers with some oxygen atoms replaced with NH functionalities), such as 3.18, has made possible the synthesis of an enormous range of O O 1. B2H6 n O O O O 2. H2O O O Cl O Cl O n 95 % yield n NH HN N O N O N ON high dilution O O O O O O base 45% 3.18 n = 1, 3.22 n = 2, 3.23 Scheme 3.6 High dilution synthesis of [2.2.1]cryptand (n  1) and [2.2.2]cryptand (n  2).9  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License The Cryptands 123 n TsO O O OTs TsO O O OTs O O + K2CO3 + nN K2CO3 N O O N N O O MeCN m MeCN O O m O O H2 N O O NH2 NH2 NH2 Yield Yields m = n = 1, 36 % 25 % m = 1, n = 2, 50 % (b) m = 2, n = 2, 40 % (a) Scheme 3.7 Simple template synthesis of selected cryptands with yields. cryptands, many with very unusual cation and anion binding properties. Some examples are shown in Figure 3.13. A particularly novel class of cryptand are the sepulchrates, which arise from a Co(III) templated condensation of various capping groups with tris(chelate) complexes of bidentate ligands. The fi rst example, reported in 1968 by Boston and Rose, involved the reaction of tris(dimethylglyoximato) cobalt(III) with the highly Lewis acidic boron trifluoride etherate (Scheme 3.8). The fact that Co(III) is a highly inert metal ion (see Chapter 10, Box 10.1) means that it is very likely that the cryptate complex is formed as a consequence of the ability of the metal ion to hold the ligand in an orientation similar to the geometry of the final product. Work by Alan Sargeson of the Australian National University has resulted in a range of related complexes derived for 1,2-diaminoethane. Sargeson coined the name ‘sepulchrates’ for the nitrogen-capped metal compounds and ‘sarcophag- ines’ for the analogous C-bridged species because the sequestered, metal complex is highly stable R N N N OH N N F N N O N O O O O O O O O O 3.24 3.25 3.26 O O N N N N N O O O O O O N N O N N O N O N O O O N O N O O O N O N N 3.27 3.28 3.29 Figure 3.13 Some examples of cryptands.  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 124 Cation-Binding Hosts F B OH O OH - O OO + N NN BF3.Et2O N NN Co Co N NN N NN O OO O OO B F N 3+ + 6 HCHO NH2 NH2 NH2 2 NH3 NHH 2 2 N NH 2 Co Co sepulchrate Li2CO3 NH2 NH NH2 NH2 NH NH2 2 2 N R 3.30 sarcophagenes NH HN NH N N 3+ N Co R=NO2, NH2, CH2Cl etc. M HN 3.31 N N N 3.32 NH HN M = Co(III), Pt(IV) R Scheme 3.8 Sepulchrates and sarcophagenes. and unreactive. This work has been extended to remarkably regio- and stereoselective preparations of Co(III) and Pt(IV) sarcophagines such as 3.32 by a Schiff base condensation technique. Schiff base macrocycles are described further in Section 3.10.6. The X-ray structure of one such compound is shown in Figure 3.14. Figure 3.14 X-ray molecular structure of the Pt(IV) sarcophagine 3.32.10  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License The Spherands 125 3.5 The Spherands Maverick, E. and Cram, D. J., ‘Spherands: Hosts preorganised for binding cations’, in Comprehensive Supramolecular Chemistry, J. L. Atwood, J. E. D. Davies, D. D. MacNicol and F. Vögtle (eds), Pergamon: Oxford, 1996, vol. 1, 213–243. In addition to Pedersen and Lehn, the 1987 Nobel Prize for Chemistry was shared by a third supramo- lecular chemist, Donald Cram, for his development of a further type of macrocyclic cation host: the spherands. While the crown ethers and even the cryptands are relatively flexible in solution, Cram realised that if a rigid host could be designed, which had donor sites that were forced to converge on a central binding pocket even before the addition of a metal cation, then strong binding and excellent selectivities between cations should be observed. Using space-filling molecular models (termed Corey–Pauling–Koltun or CPK models, Figure 3.15), Cram and co-workers designed the rigid, three-dimensional spherands 3.33 and 3.34, whose cation-binding oxygen atoms are preorganised in an octahedral array, ready to receive a metal ion. In the case of compound 3.33, three of the aryl rings are pointing upwards (out of the page) and three downwards. This results in the anisyl oxygen atoms being fixed in a nearly perfect octahedral array, while the p-methyl and anisyl methyl groups present a lipophilic surface to the solvent. This host selectively binds small cations such as Li
and, to a lesser extent, Na
, in its cavity. Indeed, 3.33 is one of the strongest complexants known for Li
. All other cations are excluded because they are simply too big to fit within the binding pocket. Spherand 3.34 has a binding pocket of similar size, formed from the tethering of the rings in pairs with diethylene glycol linkages, resulting in four rings being down and two up. The analogous fluoro compound 3.35, along with an octameric analogue, Figure 3.15 CPK models of crown ether and spherand complexes were used in early supramolecular design. Modern computational techniques, such as molecular mechanics have, to some extent, replaced such tactile representations. (Reproduced by permission of John Wiley & Sons, Ltd).  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 126 Cation-Binding Hosts Br Br 1. BuLi O O 2. Fe(acac)3 O 2 O O + O Li Cl 3. EDTA4- O O O 4. HCl O O acacH = 3.33.Li+ Scheme 3.9 Synthesis of spherand 3.33 as its Li
complex (spheraplex). have also been synthesised. While X-ray crystallography confirms that the hosts possess very similar cavities to 3.33, the fluoro species display no metal ion binding properties. Apparently the intrinsic affinity of fluoro substituents for alkali metal cations is so low that even incorporation of multiple binding sites into a highly organised host cannot bring the binding on to a measurable scale. Interest- ingly, however, a definite interaction of cation guest with a fluoro substituent is observed for cryptand 3.25, both in the solid state by X-ray crystallography and in solution by 19F NMR spectroscopy. In order to achieve the final cyclisation step in the spherand syntheses, a new synthetic ring-closure procedure was developed, which proceeds according to Equation 3.2 (acac  acetylacetonato, CH(COMe)2). The aryl lithium compound produced by action of butyl lithium is oxidised by the Fe(III) complex to give an aryl biradical, which then undergoes a template cyclisation about the Li
ion (see Section 3.9.1 for an explanation of the template effect). In the case of 3.33, this method resulted in the isolation of the complex in 28 % yield from the reaction shown in Scheme 3.9. Ar–Br BuLi  → Ar–Li Fe(acac)  3 → Ar– Ar (3.2) By way of comparison, acyclic podand hosts analogous to compound 3.33 have been produced in order to assess the importance of the rigid preorganisation afforded by a cyclic host. Comparison of the closely related podand 3.36 (which is estimated to possess over 10 000 possible conformations, only two of which can bind cations in a convergent manner) and 3.33 (which is locked in only one confor- mation) shows that the spherand binds Li
more than 1012 times more effectively. This highlights the importance of preorganisation effects in host design. Me Me Me Me Me Me O O O O O O H H Me Me Me Me Me Me 3.36 Cross-fertilisation between the crown ethers (or, more generally, corands), cryptands, spherands and podands has produced an enormous range of hybrid hosts such as cryptaspherands and hemispherands, many exhibiting all the useful features of the parent materials (Figure 3.16).  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Nomenclature of Cation-Binding Macrocycles 127 H H O O O O O O N O O N O O Et O O O O O O O O Et O O Et Et hemispherand cryptaspherand hemispherand 3.37 3.38 3.39 Figure 3.16 Hybrid cation hosts. 3.6 Nomenclature of Cation-Binding Macrocycles The naming of crown ethers is relatively simple, although for historical reasons it is only unambiguous if the exclusive use of ethylene glycol (—OCH2CH2O—) linkages is assumed; in more complicated cases, the name is usually accompanied by a pictorial representation. 1. The first number in the crown name designates the number of atoms in the ring (usually given in square brackets). 2. The second number gives the number of oxygen (or other donor) atoms. 3. Substituents are denoted with a prefix such as benzo-, dicyclohexano- etc. For example, dibenzocrown-6 is a crown ether with an 18-membered macrocyclic ring containing six oxygen atoms with two benzo substituents. A more generalised system of nomenclature for such neutral organic ligands was developed by Vögtle and Weber,6 and later modified by Cram,12 in which any monocyclic system such as a the crown ether is termed a corand (originally coronand). Open chain molecules such as 3.3 and 3.11 are called podands; bicyclic or oligocyclic systems are termed cryptands (Figure 3.17); and rigid, p-methylanisole-based systems are given the name spherand. In general, for purely oxygen donor ligands, the historical crown nomenclature is retained. Figure 3.18 shows examples of podands, corands and cryptands. Under this system, all of the crown ethers belong to the general class of corands. The name of crown-6 (3.6) becomes 18〈O626corand-6〉, meaning an 18-membered ring monocyclic (i.e. corand) compound containing six oxygen atoms linked by six spacers containing two carbon atoms each to give a total of six donor atoms. Open chain Cyclic Three-dimensional D n n D D B B D D D D n D n D n Podand Corand Cryptand D = donor atom D = O, Crown ether B = bridgehead atom Figure 3.17 Classes of cyclic and acyclic ligands.  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 128 Cation-Binding Hosts CH3 O O O O O O O O O O O O N O O O O O O O CH3 CH3 CH3 CH3 3.11 3.40 3.41 N O O O O O H N N O HN N O O N H O O N O O O 3.42 3.43 3.23 Figure 3.18 Podands 3.11 and 3.40, corands 3.41–3.43 and cryptands 3.23 related to crown-6. Compounds 3.41 and 3.42 are members of a hybrid corand–podand family termed lariat ethers. Lariat ethers with two podand arms such as 3.42 are termed BiBLE’s (bi-bracchial lariat ethers). This is sometimes abbreviated to -O6. Similarly, the azacorand compound 3.43 is called 18〈O3N326corand-6〉. A similar system of nomenclature, also common in the literature, especially for corands containing multiple nitrogen or sulfur donors, considers the corand as a large, substituted cycloalkane, hence the suffix ‘-ane’ is used along with numbers denoting the ring size and donor atoms. Thus, 18〈N626corand-6〉 (3.44) is denoted ane-N6 and its sulfur homologue, ane-S6. Other names, often descriptive, are also common in the literature. Compound 3.44 is often called hexacyclen (because of its resemblance to 1,2-diaminoethane, often abbreviated as ‘en’). Azacorand analogues of crown ethers such as 3.45, in which the ether oxygen atoms are replaced entirely by NH functions, were developed separately as ligands for transition metals and thus often have unusual names. Thus corand 3.45 is referred to as cyclam. It could also be referred to as tetrazacrown-4 or, more systematically, 14〈N42232corand-4〉. Similarly, there are trivial names such as sarcophagenes and sepulculrates 3.30–3.32, named by analogy with the cryptands.† N NH HN NH H HN NH H HN NH HN N 3.44 3.45 † The way in which new nomenclature enters the literature can be somewhat haphazard. The dangerously explosive macro- cycle Se 4N4 has been referred to as phaetodigitogenic, a term coined by Woolins et al., then of Imperial College London, which essentially implies that the compound explosively removes fingers!  10.1002/9780470740880.ch3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch3 by Uva Universiteitsbibliotheek, Wiley Online Library on [16/09/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Selectivity of Cation Complexation 129 The cryptands, which present a more spherical, rather than crown-like structure, have a somewhat different nomenclature. Each host is denoted by a series of numbers indicating the number of donor atoms in each of the bridges between the bridgehead atoms. Thus 3.22 and 3.23 are termed [2.2.2]cryptand and [2.2.1]cryptand, respectively. An enormous variety of cryptands have been synthesised, with bridges originating from carbon as well as nitrogen atoms, bridged by S, N and various other donor atoms and moieties and containing multiple bridges (e.g. the ‘soccer ball’ cryptand, 3.29). The nomenclature of these species is often highly appealing, although nonrigorous. All of these ligands have been synthesised with the binding of metal cation guests in mind, and thus the nomenclature must disti