Opening the book…
\( \let\bibref\eqref %make-command_bibref-equivalent-to-eqref\)Using spectroscopy to study molecular properties has two basic modalities. We can investigate changes to the properties of electromagnetic fields once they have passed through matter (for instance through absorption measurements), or we can study light emitted by the matter. Thus, studying emitted radiation is the complementary second half of spectroscopy. However, these two spectroscopic modalities are not as distinct as one might think, but rather two sides of the same coin. The incident and emitted fields are closely coupled, with oscillating charged particles being both the consequence of a driving electromagnetic field and the source of a charge current that re-radiates light. Here we will discuss some of the factors needed to describe emission of light from oscillating charges, starting with classical observations and describing the radiation of electromagnetic fields from a dipole emitter.
In our discussion of the classical driven oscillator, we have seen that a resonant electromagnetic field can drive a bound charged particle to oscillate at the driving frequency. The sinusoidal oscillation of this charge should similarly in the emission of radiation, even after switching off the field, since accelerated charges radiate an electromagnetic field. At distances far from the source, this radiated field is calculated as
\[\boldsymbol{\mathcal{E}}_{\text{rad}}(\mathbf{r},t') = \frac{q}{4\pi \varepsilon_0c^2}\left. \frac{d \hat{r}^2}{dt^2} \right|_{t'} %DOCeq01 %LTeq01 \]
Here \(\mathbf{r}\) is the distance from the source to the observer, \(\hat r\) is the unit vector in that direction, and \(\boldsymbol{\mathcal{E}}\) is evaluated at the retarded time \(t' = t - \left( {r/c} \right)\), to account for the time it takes for the field to propagate at the speed of light from the source to the observer. The emitted radiation are spherical waves radiated into three dimensions, which we will describe for a point dipole emitter in the following section.
We also learned from our driven oscillator model that a driven oscillator will absorb energy from the driving field when it is near resonance. To reach a steady state solution, the energy absorbed must be balanced by energy dissipation. We accounted for this phenomenologically by adding a damping constant \(\gamma \) to the equation of motion. For molecules in condensed matter, one can imagine irreversible processes by which this energy might be dissipated as heat, but for an isolated molecule or atom radiating this energy into an electromagnetic field is far more likely. Although it is possible to formulate the forces on the driven oscillator to properly account for the emission of radiation, the nuances needed to handle that problem are best reviewed elsewhere.\(\bibref{Cite1} %in-text-citation\) We can, however, calculate in a straightforward manner the power radiated into all directions by an oscillator in the absence of damping. This is essentially what one would expect when the damping is weak enough to neglect and when the driving field is strong enough to neglect its depletion due to absorption. In that limit, the radiated power is governed by the Larmor formula
\[\overline{\mathcal{P}}_{\text{rad}} = \frac{{\mu'} ^2}{6\pi \varepsilon_0 c^3} \left| \ddot Q \right|^2 %DOCeq02 %LTeq02 \]
where \(\ddot Q\) is the acceleration of the bound particle. This leads to\(\bibref{Cite2}\bibref{Cite3} %in-text-citation\)
\[\overline{\mathcal{P}}_{\text{rad}} = \frac{{\mu'}^4 E_0^2}{12 \pi \varepsilon_0 m^2 c^3} \frac{\omega^4}{\left( \omega_0^2 - \omega^2 \right)^2 + 4 \gamma^2 \omega^2} %DOCeq03 %LTeq03 \]
As before, \(\mu '\) is the dipole derivative for displacing the charged oscillator, and \(E_0\) is the amplitude of the incident driving field.
With these results in hand, it is possible to calculate the classical scattering cross section (\(\sigma_{\text{sc}}\)), which describes the power radiated by the dipole into all directions, relative to the incident intensity of the electric field driving the oscillator:
\[\sigma_{\text{sc}}(\omega ) = \frac{\overline {\mathcal{P}}_{\text{rad}}}{I_0} %DOCeq04 %LTeq04 \]
With \(I_0 = \tfrac{1}{2}\varepsilon_0 c E_0^2\), we obtain
\[ \sigma_{\text{sc}} ( \omega ) = \frac{{\mu'}^4}{12 \pi \varepsilon_0^2 m^2 c^4} \frac{\omega^4}{\left( \omega_0^2 - \omega ^2 \right)^2 + 4 \gamma^2 \omega ^2} %DOCeq05 %LTeq05 \]
Note that the leading factor in this equation has units of \(\text{m}^2\), while the second term is a unitless lineshape function. If we take \(\mu '\) to be the electron charge \(e\), we can see that this factor is proportional to the square of the so-called classical electron radius \(r_e = e^2 / \left[ 4\pi \varepsilon_0 m_e c^2 \right] \approx 3 \times 10^{- 15} \, \text{m}\). Thus, the magnitude of this scattering cross section is governed by a prefactor that is very small, on the order of \(10^{- 24} \, \text{cm}^2\). In contrast, molecular electronic absorption cross-sections are on the order of \(10^ {- 16} - 10^{- 18} \text{cm}^2\). The prefactor in eq. (8.1.5) gives the scattering cross-section directly for scattered light with frequencies that are significantly greater than the resonance frequency (\(\omega \gg \omega_0\)).
We can use eq. (8.1.5) to describe Raleigh scattering, the non-resonant scattering of light by polarizable molecules when the scattered radiation is at much lower frequencies that the electronic absorption. In that limit, \(\omega \ll \omega_0\), and we can set \(\gamma \to 0\) to give:
\[\sigma_{\text{sc}}(\omega ) = \frac{{\mu'}^4}{12\pi \varepsilon_0^2m^2c^4}\frac{\omega ^4}{\omega_0^4} %DOCeq06 %LTeq06 \]
This expression illustrates the high order \(\omega ^4\) scaling of the efficiency of Raleigh and non-resonant Raman scattering. Raleigh scattering of solar radiation from the atmosphere is considered responsible for the appearance of a blue sky.
The driven oscillator model can also be used to calculate the damping coefficient for classical radiative relaxation, although this requires a more sophisticated accounting for the forces that give rise to the radiative power loss.\(\bibref{Cite1} %in-text-citation\) This results in
\[\gamma_{\text{rad}} = \frac{{\mu'}^2 \omega ^2}{6\pi m \varepsilon_0 c^3} %DOCeq07 %LTeq07 \]
How good are the classical predictions? Using the charge of an electron for \(\mu '\), electron mass, and an oscillation frequency corresponding to visible radiation (\(\omega /2\pi = 10^{15} \, \text{Hz}\)), we find a radiative lifetime of \(\tau_{\text{rad}} = \gamma_{\text{rad}}^{-1} = 4 \times 10^{-9} \, \text{s}\). This is the same order of magnitude as the lifetimes of atomic excited states. As we have noted previously for classical models, this is satisfying in that it gets close to proper values, but this calculation bypasses many details regarding the true nature of atom-field interactions, leading to limited predictive capability. For instance, the quantum calculation of the spontaneous emission rate scales as \(\omega ^3\) as opposed to the \(\omega ^2\) scaling here.
References
| \[\tag {1} \label {Cite1}\] |
Griffiths, D. J. Introduction to Electrodynamics, 3rd ed.; Prentice-Hall, 1999. |
| \[\tag {2} \label {Cite2}\] |
An, K. Fundamentals of Laser Physics; World Scientific, 2023. DOI: 10.1142/13112 |
| \[\tag {2} \label {Cite3}\] |
Stenzel, O. Light-matter interaction: a crash course for students of optics, photonics and materials science; Springer, 2022. DOI: 10.1007/978-3-030-87144-4 |
This page titled 8.1: Classical rate of light emission was last modified on Fri, 05 Jun 2026 01:29:22 GMT and is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Andrei Tokmakoff via source content that was edited to the style and standards of the LibreTexts platform.
The book could not be loaded just now.