Molecular Devices PDF

Molecular Devices 11 ‘The whole is greater than its part.’ I. Todhunter, The Elements of Euclid, Macmillan: London, 1880 (from the original Greek). ...

Molecular Devices 11 ‘The whole is greater than its part.’ I. Todhunter, The Elements of Euclid, Macmillan: London, 1880 (from the original Greek).  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 708 Molecular Devices 11.1 Introduction Balzani, V., Credi, A. and Venturi, M. Molecular Devices and Machines, Wiley-VCH: Weinheim, 2003. 11.1.1 Philosophy of Molecular Devices The traditional goals of chemistry have centred on the idea of making molecules. If you can make mol- ecules, then you can make new substances and materials, or make more of existing ones. Chemical substances – things like dyes and paints, metals and plastics, glass and ceramics, pharmaceuticals, syn- thetic textiles etc.– are useful things to have around. You can build things with them, wear them, change aesthetic and corrosion-resistance properties, consume them, make utensils and tools with them, and do whole host of other civilised things. In each case, only one main property of the chemical substance is being employed. This property can be something like colour (light absorption or emission), hardness and mechanical strength, cohesion and flexibility, therapeutic benefit (biochemical role), electrical or thermal conductance or insulating properties, and so on. Traditionally is has been left to the discipline of engineering, and latterly electronic engineering, to combine the bulk features of a number of different (chemical) substances, or at least mechanical components, in order to produce machines or devices. We can define a machine as a functioning entity composed of a number of interacting components that collectively carry out a predefined task for (presumably) beneficial result. A machine, or device, differs from a chemical substance in that it is useful for what it does, rather than for what it is. Traditionally, because the components of a machine must be precisely orientated with respect to one another, and interact in a well-defined way, machines have been macroscopic structures that a human being can assemble through the use of keen eyesight and manual dexterity. In recent years, microscale machines (devices with components of the order of micrometres) have been produced in the electronics industry by other machines, which, metaphorically, have a vastly greater degree of manual dexterity and keener eyesight. Microscale machines take up less space, are subject to fewer external influences and, above all, are much faster and more sophisticated (in terms of the tasks they can perform) than their large ancestors. What has yet to be fully realised is the artificial production of nanoscale machines – machines the size of individual molecules – that have the potential to be yet faster and even more sophisticated. These kinds of nanoscale devices are exactly what Nature has evolved in living organisms and biochemical systems such as transport, signalling and recognition proteins, enzymes and the machinery of self- replication provide a lot if inspiration. In Section 10.1, and throughout Chapter 10, we looked at methods by which large molecules or molecular- based systems on the nanometre scale might be assembled, or induced to self-assemble. Self-assembly addresses the fundamental problem of synthesising nanoscale structures in which components are placed in the correct orientations with respect to one another. In this chapter, we continue the theme of engineering large molecular structures, but with an emphasis on the properties of the individual components themselves and the ways in which they communicate, as a means towards the construction of molecular and supramolecular analogues of ‘real world’ devices, functioning on the nanometre scale. 11.1.2 When Is a Device Supramolecular? So far we have focused on Lehn’s definition of a supramolecular compound as one involving non- covalent interactions (Section 1.1). Such a definition is entirely appropriate to host–guest chemistry, 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.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 709 and even to the templated and thermodynamic self-assembling systems discussed in Chapter 10. When we come to molecular devices, however, such a definition is not used by many workers in the field because the focus in the construction of supramolecular devices is on the functional interac- tions between the components rather than the chemical nature of their connectivity. This means that a ‘supramolecular device’ can be an entirely covalent molecule if it possesses characteristics of a supramolecular nature. Thus a supramolecular device may be defined as a complex system made up of molecular components with defi nite individual properties. These properties are intrinsic to a particular molecular component whether it is part of the device or not. Another way of saying this is that the interaction energy between the components of the supramolecular device must be small compared with other energy parameters relevant to the system. It does not matter, therefore, how the components are connected together in the device (covalently, hydrogen-bonded, coordination interac- tion etc.); all that matters is that each component should contribute something unique and identifiable with that component alone, within the system. If this rule does not hold true, and the functionality of the system is identifiable with the molecule as a whole, as opposed to the individual parts, then the complex is best thought of as a ‘large molecule’ and we will not consider it to be supramolecular. Such a definition of a supramolecular device does not exclude the ‘traditional’ concept of host and guest or receptor and substrate. Molecular recognition events between host and guest may be an intrinsic part of the operation of a supramolecular device, which might, for example, be designed to bind and then signal the presence of a guest. Common components within supramolecular devices as they are studied today are photochemically or redox active molecules, i.e. molecules capable of absorbing and/or emitting light and molecules capable of losing or gaining an electron. The definition of a supramolecular device made up from these components is illustrated in Figure 11.1. If light excitation of a molecule (♦↔ ) results in the formation Figure 11.1 Photochemical and electrochemical criteria used to classify a complex chemical species as a supramolecular device or a large molecule.  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 710 Molecular Devices of excited states that are substantially localised on one of the two components (♦ or ) or causes electron transfer from one to the other, then the molecule is said to be supramolecular. If the excited state is substantially delocalised across both components, then the complex is best thought of as being simply a large single molecule. Analogous arguments apply to redox processes. Clearly a caveat of this line of reasoning is that the nature of the interaction between the two components would be expected to be dependent crucially on the nature of the connection between them (↔), be it a bridging ligand, rigid or mobile covalent spacer unit, or a non-covalent link. 11.2 Supramolecular Photochemistry Balzani, V. and Scandola, F., Supramolecular Photochemistry, Ellis Horwood: Chichester, 1991. 11.2.1 Photophysical Fundamentals Of all the ways in which to construct a supramolecular device, the use of photochemically active components is, perhaps, the most versatile. Light-induced processes are of fundamental importance in biochemical devices such as plant photosynthetic membranes (Section 2.4). Light-absorbing com- ponents (chromophores) are readily available and lend themselves to extensive synthetic modifica- tion, and light is readily introduced to a system that is in a variety of physical states (e.g. solid, liquid or gas) or media (solutions in various solvents). Light may be used to induce events such as charge separation, initiate catalysis, interrogate a system in sensing applications, or to bring about changes in the state of a bistable device (switching). Incorporation of photochemically active components within a supramolecular complex may be expected to perturb or modulate the photochemical behaviour of the chromophore(s), giving rise to a number of interesting and potentially useful effects such as energy migration, photoinduced charge separation, perturbations of optical transitions and polarisabilities, modification of ground- and excited-state redox potentials, photoregulation of binding properties, selective photochemical reactivity etc. When a molecular chromophore is irradiated with electromagnetic radiation of a wavelength corre- sponding to the energy required to promote an electron to an accessible electronic excited state, energy is absorbed resulting in the promotion of an electron from a ground-state molecular orbital to one of higher energy. This is known as primary charge separation and results in a high-energy electron and a positively charged ‘hole’. The energy of the excited-state electron can either be dissipated as heat by relaxation by the solvent (nonradiative decay), emitted radiatively (luminescence) or used to carry out a chemical reduction (Figure 11.2a). Luminescence involving direct radiative decay, in which the electron returns immediately to the ground state from a singlet excited state, is termed fluorescence. Fluorescent emissions are usually of lower energy than the absorbed energy because the electron is promoted into a vibrationally excited state from which it relaxes non-radiatively before fluorescing back to the electronic ground state (Figure 11.2a). This is the reason why many fluorescent dyes are able to absorb high-energy UV light and fluoresce in the visible region. If the electron undergoes a change of spin state (intersystem crossing), then it accesses the triplet manifold of excited states. This transformation is formally forbidden because of the selection rule that there must not be a change in spin state during the transition, i.e. ∆S ⫽ 0. This selection rule is a reflection of the fact that there is no component of electromagnetic radiation that interacts with electron spin. However the rule may be relaxed as a consequence of spin-orbit coupling. As a result of this selection rule, the triplet excited state, once formed, is long-lived and may undergo vibrational relaxation to a lower energy level before relaxing slowly via a back intersystem crossing, emitting a kind of luminescence, termed phosphorescence, of a lower frequency to the absorbed light. Fluorescent and phosphorescent processes may be distinguished by time-resolved spectroscopic measurements because  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 Supramolecular Photochemistry 711 Figure 11.2 (a) Possible radiative events following photoexcitation, (b) Photoexcitation of an electron into a vibrationally and electronically excited state. The electronic transition takes place within a stationary nuclear framework (Franck–Condon principle). Once the transition has occurred, the molecule experiences a new force field and bursts into vibration. phosphorescence takes much longer to decay than fluorescence. In the presence of an external electron acceptor (low-lying empty orbital on an adjacent molecule or component), the excited-state electron may reduce the acceptor chemically, resulting in long-lived spatial charge separation (secondary charge separation). Eventual recombination is accompanied by emission of light of a different frequency, or by emission of heat. Finally, the energy from the excited state may be transferred to an external acceptor without electron transfer. This is termed energy transfer (ET). The resulting secondary excited state may then relax with emission of luminescence, again of a lower frequency to the original absorption. This process is the beginning of an energy transfer cascade as in photosynthesis. In some cases excited state chromophores form supramolecular complexes either with ground state chromophores on the same molecule or one nearby resulting in the formation of an excimer (excited state dimer) or, if the two chromophores are different to one another an exciplex. Formally an excimer as defined as a dimer which is associated in an electronic excited state and which is dissociative in its ground state.1 Formation of the pyrene excimer is illustrated in Figure 11.3. The results of photoexcitation may be divided into three broad categories: 1. Re-emission of the absorbed energy as light (fluorescence or phosphorescence). 2. Chemical reaction of the excited state (secondary charge separation, isomerisation, disassociation etc.). 3. Non-radiative vibrational quenching of the excitation by solvent. Within the context of supramolecular devices, re-emission of the radiation by luminescence is of inter- est in sensing and signalling applications, while chemical reactions are of interest in applications such as molecular switches and photocatalysis. Strictly speaking absorption and re-emission type processes are termed molecular ‘photophysics’ while light-induced chemical reactions or chemical processes are termed ‘photochemistry’.  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 712 Molecular Devices * * hν pyrene 340 nm pyrene monomer pyrene excimer emission λmax = 380 nm emission λmax = 550 nm Figure 11.3 The excited state of a chromophore such as pyrene can form a complex with a ground state molecule. If the result is an excited state dimer the complex is known as an excimer, while if the excited complex is formed between two different molecule it is termed and exciplex. Excimers and exciplexes emit at lower energy than the corresponding monomers. Common chromophores are transition metal complexes that may be used either as a light-absorbing (sensitising) part of a supramolecular device or as a signalling and sensing component. In the case of a mononuclear transition metal complex, a simple molecular orbital (MO) description may be derived from linear combination of the orbitals of the metal and ligands. It is convenient, and usually a reasonable approximation, to discuss both the ground- and electronically excited-state electron configurations in terms of this orbital description. This is termed the ‘localised molecular orbital approximation’. Individual MOs may be labelled M or L according to whether they are predominantly metal- or ligand-centred. An MO scheme for a mononuclear complex between a metal (M) and ligands (L) is shown in Figure 11.4. In general, in complex ground states, the σL and πL ligand orbitals are full. The πM orbitals (the t2g set in octahedral transition metal complexes) are at least partially filled (depending on the metal oxidation state) and the other orbitals are usually empty. Light absorption changes these populations. Metal-centred (MC) Figure 11.4 Energy level diagram for an octahedral transition metal complex showing the various kinds of electronic transition. MC ⫽ metal-centred, LC ⫽ ligand-centred, MLCT ⫽ metal-to-ligand charge transfer, LMCT ⫽ ligand-to metal-charge transfer.  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 Supramolecular Photochemistry 713 transitions, sometimes called d–d or ligand field transitions, are of low energy, often corresponding to the wavelength of visible light (transition metals are frequently highly coloured), and involve rearrangements of the metal d-electrons from the t2g to eg sets. Ligand-centred (LC) transitions are much higher in energy. The most common process is charge transfer (CT), which is strongly allowed since it involves a change in orbital angular momentum quantum number (d-electrons becoming p-electrons or vice versa). Metal- to-ligand charge transfer (MLCT) is particularly common since this involves promotion of electrons in the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO). The exact order of stability of the orbitals is highly dependent on the nature of the ligands and metal, and frequently the πL∗ level comes below the metal-centred eg (σM∗) set, e.g. [Ru(bpy)3]2⫹ (bpy ⫽ 2,2′-bipyridyl). In some cases, oxidation or reduction leads to undesirable decomposition or metal–ligand disassociation, and it is a fundamental requirement for the incorporation of a chromophore into a supramolecular device that its redox and photophysical behaviour should be stable and reversible. 11.2.2 Mechanisms of Energy and Electron Transfer The photoinduced transfer of energy (excitation) or electrons is of fundamental importance in molecular devices. Excitation transfer requires some kind of electronic interaction between donor and acceptor. It occurs via one of two mechanisms:2 1. The electron exchange (Dexter) mechanism which involves the overlap of wavefunctions of the donor and acceptor groups. This is a short-range excitation transfer mechanism that operates by exchange of electrons. 2. The coulombic (Förster) mechanism, a dipole-dipole mechanism that can operate over distances as long as 10 nm. The difference between the mechanisms is illustrated in Figure 11.5. While the Dexter mechanism requires direct orbital overlap, the Förster mechanism involves energy transfer by non-radiative, long- range dipole-dipole coupling to an acceptor chromophore typically up to 10 nm away. Formally this energy transfer mechanism is called ‘Förster resonance energy transfer’ (FRET) after the German photophysicist Theodor Förster. Very often when both chromophores are fluorescent, the term ‘fluo- rescence resonance energy transfer’ is used, however the energy transfer is a non-radiative process and does not involve fluorescence hence the former description is preferred. A significant challenge in the design of molecular devices is to extend the orbital overlap over a considerable distance to allow controlled, longer distance electron or Dexter-type energy transfer. The kinds of approaches that have been adopted use ‘molecular wires’ (Section 11.4.2) in which a conjugated π-system provides the necessary orbital overlap for efficient electron-exchange-type excitation transfer. An example is the linking of a [Ru(bpy)3]2⫹ type donor with a [Os(bpy)3]2⫹ acceptor by a rigid conju- gated oligophenylene spacer, Figure 11.6a. The phosphorescence of the [Ru(bpy)3]2⫹ moiety (and of the oligophenylene spacer) is quenched by extremely rapid energy transfer to the osmium unit via the electron exchange mechanism despite the fact that the components are separated by 4.2 nm. Decreasing the number of oligophenylene spacer units increases the rate and efficiency of the excitation transfer process.3 In addition photoexcitation can also result in the transfer of an excited state electron to a distant acceptor group resulting in charge separation. This process can be understood within the framework of Marcus theory and subsequent more sophisticated theoretical treatments.2, 5 The rate of electron transfer (kel) drops with distance according to an attenuation factor β el; kel ∝ exp(⫺β el rAB) where rAB is the distance between donor and acceptor components A and B. When the donor and acceptor components are separated by a vacuum β el is estimated to be ca. 2 – 5 Å⫺1. However when some kind of material substance is involved such as a bridge ‘L’ the electron transfer process can be  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 714 Molecular Devices 1 LUMO 2 HOMO 1 LUMO Förster (coulombic) mechanism LUMO energy transfer 2 HOMO 1 HOMO LUMO orbital A* - L - B A - L - B* energy A-centred excited state B-centred excited state 2 HOMO Dexter (exchange) mechanism Figure 11.5 Diagrammatic explanation of the coulombic and electron exchange energy transfer mechanisms (A and B are chromophore components and L is a bridging moiety or ligand). excitation transfer by electron exchange hν hν' N N N N 2+ 2+ N Ru N N Os N N N N N rigid spacer donor group acceptor group (a) hν photo-induced electron transfer N N N N 2+ 3+ N Ru N N Rh N n N N n = 1-3 N N hν electron donor group electron acceptor group (b) Figure 11.6 (a) The triplet excitation transfer3 between the ruthenium(II) donor and the osmium(II) acceptor in [Ru(bpy)3] 2⫹⫺(ph)7⫺[Os(bpy)3] 2⫹ (b) photo-induced electron transfer from Ru(II) to Rh(III) across varying spacer lengths.4  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 Supramolecular Photochemistry 715 mediated by mixing of the initial and final states of the system with virtual, high energy electron transfer states involving the intervening medium. This is referred to as the superexchange mechanism and has some similarities to the Dexter energy transfer mechanism. Similar systems to that shown in Figure 11.6a based on the more electron deficient Rh(III) of general formula [(Me2phen) 2Ru- bpy-(ph) n -bpy-Rh(Me2bpy)] 5⫹ (Me2phen ⫽ 4,7-dimethyl⫺1,10-phenanthroline; ph ⫽ 1,4-phenylene; n ⫽ 1⫺3) have been used to study photo-induced electron (as opposed to energy) transfer as a function of spacer length, Figure 11.6b. For the compound with n ⫽ 3 the metal-metal distance is 2.4 nm. For all of the members of the series the typical metal-to-ligand charge-transfer luminescence of the ruthenium(II) moiety is quenched, meaning that efficient intramolecular photoinduced electron transfer from the excited ruthenium(II) moiety to the rhodium(III) unit takes place. The rate of the process decreases with spacer length but substituent effects indicate this it is also highly dependent on phenylene group mutual orientation highlighting the through-bond nature of the process, which proceeds by a superexchange mechanism.4 11.2.3 Bimetallic Systems and Mixed Valence By definition, a supramolecular device comprises more than one component, and this may well involve two or more photophysically active or redox centres. The nature of the interaction between these compo- nents will depend strongly on the nature of the bridge between them. In the case of two interacting metal centres of the type discussed above, we can distinguish three classes of behaviour, denoted class I, II and III. If we consider the interaction between two valence-localised electronic isomers, in which the metal centres are of oxidation state ⫹2 or ⫹3 (e.g. M1II-MIII III II 2 and M1 -M2 ), there exist particular equilibrium ge- ometries (in terms of the primary coordination sphere M–L distances and interactions with the solvation sphere) for each ‘isomer’. In a class I system, the two metal centres are essentially completely isolated and behave in a fashion analogous to the isolated mononuclear complexes. This situation is represented in Figure 11.7a. At the geometry of the electronic minima, each isomer may be considered to be an excited state of the other, separated by a reorganisational energy, λ. Broadly speaking, λ represents the energy required to move the ligand and solvent atoms to their new equilibrium positions in response to a change in the metal oxidation state. At the crossing point, both electronic isomers have the same geometry and are of the same energy. Electron exchange between them is extremely unlikely, however, even if they acquire the necessary energy to reach the crossing point (λ/4 in this case). In most cases, however, there is some electronic interaction between the two metal centres (Figure 11.7b). At the equilibrium geometries, the metal–metal interaction has little effect on the shape of the potential energy curve, but near the crossing point there is mixing of the zero order states (the isolated case represented in Figure 11.7a). This mixing is termed ‘avoided crossing’, and is a feature of class II mixed valence species. The systems are still valence-localised, and therefore are still supramolecular in the sense explained above, but new properties arising from the intervalence metal–metal interaction can be observed with, for example, an optical interva- lence transition (IT) interconverting the two isomers. The energy for this kind of transition may be provided by irradiation. Class III behaviour, as in the famous Ru(II)-Ru(III) Creutz–Taube ion, [(NH3)5Ru(µ- pyrazene)Ru(NH3)5]5⫹, is outside the realm of supramolecular chemistry and corresponds to the situation in which no valence localisation can be observed (Figure 11.7c). In this case, there is significant metal– metal electronic coupling and the equilibrium geometries are significantly perturbed, corresponding to a single potential energy minimum for the complex in the case of very large coupling where λ ⬇ H (the interaction energy). The properties of the fully delocalised M1II1/2 -M2II1/2 system bear little resemblance to the mononuclear precursors or components. Class III systems are classified as ‘large molecules’.  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 716 Molecular Devices (a) (b) (c) hν(IT) energy λ λ/4 2H M1II-M2III M1III-M2II M1II-M2III M1III-M2II nuclear configuration M1II1/2-M2II1/2 Figure 11.7 Potential energy diagrams for a two-component mixed valence device with (a) negligible, (b) weak and (c) strong electronic coupling. The dashed curves represent unperturbed zero-order states. The horizontal axis represents generalised nuclear coordinates, including contributions from both ligand atom positions and solvent sphere. 11.2.4 Bipyridine and Friends as Device Components One of the most common types of ligand in supramolecular photochemistry is the bidentate chelate 2,2′-bipyridine (bpy, 11.1), which represents the first member of the homologous series of polypyridyls that includes 2,2′:6′,2″-terpyridine (tpy, 11.3) and carries on up to the helicate-forming sexipyridine (10.120, Section 10.8) and beyond. N N N N N N N bpy phen tpy 11.1 11.2 11.3 As molecules with extended π-systems, bpy, tpy and 1,10-phenanthroline (phen, 11.2) are all capable of absorbing light, resulting in a π–π* transition. They are thus capable of sensitising coordinated metal ions by acting as light harvesters. Conversely, if the metal ion itself is photoexcited, then relax- ation may occur by MLCT, in which the metal excited state reduces the ligand by charge transfer to the ligand-centred πL* orbital. In practice, metal complexes of bpy and phen, such as [Ru(bpy) 3] 2⫹ and [Ru(phen) 3] 2⫹, ex- hibit long-lived phosphorescent excited states, arising from ligand-centred triplet charge transfer states (3MLCT). Lifetimes are of the order of 10 2 –103 ns in fluid solution at room temperature.  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 Supramolecular Photochemistry 717 Figure 11.8 Absorption and phosphorescence electronic spectra of [Ru(bpy)3] 2⫹ (vertical axis: intensity, arbitrary units). The phosphorescence spectrum of [Ru(bpy) 3] 2⫹ is shown in Figure 11.8. Much can be learned about the nature of photoexcited states by study of electrochemically reduced or oxidised species. Electrochemical oxidation of [Ru(bpy) 3] 2⫹ involves removal of a metal-centred electron at about ⫹1.25 V to give a Ru(III) species (potential versus saturated calomel electrode (SCE) reference in MeCN). Replacement of one bpy by two chloride ligands lowers the potential to ⫹0.32 V, while two CO π-acceptor ligands raise it to ⫹1.9 V. Reduction is reversible and takes place on a π L* orbital localised on a single ligand, keeping intact the inert t 26g electron configuration at the metal. At ⫺54ºC in DMF, up to six reduction waves may be observed between ⫺1.33 and ⫺2.85 V, corresponding to the sequential addition of one and then two electrons to each ligand. These electrochemical results tell us that if the orbital configuration is the same for the starting complex and its one-electron reduced form in both the ground and excited states (Koopman’s theorem), then photoexcitation of [Ru(bpy) 3] 2⫹ must involve promotion of an electron from a metal-based HOMO to a ligand-based LUMO, i.e. MLCT, implying that the π L* level is lower in energy than the metal eg set. In contrast to the tris(bpy) and tris(phen) complexes, [Ru(tpy) 2] 2⫹ has a very short MLCT lifetime (about 250 ps) and does not luminesce. The main reason for this difference is that tpy is a less optimum ligand for Ru(II) because of its rather strained bite angle (i.e. the N-Ru-N bond angle). This structural feature results in a lower ligand field splitting energy (the energy difference between the metal-centred eg and t2g orbitals) and hence a smaller gap between the upper metal-centred (MC) level and the MLCT level (πL*; in Ru pyridyl compounds this level is below the eg set, not as shown in Figure 11.4). As a result, a convenient radiationless decay path is available via an activated crossing to an upper lying 3 MC state. At 77 K, this decay pathway is frozen and [Ru(tpy) 2] 2⫹ exhibits similar luminescence to the bpy and phen analogues. Unfortunately, despite its poor luminescent properties, tpy is a desirable ligand from a structural point of view in the design of supramolecular photochemical devices. Metal bis(tpy) complexes are achiral, which means that they do not give rise to undesirable mixtures of diastereoisomers when more than one metal centre is present, and as a bridging ligand, substituted tpy compounds are easier to make into rigid linear connectors than bpy or phen derivatives. Furthermore, substituted tpy compounds do not exhibit fac/mer isomerism as do their bidentate counterparts (Figure 11.9).  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 718 Molecular Devices enantiomers geometrical isomers X X X X X X mer fac one enantiomer N N one isomer X X N X N tpy-based linear bridging ligand N N Figure 11.9 Structural isomerism in homoleptic (same ligand) bpy and tpy complexes. In order to take advantage of the simplicity of tpy chemistry, the ligand may be modified in two ways. Firstly, addition of electron acceptors to the 4′ position (X) increases the luminescence lifetime markedly (36 ns in the case of SO2Me). On the other hand, sophisticated substitution of bpy- and phen-based units can also give linear bridging ligands, as in 11.4. N N N N 11.4 11.2.5 Bipyridyl-Type Light Harvesting Devices Balzani, V., Juris, A., Venturi, M., Campagna, S. and Serroni, S., ‘Luminescent and redox-active polynuclear transition metal complexes’, Chem. Rev. 1996, 96, 759–833. Bpy-type ligands (including phen, tpy and a vast range of substituted derivatives) may be assembled with a number of different metal centres to give versatile photochemical devices via the complexes as metals/complexes as ligands strategy (Scheme 11.1). If we envisage the synthesis of a mononuclear + N N N N N N N N N deprotection 2+ Me + 2+ 2+ 2+ N Ru N N protected Os N Ru N N Os N N N chelating N N site Cl Cl N N N N labile ligands Scheme 11.1 The complexes as metals/complexes as ligands strategy – a ruthenium(II) complex acts as a ligand for an osmium(II) complex.  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 Supramolecular Photochemistry 719 metal complex as proceeding via the interaction of a metal ion with n ligands, then the assembly of multinuclear complexes may be carried out in an analogous fashion by replacing one or both of the metal (M) or ligands (L) with metal complexes with available binding sites. The metal may be replaced by mono- or oligonuclear complexes with easily replaceable ligands such as [Ru(bpy)2Cl2] (11.5), while the ligands may be replaced by mono- or oligonuclear complexes with uncoordinated lone pairs, as in (11.6). It is generally straightforward to prepare such unsaturated build- ing blocks by simple control of reaction stoichiometry, since the complexes are relatively inert, or, if necessary, by protection via N-methylation of one of the nitrogen atoms. In this way, remarkable complexes have been built up, such as the 22-nuclear [Ru{(µ-2,3-dpp)[Ru(µ- 2,3-dpp)Ru{(µ-2,3-dpp)Ru(bpy)2}2] 2}3] 44⫹ (11.7). Compound 11.7 is an excellent example of the fact that metal complexes need to share the same bridging ligand in order to interact, thus the 12 peripheral Ru2⫹ centres, which are not connected to one another in such a way as to allow electronic interaction between them, are all oxidised at the same potential in one single 12-electron process. A typical class II mixed-valence compound based on the rigid bis(bpy) bridging ligand 11.8 has been synthesised by Vincenzo Balzani and co-workers from the University of Bologna, Italy.6 This species may be prepared as the Ru II-RuIII, OsII-OsIII and RuII-OsIII derivatives by a stepwise approach, as outlined above, starting with the less labile Os complexes in the mixed metal case. In each case, irradiation results in excitation of the M II centre, which acts as an electron donor. The MII centre is excited into a 3MLCT state of high energy, which then transfers its electron to the MIII centre, resulting in the formation of an electronic isomer of the starting compound (Figure 11.10). In the case of the Ru2 and Os2 complexes of type 11.8a and 11.8b, the product is indistinguishable from the starting material, whereas the product from the mixed-metal 11.8c is of type RuIII-OsII, which represents an excited state of the Ru II-OsIII ground state. The complex decays back to the ground state by thermal back-electron transfer. In each case, nuclear rearrangement is negligible because there is no change in the occupancy of the antibonding eg set of orbitals, which strongly affects M–L distances. The rate constant for the relaxation of the mixed metal system is appreciably slower than that any of the other processes, consistent with the lower driving force for  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 720 Molecular Devices *RuII-L-OsIII *RuII-L-RuIII N eV 3+ M(bpy)2 2.0 11.8.M2+M3+(bpy)4 N *OsII-L-OsIII a M2+ = M3+ = Ru b M2+ = M3+ = Os c M2+ = Ru, M3+ = Os 1.0 N 2+ M(bpy)2 RuIII-L-OsII N 0 RuII-L-OsIII RuII-L-RuIII RuIII-L-RuII OsII-L-OsIII OsIII-L-OsII Figure 11.10 Energy level diagrams for photoinduced electron transfer processed based on 11.8, showing excitation ( ), luminescence (- - - - ) and nonradiative decay (----). the reaction. As supramolecular devices, these compounds represent the initial steps into charge separation and electronic switching. One of the ultimate goals in the study of photoinduced electron transfer in polymetallic species of this type is the design of catalytic systems that can harness solar energy in order to carry out chemi- cal reactions that are ‘up hill’ in energetic terms. The vast majority of such reactions (e.g. splitting of water into H2 and O2) are multi-electron processes. In natural photosynthesis, sophisticated antenna groups carry excitation to the reaction centre, which produces multiple high-energy electrons in order to bring about both the synthesis of O2 from water and, more importantly, synthesis of reduced organic molecules from CO2. Work in supramolecular photochemistry has therefore been directed to systems capable of multi-electron transfer, storage and catalysis. A schematic diagram of a simple example of such a process is shown in Figure 11.11. The two photosensitisers (PS) absorb light and transfer the resulting excited-state electrons to a central electron storage and reaction centre (SR). The SR com- ponent is capable of receiving two electrons, one from each PS group, and can therefore carry out a two-electron reduction upon multiple substrates to give a covalently bound product (2 Sub⫹ → Sub2). The resulting positive holes in the PS groups are filled by external sacrificial electron donors that are not, in themselves, sufficiently reducing to reduce Sub⫹. Figure 11.11 (a) Schematic representation of a photochemical molecular device for photoinduced electron collection and multielectron catalysis. D ⫽ donor, PS ⫽ photosensitiser, SR ⫽ electron stor- age and reaction centre, sub ⫽ substrate. (b) a real prototype, compound 11.9.7  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 Supramolecular Photochemistry 721 Figure 11.12 Schematic representation of a supramolecular light-harvesting antenna device. An example of a photochemical device capable of performing this kind of chemistry is compound 11.9.7 The PS groups are Ru(II) centres and the SR is an Ir(III) species. The complex as a whole is a class II mixed-valence compound. Photoabsorption and transmission by 11.9 is not very efficient, but it is able to catalyse the electrochemical reduction of CO2 to formate via photoexcitation of the Ru(II) moieties and double electron transfer to the Ir(III) centre, transiently reducing it to Ir(I). The sacrificial donors are amines, such as dimethylaniline. The Ir(I) centre then reduces CO2 to HCO2H in the presence of H⫹. Extending this concept further, it should be possible to design large arrays of antenna components that gather energy and transfer excited electrons long distances via multiple ‘hops’ in order to concentrate them at a particular reaction centre. Current work on such systems includes photon-harvesting polymers with pendant chromophores and multiple porphyrin arrays. Dendrimer-based systems derived from metal polypyridyls are also of interest. Using the complexes-as-metals/complexes-as-ligands strategy, polymetallic, cationic light harvesting molecules with dendritic type structure (for a discussion of dendrimer chemistry see Section 14.2) may be assembled reasonably easily. The 22-nuclear species 11.7 is an example of a third-generation dendrimer of this type and such species are excellent prototype supramolecular antenna devices capable of light-harvesting via an energy-transfer mechanism, analogous to biochemical photosynthetic mem- branes (Figure 11.12). Generally, dendritic multicentre metal complexes have been constructed based on Ru(II), Os(II) as well as Re(I) and other metal centres, using bridging ligands such as 2,3-dpp (as in 11.7) and 2,5-dpp (11.10), and terminal ligands such as bpy (11.1) and biq (2,2′-biquinoline, 11.11), Scheme 11.2. N N X N N X N N N N N C 2,5-dpp biq 11.12a X = H 11.10 11.11 11.13 11.12b X = F N C  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 722 Molecular Devices Scheme 11.2 Dendrimer construction via the complexes-as-metals/complexes-as-ligands strategy – dendrimer generation increases ascending the diagram but growth is terminated by the horizontal processes. The energy of the lowest MLCT state in each metal fragment (component) is readily predicted, and varies in a systematic fashion according to the identity of the metal and ligands. It is therefore possible to design electron- or energy-transfer cascade dendrimers by varying systematically the energy of the components’ absorption from one generation to the next. Components with the lowest- energy MLCT bands (e.g. Os(µ-2,5-dpp)32⫹) are placed at the dendrimer core, and those with the highest energy absorptions terminate the dendrimer surface (e.g. Ru(biq) 2 (µ-2,3-dpp) 2⫹), resulting in controlled energy flow. Similarly, appropriate choice of fragment ordering can result in almost any combination of energy-transfer routes. All of the energy migration patterns for tetranuclear compounds shown in Figure 11.13 have been observed. Analogous systems containing six and ten components exhibit related, extended behaviour. In extending these concepts to very large dendritic systems, it is likely that a wider range of metal ions and ligands will become necessary in order to give a smoothly graduated energy-transfer cascade.  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 Supramolecular Photochemistry 723 = Ru(bpy)22+ = Os(bpy)22+ = Ru(biq)22+ Figure 11.13 Energy-transfer processes in tetranuclear (first-generation) complexes. Metal ions such as Rh(III), Ir(III), Pd(II), Pt(II) and Pt(IV) may be used, although pyridyl-type ligands are less appropriate in these cases since they do not provide the required orbital energy levels. In such cases, combinations of pyridyl and N—C⫺ carbanion ligands, such as deprotonated 2-phenylpyridine, are beginning to prove useful. F F F F + N Ir N N N F N N N + + Ir N Ir N N N N direction of energy transfer F F F N N N + 2+ 11.14 N Ir N N Ru N N N N F N N N N + + Ir N Ir N F N N F F N N F + N Ir N F F F One example is the octanuclear iridium(III) compound 11.14 which uses the C,N-carbanion ligands 11.12 and 11.13.8 The periphery of the compound comprises fluorinated 2-phenylpyridine (11.12b) iridium units, linked via bridging ligand 11.13 to similar non-fluorinated iridium(III)  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 724 Molecular Devices 2-phenylpyridine centres. This iridium-based antenna array terminates in a ruthenium(II) tris(bpy) derived unit. All of the individual ruthenium and iridium based components absorb light and hence the absorption spectrum of the complex is roughly a sum of the absorption spectra of the individual fragments, which results in efficient light harvesting. However, in the emission spectrum, the maximum is observed at 630 nm which is a wavelength characteristic only of the ruthenium(II) component. What is happening is vectorial energy transfer from the periphery of the complex to the ruthenium(II) terminus. The synthesis of 11.14 is an interesting variant on the complexes as metals/complexes as ligands strategy in which palladium-catalysed Suzuki coupling is used to give the bridging ligand 11.13 from precursors that are already complexed to the metal centres. Finally a particularly interesting strategy for creating long-lived charge-separated states as models for photosynthetic systems, in optoelectronics or as probes for DNA is the incorporation of more than one redox-active ligand within a ruthenium tris(bipy) complex. In the example shown in Figure 11.14 photoexcitation of one of the tetrathiafulvalene-derived dipyridophen- azine (TTF-dppz) ligands results in electron transfer first to the ruthenium and then onto one of the other TTF-dppz substituents resulting in a charge-separated state TTF-dppz⫺-Ru 2⫹ - dppz-TTF⫹ with a remarkably long lifetime of 2.3 µs, four orders of magnitude longer than the excited state of the ligand itself, because of the slow back-electron transfer from one ligand to another. 9 S S S S S S N N hν N S S S N N 2+ N Ru S S S N N N N e- N N e- S S S S S S Figure 11.14 A ruthenium(II) complex of an annulated donor–acceptor (D–A) ensemble, 4′,5′-bis (propylthio)tetrathiafulvenyl[i]dipyrido-[3,2-a:2′,3′-c]phenazine (TTF-dppz) exhibiting a long-lived charge-separated state TTF-dppz⫺-Ru2⫹-dppz-TTF⫹ on photoexcitation.9  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 Supramolecular Photochemistry 725 11.2.6 Light-Conversion Devices Ward, M. D., ‘Transition-metal sensitized near-infrared luminescence from lanthanides in d-f heteronuclear arrays’, Coord. Chem. Rev. 2007, 251, 1663–1677. Taking advantage of the phenomenon of luminescence, it should be possible to design a photochemical device that is capable of absorbing light of one wavelength and re-emitting it at another. This simple task is performed by any luminescent species. In terms of tuneability of the wavelength of the emis- sion and gathering the maximum amount of incident light, however, a supramolecular device offers the possibility of a modular approach. If we break the light conversion process down into a combination of absorbance (A), energy transfer (ET) and emittance (E) steps, then the functionality of the device must be of the A–ET–E type. An optimum device may be designed if the light absorber or antenna, light emitter and the degree of coupling between them may be tuned separately. The combination of coupled absorber and emitter thus produces a light-conversion device with optimal properties such as maximum light collection and emission at the desired wavelength. In component terms, the bpy moiety offers significant scope as a light harvester, as we have seen earlier. The basic bpy motif is highly tuneable, and antenna (absorber) components might be produced from a range of biaryl and heterobiaryl moieties of the bpy type. Luminescent emitters might be metal ions such as the lanthanides Eu(III) and Tb(III), which offer an extensive manifold of f-electron states. The coupling between the A and E components could then occur via a metal–ligand interaction. This concept has been realised by the construction of lanthanide cryptates such as 11.15 and 11.16. Ordinarily, Eu3⫹ is not luminescent in aqueous solution because excitation is quenched by solvent molecules without emission. Complexation by the tris(bipyridyl) cryptand ligand, however, protects the metal ion from the solvent, while simultaneously providing light-harvesting bpy units. The high charge of the Eu3⫹ cation, coupled with the macrobicyclic stabilisation offered by the ligand, means that the complexes are highly stable. The complexes are brightly luminescent, with 11.16 especially exhibiting very efficient energy conversion of the order of about 60 % of the incident light. The pres- ence of the N-oxide substituents stabilises the complex because of the hardness of the ligand, which matches that of the lanthanide ion, and increases ET coupling.10 A heterobimetallic d-f block metal strategy has also been used in the conversion of visible to near-infrared light in complexes such as 11.17 which can be readily prepared by the complexes- as-metals-complexes-as-ligands strategy. Excitation of the platinum(II) moiety at 520 nm in a variety of lanthanide derivatives of 11.17 results in each case in the characteristic lanthanide near- infrared emission.11 These kinds of complexes are of potential in biological imaging because human tissue is relatively transparent at longer wavelengths (the light of an electric torch shone through  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 726 Molecular Devices your hand appears red because the short wavelength blue light is absorbed) and in telecommunica- tions since the emission wavelengths of particularly Pr(III) (ca. 1330 nm) and Er(III) (ca. 1550 nm) closely match the ‘windows of transparency’ in the silica waveguides used for fibre-optic data transmission. 11.2.7 Non-Covalently Bonded Systems Ward, M. D. ‘Photoinduced electron and energy transfer in noncovalently bonded supramolecular assemblies’, Chem. Soc. Rev., 1997, 26, 365–375. In the systems we have examined so far, coupled photoexcitation and electron transfer (eT) or energy transfer (ET) processes have occurred in strongly bonded systems, held together by covalent bonds or coordinate interactions with a significant degree of covalency. In these cases there is a well-defined spatial relationship between the light-absorbing group and the external donor or acceptor (termed a quencher, since it quenches the luminescent re-emission of the absorbed light by providing an alterna- tive deexcitation pathway). An excellent example of the modularity of such covalent systems is 11.18, a synthetic system that combines two of the key ingredients in natural photosynthesis: a metal porphyrin chromophore and a quinone quencher as an external electron acceptor. The two components are linked covalently by a long, π-conjugated bridge. Electron transfer along the bridge results in reduction of the quinone to a semiquinone and ultimately a hydroquinone. H11C5 C5H11 N electron transfer N Zn N N H11C5 O Chromophore 11.18 O Quencher It is possible to design analogous modular (and therefore supramolecular) systems in which the components are joined by entirely noncovalent interactions such as hydrogen bonding. In this way, the concepts of self-assembly and molecular device design may be combined in order to provide a route into the preparation of much more sophisticated device arrays. The key is to construct functional com- ponents bearing mutually complementary and interacting binding sites. Such an approach can only be effective if non-covalent interactions can form an effective conduit for the eT or ET process; in order to test this possibility, two porphyrin-based systems 11.19 and 11.20 have been designed. Both systems involve photoexcitation of the Zn(II) porphyrin component, followed by electron transfer to the Fe(III) moiety, resulting in reduction to Fe(II). This electron transfer must occur via a hydrogen-bonded link in 11.19 or a covalent σ-bond framework in 11.20. Remarkably, not only are both systems effective at carrying out this process, but the photoelectron transfer rate constants of 8.1 ⫻ 109 and 4.3 ⫻ 109 s⫺1 for 11.19 and 11.20, respectively, suggest that the familiar carboxylic acid double hydrogen bond is more effective at transmitting the excited electron than the σ-framework in 11.20.  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 Supramolecular Photochemistry 727 R R R = 3,4,5-C6H2(OMe)3 N N Cl O H O N N R Zn Fe R N N O H O N N R 11.19 R Ph Ph R = OMe N N Cl N N R Zn Fe R N N N N Ph 11.20 Ph Consistent with the principles outlined in Chapter 1, strong hydrogen-bonded association may be achieved by multiple complementary interactions, as in the barbiturate-appended porphyrin system shown in 11.21 (cf. molecular rosettes, Section 10.6.4). This receptor binds a complementary dansyl (dimethylaminonapthalene-sulfonyl) group with an association constant in excess of 106 M⫺1. In this case, electron transfer is from the photoabsorbing dansyl group to the porphyrin acceptor; quenching is almost total, suggesting very fast electron transfer relative to the rate of fluorescence.12 O OEt Pr N Et Et H N O O N O H N N N H O S H Bu R Zn N O N N N N N H H N H O O N Et Et N O H 11.21 N Pr O OEt A remarkable example of a photon acceptor is 11.22, which, instead of possessing a molecular chromophore, relies upon a nanoparticle of TiO2 semiconductor about 2.2 nm in diameter.13 The semiconductor particle is held in place by absorption of the alkyl chains of the diamidopyridine-based receptor to its surface. The nanocrystallite of TiO2 has a small energy gap (band gap) between its valence electron states and its conduction band of empty orbitals. Photoexcitation at 355 nm can pro- mote an electron from the valence to the conduction band, and the excited electron can then undergo electron transfer across the hydrogen-bonded connector to the viologen (4,4′-bipyridinium) acceptor. Control experiments show that the presence of the hydrogen bond is essential for electron transfer to take place, and without it through-space or collision-activated transfer does not occur.  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 728 Molecular Devices O - N H O hν TiO2 N H N N N N H O H + + O N N + O electron transfer 11.22 Even π-stacking interactions can be used to mediate electron transfer between components in photo- chemical devices. Compound 11.23 uses the intrinsic curvature of the glycoluril building block (Section 6.2.4) to position a quinone acceptor some 9 Å above a porphyrin chromophore in a U-shaped arrange- ment. In CCl4 solution, the inner surface of the ‘U’ is filled with solvent molecules and the electron must traverse the very long through-bond pathway from chromophore to acceptor, resulting in very little elec- tron-transfer quenching. In contrast, addition of an aromatic guest molecule such as hexyl-3,5-dihydroxy- benzoate results in guest binding within the U-shaped cavity of the molecule, which acts as a solution host for aromatic guests. The guest is bound by a combination of π – π stacking interactions and hydrogen bonds to the amide carbonyl groups at the bottom of the molecular cleft. Electron transfer can then be mediated via the π-system of the guest and, as a result, 75 % of the luminescence of the porphyrin frag- ment is quenched by electron transfer via the aromatic guest to the quinone acceptor (Figure 11.15). This kind of electron transfer via a π-stacked system is reminiscent of the electron-transfer processes within the DNA π-stacked double helix. A particularly novel kind of two-component energy transfer device is the pH switched ‘plug and socket’ system shown in Scheme 11.3. When unbound there is no interaction between the binapthyl unit attached to the crown ether and the anthracene-appended secondary amine. However, proton- ation of the amine causes its complexation by the crown ether (as in Section 3.12.1) bringing the two chromophores into close proximity just like putting a plug into a socket. Energy transfer can then occur from the binapthyl group to the anthracenyl unit and hence irradiation of the binapthalene results in fluorescence of the anthracene-derived chromophore. If the methyl substituent is replaced by a bulky group such as phenyl then no complexation occurs because the plug does not fit the socket and hence no anthracenene fluorescence is observed on binapthalene excitation.14 This concept has been further extended to give a three-molecule system in which if the two terminal molecules are a Figure 11.15 Photochemical electron transfer mediated by π-stacking interactions with an aromatic guest.  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 Supramolecular Photochemistry 729 hν ET O Me O O O N O H H+ plug in H O + O Me + O N O n-Bu3N plug out hν′ O H O O O O 11.24 Scheme 11.3 A supramolecular ‘plug-and-socket’ system.14 O O O O O + N H hν hν H O O O O extension cable O N O O N + N plug N 2+ O Ru N O O N N + e- O N O socket O O O 11.25 Figure 11.16 A molecular extension cable as part of the ‘plug and socket’ concept.15 plug and socket, a central component is a molecular analogue of an ‘extension cable’. The system is shown in Figure 11.16 with ammonium ion binding by the crown ether portion of the ruthenium(II) tris(bipyridyl) appended chromophore linking in a receptor site for a bipyridinium electron acceptor. Thus photoexcitation of the ruthenium-containing ‘socket’ results in electron transfer across the molecular extension cable to the bipyridinium ‘plug’.15 Because of the sensitivity of fluorescence detection this kind of FRET process (Section 11.2.2) has also been used to study self-assembly mechanisms at extremely low concentrations, e.g. of resor- carene hexameric capsules. We have already seen the self-assembly of resorcarene hexamers in Section 10.6.3. This kind of self-assembly is generally studied by NMR spectroscopy in the millimolar range. Appending fluorophores to the resorcarene capsules enables much lower concentrations to be addressed. Rebek Jr. and co-workers (Scripps, USA) have tagged resorcarenes with two different chromophores, namely pyrenyl (donor) and perylenyl (acceptor) groups. Mixing the two pure hexamers at nanomolar concentration results initially in observation only of the fluorescence of the donor. Over time, however, the self-assembled hexamers disproportionate to give mixed hexamers in which donor and acceptor are both present in the same assembly, resulting in the growth of acceptor emission by a FRET process,  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 730 Molecular Devices N+ 11.26 Figure 11.17 Model showing the FRET process between pyrenyl donor and perylenyl acceptor in hexameric resorcarene capsules containing benzene guests (reproduced from Reference 16) along with the structure of the fluorescent guest 11.26. Figure 11.17. The emission can be used to monitor the kinetics of capsule assembly showing it to be relatively fast under these dilute conditions (capsule half life ca. 10 mins depending on solvent). Addition of methanol to the chloroform solution to disrupt the hexamer assembly also results in loss of the FRET emission. FRET was also observed from the outside to the inside of the cavity using a fluorescent guest 11.26.16 11.3 Information and Signals: Semiochemistry and Sensing 11.3.1 Supramolecular Semiochemistry Fabbrizzi, L., Poggi, A., ‘Sensors and switches from supramolecular chemistry’, Chem. Soc. Rev. 1995, 197–202. Semiochemistry is the name given to the area of supramolecular chemistry that is concerned with what might be termed, broadly, signalling devices. The term comes from ‘semiotics,’ meaning the study of signs or symbols and their use or interpretation. Thus, in biochemistry, a semiochemical is a substance such as a pheromone that conveys a signal from one organism to another. We looked briefly at biologi- cal semiochemistry, a process fundamentally based in molecular recognition, in Section 2.8. The most obvious area of application for semiochemistry is in the production of molecular sensor devices: com- pounds that can carry out the twin tasks of both molecular recognition and sensing. The basic concept of a molecular sensor is illustrated in Figure 11.18. The substrate (guest, analyte) must be attracted to a receptor portion of the sensor. This is a straightforward molecular recognition event, and the receptor can be any of the systems, such as crown ethers, cryptands, cavitands and so on, discussed in Chapters 3–6. The binding must be selective for the target substrate in the presence of a range of other potential guest species, depending on the system and its environment. The mere act of binding, however, is not enough. The receptor must also be in communication with a signalling unit that is responsive to the guest binding, which generates a signal in the form of an emission of electromagnetic radiation (photo- chemical sensing), a current (electrochemical sensing) or an otherwise externally measurable change (e.g. in colour or pH). This transduction implies automatically that the spacer that joins the signalling and receptor units together must permit communication between the two of them, and that the binding event must trigger intrinsically a change in the properties of the bound complex compared to the free  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 Information and Signals: Semiochemistry and Sensing 731 Substrate Spacer Signalling unit Receptor Figure 11.18 Cartoon representation of a chemical sensor. guest or receptor, which results in signal generation. Criteria for the construction of useful sensors may be summarised as: stability guest (analyte) selectivity guest affinity efficient signal transduction emission of detectable intensities of UV or visible radiation or other quantifiable signal kinetically rapid sensitisation ease of delivery to the target system availability. One of the potential applications of these kinds of hosts is in analytical chemistry, in which the host is used to bind and recognise small quantities of analytes. If analyte concentration is particularly low, then high affinities are necessary in order to produce a measurable concentration of the complex and hence give a response. Unfortunately, high affinity and high selectivity are often antagonistic. Hosts that exhibit high binding constants often bind well to a wide variety of guest species, whereas hosts that are highly selective often exhibit low binding constants. This behaviour is frequently due to the fact that selectivity often arises from a strong, non-selective host–guest affinity (e.g. solvophobic binding or electrostatic attraction) coupled with a weaker repulsive interaction (e.g. unfavourable steric effects), which varies in magnitude according to the precise nature or geometry of the guest. In its broadest sense, however, semiochemistry is more than just sensing. It also incorporates aspects of signal generation, processing, transfer, amplification, conversion and detection. In other words, all of the kinds of concepts associated with modern computers and electronic devices. The construction of molecu- lar electronic devices is discussed in detail in Section 11.4. For the moment, we will concern ourselves with signal generation in chemical sensors. In other words how a supramolecular host-guest binding and recognition event can be transduced into a measurable output, be it either a change in colour and hence absorbance, a change in the intensity or wavelength of emitted light or a change in redox potential. 11.3.2 Photophysical Sensing and Imaging Gunnlaugsson, T., Glynn, M., Tocci, G. M., Kruger, P. E., Pfeffer, F. M., ‘Anion recognition and sensing in organic and aqueous media using luminescent and colorimetric sensors’, Coord. Chem. Rev. 2006, 250, 3094–3117. When we come to the specific design of artificial molecules capable of signalling the presence of par- ticular analytes to a human observer, it is not sufficient to have only molecular recognition, but also transduction of that recognition event in a process that ultimately leads to an observable signal. A par- ticularly sensitive method of signal production is the emission of visible light; either visible to the human  10.1002/9780470740880.ch11, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/9780470740880.ch11 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 732 Molecular Devices hνfluo hνfluo X hνabs hνabs O O O O N O N O O O electron transfer O O K+ 11.27 Scheme 11.4 Cation sensing by anthracene-substituted azacorand 11.27.19 eye, or more commonly to a sensitive detection device such as a fluorescence spectrometer. Fluorescence detection is particularly sensitive an