Chemiluminescence in Analytical Chemistry

3

The Nature of Chemiluminescent Reactions

Stephen G. Schulman and Joanna M. Schulman

University of Florida, Gainesville, Florida

Yener Rakiciog˘lu

Istanbul Technical University, Istanbul, Turkey

1.	INTRODUCTION		68
2.	CHEMICAL CONSIDERATIONS		68
3.	PHOTOPHYSICAL CONSIDERATIONS		70
4.	STRUCTURAL AND ENVIRONMENTAL CONSIDERATIONS		72
5.	THE INFLUENCE OF MOLECULAR STRUCTURE ON
	FLUORESCENCE		72
6.	THE INFLUENCE OF THE ENVIRONMENT ON THE
	FLUORESCENCE SPECTRUM		73
	6.1	Solvent Effects	73
	6.2	Quenching of Fluorescence	74
	6.3	Energy Transfer	75
7.	CHEMILUMINESCENCE REACTION PARAMETERS		76
8.	KINETIC CONSIDERATIONS		77
	8.1	Type I Chemiluminescence Reaction	77
	8.2	Type II Chemiluminescence Reaction	77
9.	SUMMARY		79

67

68	Schulman et al.

1. INTRODUCTION

Chemiluminescence (CL) is the emission of light by molecules that are electronically excited by virtue of their precursors’ participation in a highly exergonic chemical reaction, almost invariably an oxidation reaction. Such reactions occur in vitro and in vivo and in the latter case they are mediated by enzymes and the resulting chemiluminescence is called bioluminescence. Chemiluminescence is actually fluorescence (spin-allowed, radiative deactivation of the lowest excited singlet state) from the electronically excited product of the chemical reaction. It differs from the process normally called fluorescence (photofluorescence) only in that in fluorescence, the light-emitting molecules are electronically excited by the absorption of light whereas in CL, no external light source is employed. Although, in principle, CL could arise from the lowest triplet state as a phosphorescent emission, the long inherent lifetimes of triplet states and their near-total quenching by photoreaction in fluid solutions has, to our knowledge, precluded such a phenomenon from ever having been observed.

In this chapter will be considered:

1.The means by which the light-emitting molecule is formed in its emissive state

2.The nonradiative processes immediately leading to and competing with light emission

3.The kinetics of CL reactions

This discussion will be confined to strictly chemical reactions and will not include electrogenerated chemiluminescence.

2. CHEMICAL CONSIDERATIONS

A sine qua non for CL to occur is that the precursor(s) of the light-emitting species must participate in a reaction that releases a considerable amount of energy. For visible emission (say 400–750 nm in wavelength) 40–70 kcal/mol is required. Actually, the enthalpies of reaction have to be slightly greater than this owing to energy losses by thermal relaxation in the ground (after emission) and lowest excited singlet (after excitation) states of the fluorophore. Normally, only certain oxidation-reduction reactions generate this much energy. In addition, at least some of the energy produced must be channeled into a reaction pathway, in which at least one of the upper vibrational levels of the reactants, probably corresponding to the transition state of the initial reaction, has the same energy and a comparable geometrical structure as an upper vibrational level of the lowest excited singlet state of a potentially emissive product of the reaction. Higher excited singlet states, although of consequence in the generation of photofluorescence,

The Nature of Chemiluminescent Reactions

69

need not be considered here because their energies are so high, relative to the thermal energies of the reactants, that they would be impossible to populate, even in a very exergonic chemical reaction. The geometric identity or near-identity of the interconverting species ensures that most of the energy of activation is channeled into free energy of activation rather than being wasted as entropy of activation.

Invariably, there will also be a pathway for ‘‘dark’’ reaction. The dark reaction may lead directly to the ground, electronic-state species that also results from the fluorescence of the excited product, but this is not necessarily so. It is possible to have a competing dark reaction that leads directly to ground-state products different from that produced by fluorescence. If ∆Ha is the sum of the enthalpies of activation for all dark reaction competing with the chemiluminescent pathway, whose enthalpy of activation is ∆H*a , CL will be probable when ∆H*a ∆Ha. This occurs when the lowest excited singlet state of the flurophore has the same geometrical configuration as the ground electronic state of the reactants at lower energy than when the dark products in the ground electronic state have the same configuration as the reactants in their ground electronic state. This is illustrated in Figure 1. Often, catalysts, which are usually transition metal

Figure 1 Relative positions of the potential energy (E) surfaces of the electronic states involved in a hypothetical chemiluminescent reaction as a function of internuclear separation (r). P and P* represent the ground and lowest electronically excited singlet states of the product of the reaction, respectively. R represents the ground electronic state of the reactant. ∆H is the enthalpy of the ‘‘dark’’ reaction while ∆Ha is its enthalpy of activation.

∆H*a is the enthalpy of activation of the photoreaction. hv denotes the emission of chemiluminescence.

70	Schulman et al.

ions in chemiluminescent reactions and enzymes in bioluminescent reactions, are employed to make light-emitting reactions proceed at a convenient rate. This is accomplished by the catalyst forming a transient complex with the reagents, whose ∆H*a is lower than that of the uncatalyzed reaction. The catalyst is usually a co-oxidant; i.e., it assists electron transfer. The catalysis of the luminol– hydrogen peroxide reaction by Co(II) and of the luciferin-oxygen reaction by the enzyme luciferase are important examples.

3. PHOTOPHYSICAL CONSIDERATIONS

Subsequent to the formation of the potentially CL molecule in its lowest excited singlet state, a series of events carry the excited molecule down to its ground electronic state. Since the electronically excited molecule is initially also vibrationally excited, rapid, stepwise (10 13–10 12 s) thermal deactivation of the excited molecule, called thermal or vibrational relaxation, in which the molecule loses vibrational energy by inelastic collisions with the solvent, occurs. The process of vibrational relaxation carries the molecule to the lowest vibrational level of the lowest excited singlet state. Certain molecules may return radiationlessly all the way to the ground electronic state in the same time frame by thermal deactivation, a process known as internal conversion. However, in some molecules, for a variety of reasons, the return from the lowest vibrational level of the lowest excited singlet state to the lowest vibrational level of the ground state by internal conversion and vibrational relaxation is forbidden (i.e., of low probability or long duration). In these molecules return to the ground electronic state occurs by one of two alternative pathways, the simpler being the direct emission of ultraviolet or visible radiation whose frequency or wavelength is governed by the energy gap between the lowest excited singlet state and the ground electronic state. The radiative transition between excited and ground states of the same spin multiplicity occurs in a time frame of 10 11–10 7 s after excitation, and is called fluorescence. Owing to the fact that the ground electronic state of a molecule has several vibrational levels associated with it, fluorescence emission does not occur at a single wavelength, but rather over a range of wavelengths corresponding to several vibrational transitions as components of a single electronic transition.

Several processes may compete with fluorescence for deactivation of the lowest excited singlet state. As a result only a fraction of the molecules formed in the lowest excited singlet state, φf, actually fluoresce. φf is called the quantum yield or fluorescence efficiency. It is usually a fraction but may be unity in some exceptional cases and is related to the probabilities (rate constants) of fluorescence (kf ) and competitive processes (kd ) by

	φf	kf	(1)
		kf kd

The Nature of Chemiluminescent Reactions

71

Thus, the greater the numbers or rates of processes competing with fluorescence for deactivation of the lowest excited singlet state, the lower the value of φf . The quantum yield of fluorescence is important in determining how intense chemiluminescence can be for a particular reaction.

Another important property of fluorescing molecules is the lifetime of the lowest excited singlet state (τf ). If the mean rate of fluorescence is the number of fluorescence events per unit of time, the mean lifetime of the excited state is the reciprocal rate, or the mean time per fluorescence event. The quantum yield of fluorescence and the lifetime of the excited state are related by

τf φf τN

(2)

where τN is called the natural lifetime of the excited state and represents the lifetime the fluorescing molecule would have if fluorescence was the sole pathway for deactivating the lowest excited singlet state. While τf and φf are determined largely by the kinds and rates of processes competing with fluorescence, τN is a function only of molecular structure.

The second pathway for deactivation of the lowest excited singlet state competes temporally with fluorescence and is called intersystem crossing from the lowest excited singlet to the lowest triplet state. Intersystem crossing entails a change in spin angular momentum, which, classically, violates the law of conservation of angular momentum. Although it is about a million times less probable than the corresponding singlet-singlet radiationless process, internal conversion, its rate is comparable to that of the spin allowed radiative transition (fluorescence). Subsequent to intersystem crossing, the molecules populating the lowest triplet state undergo vibrational relaxation to the lowest vibrational level of the lowest triplet state. Molecules in the lowest triplet state can return to the ground state radiationlessly by triplet-singlet intersystem crossing or by the emission of light. The emission of light accompanying the transition from the triplet state to the ground singlet state is also a forbidden transition, is characterized by a long duration (10 4–10 s), and is called phosphorescence. Because triplet states are so long-lived, radiationless chemical and physical processes in fluid solution compete effectively with phosphorescence for deactivation of the lowest excited triplet state. Except for the shorter-lived phosphorescences, collisional deactivation by solvent molecules, quenching by paramagnetic species (e.g., oxygen), photochemical reactions, and certain other processes preclude the observation of phosphorescence in fluid media unless the potentially phosphorescent molecules are protected from their environment by inclusion in micelles. Owing to the need for diffusion of the reactants in a chemiluminescent reaction into contact with each other, CL requires a fluid environment for its observation. Hence, phosphorescence has never been observed in a CL reaction. Although CL has a long duration (several seconds to several hours), this is a function of the slow reaction kinetics rather than the decay characteristics of the fluorescence. Fluorescent decay lasts only nanoseconds.

72	Schulman et al.

4.STRUCTURAL AND ENVIRONMENTAL CONSIDERATIONS

Molecules that enter into CL reactions are generally reduced species that can be easily oxidized. Molecules containing amino and hydroxy groups fall into this category, as do polycyclic aromatic ring systems. Under strongly alkaline conditions, hydroxy groups and even some arylamino or arylamido groups can be deprotonated making them even more susceptible to oxidation. On the other hand, electron-withdrawing groups tend to stabilize electronic charge and make oxidation more difficult in molecules to which they are affixed.

From the chemical point of view, the solvent in which the CL experiment is carried out can have a dramatic influence on the efficiency of the CL reaction as solvation can alter the shapes, the depths, and the densities of the vibrational states of the potential surfaces representing the ground states of products and reactants and the lowest excited singlet state of the potential fluorophore. The alteration of the intersections of these potential energy surfaces can affect the enthalpies of reaction and the enthalpies of activation for dark and lumigenic reactions. In some cases, these changes will favor CL (if ∆H*a decreases relative to ∆Ha) and in some cases, they will make it thermodynamically unfavorable for CL to occur.

Other influences of the structure and the environment are manifest in the rates of processes competing with fluorescence for deactivation of the lowest excited singlet state. These are the processes and properties that influence the fluorescence process and will be discussed briefly here.

5.THE INFLUENCE OF MOLECULAR STRUCTURE ON FLUORESCENCE

Among other factors, the quantum yield of fluorescence determines the intensity of light emission in a CL. This, as well as the position in the spectrum occupied by the fluorescence band, is largely a function of the molecular structure.

Fluorescence is most often observed in highly conjugated or aromatic organic molecules with rigid molecular skeletons. The less vibrational and rotational freedom that the molecules has, the greater is the probability that the energy gap between the lowest excited singlet state and the ground electronic state will be large so that fluorescence will predominate over nonradiative deactivation and assure a high fluorescence efficiency.

Aromatic molecules containing freely rotating substituents usually tend to fluoresce less intensely than those without these substituents. This results from the introduction into each electronic state of rotational and vibrational substates by the exocyclic substituents. At very low temperatures, the quantum yields of

The Nature of Chemiluminescent Reactions

73

fluorescence tend to be greater than at ambient temperatures as a result of restricted vibrational and rotational freedom and, consequently, a lower efficiency of internal conversion. It is, thus, likely that if appropriate solvents that remain fluid at low temperatures can be used to conduct CL experiments, higher light yields will result. There is a tendency for many substituents to increase quantum yields of fluorescence. This arises from the increased rate of radiative decay accompanying extension of an aromatic system by substitution with a strongly interacting group such as ENH2 or EOH. However, with certain substituents, the fluorescence quantum yields of aromatic molecules are diminished. This is especially so in the case of substitution by heavy atoms such as bromium and iodine, six-membered heterocyclic rings (e.g., quinoline), and other groups having sp2 hybridized nonbonding electrons. Each of these substituents has the ability to cause mixing of the spin and orbital electronic motions of the aromatic system. Spin-orbital coupling destroys the concept of molecular spin as a well-defined property of the molecule and thereby enhances the probability or rate of singlettriplet intersystem crossing. This process favors population of the lowest triplet state at the expense of the lowest excited singlet state and thus decreases the fluorescence quantum yield. Consequently nitro compounds, bromo and iodo derivatives, aldehydes, ketones, and N-heterocyclics tend to fluoresce very weakly or not at all and are not likely to function as the fluorophores in CL reactions.

Molecular structure can have a profound effect on the position in the spectrum where fluorescence occurs, as well as on its intensity. It can be shown by quantum mechanics that the more extended a conjugated system is, the smaller will be the separation in energy between the ground state and the lowest excited singlet state. This is evident in the fact that benzene, naphthalene, and anthracene, having one, two, and three rings, fluoresce maximally at 262 nm, 320 nm, and 379 nm, respectively.

Similarly, the affixment of conjugate substituents onto aromatic systems extends the conjugation of the latter and causes fluorescence maxima of the substituted derivatives to lie at wavelengths longer than those of the parent compound. Hence, the fluorescence of aniline lies at 340 nm while that of the parent hydrocarbon benzene lies at 262 nm.

6.THE INFLUENCE OF THE ENVIRONMENT ON THE FLUORESCENCE SPECTRUM

6.1 Solvent Effects

Although the capacity for fluorescence is primarily a function of molecular structure, the solvent of a potentially fluorescing molecule can have a dramatic effect on the fluorescence.

74	Schulman et al.

Solvent interactions with solute molecules are mainly electrostatic and it is usually the differences between the electrostatic stabilization energies of ground and excited states that contribute to the relative intensities and spectral positions of fluorescence in different solvents.

The changes in π electron distribution that take place upon transition from the lowest excited singlet state to the ground state during fluorescence cause changes in the dipolar and hydrogen-bonding properties of the solute. If the solute is more polar in the excited state than in the ground state, fluorescence will occur at longer wavelengths in a polar solvent than in a nonpolar solvent because the more polar solvent will stabilize the excited state relative to the ground state. Moreover, the fact that photoluminescence originates from an excited state that is in equilibrium with its solvent cage and terminates in a ground state that is not causes the fluorescence to lie at longer wavelengths the more polar or more strongly hydrogen-bonding the solvent.

In molecules having atoms with trigonally hybridized nonbonded electron pairs (e.g., carbonyl compounds and certain N-heterocyclics), the lowest excited singlet state is formed by promoting a lone-pair electron to a vacant π orbital (this is called an n, π* state). These molecules tend to show very little fluorescence in aprotic solvents such as aliphatic hydrocarbons because the n, π* excited singlet state is efficiently deactivated by intersystem crossing. However, in protic solvents such as water or ethanol, these molecules become fluorescent. This results from the destabilization of the lowest singlet n, π* state by hydrogen bonding. If this interaction is sufficiently strong, the fluorescent π, π* state drops below the n, π* state allowing intense fluorescence. Quinoline, for example, fluoresces in water but not in cyclohexane.

Solvents containing atoms of high atomic number (e.g., alkyl iodides or bromides) also have a substantial effect on the intensity of fluorescence of solute molecules. Atoms of high atomic number in the solvent cage of the solute molecule enhance spin-orbital coupling in the lowest excited singlet state of the solute. This favors the radiationless population of the lowest triplet state at the expense of the lowest excited singlet state. Thus in heavy-atom solvents, all other things being equal, fluorescence is always less intense than in solvents of low molecular weight.

6.2 Quenching of Fluorescence

Fluorescence may be decreased or completely eliminated by interactions with other chemical species. This phenomenon is called quenching of fluorescence. Obviously, if the fluorescence of a fluorophore generated in a CL reaction is quenched the observation of chemiluminescence will be precluded.

Two kinds of quenching are distinguished. In static quenching, interaction between the potentially fluorescent molecule and the quencher takes place in the

The Nature of Chemiluminescent Reactions

75

ground state, forming a nonfluorescent complex. The efficiency of quenching is governed by the formation constant of the complex as well as by the concentration of the quencher. The quenching of the fluorescence of salicylic acid by Cu(II) is an example.

In dynamic quenching (or diffusional quenching) the quenching species and the potentially fluorescent molecule react during the lifetime of the excited state of the latter. The efficiency of dynamic quenching depends upon the viscosity of the solution, the lifetime of the excited state (τo) of the luminescent species, and the concentration of the quencher [Q]. This is summarized in the SternVolmer equation:

	φ/φo	1	(3)
		1 kQ τo [Q]

where kQ is the rate constant for encounters between quencher and potentially luminescing species and φ and φo are the quantum yields of fluorescence in the absence and presence of concentration [Q] of the quencher, respectively.

kQ is typical of diffusion-controlled reactions ( 1010 M 1 s 1) while τo for a fluorescent molecule is typically 10 8 s or less. Hence, kQ τo 102 M 1 and for dynamic quenching to be observed (say 1% quenching), [Q] must be greater than or equal to 10 4 M.

6.3 Energy Transfer

Energy transfer entails the excitation of a molecule that during the lifetime of the excited state passes its excitation energy to another molecule. The loss of excitation energy from the initial excited species (the donor) results in quenching of the luminescence of the energy donor and may result in luminescence from the energy recipient (acceptor), which becomes excited in the process.

Energy transfer can occur by either of two acceptor-concentration-depen- dent processes. In the resonance excitation transfer mechanism or dipole (Fo¨rster) mechanism, the donor and acceptor molecules are not in contact with one another and may be separated by as much as 10 nm (although transfer distances closer to 1 nm are more common). In the classical sense, the excited energy donor molecule may be thought of as a transmitting antenna that creates an electrical field in its vicinity. Potential acceptor molecules within the range of this electrical field function as receiving antennae and absorb energy from the field resulting in their electronic excitation.

The rate of resonance energy transfer decreases with the sixth power of the distance between the donor and acceptor dipoles according to

		1		Ro		6
	kET					(4)
		τD		R