Photochemical Materials: Absorbers, Emitters, Displays, Sensitisers, Acceptors, Traps and Photochromics

Abstract

In this chapter we discuss some of the typical materials used in photochemistry. We describe, in general terms, how their suitability for application as absorber, emitter, sensitiser, energy acceptor or quencher, depends on the energy states within the material and the routes of interconversion between these states, and also how suitability as a redox or chemical sensitiser/acceptor/trap is determined by specific chemical reactivities. We describe the application of photochemical principles to the design of light sources and displays, and describe the photochemical principles and applications of photochromics and molecular switches. A table giving the structures, characteristics, and uses, of a number of compounds widely used in photochemistry is provided at the end of the chapter.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

4.1 Introduction

Whether a material will act as a passive absorber, an emitter, or sensitiser, depends, for the most part, on how excitation energy is deactivated in that material. If deactivation is via a fast non-radiative process, the material will act as a passive absorber. If deactivation goes via an emissive excited-state, then the material may be a useful emitter. If deactivation proceeds via a relatively long lived excited-state, then the excited state may be useful as an energy transfer sensitiser, since the long lifetime may allow transfer of the excited-state energy to another species. If deactivation goes via a chemical reaction which leads to products of interest e.g. singlet oxygen, radicals, or redox active species (i.e. strong oxidants or reductants), then the material may be a useful photochemical sensitiser for these species.

Whether a material is useful as an excited-state acceptor, depends on the energy level of excited states within the material and the way those states deactivate once populated. A useful redox or chemical acceptor/trap will show efficient and specific reactions: redox traps will undergo a specific reduction or oxidation; singlet oxygen acceptors have specific reactions with singlet oxygen; and radical quenchers or traps have specific reactions with radicals.

In the following discussion, a number in bold indicates an entry for that compound in the table of commonly used compounds, their uses and properties, Table 4.1, found at the end of the chapter.

4.2 Passive Absorbers

The major uses of passive absorbers are as colorants and sunscreens, although in passive solar heaters the heat generated during excited-state deactivation is the important photo-product. Dyes and pigments are used as passive absorbers in established technologies such as: paints, plastics, textiles, paper, printing, imaging, and foodstuffs. In sunscreen formulations, UV absorption rather than visible absorption is required. High technology applications of passive absorption include: biochemical “stains” where not only colour intensity, but also selective binding to particular biochemical substrates, is required; recording dyes for optical storage in CDs and DVDs; and imaging and digital printing where a combination of the need for accurate colour reproduction, chemical and photochemical stability, and compatibility with printing processes, requires highly specialised dye design [1–3]. In the space available here we can give only a very brief account of what is commonly termed colour chemistry, but the interested reader is directed to references [1–3] which between them provide an excellent introduction and overview of this very important aspect of applied photochemistry.

Important characteristics for colorants are: intensity, brightness and stability. Relative costs mean that only colorants with high intensity absorptions are commonly used. This restricts the types of transition involved to: molecular charge transfer bands, π–π* bands, n–π* bands with a high degree of π–π* coupling, or direct bandgap semiconductor transitions. The brightness of a colorant depends primarily on absorption band width; narrow bands give bright colours. The absorption band width is influenced by the width of ground and excited state potential energy curves, how similar the molecular geometries in ground and excited state are, and also the presence of close or overlapping transitions. Photostability/lightfastness is very important for all colorants; poor lightfastness causes loss of colour in imaging and printing, and leads to conservation problems for artwork and artefacts. Stability towards thermal degradation is important in colouring plastics which may be moulded at high temperatures, in colorants for outside structures, metalwork and cars, as well as for sublimable dyes. Stability towards hydrolysis is important in imaging, printing and artist’s materials. Washfastness is a particularly important feature for textile dyes.

Colorants are used as dyes or pigments. Dyes are colouring materials which are soluble either in the medium in which they are incorporated, or, as the term is commonly used in textiles technology, in a dyeing solution applied to the medium in which they are incorporated. (Disperse ‘dyes’ are applied as a fine dispersion to synthetic textiles, but dissolve in the textile to give a ‘solid solution’). Pigments are colouring materials which are insoluble in the medium in which they are incorporated. The principle applications of pigments are in paints, printing inks and plastics, although they are also used to colour cement, ceramics, concrete, cosmetics, glass, paper and rubber. Usually the application of pigments involves their incorporation into a liquid medium, a wet paint, or a molten thermoplastic material, by a dispersion process in which the pigment aggregates are broken down into very small primary particles or aggregates. When the medium solidifies, the individual pigment particles become fixed in the solid polymeric matrix. Apart from imparting colour, pigments also provide opacity by scattering light. The size and size distribution of the pigment particles is important in terms of both colour and opacity; particle sizes of ~0.2–0.3 μm are often used since these provide maximum opacity.

4.2.1 Inorganic Colorants

Most inorganic colouring materials are pigments. Surface treatments are often applied to inorganic pigments, e.g. coating with surfactants may improve ease of dispersion, while coating with inert inorganic oxides, such as silica, can give improved lightfastness and chemical stability. Among the most important inorganic pigments are titanium dioxide (white); carbon black; metal oxides e.g. iron and manganese (yellow, brown, red, black) and chromium (green); cadmium/zinc sulfides/selenides (yellow/orange/red); cobalt aluminate (cobalt blue); ultramarine blue; and Prussian blue. The origin of colour in these materials is varied.

TiO₂ (9.1) and Cd-Zn/S-Se (9.3–9.5) are semiconductors. TiO₂ has a band gap of 3.0–3.2 eV, i.e. it absorbs in the UV region. It is used as a white pigment because it has a very high refractive index, and therefore a high scattering efficiency and excellent opacity for all visible wavelengths. A surface coating is generally required to prevent photoreactions on the TiO₂ surface from damaging the dispersion medium, a phenomenon known in the paint industry as chalking. With a band gap of 1.6 eV CdSe absorbs all visible photons and therefore appears black; at 2.6 eV CdS absorbs blue photons and therefore appears yellow. Changing the S/Se ratio shifts the band gap from 2.6 to 1.6 eV, taking the colour through the yellow-orange-red–black range. Replacement of Cd by Zn gives greenish-yellow tones.

In many oxide pigments the colour is due to ligand–metal charge transfer (LMCT) transitions (see Chap. 3). Natural iron oxide based pigments include: yellow ochre, red haematite, and the browns, sienna and burnt sienna; umber and burnt umber are iron oxide with manganese dioxide. Synthetic red iron oxides are anhydrous Fe₂O₃, while synthetic yellow pigments are iron(III) oxide/hydroxides, FeO(OH), and black pigments are non-stoichiometric Fe(II)/Fe(III) oxides. They have excellent durability, high opacity, low cost, and low toxicity. Cr(III) oxide, which gives a dull green pigment with outstanding durability, is another important oxide pigment. Lead chromate pigments, which have now been almost completely replaced because of their toxicity, are important historically. Pure PbCrO₄ gives a rich yellow pigment, the colour of which originates from a charge transfer transition on the chromate CrO₄ ²⁻ ion; the role of lead is to make a highly insoluble pigment. Lemon shades are obtained by the addition of lead sulfate, while addition of molybdate gives orange red tones.

Cobalt blue is a cobalt aluminate, CoAl₂O₄, with a spinel crystal structure in which Co atoms sit in a tetrahedral environment (the same coordination geometry which gives self-indicating silica gel a blue colour when dry—the pink colour when wet is due to cobalt in octahedral coordination). Here the colour originates from metal-centred d–d transitions on Co(II). A blue/green analogue, CoCr₂O₄, has Co in tetrahedral sites and Cr in octahedral sites in the crystal structure.

Ultramarines are complex sodium aluminosilicate structures, containing trapped sulfur anions, S²⁻ and S³⁻, which absorb in the red spectral region, e.g. the S³⁻ anion has an absorption maximum, λ_max, at ~600 nm [4]. Originally from the mineral lapis lazuli (blue stone), brought to Europe from Afghanistan, where it is still mined today, and described as ultramarine (beyond the sea). It was very expensive, partly because of the source, but also due to difficulty in preparation, and in Western religious art it was often reserved for the mantle of the Virgin Mary [5]. Today, it is made synthetically as French ultramarine, after a synthetic route was discovered by Jean Baptiste Guimet in 1826.

Prussian blue is a mixed Fe(II)/Fe(III) complex polymeric species in which Fe(II) is octahedrally coordinated by C, and Fe(III) is octahedrally coordinated by N, to give a structure containing Fe(II)-C–N-Fe(III)-N–C-Fe(II)-linkages, in which the colour originates from electron transfer between the two metal oxidation states. It was discovered in 1704, and used in ‘blueprints’ and also in the cyanotype photographic process developed by Herschel (see Chap. 11). The cyanotype process is made possible by the photochemical reduction of Fe(III) citrate (or oxalate) to Fe(II), which reacts with ferricyanide present in the coating formulation to give Prussian blue. A similar photoreduction of Fe(III) oxalate to Fe(II) is used in the ferrioxalate actinometer (see Chap. 14).

4.2.2 Organic Colouring Materials

Here, pigments and dyes are both important. There is an enormous range of organic dyes available, and a number of classification methods have been used, but that based on the electronic nature of the transition is most relevant here [1].

Donor–acceptor colorants. These are by far the most important group of organic colorants, they include azo-, anthraquinone- and carbonyl-based dyes. Colour originates from transitions in which there is a significant shift in electron density from the donor to the acceptor parts of the molecule. Azo dyes account for over 50 % of all commercial dyes. There is a vast range available. They can contain multiple azo, (–N=N–), groups, but monoazo dyes are the most important (e.g. methyl orange 11.13). These contain an electron donating group (often hydroxy or amine) and an electron accepting (often aryl) group on either side of the azo bond. The electronic transition occurs across the azo bridge from the electron donor to the electron acceptor group. These dyes cover the whole spectral range but yellows, oranges and reds are most important. Synthesis is relatively straightforward from cheap starting materials, so they are very cost effective. Azo dyes can also exhibit cis–trans photoisomerisation across the azo bond, which results in photochromism (see later), although this is not common in modern dyes. Addition of a hydroxy group adjacent to the azo bond generally improves light stability, since proton transfer between O and N atoms in the excited state can act as a rapid mechanism for loss of excitation energy. Metal complex azo dyes are also very common, with copper, cobalt and chromium being common metal ions used. The metal ion leads to improved lightfastness. There are probably a number of factors at play in this, e.g. reduction of electron density at the chromophore and therefore a reduction in ease of photooxidation, excited-state deactivation by low lying d–d levels, and steric protection of the chromophore. Metal dyes also show improved washfastness because of the larger size of the metal complex, and stronger fibre interactions. However metal complexes are generally duller colours than the parent azo due to their broader absorption bands, which result from additional overlapping d–d transitions and charge-transfer bands, and sometimes the presence of isomers with slightly different absorption spectra.

Anthraquinone dyes are the second most important group of organic colorants. The basic structure is 9,10-anthraquinone (11.3) but with electron donor groups in 1, 4, 5, and 8 positions. The absorption maximum can be shifted by choice of type and number of substituent. The historically important natural dye alizarin is 1,2-dihydroxyanthaquinone (11.4), originally obtained by extraction from the root of the madder plant. Hydrogen bonding from the carbonyl oxygen to an adjacent OH, or NH group is an important factor in light stability, since reversible proton transfer in the excited state can act as a rapid mechanism for loss of excitation energy (similar to putting a hydroxy group ortho to the azo group in azo dyes). It is possible to get the whole spectral range, but violets, blues and green are particularly important since these complement the yellow, oranges, and reds best obtained in azo dyes. Anthraquinone dyes have good brightness and fastness; however they cost more than azo dyes because the molar absorption coefficient is lower and they are more expensive to make. They are not used much as pigments because pigment blues and green are best provided by phthalocyanines.

Indigo and 6,6-dibromoindigo (tyrian purple) (11.11) are two of the oldest known dyes [6]. Natural indigo is extracted from plants, indigo (Indigofera tinctoria) in the Far-East or woad (Isatis tinctoria) in Europe; tyrian purple is extracted from shellfish. In both cases the dyestuff itself is not present in the natural source but is generated by fermentation and/or air oxidation. Nowadays, only indigo is important commercially, and it is almost exclusively synthetic. It is used mainly for denim (de-Nîmes, where the thick cotton cloth was originally made), where fading without change of shade only, at points of textile stress and wear is a key fashion feature.

Polyenes . These are systems of extended conjugation in which the molecular orbitals extend over a large part of the molecule and the π–π* transition does not result in a major shift in electron density from one part of the molecule to another. As the degree of conjugation increases, the spacing of electronic energy levels decreases, and λ_max shifts to longer wavelengths. Phthalocyanines (5.5, 5.6) are the most important synthetic commercial polyene colorants. Metallophthalocyanines (5.6) give brilliant intense blue/green colours of high stability. The intensity of colour is due to very high molar absorption coefficients, while the brilliance is due to very narrow absorption bands due to the rigidity of the molecular structure. They are most commonly used as pigments. They are too large to penetrate into many fibres, but can be used for polyester and as reactive dyes for cotton, where they are covalently bound onto the cotton surface. Naphthalocyanines have absorption maxima in the NIR which makes them interesting dyes for optical data storage and ‘invisible’ security printing. Carotenoids are important natural polyene colorants, and are used commercially in foodstuffs, the most important being β-carotene (11.5).

Cyanines, squaraines, rhodamines, xanthenes. Cyanines (4.6, 4.7) are similar to polyenes, but with remarkably low energy transitions for such small molecules. This is because they have an odd number of atoms carrying p orbitals in the conjugated chain (polyenes have an even number), and the terminal N, rather than C, atoms results in two extra electrons per chain compared to polyene dyes. The resulting MO structure has a relatively high energy HOMO with significant non-bonding character, as opposed to the bonding HOMO of polyenes, and there is, therefore, a relatively low energy HOMO → LUMO transition (see Chap. 1). These molecules give intense bright colours due to their high molar absorption coefficients and quite narrow absorption bands. The λ_max is dependent upon chain length, and dyes absorbing from the blue to IR spectral regions are available. ‘Hidden cyanines’, in which the cyanine structure is not so obvious, e.g. phenolphthalein, rhodamines (4.1, 4.2), squaraines (11.19, 11.20) and other dyes such as methylene blue, (11.12) nile blue (11.15) and triarymethane dyes (11.7), and xanthenes (fluoresceins) (4.3–4.5), which can be considered oxygen analogues of cyanines, also give intense bright colours. Rigid structures such as rhodamines and fluoresceins are often highly fluorescent, whereas ‘looser’ structures like phenolphthalein are not. Dyes spanning the full spectral range can be made. Although historically important dyestuffs, particularly the triarymethane dyes, cyanines and hidden cyanines are not so chemically and photochemically stable as other dyes and so are not used as absorptive colorants, except in specialist applications such as biological ‘stains’ for microscopy, and as colorimetric pH indicators. The highly fluorescent variants e.g. rhodamines, are widely used as laser dyes, and also as biological stains and fluorescent markers. They are also used as fluorescent colorants, e.g. in artist’s inks, where the resultant brilliant colours are a product of both absorption and emission, but in this application their poor light stability is a significant disadvantage.

4.2.3 Sunscreens

Sunscreens are UV absorbing materials suspended in a cream or spray. The two main types of UV absorbers used are: (1) colloidal semiconductor particulates, notably ZnO (9.2) and silica- or alumina-coated TiO₂ (9.1), and (2) organic absorbers. Key photochemical requirements are high UV absorption with no visible absorption, (although scattering of visible light in opaque white sunscreens is acceptable), and high photostability with a complete lack of any photoreactions which may lead to phototoxicity. Chemical requirements are low toxicity, and compatibility with cream formulations. Ideally, excitation energy is rapidly lost as heat, and typical mechanisms for this in organic sunscreens are proton transfer in hydroxyl benzophenone (11.6) based sunscreens and molecular twisting in cinnamates (11.17).

4.3 Emitters

There are many type of emitters: gas phase atoms and ions; molecules in the gas, solution, or solid phase; metal ions as an integral part of a solid state lattice; energy traps or ‘activator’ sites, which may be a specific ion or site defect, in a semiconductor or other solid; conjugated polymers; semiconductors as either a solid, or as a colloidal dispersion in some other medium or solution; and hot metals. In all but the latter three cases emission is from a localised emitter, with both states involved in the transition localised in a small region of space, on an atom, ion, or molecule. However, for semiconductors, some emissive polymers, and hot metals, the states involved in the transition extend across a relatively large region of space and a large number of atoms.

The characteristics of the light emitted are determined by the following.

• (1) The energy levels of the emitter in the medium used. These determine the emission wavelength(s), and shape of the emission band(s). For semiconductors, and some emissive polymers (3.1–3.9), emission is a property of the whole material, and control of the size of the semiconductor particle or chain length of the polymer may be important, e.g. in semiconductor quantum dots (9.3–9.7). For semiconductor materials on the macroscale the valence and conduction bands have band widths associated with the continuum of orbitals. However, on going from the macroscale to nanometre (nm) dimensions two effects occur due to the removal of atoms (and hence orbitals); firstly, the bands cease to be a continuum and individual orbitals, and hence quantised energy levels are observed (hence the term quantum dot); secondly, orbitals are removed from the edges of the valence and/or conduction bands, which increases the band gap. The size of the quantum dot dictates the absorption and emission characteristics; the smaller the quantum dot, the larger the band gap, and hence the more blue-shifted (shorter wavelength) the emission (Fig. 4.1) [7].

  In metallic solids, the presence of a continuum of states means that there are many transitions possible, and so when heated, as in an electric tungsten filament bulb, these materials emit a continuous spectrum which is very different to the line or band emission of atomic or molecular species. Other hot solids, and hot high pressure gases, often emit a mix of line or bands on top of a continuum (see for example the spectrum of high pressure Hg or Xe lamps in Chap. 14).

Fig. 4.1

Schematic representation of the effect of size on semiconductor properties, i.e. changes on going from the macro-scale (continuum of energy levels) to the nano-scale (quantised energy levels). The average energy position of the bands do not change (represented by the lines dissecting the relevant bands and orbitals) however, the band gap increases with decreasing size as extreme energy levels are removed. Figure adapted from Ref. [8]

• (2) The nature of the emissive transition. Typically, allowed transitions, such as fluorescence or direct bandgap semiconductor emission, have radiative lifetimes between a few and a hundred ns. Forbidden transitions, such as molecular phosphorescence, have radiative lifetimes longer than μs, usually much longer.

• (3) Competition with other deactivation routes for the excitation energy. The competition between emission and other deactivation routes determines the emission quantum yield. Competing deactivation processes are important for almost all emitters, even those for which emission is a highly probable process; there are few emitters with an emission quantum yield approaching 1. Competing deactivating processes are particularly important for forbidden transitions, where radiative rates are relatively slow. Most reasonably efficient room temperature phosphorescent compounds are heavy metal complexes where the degree of forbiddeness is reduced by heavy-atom spin–orbit coupling, and the radiative lifetime is in the μs–ms range. Most organic molecules, which do not have such coupling, show phosphorescence only in a low temperature glass, with radiative lifetimes in the ms to tens of seconds range.

• (4) The method of population of the emissive state. If population is achieved directly from the initial excitation source, then as soon as that source is removed the emission intensity will decrease with a rate determined by the deactivation processes of the emissive state. However if the excited-state is populated by processes which occur after the initial excitation process, e.g. energy transfer, energy migration, exciton migration, or delayed fluorescence, then the emission intensity will decrease at a rate determined by both deactivation and population processes. In some cases, where the rate of population of the emissive state can be made very slow, e.g. by energy or exciton migration within a solid, very long lived phosphors (9.3) are possible.

• (5) Population inversion . If the emissive state can be formed such that a population inversion is obtained it may be possible to obtain laser emission (e.g. Nd^III ions (10.5) in the Nd/YAG laser, see Chap. 14).

Emitters can be classified in terms of those which emit in the gas phase, the solution phase, or the solid state, and also by mode of excitation, i.e. electroluminescence, thermoluminescence, chemiluminescence, radioluminescence and photoluminescence. We are most concerned with photoluminescent materials, but thermoluminescence and electroluminescence, in both gas and solid phases, are important technologies.

4.3.1 Solid State Thermoluminescence

The first commercial electric lamps of Edison and Swann involved incandescence produced by passing an electric current through a carbon filament. Modern incandescent lights use the same principle, but with the fragile carbon thread replaced by tungsten. The incandescent lamp has the advantage of good spectral output. Like the Sun, the broadband emission (Fig. 4.2) is, to a good approximation, that of a black body radiator [9]. However, although the cost of the incandescent lamp is low, the efficiency of energy conversion is limited (5–10 %) since most of the electrical energy used is lost as heat, and the lifetime of the lamps is limited. Although modest increases in efficiency are possible, there is a global tendency to phase out incandescent lamps for domestic use in favour of more energy efficient forms of lighting.

Fig. 4.2

Emission spectra of: a 5000 K fluorescent lamp, b 2800 K tungsten filament lamp. Figure adapted from Ref. [10]

Other common incandescent emitters are the ‘glowbar’ used in IR spectroscopy which is also a good approximation to a black body emitter, and burning magnesium and the incandescent lanthanide oxides used in the mantles of gas lamps; for these emission is a combination of a broad band continuum with specific emission lines superimposed.

The measurement of thermally-activated luminescence from radiation-induced defects in minerals, particularly those in ceramics, is usually termed thermoluminescence when used in archeological dating, although the role of thermal excitation here is not the generation of emissive states but rather to allow defect relaxation, and concomitant emission, to occur. The general principle is as follows. When ceramics are fired, the high temperatures allow complete defect relaxation in the ceramic minerals, such that immediately after firing the thermoluminescence signal would be zero. Over time, natural radiation damage causes defect formation within the ceramic, and thus the intensity of thermoluminescence at any time after firing is related to the rate of defect formation and, the key archeological factor, the time since the firing.

4.3.2 Gas Phase Plasma Emission

All atoms and ions have some excited states which relax by emission in the UV/Vis spectral region. This forms the basis of atomic emission/fluorescence spectroscopy. Excitation may be: thermal in a flame; electrical in an arc; or by radio- or microwave-frequency radiation. This produces a plasma, which leads to production of excited states of the atoms and molecules of the gas. These then emit light. Apart from its use in analysis via atomic emission spectrometry and flame photometry, and the generation of colours in fireworks and flares, thermal excitation is not widely used to generate gas phase excited-states. (The yellow/white emission of candle/spirit lamp and low air flow Bunsen flames is due to incandescent carbon particles rather than specific atom or molecular emission—these generate the blue component of the light of the flame.) However, electrically excited plasmas in gas discharge lamps give us many of our light sources. These have been known since the nineteenth century. Important gas discharge lamps used in experimental photochemistry are: (i) low, medium and high pressure mercury lamps which are excellent sources of UV radiation; (ii) the Xe arc lamp, which is widely used as an intense continuous or μs pulsed UV/vis source; (iii) Xe and Kr flash lamps which give UV/Vis emission with ca. μs pulse duration, (also used as photographic flashlamps); (iv) gas phase lasers; (v) the nitrogen and oxygen pulse spark lamps used in single photon counting; and (vi) the deuterium lamp which is used as a UV source in UV/Vis spectrophotometers and other instruments (see Chap. 14 for further information on light sources).

Gas discharge lamps exhibit higher energy conversion efficiency than incandescent lamps and are widely used for general industrial and domestic illumination. They can be divided in terms of systems in which there is local thermal equilibrium (LTE plasmas) and those in which there is not (non-LTE plasmas) [11]. LTE plasmas, sometimes termed high intensity discharge (HID) lamps, operate at high gas pressures. A good example is the high-pressure sodium lamp, produced by adding sodium metal to the mercury lamp and used as an intense yellow light source for many outside applications, such as street lighting and freight yards. For these applications, the colour of the illumination is less important than the cost. An intense white HID lamp can be obtained by adding metal halide salts, and is frequently used in outdoor floodlighting. The most important non-LTE plasmas are low pressure mercury or sodium lamps. Gas discharge lamps have the disadvantage that the light is emitted in discrete lines rather than a broad band obtained through incandescence, as can be seen in the spectra of mercury lamps discussed in Chap. 14 (see Fig. 14.3). In addition, in many cases, much of the light is in the UV (such as the 254 nm line in the mercury discharge lamp) and is not of itself useful for lighting, although it can be converted to visible light by a phosphor, e.g. the white phosphor coating on the inside of a fluorescent lamp.

4.3.2.1 Fluorescent Lamps and Phosphors

During the early part of the twentieth century, considerable effort was made to convert the UV component of gas discharge lamps into broadband visible light through the use of solid state luminescent materials, commonly termed phosphors. Phosphors have been studied since the middle of the nineteenth century, and are, typically, inorganic semiconductors, such as zinc sulfide (9.3), zinc silicate, or calcium tungstate [12, 13], frequently doped with other species. The first practical “fluorescent” lamps, invented in Germany in the 1920s and commercialised in the US in the following decade, involved low pressure mercury discharge lamps with the walls of the tube coated with appropriate phosphors. The same phosphors could be excited by high energy radiation, such as electron beams (cathode rays), and the search for new luminescent materials for fluorescent lamps has paralleled that for cathode ray tubes (CRTs) for television or other display applications. Much of the early research on phosphors was empirical [12]. However, by the middle of the twentieth century there was sufficient knowledge on electronic structure, luminescence spectra and photophysics of inorganic materials to apply a more rational approach to the design of phosphors. Luminescence of inorganic materials can be divided into that of localised centres, such as metal ions, and emission involving delocalised semiconductor bands [13]. In general, emission involving semiconductor bands will give a rather broad spectrum, while that from localised centres will involve sharp, well defined spectral lines. For display applications, it is often advantageous to have rather sharp spectra, such as is found with the f–f transitions of lanthanide ions (10.1–10.6). A major breakthrough in luminescent materials for CRTs came with the development of a good red phosphor involving europium(III) in yttrium vanadate (10.1) [14]. Full colour displays are obtained by combining emission from different coloured phosphors, typically red, green and blue (RGB) using the Commission Internationale de l’Éclairage (CIE) chromaticity diagram (Fig. 14.3) [15]. An entertaining account of the development of this diagram is given elsewhere [16].

Since the seminal work on europium(III) based phosphors, lanthanides have become some of the most important materials for both lighting and displays [17]. With the normal fluorescent lamps containing mercury vapour, electronic energy transfer is crucial for efficient conversion of UV into visible light. This can involve both Förster [18] and Dexter [19] mechanisms. As discussed in Chap. 1, the Förster mechanism involves dipole–dipole interactions which require good overlap between the emission spectrum of the energy donor and the absorption spectrum of the acceptor, while the Dexter mechanism involves electron exchange, and needs close proximity between the donor and acceptor. Frequently, cerium(III) is used as an intermediate in fluorescent lamps for good photochemical reasons. The electronic transitions in most lanthanide ions involve two f orbitals, and are very weak, since they are forbidden by the Laporte selection rule (Chap. 1). However, the electronic transitions in Ce(III) involve transition of an electron between 4f and 5d orbitals, which is allowed by both spin and Laporte selection rules. This gives the Ce(III) ion a good molar absorption coefficient, which favours energy transfer from excited mercury atoms. In addition, the lifetime of the Ce(III) fluorescence is short, which avoids problems of saturation of excited states at high light intensities. The consequence is that excited Ce(III) in the solid phosphor layer can transfer energy efficiently to neighbouring terbium(III) or other lanthanide ions. Although this discussion is an oversimplification, and there are many parts of the sensitisation mechanism that are still poorly understood, it does indicate some of the photochemical ideas used in optimising the efficiency of fluorescent lamps. A schematic description of energy transfer from excited mercury atoms (Hg^*) to Tb(III) (10.2) via Ce(III) is given below.

$$ {\text{Hg}}^{*} + {\text{ Ce}}\left( {\text{III}} \right) \to {\text{Hg }} + {\text{ Ce}}\left( {\text{III}} \right)^{*} $$

(4.1)

$$ {\text{Ce}}\left( {\text{III}} \right)^{*} + {\text{ Tb}}\left( {\text{III}} \right) \to {\text{Ce}}\left( {\text{III}} \right) \, + {\text{ Tb}}\left( {\text{III}} \right)^{*} $$

(4.2)

$$ {\text{Tb}}\left( {\text{III}} \right)^{*} \to {\text{Tb}}\left( {\text{III}} \right) + h\nu $$

(4.3)

In comparing the properties of different light sources, three parameters are commonly used. The first is the luminous efficacy , which is the response of the average eye to light over the visible spectral region (380–780 nm), and is measured in lumens (visible light) produced per watt of electrical power (lm/W). The second property, the correlated colour temperature , is a measure of the appearance of the light source and is given as the temperature of the black body which most closely represents the light source. The spectrum of sunlight depends upon altitude and time of day, but at noon typically has a correlated colour temperature of 5000–7500 K. Finally, the colour-rendering index is an indication of how accurately the light source can reproduce colours of objects compared with natural light (daylight), and is measured by comparison of the apparent colours of reference pigments using the light source and a black body standard [9].

A 35 W fluorescent lamp has a luminous efficacy of about 104 lm/W, compared with 16 lm/W for a 100 W incandescent lamp [20]. However, fluorescent lamps have a number of disadvantages. In particular, they need an electric ballast to produce the initial discharge in the mercury vapour, which requires a longer start-up than an incandescent lamp. This limits applications in areas, such as traffic lights, which need fairly rapid switching. In addition, white light is important for illumination, particularly for indoor applications, and although the phosphor in a fluorescent lamp can be developed to give a good spectral distribution, often by using a mixture of materials, there are always some lines present in the emission spectrum from the mercury lamp. This can influence the perceived appearance of colours. Most people will be familiar with the experience of buying clothes in a shop lit with fluorescent lighting and then finding the colour appears different in sunlight. This is because of differences in the emission of the light sources; spectra of typical incandescent and fluorescent lamps are compared in Fig. 4.2; the colour-rendering index is 100 for the incandescent lamp and 85 for the fluorescent one.

Fluorescent lamps have much higher luminous efficacy than incandescent ones. Although they are more expensive, their usable lifetimes are an order of magnitude longer. There are, thus, very real economic advantages of these systems, particularly the recently developed energy-saving compact fluorescent lamps (CFL), over tungsten filament lamps. However the majority of fluorescent lamps still use the toxic heavy metal mercury, although other gases are being tried. This has a long term environmental impact and has stimulated the development of other sources of lighting, in particular light-emitting diodes (LEDs).

4.3.3 Electroluminescence and Optoelectronic Displays

4.3.3.1 Cathode Ray Tubes (CRTs)

For much of the twentieth century, the dominant optoelectronic display device was the cathode ray tube, where an electron beam is scanned across a phosphor screen in a vacuum tube. Nowadays, conventional CRTs are being superseded in the display area by flat-panel displays for applications in television, laptop computers, mobile phones, etc. The main competitors in this area at the moment are liquid crystal displays (LCDs) [9, 21], plasma display panels (PDP) [22] and LEDs [23]. Plasma display panels generate a plasma discharge in a mixture of Xe and Ne to produce vacuum UV radiation. This is then used to excite a phosphor screen. This has the advantage that it is possible to prepare large area screens with high brightness [22]. The limitation is that it requires relatively high power consumption. Related technologies being developed for flat screen displays use field emission with a cold cathode and microcavity plasmas [22].

4.3.3.2 Liquid Crystal Displays (LCDs)

These are not, intrinsically, photochemical systems, however, they are optoelectronic systems, and backlighting is fundamental to both their functioning and overall energy efficiency. The most common LCDs involve an oriented twisted nematic phase of an organic material sandwiched between two optically transparent electrodes (typically indium tin oxide, ITO) [21]. This is illuminated by polarised light and the anisotropic liquid crystal changes the plane of polarisation. When observed through a second polariser, light is transmitted when this is parallel to the plane of polarisation or blocked when it is perpendicular. The orientation of the liquid crystal can be changed by application of an electrical potential. In displays this is used to switch between the light (transmitting) and dark (non-transmitting) states, with each pixel corresponding to a liquid crystal cell. A schematic of an LCD based on a typical twisted nematic liquid crystal is shown in Fig. 4.3. The initial monochrome liquid crystal displays were used in watches and calculators [21]. Full colour high definition LCDs are now common, and represent a multibillion pound sector of the display industry. Colour is achieved by dividing the pixels into three, with red, green and blue filters in separate layers. One of the early limitations was that the angle of vision of the displays was limited. This has been partially overcome by using thin film transistors (TFT) to address the cells in active matrix LCDs [21, 24], and has led to the high definition LCD screens currently in use for television, mobile phone and laptop applications. These displays have the advantage that they need very little power for switching. However, one limitation is that the need for backlighting increases the energy consumption, since light is lost on going through the polarisers and colour filters. The battery lifetime of laptop computers is frequently limited by energy usage by LCD backlights. More energy efficient lighting systems, particularly involving light emitting diodes, are under development and are helping overcome this problem [25].

Fig. 4.3

Schematic diagram of one pixel of a twisted nematic liquid crystal display. Polarised light is twisted on passing through the twisted nematic liquid crystal. The orientation of this is changed by applying an electric field. Figure adapted from Ref. [166]

4.3.3.3 Electroluminescence: LEDs, OLEDs and PLEDs

LEDs. Electroluminescence is the emission of light from materials upon application of an electric field [24]. Although electroluminescence from carborundum (a form of amorphous silicon carbide) was first reported over a century ago [26], the development for practical applications in lighting and displays started in the early 1960s with the first report of a red light-emitting diode involving a GaAsP semiconductor [27]. The luminous efficacy of this early inorganic semiconductor LED was only 0.15 lm/W. The initial use focused on red indicator lights and low definition displays on calculators and watches. Since then the area has shown enormous developments in terms of luminous efficacy, available spectral output, intensity, cost and stability [25, 26], and high power LEDs, typically involving elements from Groups III (3) and V (15) of the Periodic Table, now occupy a major role as high-brightness visible light sources. The structure of a typical inorganic semiconductor LED is shown in Fig. 4.4a. Apart from their efficacy, which approaches that of fluorescent lamps, inorganic semiconductor LEDs have very fast switching times (ns), which makes them excellent candidates for technological applications in areas such as traffic signals and red car stop lights [25], and also pulsed sources for photochemical studies (see Chap. 14). A major advance for practical applications was the development of stable, intense blue LEDs based on wide band-gap nitride semiconductors, such as InGaN/AlGaN [28]. As well as expanding the spectral range of LEDs, which now extend into the UV, it has led to the development of high intensity white LEDs, with applications ranging from car headlights to solid state lighting for both domestic and industrial use [25, 26, 29, 30]. Various methods can be used to produce white semiconductor LEDs [25]. One involves combining red, green and blue LEDs. Although this gives high efficiency, and is used in backlighting for liquid crystal displays, the problem is matching the colours, particularly as they will age at different rates. Alternatives are to use a UV emitter to excite red, green and blue phosphors in a similar fashion to their excitation in fluorescent lamps, or to use a blue LED to excite a yellow phosphor, such as YAG:Ce, mixing the blue with the yellow to give white light [25]. There is intense competition to develop good white LEDs using these ideas for solid state lighting [29, 30], and they are likely to shortly become one of the main sources of lighting. One problem which limits their unit cost is that it is only possible to prepare efficient inorganic semiconductor LEDs with a relatively small surface area. Lamps will typically require a number of these LEDs. As described in the next section, this problem can be overcome by replacing the inorganic semiconductors with organic ones.

Fig. 4.4

a Structure of an inorganic semiconductor LED; b structure of an OLED; c charge injection and excited state formation in an OLED. Figure adapted from Ref. [167]

OLEDsand PLEDs. Electroluminescence in aromatic molecules (1.1–1.7), such as single crystals of anthracene (1.1), has been known since the 1960s [21, 31]. However, very high voltages were needed to generate light emission. In 1987, Tang and Van Slyke at Eastman Kodak showed that it was possible to obtain efficient green light electroluminescence from an organic light-emitting diode (OLED) containing a thin, vacuum deposited film of tris(8-hydroxyquinoline)aluminium(III), Alq₃ (6.1) [51, 32]. The device had the metal complex sandwiched between an optically transparent ITO anode and a Mg:Ag cathode (Fig. 4.4b). The basic mechanism of electroluminescence in organic compounds [31] involves electron injection into the lowest unoccupied molecular orbital (LUMO) at the cathode and positive charge (hole) injection at the anode (Fig. 4.4c) to produce the excited singlet state of Alq₃, which then fluoresces. As expected from this mechanism the electroluminescence spectrum is very similar to the photoluminescence. Luminous efficacy of the initial device was 1.5 lm/W, but subsequent developments, in particular, using charge injection or blocking layers at the electrodes [23], led to rapid increases in efficiency of OLEDs, which are now major materials in the display area. One advantage for certain applications is that they can be prepared on flexible supports [31]. About 3 years after the first report of an OLED, Friend and coworkers at Cambridge showed that it is possible to use conjugated polymers (CPs) such as poly(p-phenylenevinylene) (3.2) as the light-emitting layer to give a polymer light emitting diode (PLED) [14, 33, 34]. This has the advantage in device preparation that these can be deposited from solution by using standard printing techniques, such as ink-jet printing, which makes them particularly suitable for preparation of large area displays. It is also feasible to perform large scale commercial production of displays through reel-to-reel printing methodologies. Structures of some commonly used conjugated polymers are shown in 3.1–3.9. A more detailed description of conjugated polymers for electroluminescent applications is given elsewhere [35].

One limitation of electroluminescence in organic materials is that charge injection produces both singlet and triplet excited states. Statistically these will be produced in the ratio 1:3. Since only the singlet state is normally luminescent, this reduces the maximum efficiency possible from these devices to 25 %, although there are suggestions that for PLEDs it may be possible to have higher singlet:triplet ratios. Forrest and co-workers showed that it is possible to overcome this problem, and obtain increased luminescence efficiency, using room-temperature metal organic phosphorescent materials [45, 36]. The initial room temperature phosphor was a platinum porphyrin (5.4). Since then a wide range of other complexes [37, 38], in particular phenylpyridine complexes of iridium(III), have been developed (6.6–6.8). Both OLEDs and PLEDs are now on the market in high definition displays. The next step will be the use of both small molecule organics and conjugated polymers in white organic light emitting diodes (WOLEDS) for artificial lighting [37, 39]. These have the advantages of broad spectral output, giving good colour rendering indices, and good CIE coordinates (See Chap. 14). Key practical problems at the moment include light out-coupling efficiency, practical operating voltages and long-term operational stability. However, devices are now available with lifetimes greater than that of typical incandescent sources (≈1000 h), and the attractive design possibilities available with WOLEDS suggests that they will shortly make an important contribution to the technologies available for artificial lighting.

4.3.4 Radioluminescence

Radioluminescence is produced by the bombardment of a material with ionising radiation such as beta particles. One application is in tritium light sources, in which gaseous tritium is encased in a glass tube lined with a phosphor, e.g. zinc sulfide (9.3). Since no electrical excitation source is required, tritium light sources are self-powered, and find application in emergency exit lighting and wristwatch illumination. The colour of the emitted luminescence depends on the phosphor. Radioluminescence is also used in scintillation counters to measure radiation levels. The scintillator consists of either an organic crystal, e.g. anthracene (1.1), or a mixture of lumophores (e.g. p-terphenyl, 2,5-diphenyloxazole) dispersed in an organic solvent or plastic film, that luminesces when struck by ionising radiation. A photomultiplier tube is used to quantify the emitted photons. Depending on the lumophore used, the emission may either be fluorescence, phosphorescence or delayed fluorescence. Scintillators are used in homeland security radiation detectors, medical diagnostics and high energy particle physics experiments.

4.3.5 Chemiluminescence

Chemiluminescence is the emission of light as the result of a chemical reaction. The classic example is luminol oxidation. When a basic solution of luminol (5-amino-2,3-dihydro-1,4-phthalazinedione) is mixed with hydrogen peroxide in the presence of a suitable catalyst (e.g. Cu²⁺ or ferricyanide), it becomes oxidised, forming an unstable peroxide radical in its excited state. The excited state radical relaxes radiatively emitting a blue glow. Luminol is used by forensic crime scene examiners to detect the presence of blood traces with a sensitivity at the parts per million (ppm) level, since the iron present in haemoglobin also catalyses this reaction. However, perhaps the most familiar application of chemiluminescence is in glow sticks. In this case, a chemical reaction is used to sensitise the emission from a lumophore. The glow stick contains a solution of a suitable fluorescent dye and diphenyl oxalate and a glass ampule containing hydrogen peroxide. When the glow stick is ‘snapped’, the glass ampule is broken, releasing the hydrogen peroxide which reacts with diphenyl oxalate forming a peroxyacid ester intermediate. This unstable intermediate decomposes spontaneously to carbon dioxide, releasing its excess energy and exciting the dye, which then relaxes by emitting a photon. The colour of the glow stick depends on the dye used.

A number of biological species, including fire flies, glow worms and marine organisms, emit light. This is termed bioluminescence and involves chemiluminescent reactions within the organism. Bioluminescence is a useful tool in genetic engineering and bioluminescence imaging can be used to study biological process in vivo.

4.3.6 Photoluminescent Materials: Fluorescent and Phosphorescent Emitters

4.3.6.1 Fluorescent Dyes and Pigments

Optical brighteners . These are used as additives in papers, plastics, fabrics, and fabric detergents, to convert the small fraction of sunlight which lies in the near UV into visible (usually blue) light. The photochemical requirements are: (i) efficient near UV absorption with no absorption tailing out into the visible, otherwise a yellow coloration will be produced; (ii) a high emission quantum yield in the blue spectral region; and (iii) high photochemical stability. Chemically, the compounds must adhere to the required substrate, and so optical brighteners with different chemical substituents may be required for cotton and nylon/silk fabrics. They must also be chemically and thermally stable. Triazole-stilbene derivatives are some of the most common optical brighteners (2.1–2.3). However, these tend to fade on long-term exposure to UV light, and when long-term stability is important non-stilbene compounds are favoured. Many optical brighteners are available from fine chemical suppliers, but relatively few are in the standard research chemical catalogues.

Fluorescent probes . Fluorescence probes are materials which indicate, by fluorescence, the presence of some specific chemical species or environment. There are two types: passive and active. Passive probes indicate the presence of the material of interest simply by physically or chemically binding to it. They thus indicate the presence of an analyte substrate by their own fluorescence (although this may also be altered by the presence of the analyte), the intensity of which can be used to determine the presence of the analyte either qualitatively or in some cases quantitatively. A typical use of such probes is in fluorescence microscopy, and there are a wide range of commercially available probes for this purpose (e.g. 4.3). Active probes undergo a photochemical change in the presence of the chemical or environmental feature to which they are sensitive. This results in a change in some emission property such as wavelength, intensity, polarisation, or lifetime (e.g. 1.4, 4.8). Optical probes are discussed in detail in Chap. 12.

Laser dyes. The role of a laser dye is to absorb the pump radiation efficiently, and convert this absorbed energy into an excited-state which can itself act as the upper state in a laser transition (see Chap. 14). Photo- and chemical stability are important, but the key photochemical features are: (i) absorption at the excitation wavelength; (ii) a lifetime sufficiently long to allow a population inversion to build, (although the lifetime required may be quite short, depending upon the duration of the pump pulse); and (iii) a low absorption at the lasing wavelength. In assessing and controlling the latter it is the absorption profile of the solution during the population inversion which is relevant, so it is the transient absorption as much as the ground state absorption which is important. Absorption due to long lived triplet states is a particular problem and so a very low triplet quantum yield is desirable. Triplet quenchers are sometimes used to reduce the lifetime and hence quasi-steady-state concentration of triplets during the population inversion. There are now many laser dyes available which cover the full spectral range from the IR to the UV, and the best sources of up to date information are manufacturer’s catalogues. However there are a few dye classes which are particularly useful as laser dyes, where the basic chemical skeleton is retained but changes in substituent are used to shift absorption and emission wavelengths, e.g. basic structures of 2.1, 4.1, 4.2, 4.6.

Emissive polymers. One of the major current areas of research is on emissive conjugated polymers (CPs) (3.1–3.9) due to their exceptional optical, electronic and mechanical properties [40]. Ionic conjugated polyelectrolytes (CPEs) tend to be highly sensitive to changes in their physical and chemical environment and, hence, one of their major potential uses is in biological and chemical sensors [41]. They have the advantage of high sensitivity with short luminescent lifetimes and the possibility of amplified fluorescence energy transfer or quenching processes (see Chap. 12) [42]. Recent research on the application of CPEs in biological and chemical sensors has led to a greater structural diversity and new synthetic protocols for their preparation [43]. Neutral CPs are also important technological materials, with applications in optoelectronic devices, such as PLEDs [34] and solar cells [44]. Part of the interest stems from the possibility of implementing solution-based deposition methods as part of the device fabrication process. Incorporation of ionic side groups increases solubility in polar organic solvents and water, which may allow more environmentally friendly manufacturing processes. Current interest in solar cells based on CPs stems from the promise of low cost fabrication. Research into cationic conjugated polymers (CCPs) has recently been mainly focussed on sequence specific DNA assays, designed by utilising the electrostatic interactions between cationic conjugated polymers and negatively charged DNA. These assays commonly exploit the ability of CCPs for efficient excitation energy transfer using, for example, protein nucleic acids (PNA) and Förster resonance energy transfer (FRET) to luminescent acceptors, such as fluorescein.

4.3.6.2 Phosphorescent Dyes and Materials

Phosphorescent molecular species in solution or matrices. Many organic molecules and dyes are phosphorescent in rigid low temperature glasses, where processes which might deactivate long lived triplet states are inhibited. Phosphorescent yields and lifetimes of many organic molecules, (and some inorganic complexes), are given in [45], quantum yields can be very high and triplet lifetimes can easily be up to a few tens of seconds at 77 K (the determination of these parameters is described in Chap. 15). However few organic molecules are phosphorescent at room temperature. Those commonly available organic compounds that, such as bromonaphthalene, usually carry heavy atom substituents, but even so they are usually only very weakly phosphorescent at room temperature. One particular group of organics which are strongly phosphorescent at room temperature are the thiocarbonyls (11.21). Although thiocarbonyls show an interesting, rich, and unusual, photochemistry they are relatively photochemically and chemically unstable, which limits their uses.

A number of inorganic complexes show room temperature (and low temperature) phosphorescence, with two groups widely used as phosphorescent materials namely: Pt and Pd porphyrins (5.4), and Au (either in the monovalent or trivalent oxidation states), Ru(II), Rh(III) and Ir(III) complexes in which the metal is usually complexed by nitrogen, sulphur or heterocyclic rings (6.2–6.8). Further examples may be found in Ref. [38]. In these complexes the presence of the heavy transition metal gives those orbitals with some metal character spin–orbit coupling, which enhances both radiative and non-radiative singlet–triplet interconversion rates. The most important effect is a relative increase in the rate of the T₁ to S₀ radiative process; thus, while triplet quantum yields are much higher than those in organic molecules, emission lifetimes are much shorter, typically ~1 μs–1 ms.

Delayed fluorescence. It is possible to generate relatively long lived fluorescence emission if there is a mechanism for repopulation of the singlet from the triplet. This is termed delayed fluorescence, and there are two common mechanisms.

• (1) Thermal repopulation, where the triplet level is close enough to the singlet level that the singlet state can be thermally populated from the triplet. Sometimes called E-type delayed fluorescence, after the compound eosin (4.4) which exhibits this behaviour. The lifetime of E-type delayed fluorescence is equal to the triplet lifetime. Other materials which exhibit E-type delayed fluorescence are palladium porphyrins (5.4) and thiocarbonyls (11.21).

• (2) Triplet–triplet annihilation in which two triplet states combine to give an excited singlet which generates fluorescence and a ground state singlet, called P-type delayed fluorescence after the molecule pyrene (1.4) which exhibits this behaviour. The lifetime of P-type delayed fluorescence is equal to one half that of the corresponding triplet.

Although named after the archetypical molecular examples, both types of delayed fluorescence are reasonably common. Delayed fluorescence is discussed in more detail in Chap. 1.

4.4 Sensitisers, Donors, Acceptors, Quenchers and Traps

Sensitisers are absorbing materials which can be used in photochemical processes in very specific ways; to make specific reactions occur, or force a specific reaction route by the generation of specific singlets, triplets, radicals or redox active chemicals via energy transfer, electron transfer, or radical reactions. When used as donors (D) in energy transfer reactions they can be used to generate singlets or triplets of acceptor (A) molecules which might not available by direct excitation, or to generate acceptor triplet states without the need for direct excitation of the acceptor. Sensitisers can also be used to probe molecular arrangements and distributions since, if a sensitiser is expected to generate a specific product by reaction with an acceptor, the presence or absence of sensitisation, as well as the efficiency of the process, can be used as a measure of either the proximity of the sensitiser and acceptor molecules, or the viscosity of the medium (see Chap. 12).

The term acceptor is used to describe: (1) a compound, or atom, which accepts energy from a higher energy donor in excited-state energy transfer, and which becomes, as a consequence, excited; or (2) a compound which reacts with an excited-state (or less frequently some other chemical species, notably singlet oxygen), to give a recognisable specific product. In both of these processes the acceptor undergoes a recognisable change, indicating its role in the reaction.

The term quencher is broader, and includes any material which acts to reduce emission, or the yield of a photochemical reaction, by interaction with an excited-state. This interaction may be physical, or chemical, and either reversible or irreversible, and nothing about the nature of the quenching process or any change in the state of the quencher is necessarily inferred.

The term trap has two uses. (1) In solid state chemistry a trap is a site, a part of the structure, into which energy, or an electron (or hole) can migrate, be trapped, and lost to the system. (2) The term trap is also used for chemical species which give specific, and usually measurable, reactions with species of interest, notably free radicals, i.e. free radical traps (8.1–8.3). Here, trap sense (2) is very similar to acceptor sense (2), but there is a subtle difference because of the different type of mobility free radicals and chemical species exhibit. Once formed, chemical species migrate by molecular diffusion. Free radicals also migrate by molecular diffusion but they also undergo transport and population growth through a series of propagation and branching reactions, and so the free radical itself is mobile, even though the molecular carrier itself is exchanged, in a similar way to the mobility of ‘energy’ in energy transfer migration in solids.

4.4.1 Excited State Sensitisers and Acceptors

For organic molecules, and most inorganic complexes, molecular singlet and triplet states are most important, and therefore singlet and triplet sensitisation are most commonly encountered. For some other groups of materials, notably those involving atomic transitions, such as gas phase atoms or lanthanide ions in solids or solution, sensitisation involving states of other spin multiplicities is important. In singlet sensitisation the required reaction is the transfer of singlet energy from the sensitiser, which is the donor, to another molecule, the acceptor. Any excited singlet state higher in energy than the acceptor singlet can thermodynamically act as a sensitiser, but as discussed in Chap. 1, other conditions must be right for the energy transfer process. Energy transfer requires energy matching between donor and acceptor states. In practise, for molecules of moderate size the high density of states (DOS) means that almost any sensitiser of higher energy than the acceptor will act as a sensitiser by collisional or Dexter energy transfer, provided the donor and acceptor are, or can become, close enough within the donor lifetime. For Dexter transfer involving states localised on atoms or ions, which do not have a high DOS, energy matching becomes a more stringent condition. For atomic states in a solid state lattice, energy matching can sometimes be promoted by coupling with the lattice vibrations (phonons). FRET involves coupling of molecular dipoles, and can occur over a much longer range because orbital overlap is not required. However, overlap of donor emission with an acceptor absorption band for an allowed transition is necessary for efficient energy transfer.

In general, the most effective singlet sensitisers will be those with long lifetimes, and as a consequence, high fluorescence quantum yields. The short lifetime of molecular singlet states means that for efficient Dexter singlet–singlet energy transfer the acceptor must be adjacent to the donor, or, if contact is diffusion controlled, the acceptor must be present at high concentration in a low viscosity medium.

Molecular triplet energy transfer is usually via Dexter energy transfer. However, because of the long lifetime of some triplets, FRET is also possible from a triplet donor to a singlet acceptor, where the long donor lifetime compensates for what must be, because the radiative transitions involves are spin forbidden, a slow energy transfer rate constant.

Generally, if the energy difference between D and A triplet states is greater than a few kJ mol⁻¹, energy transfer in solution will occur at every encounter between D and A and therefore the rate constant is close to the diffusion controlled value. However, if the molecular structure of one or both D and A is significantly different in the triplet state as compared to the ground state then the reaction requires major molecular structural reorganisation, and this can slow the energy transfer rate considerably. Balzani et al. have analysed the effect of structural rearrangement in a similar way to that used in the Marcus theory of electron transfer reactions [46].

For the determination of the triplet energy of an acceptor, a series of sensitisers with differing triplet energies and known transient difference spectrum are required; the experimental approach is described in Chap. 15 which also gives triplet state properties for selected compounds (Table 15.2). The porphyrins (5.1, 5.2), phthalocyanines (5.3, 5.5) and naphthalocyanines (5.7) make a useful series of relatively low energy triplet sensitisers because of their structural similarity (5). Unfortunately only a few of these compounds are phosphorescent and therefore flash photolysis is required for direct kinetic studies of most triplet sensitisation.

Triplet sensitisers can generally be placed in one of three categories:

• 1. High to moderate energy polyaromatic and polyaromatic derivatives (such as 1.1–1.7), or other organics, which are not phosphorescent at room temperature, but are often phosphorescent at 77 K. Most have triplet lifetimes of ~ms duration and well-characterised triplet transient difference spectra. Many are commercially available.

• 2. The relatively low energy porphyrins (5.1, 5.2), phthalocyanines (5.3, 5.5), naphthalocyanines, and their metallated derivatives, some of which are phosphorescent at room temperature, but many of which are not phosphorescent at either room temperature or 77 K. Lifetimes are typically 100 μs to a few ms, and most have well-characterised triplet transient difference spectra. Many are commercially available.

• 3. Moderate to low energy Au, Pt, Pd, Ir and Ru complexes which are phosphorescent at room temperatures with lifetimes typically 1–20 μs, some of which are commercially available (6.2, 6.6–6.8).

The use of triplet state acceptors has generally two roles. (1) Where the triplet state is to be removed and the transferred energy degraded as heat. Any passive absorber with lower triplet state energy than the donor will act in this way. Such a triplet state acceptor is also a photochemical stabiliser. (2) Where the triplet energy is to be trapped in a triplet state to be used for measurement or some specific function, e.g. energy transfer to an acceptor of known transient triplet absorption spectrum or emission. Here the acceptor triplet state photophysics and photochemistry must be known. Identification of energy transfer to such another known triplet state is often used as confirmatory evidence that a triplet state species is involved in the reaction under study.

Sensitisation is often used to generate electronic excited states in lanthanide (Ln) complexes and Ln-containing solid-state phosphors (e.g. 10.1–10.6). Ln-materials emit over the entire visible spectrum: red (Eu³⁺, Pr³⁺, Sm³⁺), green (Tb³⁺, Er³⁺) and blue (Tm³⁺, Ce³⁺). They are therefore interesting for a wide variety of applications including solid-state lighting, lasers, and optical communications and storage. The optical transitions of Ln³⁺ ions take place predominantly within the 4f manifold, where the electrons are largely shielded from external crystal field effects by the filled 5s and 5p levels. Consequently, Ln³⁺ ions give rise to much narrower, atomic-like line absorption and emission spectra compared to organic small molecules or polymers. The Ln³⁺ electronic configuration gives rise to ground and excited states with a variety of multiplicities other than singlets and triplets (e.g. quartet, quintet etc.); consequently some ions are fluorescent (ΔS = 0), others are phosphorescent (ΔS ≠ 0), and some are both.

However, f–f transitions are formally electric dipole forbidden by the Laporte selection rule, (a change in orbital angular momentum of ±1 is required to accommodate the loss of photon spin upon absorption), although they are allowed by electric quadrupole, magnetic dipole and forced electric dipole mechanisms to some extent. Direct excitation of the Ln³⁺ ion is therefore not easily achieved, due to the low molar absorption coefficients associated with these transitions (ε ~ 5–10 mol⁻¹ dm³ cm⁻¹). In Ln-complexes, indirect excitation via a sensitising ligand or antenna is used to overcome this limitation [46]. The sensitising ligand absorbs light, initially forming its excited singlet state. The excitation energy is transferred to the ligand’s triplet state via intersystem crossing (the efficiency of this process being improved due to enhanced spin–orbit coupling induced by the heavy atom effect of the Ln³⁺ centre). Population of the Ln³⁺ excited emissive state is subsequently achieved through intramolecular energy transfer from the ligand triplet state. This process therefore requires that the ligand triplet state is higher in energy than the Ln³⁺ excited state being sensitised.

4.4.2 Singlet Oxygen Sensitisers, Quenchers and Acceptors

Most singlet oxygen sensitisers are triplet states. Singlet oxygen sensitisation is very similar to triplet sensitisation, except the spin state of the acceptor, ground-state oxygen, is a triplet, and the products are two singlets, i.e. singlet oxygen and the singlet ground state sensitiser [47, 48]. The energetic and spin conservation rules are the same as triplet energy transfer, but the spin statistics are different because the two reacting species are triplets. The spin angular momentum quantum number along any reference axis of each triplet state (i.e. the triplet sensitiser and ground state oxygen) can take one of three values; −1, 0, +1; thus when any two triplets combine in the encounter pair there are 3 × 3 possible combinations. 1/9^th of the encounters will give an overall singlet encounter pair, 3/9^th a triplet, and 5/9^th a quintet encounter pair. Of these three: only the singlet encounter pair can lead to two singlet state products; the triplet pair can, energetics allowing, give two electron transfer radical doublet states; while in the quintet case there are no spin-allowed energy transfer or electron transfer products possible, so that, in the absence of spin relaxation, the quintet encounter pair can only separate back into reactants. Thus singlet oxygen generation can be expected to occur with a maximum rate of around 1/9^th the encounter rate, and this is usually borne out experimentally.

Most species which generate a triplet state of significant lifetime and energy higher than that of singlet oxygen (94.3 kJ mol⁻¹, 0.98 eV, corresponding to a transition in the NIR at ~1270 nm—see Fig. 15.3) have the potential to be singlet oxygen sensitisers, since oxygen quenching of such triplet states generally goes predominantly via energy transfer (although electron transfer to give superoxide radical can be a significant, and interfering, reaction). For many such compounds the singlet oxygen yield in fluid air-equilibrated solution can be expected to be similar to the triplet yield. However the term singlet oxygen sensitiser is usually reserved for compounds which have: high triplet quantum yields; high oxygen quenching rate constants in which singlet oxygen predominates as the reaction product; reasonably high molar absorption coefficients and are thus efficient absorbers; low singlet oxygen quenching rates; and a well-characterised and quantified photochemistry. The measurement of the singlet oxygen yield is discussed in Chap. 15.

The role of a singlet oxygen quencher is usually just to remove singlet oxygen, often to inhibit singlet oxygen induced photodegradation. There are two main mechanisms by which this can be achieved.

• (1) Energy transfer into a low energy triplet acceptor in which molecule the triplet energy is rapidly degraded to heat; typical examples are Ni complexes (11.10).

• (2) Electron transfer into a molecule in the encounter complex followed by rapid reverse electron transfer before dissociation of the encounter complex. Typical examples of this type of quencher are hindered amines such as DABCO (11.8).

Quenching rate constants for triplet energy transfer quenchers are often faster than for electron transfer quenchers. However the requirement for a triplet state lower than singlet oxygen invariably implies singlet state energies in the visible region, and thus singlet oxygen triplet energy acceptors are coloured to varying degrees. Apart from anything else this means they will also act as competitive absorbers in almost any mechanistic study. Amine electron transfer quenchers are usually colourless, often with longest wavelength absorptions in the mid UV.

The role of a singlet oxygen acceptor is usually to show evidence of singlet oxygen in mechanistic studies. The development of detectors for the direct detection of singlet oxygen emission (Chaps. 14, 15) has alleviated the need for indirect measurements where the singlet oxygen yield is reasonably high, but singlet oxygen acceptors are still useful especially when the singlet oxygen yield is so low as to be undetectable directly. Three common approaches are used.

• (1) Kinetic studies where the triplet state acceptor can be identified. A good example of this is shown by use of β-carotene (11.6). β-carotene itself has a very low quantum yield of triplet state formation by direct excitation, the triplet energy is lower than that of singlet oxygen and the triplet lifetime and transient absorption spectrum are known. Thus if the system under study allows the photochemical formation of singlet oxygen then this can be studied using ns laser flash photolysis by following the kinetics of energy transfer from singlet oxygen to β-carotene and formation of β-carotene triplet. β-carotene triplet will also be populated by energy transfer from any triplet state used in the initial formation of singlet oxygen but consideration of the kinetics shows that, because of the combination of rapid quenching of triplet state singlet oxygen sensitisers by oxygen in aerated solution and the relatively long lifetime of singlet oxygen, it is possible to separate out these two processes. Quenching of singlet oxygen by carotenoids is discussed in detail in Chap. 8.

• (2) Where the rate of loss of acceptor can be followed, either spectroscopically, or by chromatographic analysis such as GC or HPLC (in which case the specific involvement of singlet oxygen can often be confirmed by product analysis). Spectroscopic detection requires an acceptor of known absorption or emission characteristics. If ns laser flash photolysis is available then the kinetics of decay of the acceptor absorption or emission can be followed, and kinetic analysis can be used to confirm a singlet oxygen process. Diphenylisobenzofuran (11.9) and rubrene (1.7) have been widely used as singlet oxygen acceptors in these types of experiments, with reaction with singlet oxygen followed by either loss of absorption or of fluorescence.

• (3) Where the involvement of singlet oxygen can be identified by product analysis. Here the product from reaction between singlet oxygen and the acceptor gives rise to a stable molecular species which can be identified either spectroscopically or by chemical analysis such as GC or HPLC.

It is worth noting that the lifetime of singlet oxygen is highly dependent on solvent [49]. The lifetime is particularly short in solvents with OH bonds, which provide high frequency vibrations into which the electronic energy of singlet oxygen can be transferred, or in solvents, such as amines where electron transfer is possible, and it is particularly long in halogenated solvents which only have low frequency vibrations. Due to the effect of OH oscillations in providing a quenching mechanism, the lifetime is also highly dependent upon solvent deuteration. Thus a comparison of rates or yields in H₂O and D₂O, or normal and deuterated alcohols, is a useful tool in helping unravel or identify a singlet oxygen mechanism.

4.4.3 Redox Sensitisers

An excited-state is simultaneously both a better oxidant and better reductant than the ground-state molecule. The difference in redox potentials between ground and excited-state is given, to a reasonable approximation, by the excited-state energy in eV. Redox sensitisers create charge transfer from an excited state. This can be either unimolecularly across a molecule or semiconductor, or bimolecularly via a process in which the excited-state undergoes a charge transfer reaction with the solvent, a redox quencher, or a semiconductor. The generated charge transfer products can then be used in further reactions. Photo-redox reactions lie at the heart of most photochemical process for solar energy conversion and redox sensitisation of solution phase reactions and electron injection into semiconductors have been widely studied with this application in mind. Although there has been a recent burst of interest in compounds which will inject electrons into semiconductor conduction bands because of their potential use in the dye sensitised solar cells (6.3–6.5, 11.19, 11.20) described in Chap. 7, the process has been of technological importance since the discovery of spectral sensitisation of silver halides in the late eighteen hundreds and the subsequent use of cyanines (e.g. 4.7) and other dyes as irreversible electron injection sensitisers in panchromatic photographic films (see Chap. 11). Photo-redox processes are also important in imaging science, semiconductor photocatalysis (see Chap. 6), and photo-redox based actinometers such as the ferrioxalate (see Chap. 14) and uranyl actinometers [45].

4.4.4 Radical Sensitisers, Quenchers and Traps

The products of redox sensitisation are usually radical ions. Neutral radicals can be generated unimolecularly by homolytic cleavage of an excited state molecule, or by bimolecular homolytic cleavage, the most common example of such being hydrogen abstraction from solvent. Energetically, electron transfer reactions become significantly less favourable as the polarity of the solvent is decreased, whereas the energetics of neutral radical reactions are relatively solvent independent. Photochemical radical initiation processes are important in gas, solution and solid-state radical reactions.

For aqueous phase studies, where, because of solvent polarity, radical ions are of most interest, the Ru(II)trisbipyridyl (6.2)/persulfate reaction pair [50] can be used. For organic solvents, benzophenone (11.5) in the presence of a hydrogen abstractable solvent is a widely studied/used radical photoinitiator. Radical initiators are very important in photopolymerisations, and there are many commercially available photoinitiators (7.1–7.3). Where the molar absorption coefficient of the photo-produced radical is known, then flash photolysis allows a determination of the yield of subsequent radicals.

The essential feature of a radical quencher or trap is the availability of a radical state of low energy which is therefore relatively stable to further reaction. There are a wide range of radical quenchers available commercially as stabilisers, particularly polymer stabilisers. Electron spin resonance (ESR) is the obvious method of radical characterisation but some radical traps can be identified spectrophotometrically. Radical sensitisers, quenchers and traps are discussed in further detail in Chap. 8.

4.5 Photochromism and Molecular Switches

4.5.1 Chromism and Photochromism

Chromism is the reversible change in colour of certain materials upon application of external stimuli such as heat (thermochromism), light (photochromism), electrical current (electrochromism) or solvent polarity (solvatochromism) [24]. In this section we will concentrate on photochromism—light-induced colour changes. These have a variety of actual and potential applications, one of the most important of which is in photochromic ophthalmic lenses which darken in bright sunlight and become colourless in normal light. Photochromism involves a molecular system interconverting between two forms which have different absorption spectra. The process is reversible and the back reaction can either be induced by heat (designated T-type) or photochemically, using light of a different wavelength from the forward process, (P-type). The concept of photochromism is indicated schematically for a unimolecular process in Fig. 4.5, where light absorption by species A (normally absorbing in the UV) produces the longer wavelength absorbing species B through some photoinduced process; it is also possible to have bimolecular photochromic processes.

Fig. 4.5

Absorption spectra of a unimolecular two-state photochromic system. Figure adapted from Ref. [51]

The first reports of photochromism in the scientific literature date back to the middle of the nineteenth century with the observation of bleaching of orange coloured solutions of tetracene in daylight and the regeneration of the coloured solution in the dark [51]. The reaction involves a photodimerisation [52]. The photochromic behavior of tetracene contrasts with that in Fig. 4.5, since the photoproducts absorb at shorter wavelengths than their precursor. This is termed negative photochromism. In addition, the forward reaction is a bimolecular process. A more common scenario is that the initial photochromic species absorbs in the UV and on photolysis produces a coloured photoproduct absorbing in the visible region of the spectrum (positive photochromism), and the process involves interconversion of a single molecule between two chemically distinct forms.

The term photochromism is attributed to the distinguished Israeli scientist Yehuda Hirshberg [51, 53], who correctly identified the importance of chemical transformations in these systems. Some of the earlier literature used the term “phototropy” for the observed colour changes, suggesting that purely physical phenomena are involved [51]. However, it is now recognised that all important photochromic processes involve reversible chemical changes, and the term phototropism is reserved for the effect of light on the growth of plants, which may be directed either towards or away from the sun or other light sources [51]. Interest in photochromism in the early part of last century was rather limited [54], but was stimulated in the 1950s by the potential strategic importance of materials which could undergo reversible changes with light for various applications [55], including photochromic glasses which would darken rapidly following intense light pulses, such as those produced in nuclear explosions. These have been termed optical power-limiting substances [51]. Various reversible organic and inorganic photoprocesses were considered as possible systems for these applications, including formation of triplet excited states of aromatic molecules, isomerisations, electron and atom transfer. Subsequent developments concentrated on non-military uses, and the first serious practical application came with the development by Corning Glass in the U.S.A, of photochromic silicate glasses sensitised by silver halides, modulated by the presence of small amounts of copper(I) salts [55, 56]. The general reaction scheme can be summarised as:

$$ {\text{Ag}}^{ + } + {\text{ X}}^{ - } + h\nu \rightleftarrows {\text{Ag}}^\bullet\, + {\text{ Cl}}^\bullet $$

(4.4)

$$ {\text{Ag}}^{ + } + {\text{ Cu}}^{ + } + h\nu \rightleftarrows {\text{Ag}}^\bullet \, + {\text{ Cu}}^{ 2+ } $$

(4.5)

The silver halide system is similar to that involved in the silver-based photographic process (see Chap. 11), but irreversible formation of photoproducts is inhibited by the fact that the silver halides are present as nanometre sized particles dispersed in a non-conducting silicate matrix. This prevents the permanent photochemical reactions which take place in the photographic system to form the silver based latent images. Work on the silver halide glasses led to the development of the first viable photochromic lenses, which went on the market in the mid-1960s. The lenses have good optical properties, show excellent reversibility for their photochromic processes and reasonable darkening and bleaching times. However, the system involves silicate glass lenses, and in the following decade the ophthalmic market was moving towards plastic lenses [57]. While the silver halide system is excellent for silicate-based glasses, it is less suited for inclusion in the organic polymer systems used in plastic lenses. For this, organic photochromic systems involving thermal back reactions (T-type) are much more suitable [57, 58].

4.5.2 Organic Photochromic Systems

A variety of photochemical processes in organic molecules lead to photochromic changes, including pericyclic reactions, cis–trans (E/Z) isomerisations, intramolecular hydrogen transfer, photodissociation processes and electron transfer [51]. The area has been reviewed extensively [54, 59–62], and some typical examples of photochromic materials are given in Table 4.1 (12.1–12.6). The most important T-type ones for technical and industrial applications in areas such as ophthalmic lenses involve spiropyrans (12.1), spirooxazines (12.2) and naphthopyrans (chromenes, 12.3). In all three cases, light absorption leads to production of a coloured (merocyanine) form, where extended conjugation is achieved through ring opening. The absorption spectra of both the colourless and coloured forms can be modified by appropriate substitution of the aromatic rings. This allows colour tuning to produce the best properties for optical usage. There are a number of factors which need to be controlled, including the transmission (absorption) spectra of the coloured form, the light response, speed of recovery of the colourless form, the number of cycles the system can undergo and the long term stability of the system [51]. While the spiropyrans were some of the first systems to be studied, the spiroxazines show much lower fatigue on extended use [58], and the first commercial plastic photochromic lenses, which were introduced in the 1980s, involved an indolinospironaphthoxazine incorporated in a polycarbonate matrix [57]. More recently, the naphthopyrans have become the commercially most important class of photochromic materials for this type of application [58]. However, they still have some failings in terms of long term applications and there is considerable interest in the development of new photochromic materials involving these cyclisation/ring opening processes.

The way that the photochromic material is incorporated into the lenses is of importance for the commercial application of these materials. This can be achieved by injection-moulding in a thermoplastic or precursor monomer or resin system, surface coating, diffusion into lens surfaces (imbibition) or formation of laminate structures, where a photochromic layer is placed between two halves of the lens structure [58].

Another important type of photochromic reaction involves cis–trans photoisomerisation [63]. With azobenzenes (12.4), the trans (anti) form has a strong absorption, attributed to a π,π ^* transition in the near UV region and a weaker n,π ^* band in the blue region of the spectrum. Upon photoexcitation with light of appropriate wavelengths (~340 nm for the unsubstituted derivative) the π,π ^* band shifts to the blue and the longer wavelength n,π ^* band increases in intensity due to formation of the cis(syn) form. Although photochromism will lead to a photostationary state, up to 90 % of the cis form can be produced. The reverse cis–trans reaction can take place either thermally or by irradiation with longer wavelength light [54, 63, 64]. This possibility of interconverting between two structures using light of different wavelengths is termed photoswitching. The trans isomer of azobenzene is planar but, due to steric hinderance, the cis form is bent. In addition to the colour change, this leads to changes in dipole moment, polarisability and, in the solid state, packing in crystal structures. This will also lead to modifications in the properties of the surrounding medium, which can enhance the applications of photochromic materials. For example, if azobenzenes (or other photochromic materials) are incorporated into a polymeric matrix their photochromic reaction can affect properties, such as shape, refractive index, phase, solubility and surface wettability [65]. This is termed a photoresponsive system. These have a number of important applications which are discussed later.

Reversible trans–cis isomerisations with alkenes (Fig. 4.6) are also relevant for photochromism and photoswitching. With the simple systems, normally only photoinduced processes are involved because of the high energy barrier between the two forms. These alkene-based photoswitches can be useful in molecular devices. With polyenes, both thermal and photochemical processes are possible, and these can be used as P-type and T-type photochromics. A rare, naturally occurring photochromic system involving cis–trans isomerisation process occurs with bacteriorhodopsin, which is found in halobacteria [66]. Its structure and photochemical processes are very similar to the visual pigment rhodopsin present in the retina of the eye. In both cases, the structure involves the polyolefin, retinal, linked to a protein through a Schiff’s base (see Fig. 1.1). With bacteriorhodopsin, photochromism involves interconversion between the all-trans form absorbing at 570 nm and the 13-cis isomer absorbing around 410 nm. The system can be recycled many times without any signs of fatigue and shows excellent long-term stability, which makes it a good candidate for use in optical memories and data processing.

Fig. 4.6

Cis–trans isomerisation and cyclisation in stilbene

With the cis isomer of diarylethenes, a second photochromic process can occur: photocyclisation [61, 67]. In the simplest case, cis-stilbene, the initially formed dihydrophenanthrene is rapidly oxidised to phenanthrene in an irreversible process (Fig. 4.6), making it unsuitable for photochromic applications. However, this can be overcome by replacing the phenyl rings by heterocyclic groups, such as thiophene (12.5). These diarylethenes are important P-type photochromic systems showing good thermal stability, resistance to fatigue, and are important as photo switches. Relatively large spectral shifts are seen between the shorter wavelength absorbing open structure and the long wavelength closed form. The spectral properties can be tuned by introducing substituents into the heterocyclic rings. The structural changes on ring closure affect properties such as fluorescence, refractive index, polarisability and electrical conductivity. A related P-type photochromic system involves the fulgides and fulgimides (12.6). Again, the photochromism involves a colourless open form, sometimes referred to as the E-form, and the product of photocyclisation, termed the C-form [68]. There is an additional photochemical pathway leading to the colourless Z-form. This competing process decreases the efficiency of the photochromic system, but can be minimised by appropriate design of the molecules.

While many other organic photochromic systems exist, the above are the most important types currently used for practical applications.

4.5.3 Three State and Gated Photochromics and Two-Photon Systems

The previous section describes photochromic systems in which interconversion between two forms can be induced by absorption of light. However, more complex scenarios also exist and some have particular practical importance. With 2-(2′,4′-dinitrobenzyl)pyridine (DNBP), photochromism involves phototautomerisation with hydrogen transfer [69, 70]. However, this can either be transferred to the pyridine nitrogen giving the blue NH form or to the oxygen of the nitro group to give the yellow OH form (Fig. 4.7). These can revert thermally or photochemically to the most stable colourless CH form.

Fig. 4.7

Three photochromic forms produced from 2-(2′,4′-dinitrobenzyl)pyridine (DNBP). Figure adapted from Ref. [69]

For certain applications of photochromics, it is useful to be able to convert one or more of the forms reversibly into a stable non-photochromic structure. These systems are termed gated photochromics [51] and are of particular importance for optical data storage. Figure 4.8 shows an example of a gated photochromic involving diarylethenes [71]. According to the Woodward-Hoffmann rules, the photocyclisation is a conrotatory process and is only possible through the antiparallel form. In hydrophobic solvents, such as cyclohexane, the parallel open form is stabilised by hydrogen bonding and cannot photocyclise. However, upon addition of a hydroxylic solvent, such as ethanol, or heating, the hydrogen bonds are broken leading to formation of the antiparallel open form which can undergo the photochromic reaction.

Fig. 4.8

Solvent gated photochromism in a diarylethylene. Reprinted with permission from Irie et al. [71]. Copyright (1992) American Chemical Society

Chromism may also be induced by two separate external stimuli. This is termed dual-mode photochromism [51]. A particularly versatile example involves the flavylium system, the basic structure of anthocyanin dyes. With these, because of the complex acid–base behavior, interconversion between the various coloured species formed can be controlled by the dual application of light and pH changes [72]. It is possible in this way to have a pH gated photochromic system.

With photochromic systems, as with other areas of photochemistry, we are normally using monophotonic processes in which a molecule absorbs one photon. However, it is possible to have two-photon or multi-photon photochromic systems. These have certain attractive properties. Two possibilities exist [51]. In the first (sequential) case, a molecule absorbs one photon to form its excited state. This (or a subsequent species) may then absorb a second photon to give the product:

$$ {\text{A }} + h\nu \to {\text{A}}^{*} $$

(4.6)

$$ {\text{A}}^{*} + h\nu \to {\text{B}} $$

(4.7)

An example of this sequential two-photon photochromism has been reported with a naphthopyran derivative [73]. This has the advantage, when it is used for optical data storage, of non-destructive readout capacity.

In the second case, a molecule simultaneously absorbs two photons via a virtual level to produce the excited state, which is subsequently transformed into the photo-product:

$$ {\text{A }} + { 2}\,h\nu \to {\text{A}}^{*} \to {\text{B}} $$

(4.8)

Since it is only necessary that the sum of the energies of the two photons is sufficient to produce the excited state, the exciting light can be of longer wavelength than the absorption band of A. This means that NIR light can be used, minimising photochemical degradation. In addition, the probability of simultaneous interaction of two photons and one molecule is very low so an intense light source is necessary, typically a pulsed laser, and the effect can be limited optically to a small region of the sample. If the photochromic system is incorporated into a polymeric host this opens the possibility of achieving 3D data storage through focusing the laser at different points in the sample [74].

4.5.4 Some Applications of Photochromic Materials

By far the biggest application of photochromic systems is in ophthalmic lenses. These now normally involve T-type spiroxazine or napthopyran photochromics in thermoplastic polymers. The lens colours under the UV component of sunlight, but not significantly under artificial light, which lacks this part of the spectrum. As the optimal systems involve neutral colours grey or brown, frequently mixtures of photochromics are used [75]. Design of commercial formulations is complicated by the need for the various components to fade and undergo fatigue at the same rate, and there is currently considerable interest in the development of dyes which are intrinsically neutral in colour.

T-type photochromic thermoplastic systems are also finding non-opthalmic specialty applications in areas such as colouring drinks bottles, toys (including dolls which develop suntans) and crash helmet visors for motorcyclists. Photochromic systems are also used in formulations for surface coatings, and have been used for security printing, such as in passports. In addition, they show potential for personal care use, such as in cosmetics and hair dyes. A good description of these applications is given in Ref. [58].

Interesting effects can be produced in textiles by using photochromic colorants. Because of stability problems in processing, these are often either incorporated into a polymer matrix inside textile fibres [76] or microcapsules containing the photochromic material are coated onto textile surfaces [77]. While products, such as T-shirts which change colour in sunlight, are available on the market, at present the development of this area is limited due to difficulties in obtaining cost-efficient, durable products [58].

Photochromic transformations in matrices such as polymers can lead to changes in the bulk properties of the matrix. Such photoresponsive systems can have various applications. We will indicate two of these. If a photochromic system, such as an azobenzene, is incorporated into a liquid crystalline polymer system, photoconversion can lead to changes in the ordering and orientation of the liquid crystalline mesophase [65]. This leads to changes in various physical properties, including the optical anisotropy, which can be used in display and other applications. A second case involves photo-responsive biomaterials [65]. Incorporation of photochromic molecules can be used in areas such as photo-regulation of biological properties, controlled drug release and photo-regulated membrane permeability.

The area of information technology (IT) has been based upon the electronic properties of semiconductors. Gordon Moore, one of the founders of Intel, published an article in 1965 which indicated that the capacity of computer processing will double about every 18 months [78]. This empirical law is still valid, but is reaching its limits, in particular because as electronic memories become smaller, they start to have problems of heating and cross-talk, and there is a need for development of new systems. Three characteristics are required for a memory, the ability to write, read and erase information. Optical (photonic) systems using photochromic materials can achieve these requirements while overcoming many of the problems of limitations of purely electronic systems, since the ultimate data density achievable is limited by the area which can be resolved, which depends upon light wavelength, as discussed in Chap. 1. Photonic systems also have the advantage that they can be multiplexed by using more than one property, e.g. wavelength, polarisation and phase, while memories can be further enhanced using 3D data storage through two-photon absorption [74, 79]. A further possibility is to obtain sub-diffraction limited systems through near-field optics [80]. Until recently, erasable memory systems have tended to use inorganic materials using magneto-optic effects or phase change for data recording. While these may have organic pigments to enhance spectral properties [24], the IT industry had been wary of purely photonic organic systems because of doubts on long-term stability. However, a number of good, stable, low-fatigue photochromic systems have now been developed and show considerable promise for purely optical data storage. The desirable properties of photochromic systems for these applications are good thermal stabilities of the two photochromic forms, fast response, resistance to fatigue, high sensitivity and non-destructive read-out. The P-type photochromics, diarylethenes and fulgides [61, 67, 68, 81], fulfill many of these properties. One limitation of photochromic systems is that reading one photochromic form, either through absorption or emission spectra, can convert it back to the other form. However, as noted above, photochromism also leads to changes on other properties, such as the refractive index of the medium, and this can be used to address the system.

A somewhat different application of P-type photochromics is their use as ‘smart’ receptors in sensing cations, anions and biologically relevant systems [82]. This is based on photoinduced switching between two forms, only one of which is tailored to bind to the analyte through host–guest interactions. The possibility of switching between the two forms provides the attractive potential of reusing these sensors. A more detailed discussion of the general area of optical sensors and probes is given in Chap. 12.

4.5.5 Photoswitches: Molecular Logic, Rotors and Machines

The ideas of molecular memories and data storage described in the previous section can be extended to molecular computing. IT systems are based on logic gates with specific input–output behavior. These typically involve binary systems, where the input can be 0 or 1, and the output is, equally, 0 or 1. Photochromic systems fulfill the requirements of such a two-state system, and have been used in molecular logic devices [83]. These can be extended to applications in more complex logic functions by using a second input, such as addition of a metal ion or a change in pH. Although the area is in its infancy, photochromic systems show excellent possibilities for application in molecular scale computing.

The distinguished physicist Richard Feynman in a famous talk to the American Physical Society entitled “There’s plenty of room at the bottom” [84] issued the challenge that it should be possible to make machines out of molecules. In addition to the intellectual and synthetic challenges of designing and making such systems, they also have potential for applications as pumps and motors in a variety of chemical and biomedical applications. There is now considerable research activity devoted to the use of molecular switches to produce such molecular machines [81, 85–87]. The basic requirement of a molecular machine is that it should involve “an assembly of a discrete number of molecular components (that is, a supramolecular structure) designed to perform specific mechanical movements as a consequence of appropriate external stimuli” [81]. Light is a particular valuable external stimulus [88], and, as shown in Fig. 4.6, photoswitching through cis–trans isomerisation does provide a possible basis for molecular rotor. However, for a true rotor it is necessary to have a unidirectional 360º rotation. This can be achieved by having a chiral photochromic system [86], as indicated in Fig. 4.9.

Fig. 4.9

Schematic view and structure of a molecular motor. Reprinted with permission from Feringa [86]. Copyright (2001) American Chemical Society

This forms the basis for the development of true molecular motors and machines.

4.6 Conclusions

This chapter has discussed some of the most important and commonly encountered photochemical materials, whose properties and subsequent applications are primarily dependent on their absorption and emission characteristics. The most important factors are; (i) the available energy states of a given material and the routes of interconversion between these states and (ii) the excited state deactivation pathways. These factors dictate whether a material will act as a passive absorber, an emitter, or sensitiser. Absorbers, both organic and inorganic, find use in areas such as colorants, sunscreens, paints, pigments and dyes; high molar absorption coefficients are required to produce intense colours, while narrow absorption bands give rise to bright colours. For emitters, a high emission quantum yield in the required medium for the intended use is of obvious importance. The emission quantum yield is dependent on competition with other deactivation routes, while the emission wavelength (and therefore colour) and band structure depend on the relative energy levels of the emitter in any given medium. The emission lifetime is dependent on the probability of the radiative transition, i.e. whether it is ‘allowed’ (typically 10–100 ns) or ‘forbidden’ (μs or longer). The application of efficient emitters in light sources and display technology has been discussed. Excited state and radical sensitisers are useful for a variety of applications, including photodynamic therapy (e.g. singlet oxygen sensitisation, see Chap. 9) and photopolymerisation and device fabrication (see Chap. 13) and examples of the most commonly exploited sensitisation mechanisms have been provided. Photochromism and photochromic materials, including molecular switches, have also been discussed at length. For photochromic materials it is the absorption characteristics of both isomers that are most important for potential applications (change of colour, colourless to coloured or vice versa).

The aim of this chapter was to provide an overview of the most common and useful photochemical materials and their primary photophysical properties that allow for use, or potential use, in a given application. In this vein, we have presented a comprehensive table detailing the most common structures, physical and photophysical properties, and specific applications of each of the classes of photochemical materials discussed.

Table 4.1 A collection of data, structures, characteristics, uses and noteworthy properties of some commonly used photochemical materials