As early as the 1800s, it was known that excited-state atoms (that is, atoms which have absorbed energy via light, combustion, electricity, etc.) return to the ground state by emitting photons (or fluorescing) at specific frequencies unique to each atomic species. This was problematic for the physicists of the time, whose models predicted that atoms should be able to have any energy, and thus emit light at any wavelength(not to mention that the lifetime of an atom would be a few billionths of a second in those models used before the rise of quantum physics)!
The set of allowed emission wavelengths for a given atom is known as its spectrum, emission spectrum, or fluorescence spectrum. Shown above is the visible part of the spectrum for hydrogen known as the Balmer series. From reddest to bluest, the Balmer series has spectral lines at 656.3nm, 486.1nm, 434.0nm, and 397.0nm. Shown below is the full hydrogen spectrum on a logarithmic scale - the Balmer series is denoted by Ba-\alpha
In 1913, Niels Bohr developed a new model for the hydrogen atom based on a unique assumption: that the discrete nature of atomic spectra indicates that atoms cannot have just any energy: atoms must have defined, unique energy levels (corresponding to different values of n in the figure). This is incredibly counterintuitive - like claiming a battery can be fully charged or empty, but no amount in between!
Bohr's model was largely ignored by his contemporaries for this very reason. But its results were incredible: the gaps between energy levels matched up exactly with the energies (hv in the figure, where h is Planck's constant and v is the frequency of the emitted light) of the hydrogen fluorescence spectrum.
In Bohr's model, fluorescence occurs when atoms make an instantaneous transition - a quantum leap - from a higher energy state to a lower one without ever occupying an energy in between. The emitted light "carries off" the energy lost by the atom in the process!
The figure displays the mechanics of fluorescence as we understand it thanks to Bohr's model: an atom is excited to a higher energy state (labeled by S2), then after some fluorescence lifetime, it spontaneously decays to a lower state (either S1 or S0 here), emitting a photon with energy equal to the difference in energy between its initial and final state.
We've already discussed how emission spectra are unique for each element, helping us to identify unknown materials. This is particularly helpful in astrophysics: we can use the wavelengths of light from distant galaxies to determine their elemental composition. Moreover, we can use the emission spectrum for an atom to extrapolate its allowed energies.
But what if the spectral wavelengths of a material could give insight on that material's physical properties? It turns out this is exactly the case in molecular spectroscopy: Molecules, like atoms, are only allowed to have specific energies. Unlike atoms, they have some additional restrictions: they can only vibrate and rotate (the below figure models a molecule as two rotating masses on a vi vibrating spring) at certain frequencies! This is also very counterintuitive - on our scales, it would be analogous like claiming a car's wheels could only move at a few allowed speeds.
We can determine what these allowed frequencies are by - you guessed it - measuring the emission spectrum of the molecule, just like for an atom.
Molecules have a much more complicated energy level structure than atoms. In this figure, we see a portion of the energy structure for diatomic iodine (I2), with the vertical axis representing the energy stored in the molecule and the horizontal axis representing the distance between its two nuclei. Each curve represents an allowed electronic energy state and each horizontal line represents an allowed vibrational mode (i.e., a frequency at which the molecule is allowed to vibrate).
Molecules in higher vibrational states have a higher maximum speed as they oscillate, meaning they can stretch (or compress) further from the equilibrium internuclear distance (the point where the electronic energy curve is at its lowest). The point at which a molecule can't stretch any further is known as a turning point - shown on the figure where the vibrational mode lines intersect their respective electronic energy curve.
There's a lot going on in this graph, so let's visualize a simple example: an iodine molecule in its ground state - that is, in the n'' electronic state and the 0 vibrational level. That molecule bounces back and forth between turning points (indeed, there is still some vibration even at the lowest possible energy!) - compressing until the distance between its atoms reaches the inner turning point (the leftmost intersection of the n'' curve and the v=0 line), expanding until it reaches the outer turning point (the rightmost intersection of the n'' curve and the v=0 line), then repeating the process. The total energy of the molecule never deviates, only changing forms - completely in the form of potential energy at the turning points, and completely in the form of kinetic energy at the equilibrium internuclear distance. Thus the total energy is equal to the height of the turning points.
Now let's talk about transitions (illustrated by the arrows in the figure). Iodine in the ground state strongly absorbs 532nm light (a shade of green), making a simultaneous transition to the n' electronic state and the 9 vibrational state. Like atoms, molecules in excited electronic states spontaneously decay to the lowest electronic state, emitting light in the process. But molecules do not have to transition to the lowest vibrational state - when decaying, they may transition to any vibrational state lower than their initial state. Critically, decaying to a higher vibrational state will release less energy than decaying to a lower vibrational state! Thus, many fluorescence wavelengths are observed for a single electronic transition - ranging from the same shade of green used to excite the molecule (the leftmost fluorescence arrow in the figure) to dull shades of red (the rightmost fluorescence arrow).
The spacing between adjacent fluorescence wavelengths for a molecule allows us to determine the allowed vibrational frequencies for that molecule - which can be used in a couple helpful calculations to determine important molecular properties and reproduce the molecular potential curves depicted in the figure. This second result takes some difficult calculations, but is of utmost importance in spectroscopic theory.
As it turns out, if we know the potential curves for a molecule, we know everything there is to know.
Balmer Series and Hydrogen Spectrum: Wikipedia contributors. (2025, March 1). History of spectroscopy. In Wikipedia, The Free Encyclopedia. Retrieved 20:36, April 15, 2025, from https://en.wikipedia.org/w/index.php?title=History_of_spectroscopy&oldid=1278238922
Bohr Atomic Model: Wikipedia contributors. (2025, April 3). Bohr model. In Wikipedia, The Free Encyclopedia. Retrieved 20:37, April 15, 2025, from https://en.wikipedia.org/w/index.php?title=Bohr_model&oldid=1283764177
Rotating Spring Molecular Model: Contributors (2014, November 5). Example: The Moment of Inertia of a Diatomic Molecule. In Projects and Practices in Physics. Retrieved 20:40, April 15, 2025 from https://msuperl.org/wikis/pcubed/doku.php?id=183_notes:examples:the_moment_of_inertia_of_a_diatomic_molecule&rev=1415220029
Iodine Energy Diagram: Bauer, Thomas J (2010, December 2). Laser Induced Fluorescence in Iodine. Retrieved 20:45, April 15, 2025 from http://academics.wellesley.edu/Physics/Tbauer/LIF/