1.7.1 The Width of Spectral Lines Throughout this chapter we have drawn spectral absorptions and emissions Hot as infinitely sharp lines but as more or less broad peaks; we have seen that one reason for this is that the mechanical slits in spectrometers are not infinitely narrow and thus allow a range of frequencies, rather than a single frequency, to fall on the detector, hence blurring the pattern. While im- Provements in spectrometer design can improve the resolving power of an instrument, however, there is nonetheless a minimum width inherent in any atomic or molecular transition—the natural line width—beyond which no instrument, however superior, will show a sharpening, This width arises essentially because the energy levels of atomic and molecular systems arenot precisely determined, but have a certain fuzziness or imprecision, Sev- eral factors contribute to this. 1. Collision broadening. Atoms or molecules in liquid and gaseous phases are in continual motion and collide frequently with each other. These collisions inevitably cause some deformation of the particles and hence perturb, to some extent, the energies of at least the outer electrons in cach. This immediately gives a possible explanation for the width of visible and ultra-violet spectral lines, since these deal largely with tran- sitions between outer electronic shells. Equally vibrational and rotational spectra are broadened since collisions interfere with these motions too. In general, molecular interactions are more severe in liquids than in gases, and gas-phase spectra usually exhibit sharper lines than those of the corresponding liquid. In the case of solids, the motions of the particles are more limited in extent and less random in direction, so that solid-phase spectra are often sharp but show evidence of interactions by the splitting of lines into two or more components. 2. Doppler broadening. Again in liquids and gases the motion of the parti- cles causes their absorption and emission frequencies to show a Doppler shift; since the motion is random in a given sample, shifts to both high and low frequencies occur and hence the spectral line is broadened. In general, for liquids collision broadening is the most important factor, whereas for gases, where collision broadening is less pronounced, the Doppler effect often determines the natural line width 3. Heisenberg uncertainty principle. Even in an isolated, stationary mol- ecule or atom the energy levels are not infinitely sharp, due to the operation of a fundamental and very important principle, the Uncer- tainty Principle of Heisenberg. In effect this says that, if'a system exists in an energy state for a limited time dr seconds, then the energy of that state will be uncertain (fuzzy) to an extent 5£ where SE x bt = h/2n = 10-4 Fs (1.10) where h is again Planck's constant. Thus we see that the lowest energy state of a system is sharply defined since, left to itself, the system will remain in that state for an infinite time; thus 5t = 00, and dE = 0. But, for example, the lifetime of an excited electronic state is usually only about 10°"s, which gives a value for SE of about 10724/10-* = 10°28 J. A transition between this state and the ground state will thus have an energy uncertainty of 5E, and a corresponding uncertainty in the associated radiation frequency of 6E/h, which we can write as: bE 1 Figs (L11) =F * Sahat * TabiThus for our example of an excited electronic state lifetime of 10-* s, 6v = 10° Hz. This apparently large uncertainty is, in fact, small com- pared with the usual radiation frequency of such transitions, 10'*-10° Hz, and so the natural line width is said to be small; in fact, the apparent widths of electronic transitions are far more dependent on collision and Doppler broadening than on energy uncertainties. On the other hand an excited electron spin state may exist for some 10°" 's which, from Eq, (1.11), leads to a frequency uncertainty of some 10” Hz for a transition. This, compared with the usual frequency of such transitions, 10°-10° Hz, represents a very broad transition indeed, and here the Heisenberg uncertainty relation is by far the most important effect.