Energy Levels and Molecular Transitions

A molecule absorbing light is not a smooth, continuous process. Every photon a molecule can absorb corresponds to a jump between two discrete, quantized energy levels. Nothing in between is allowed. This single fact, quantization, is the reason spectrochemical analysis works at all. A compound's spectrum is a fingerprint of the exact energy gaps its electrons, bonds, and rotating framework are permitted to have. Understanding why matters immensely in practice. When a chemist runs a UV-Vis scan on a copper(II) complex, or an infrared spectrum on an unknown organic solid, or a rotational microwave spectrum on a gas-phase pollutant, they are reading off transitions between quantized levels. These are electronic, vibrational, and rotational, respectively. The wavelength or wavenumber at which absorption occurs is dictated entirely by $\Delta E = h\nu$ . This connects a measurable optical quantity directly to a molecular energy difference of a very particular size. The big picture is a hierarchy of energy scales. Electronic transitions require the most energy (UV-visible photons, ~150–600 kJ/mol). Vibrational transitions require roughly one-tenth to one-hundredth as much (infrared photons, ~1–40 kJ/mol). Rotational transitions require the least (microwave photons, often under 1 kJ/mol). Because these scales differ by orders of magnitude, they show up in completely different spectral regions. A single electronic transition in a real molecule is typically accompanied by a fine structure of vibrational sub-levels. Each of these carries its own rotational fine structure. This is a nested structure. It is sometimes visible as banded, textured UV-Vis spectra of small gas-phase molecules like I $_2$ vapor. This article builds the quantized-level picture from first principles. It covers all three types of motion. It derives the governing energy formulas with the algebra shown explicitly. It also works through three realistic numerical examples. These are a Bohr-model estimate, a real diatomic vibration, and a rotational spectroscopy calculation. This way, every number here can be checked by hand.

Key takeaways

Molecular energy is quantized: electrons, bond vibrations, and molecular rotation can only take specific, discrete energy values, not a continuum.
A transition between two levels absorbs or emits a photon whose energy exactly equals the energy gap: $\Delta E = h\nu = hc/\lambda$ .
Electronic transitions (UV-Vis) have the largest gaps; vibrational (IR) are roughly 10–100× smaller; rotational (microwave/far-IR) are smaller still.
The vibrational energy ladder is (to first approximation) evenly spaced, $E_v = h\nu_0(v+\tfrac12)$ ; the rotational ladder is not, $E_J = BJ(J+1)$ .
Real spectra show fine structure because each electronic state contains a vibrational manifold, and each vibrational level contains a rotational manifold — three nested quantization scales in one molecule.
Selection rules (change in quantum number, symmetry, dipole moment change) determine which of the many mathematically allowed transitions are actually observed with appreciable intensity.
Key idea: every spectral line is a bookkeeping entry — $\Delta E$ between two quantized states — and spectrochemical analysis is the art of reading that bookkeeping backward to learn structure.

Where this lives today

Spectrochemical instruments built directly on this quantized-level picture are workhorses of modern chemistry, biology, and industry. UV-Vis spectrophotometers measure electronic transitions to quantify everything from protein concentration (absorbance at 280 nm from aromatic residues) to trace metal complexes in environmental water testing; the wavelength of maximum absorbance is a direct readout of an electronic energy gap, as introduced in UV-Vis Chromophores and Electronic Transitions. Fourier-transform infrared (FTIR) spectrometers exploit vibrational energy gaps to identify functional groups in unknown organic compounds, verify polymer composition in quality control, and authenticate pharmaceutical ingredients by their vibrational fingerprint region. Rotational (microwave) spectroscopy, though less common in routine analytical labs, is essential in atmospheric chemistry and radio astronomy — the CO $J=1\to0$ rotational line at 115 GHz (2.6 mm wavelength) used in Example 3 is one of the most important tracers for mapping molecular gas clouds in interstellar space, since it is bright and easily observed even though CO is a minor constituent of those clouds. Raman spectroscopy provides a complementary window onto the same vibrational (and sometimes rotational) energy levels through inelastic light scattering rather than direct absorption, useful for probing symmetric vibrations that are IR-inactive — exactly the homonuclear-diatomic and totally-symmetric-mode cases flagged above as spectroscopically silent in ordinary absorption. Laser-induced fluorescence and time-resolved spectroscopy exploit the vibronic fine structure discussed here to study excited-state dynamics on picosecond-to-femtosecond timescales, foundational to modern photochemistry and photovoltaic materials research. In every one of these techniques, the underlying logic is identical to the one built up in this article: measure a photon energy, and back out a quantized energy gap that reports on structure — a principle that generalizes directly to the broader framework of Absorption and Emission Processes.

Ask the tutor: How do selection rules decide which of these mathematically allowed transitions actually show up with strong intensity in a real spectrum?

The physical picture

A free particle can have any kinetic energy at all. A particle confined — an electron bound to a nucleus, two atoms tied together by a chemical bond, a rigid molecule tumbling about its center of mass — cannot. Confinement forces the allowed wavefunctions to satisfy boundary conditions, and only certain standing-wave-like solutions fit; each solution carries a specific, discrete energy eigenvalue. This is the single mechanism behind all three types of level structure treated here.

Electronic levels. The electrons in an atom or molecule occupy quantized orbitals with energies set by the Coulomb attraction to nuclei, modified by electron-electron repulsion and (in molecules) bonding/antibonding interactions between atomic orbitals. The energy gaps between electronic states are large — comparable to typical bond energies — because they reflect the strong electrostatic forces holding electrons near nuclei. These gaps correspond to photon energies in the ultraviolet and visible region, roughly 150–600 kJ/mol (wavelengths from about 200 nm to 700 nm).

Vibrational levels. Once two atoms are bonded, the bond behaves — for small displacements — like a mass on a spring: a harmonic oscillator. Quantum mechanically this gives evenly spaced energy levels. The "spring constant" here is the bond's force constant, which is much softer than the Coulombic forces binding electrons, so the vibrational energy gaps are far smaller than electronic ones — infrared photons, roughly 1–40 kJ/mol.

Rotational levels. A rigid molecule tumbling in space is modeled as a rigid rotor; angular momentum is quantized, and this produces rotational energy levels that are not evenly spaced — they spread apart quadratically with rotational quantum number $J$ . These gaps are the smallest of the three, typically under 1 kJ/mol, falling in the microwave/far-infrared region. Because $\Delta E_{el} \gg \Delta E_{vib} \gg \Delta E_{rot}$ , a molecule undergoing an electronic transition typically changes its vibrational state simultaneously (vibronic coupling), and each vibrational sub-band is itself split into a comb of rotational lines. In condensed phases (liquids, solutions) collisional broadening usually smears the rotational fine structure into a smooth envelope, which is why solution UV-Vis spectra show broad bumps rather than sharp lines, while gas-phase high-resolution spectra reveal the full rotational comb.

Figure

The three energy scales, drawn to relative scale, and the nested fine-structure they produce

Figure

Harmonic-oscillator vibrational ladder vs. rigid-rotor rotational ladder

The core ideas

The universal bridge between a photon and a molecular energy gap is the Bohr/Planck condition:

\Delta E = E_{upper} - E_{lower} = h\nu = \frac{hc}{\lambda} = hc\bar{\nu}

where:

$\Delta E$ is the energy difference between two quantized molecular states (J, or often expressed per mole in kJ/mol),
$h = 6.62607 \times 10^{-34}\ \text{J}\,\cdot\,\text{s}$ is Planck's constant,
$\nu$ is the photon frequency (Hz = s $^{-1}$ ),
$c = 2.99792 \times 10^{8}\ \text{m/s}$ is the speed of light in vacuum,
$\lambda$ is the photon wavelength,
$\bar\nu = 1/\lambda$ is the wavenumber (cm $^{-1}$ ), the unit of choice in vibrational spectroscopy because it is directly proportional to energy.

Reading it: whichever spectral region is probed — UV-Vis, IR, or microwave — the observed absorption or emission frequency is a direct, linear readout of an energy gap between two quantized states. There is no other way for a photon to be absorbed resonantly: the energy must match exactly (subject to natural, Doppler, and collisional linewidth). For a diatomic molecule modeled as a quantum harmonic oscillator, the vibrational energy levels are

E_v = h\nu_0\left(v + \frac{1}{2}\right), \qquad v = 0, 1, 2, \ldots

where:

$v$ is the vibrational quantum number,
$\nu_0 = \dfrac{1}{2\pi}\sqrt{k/\mu}$ is the classical vibration frequency,
$k$ is the bond force constant (N/m),
$\mu = \dfrac{m_1 m_2}{m_1+m_2}$ is the reduced mass of the two bonded atoms.

Reading it: the ladder is evenly spaced, $\Delta E_{vib} = h\nu_0$ , independent of $v$ — a hallmark of the harmonic approximation. Real bonds are anharmonic and the spacing shrinks slightly as $v$ increases, but the harmonic picture is an excellent zeroth approximation and the one used to interpret ordinary infrared spectra. For a diatomic modeled as a rigid rotor, the rotational energy levels are

E_J = B J(J+1), \qquad B = \frac{h^2}{8\pi^2 I}, \qquad I = \mu r^2

where:

$J = 0, 1, 2,\ldots$ is the rotational quantum number,
$B$ is the rotational constant (J, or often expressed in cm $^{-1}$ as $\tilde B = h/8\pi^2 c I$ ),
$I$ is the moment of inertia of the molecule about its center of mass,
$r$ is the bond length.

Reading it: unlike vibration, the rotational spacings grow with $J$ : the gap from $J$ to $J+1$ is $\Delta E = 2B(J+1)$ , so a rotational spectrum is a series of nearly equally spaced lines in frequency (each new line spaced $2B$ from the last) because $\Delta E$ between adjacent levels increases linearly in $J$ , giving equal frequency spacing $2B/h$ (in ordinary spectroscopic units) even though $E_J$ itself grows quadratically.

Derivation AThe Bohr hydrogen-atom energy levels — the prototype for "quantized electronic energy"

Goal. Show explicitly why an electron bound to a nucleus can only have discrete energies, using the simplest possible model (hydrogen-like atom), and derive $E_n = -\dfrac{Z^2 e^4 m_e}{8\varepsilon_0^2 h^2 n^2}$ .

Step 1 — force balance. Treat the electron as moving in a circular orbit of radius $r$ around a nucleus of charge $+Ze$ . The Coulomb attraction supplies the centripetal force:

\frac{Ze^2}{4\pi\varepsilon_0 r^2} = \frac{m_e v^2}{r}

Rearranging gives the kinetic energy in terms of $r$ :

m_e v^2 = \frac{Ze^2}{4\pi\varepsilon_0 r}

Step 2 — Bohr's quantization postulate. Bohr imposed that angular momentum is quantized in integer multiples of $\hbar = h/2\pi$ :

m_e v r = n\hbar, \qquad n = 1, 2, 3,\ldots

Solving for $v = n\hbar/(m_e r)$ and substituting into the force-balance equation from Step 1:

m_e \left(\frac{n\hbar}{m_e r}\right)^2 = \frac{Ze^2}{4\pi\varepsilon_0 r}

\frac{n^2\hbar^2}{m_e r} = \frac{Ze^2}{4\pi\varepsilon_0}

Step 3 — solve for the allowed radii. Rearranging for $r$ :

r_n = \frac{4\pi\varepsilon_0 n^2 \hbar^2}{Z e^2 m_e}

This shows immediately that only discrete radii $r_1, r_2, r_3,\ldots$ are permitted — one for each integer $n$ — because $n$ can only take integer values by the quantization postulate.

Step 4 — total energy at each allowed radius. The total energy is kinetic plus Coulomb potential energy:

E_n = \frac{1}{2}m_e v^2 - \frac{Ze^2}{4\pi\varepsilon_0 r_n}

Using $m_e v^2 = Ze^2/(4\pi\varepsilon_0 r_n)$ from Step 1, the kinetic term is exactly half the magnitude of the potential term, giving

E_n = \frac{Ze^2}{8\pi\varepsilon_0 r_n} - \frac{Ze^2}{4\pi\varepsilon_0 r_n} = -\frac{Ze^2}{8\pi\varepsilon_0 r_n}

Substituting $r_n$ from Step 3:

E_n = -\frac{Ze^2}{8\pi\varepsilon_0}\cdot\frac{Ze^2 m_e}{4\pi\varepsilon_0 n^2\hbar^2} = -\frac{Z^2 e^4 m_e}{32\pi^2\varepsilon_0^2\hbar^2 n^2}

Writing $\hbar^2 = h^2/4\pi^2$ and simplifying the constants gives the standard form:

E_n = -\frac{Z^2 e^4 m_e}{8\varepsilon_0^2 h^2}\cdot\frac{1}{n^2}

Result. Only integer $n$ is allowed, so only a discrete set $E_1, E_2, E_3,\ldots$ of energies exist; a transition between $n_i$ and $n_f$ absorbs or emits a photon of energy $\Delta E = E_f - E_i$ , exactly the Rydberg-formula structure. This is the historical prototype for "quantized electronic energy levels," and the qualitative lesson — confinement plus a quantization rule forces discreteness — carries over unchanged to multi-electron atoms and molecules (solved instead with the full Schrödinger equation).

Go deeper · Derivation B · 1 minThe rigid-rotor energy levels from the time-independent Schrödinger equation

Derivation BThe rigid-rotor energy levels from the time-independent Schrödinger equation

Goal. Derive $E_J = BJ(J+1)$ for a diatomic molecule rotating freely, starting from the rotational kinetic energy operator, and show why the spacing grows with $J$ .

Step 1 — set up the Hamiltonian. A rigid rotor has no potential energy (rotation costs nothing but kinetic energy at fixed bond length), so the Hamiltonian is purely the rotational kinetic energy operator:

\hat H = \frac{\hat L^2}{2I}

where $\hat L^2$ is the total angular momentum operator and $I = \mu r^2$ is the moment of inertia, with $\mu$ the reduced mass and $r$ the fixed bond length.

Step 2 — angular momentum eigenvalues. The eigenfunctions of $\hat L^2$ are the spherical harmonics $Y_J^{m_J}(\theta,\phi)$ , which satisfy

\hat L^2\, Y_J^{m_J} = J(J+1)\hbar^2\, Y_J^{m_J}, \qquad J = 0, 1, 2,\ldots,\quad m_J = -J,\ldots,+J

This quantization of $L^2$ arises from requiring the wavefunction to be single-valued and finite on the sphere — the spherical analogue of "fitting a standing wave" — and it is this boundary condition, not an ad hoc postulate, that forces $J$ to be a non-negative integer.

Step 3 — energy eigenvalues. Substituting the eigenvalue into the Hamiltonian:

E_J = \frac{J(J+1)\hbar^2}{2I}

Writing $\hbar^2 = h^2/4\pi^2$ :

E_J = \frac{h^2}{8\pi^2 I}J(J+1) \equiv B\,J(J+1), \qquad B \equiv \frac{h^2}{8\pi^2 I}

Each level has degeneracy $2J+1$ (the number of $m_J$ values), which matters for populating levels at thermal equilibrium but does not change the energy formula itself.

Step 4 — spacing between adjacent levels. The transition energy from level $J$ to level $J+1$ (the allowed rotational transition, selection rule $\Delta J = \pm 1$ for a polar rigid rotor) is

\Delta E_{J\to J+1} = B(J+1)(J+2) - BJ(J+1) = B(J+1)\big[(J+2)-J\big] = 2B(J+1)

Result. The spacing grows linearly with $J$ : going from $J=0\to1$ costs $2B$ , from $J=1\to2$ costs $4B$ , from $J=2\to3$ costs $6B$ , and so on. This is fundamentally different from the vibrational ladder (constant spacing $h\nu_0$ ) and is the diagnostic feature that lets a rotational spectrum be told apart instantly from a vibrational one: a rotational absorption spectrum is a series of lines equally spaced in frequency by $2B/h$ , not equally spaced in energy per se, while a harmonic vibrational spectrum has a single fundamental line (in the simplest approximation).

Worked examples

Example 1 (foundational) — Bohr-model electronic transition energy in atomic hydrogen ( $Z=1$ ). Calculate the energy and wavelength of the photon emitted when a hydrogen electron falls from $n=4$ to $n=2$ (part of the Balmer series), using the Bohr formula derived above,

E_n = -\frac{2.180\times10^{-18}\ \text{J}}{n^2}

(the constant $2.180\times10^{-18}$ J is $Z^2e^4m_e/8\varepsilon_0^2h^2$ evaluated for $Z=1$ ; this is the well-known Rydberg energy, $R_H hc$ ). Step 1. Compute $E_4$ and $E_2$ :

E_4 = -\frac{2.180\times10^{-18}}{16} = -1.3625\times10^{-19}\ \text{J}

E_2 = -\frac{2.180\times10^{-18}}{4} = -5.450\times10^{-19}\ \text{J}

Step 2. The photon energy released is

\Delta E = E_2 - E_4 = (-5.450 + 1.3625)\times10^{-19} = -4.0875\times10^{-19}\ \text{J}

The electron loses this much energy (moving down in $n$ ), so the emitted photon carries $|\Delta E| = 4.0875\times10^{-19}$ J. Step 3. Convert to wavelength using $\Delta E = hc/\lambda$ :

\lambda = \frac{hc}{\Delta E} = \frac{(6.626\times10^{-34}\ \text{J}\,\cdot\,\text{s})(2.998\times10^8\ \text{m/s})}{4.0875\times10^{-19}\ \text{J}}

\lambda = \frac{1.9864\times10^{-25}}{4.0875\times10^{-19}} = 4.860\times10^{-7}\ \text{m} = 486.0\ \text{nm}

This is the well-known blue-green H $_\beta$ Balmer line, directly observable in a hydrogen discharge lamp — a good sanity check that the arithmetic is right, since 486 nm is a textbook-tabulated value. Limiting-case check. As $n_i \to \infty$ for a transition landing on $n_f = 2$ , $E_{n_i}\to 0$ and $\Delta E \to |E_2| = 5.450\times10^{-19}$ J, giving $\lambda \to hc/5.450\times10^{-19} = 364.5$ nm — the Balmer series limit (the series converges to a finite shortest wavelength, not zero), matching the known convergence limit of the Balmer series.

Example 2 (applied) — Vibrational force constant of gaseous HCl from its IR fundamental. Gaseous HCl shows a strong infrared fundamental absorption at $\bar\nu = 2886\ \text{cm}^{-1}$ . Find the force constant $k$ of the H–Cl bond and the vibrational quantum spacing in kJ/mol. Step 1 — reduced mass. Using $m_H = 1.008$ u and $m_{Cl} = 34.97$ u (the dominant $^{35}$ Cl isotope):

\mu = \frac{m_H m_{Cl}}{m_H + m_{Cl}} = \frac{(1.008)(34.97)}{1.008+34.97}\ \text{u} = \frac{35.25}{35.98}\ \text{u} = 0.9797\ \text{u}

Converting to kg (1 u $= 1.6605\times10^{-27}$ kg):

\mu = 0.9797 \times 1.6605\times10^{-27}\ \text{kg} = 1.6266\times10^{-27}\ \text{kg}

Step 2 — convert wavenumber to frequency. $\bar\nu = 2886\ \text{cm}^{-1} = 2.886\times10^5\ \text{m}^{-1}$ , so

\nu_0 = c\bar\nu = (2.998\times10^8\ \text{m/s})(2.886\times10^5\ \text{m}^{-1}) = 8.652\times10^{13}\ \text{Hz}

Step 3 — solve for $k$ from $\nu_0 = \frac{1}{2\pi}\sqrt{k/\mu}$ . Squaring and rearranging:

k = \mu (2\pi\nu_0)^2

2\pi\nu_0 = 2\pi(8.652\times10^{13}) = 5.437\times10^{14}\ \text{rad/s}

k = (1.6266\times10^{-27}\ \text{kg})(5.437\times10^{14}\ \text{s}^{-1})^2 = (1.6266\times10^{-27})(2.956\times10^{29})

k = 480.9\ \text{N/m}

This matches the accepted textbook value for HCl (~480–516 N/m depending on source and anharmonic correction), a reasonable check on the arithmetic. Step 4 — vibrational energy quantum in kJ/mol. One quantum is $\Delta E = h\nu_0$ :

\Delta E = (6.626\times10^{-34}\ \text{J}\,\cdot\,\text{s})(8.652\times10^{13}\ \text{s}^{-1}) = 5.734\times10^{-20}\ \text{J per molecule}

Multiplying by Avogadro's number:

\Delta E_{molar} = (5.734\times10^{-20}\ \text{J})(6.022\times10^{23}\ \text{mol}^{-1}) = 3.453\times10^4\ \text{J/mol} = 34.53\ \text{kJ/mol}

Limiting-case check. If the bond were infinitely stiff ( $k\to\infty$ ), $\nu_0\to\infty$ and the vibrational quantum would diverge — consistent with the physical idea that a rigid, unbreakable bond cannot vibrate at any finite, accessible photon energy; conversely as $k\to0$ (a floppy, nearly free rotor-like bond) $\Delta E\to 0$ , recovering a continuum-like limit where quantization becomes practically unobservable.

Example 3 (exam/competition level) — Predicting the rotational spectrum of gas-phase CO and identifying an unknown line. Carbon monoxide, $^{12}$ C $^{16}$ O, has bond length $r_e = 112.8\ \text{pm}$ and is a rigid rotor whose pure rotational absorption spectrum is measured in the microwave region. (a) Compute the rotational constant $B$ in both J and cm $^{-1}$ . (b) Predict the frequency of the $J=2\to3$ transition. (c) A student observes an unassigned absorption line at $\tilde\nu = 11.51\ \text{cm}^{-1}$ — identify which transition it is. Step 1 — reduced mass. Isotope masses: $m_C = 12.000$ u, $m_O = 15.995$ u.

\mu = \frac{(12.000)(15.995)}{12.000+15.995}\ \text{u} = \frac{191.94}{27.995}\ \text{u} = 6.8563\ \text{u}

\mu = 6.8563 \times 1.6605\times10^{-27}\ \text{kg} = 1.1385\times10^{-26}\ \text{kg}

Step 2 — moment of inertia. $r_e = 112.8\ \text{pm} = 1.128\times10^{-10}$ m.

I = \mu r_e^2 = (1.1385\times10^{-26}\ \text{kg})(1.128\times10^{-10}\ \text{m})^2

I = (1.1385\times10^{-26})(1.2724\times10^{-20}) = 1.4487\times10^{-46}\ \text{kg}\,\cdot\,\text{m}^2

Step 3 — rotational constant.

B = \frac{h^2}{8\pi^2 I} = \frac{(6.626\times10^{-34})^2}{8\pi^2(1.4487\times10^{-46})}

Numerator: $(6.626\times10^{-34})^2 = 4.3904\times10^{-67}$ . Denominator: $8\pi^2 = 78.96$ , so $78.96 \times 1.4487\times10^{-46} = 1.1441\times10^{-44}$ .

B = \frac{4.3904\times10^{-67}}{1.1441\times10^{-44}} = 3.837\times10^{-23}\ \text{J}

Converting to cm $^{-1}$ (divide by $hc$ , with $hc = 1.986\times10^{-23}$ J·cm):

\tilde B = \frac{3.837\times10^{-23}\ \text{J}}{1.986\times10^{-23}\ \text{J}\,\cdot\,\text{cm}} = 1.932\ \text{cm}^{-1}

This matches the well-known literature value for CO, $\tilde B \approx 1.9313\ \text{cm}^{-1}$ . Step 4 — the $J=2\to3$ line. Using $\Delta E_{J\to J+1} = 2B(J+1)$ derived earlier, with $J=2$ :

\Delta \tilde\nu = 2\tilde B(J+1) = 2(1.932)(3) = 11.59\ \text{cm}^{-1}

Step 5 — identify the unknown line. The observed line is at $11.51\ \text{cm}^{-1}$ . Using $\Delta\tilde\nu = 2\tilde B(J+1)$ and solving for $J$ :

J + 1 = \frac{\Delta\tilde\nu}{2\tilde B} = \frac{11.51}{2(1.932)} = \frac{11.51}{3.864} = 2.979 \approx 3.0

So $J+1 = 3$ , meaning $J = 2$ : the observed line is the $J=2\to J=3$ transition (small deviation from the predicted 11.59 cm $^{-1}$ is consistent with centrifugal distortion, a real correction beyond the rigid-rotor approximation, which slightly lowers high- $J$ transition energies below the ideal rigid-rotor prediction). Limiting-case check. As $J\to0$ the spacing $2B(J+1)\to 2B = 3.86\ \text{cm}^{-1}$ , the smallest possible rotational quantum for CO — the $J=0\to1$ transition — consistent with the known microwave-region CO absorption at 3.845 cm $^{-1}$ (115.27 GHz), a widely used astrophysical tracer line.

Limiting cases, scope & common traps

Situation	What happens	Why
$v \to$ large (high vibrational quantum number)	Harmonic ladder breaks down; levels converge and eventually the bond dissociates	Real bonds are anharmonic (Morse potential); the $E_v = h\nu_0(v+\tfrac12)$ formula is only the low- $v$ limit
$J \to$ large (high rotational quantum number)	Spacing $2B(J+1)$ over-predicts the true energy; centrifugal distortion softens the ladder	Bond stretches slightly under rotation, increasing $I$ and lowering $B$ effectively
Molecule in solution vs. gas phase	Rotational fine structure disappears into a smooth envelope	Frequent solvent collisions interrupt free rotation (collisional/pressure broadening)
$k \to \infty$ (infinitely rigid bond)	Vibrational quantum $\to \infty$ ; no vibration observable at accessible photon energies	$\nu_0 \propto \sqrt{k/\mu}$ diverges
Symmetric linear molecules (e.g. CO $_2$ , N $_2$ )	Some vibrational modes are IR-inactive despite being real molecular motions	No net change in dipole moment for that mode — selection rule fails, not the quantization
Homonuclear diatomics (N $_2$ , O $_2$ , H $_2$ )	No permanent-dipole pure rotational spectrum at all	A rotating dipole is required to couple to the oscillating electric field of light; zero dipole moment means zero rotational transition intensity, even though the levels themselves still exist

Common trapCommon misconception: "quantized" means "widely separated."

Students sometimes assume that because energy is quantized, the levels must be far apart — visibly discrete like rungs on a ladder. In reality, rotational levels in ordinary molecules are separated by only fractions of a kJ/mol, so close together that thermal energy at room temperature ( $RT \approx 2.5$ kJ/mol) populates dozens of them simultaneously. Quantization refers to the existence of discrete allowed values, not to any particular spacing — the spacing itself ranges from enormous (electronic, tens of eV) to minuscule (rotational, fractions of meV), and this size difference, not the mere fact of discreteness, is what places electronic, vibrational, and rotational spectroscopy in entirely separate regions of the Regions of the Electromagnetic Spectrum.

Common trapCommon misconception: every mathematically allowed

\Delta E

produces an observed spectral line.

The Bohr and rigid-rotor formulas generate an infinite ladder of levels, and in principle a photon could bridge any two of them. Selection rules — conservation of angular momentum, symmetry of the transition dipole moment, and (for vibration) the requirement that the dipole moment change during the motion — forbid or drastically weaken most of these transitions. A homonuclear diatomic like N $_2$ has perfectly well-defined quantized rotational and vibrational levels, yet shows no ordinary IR or microwave absorption spectrum at all, because it has no dipole moment to couple to the light's oscillating field. Predicting level energies and predicting which lines actually appear are two separate calculations.

The chemist behind it

The quantitative bridge between spectral lines and atomic structure is inseparable from Niels Bohr, who in 1913 proposed that an electron in a hydrogen atom could occupy only certain stationary orbits, each with a fixed, quantized energy, and that light was absorbed or emitted only when the electron jumped between them. Bohr was a Danish physicist working at a moment when spectroscopists already possessed exquisitely precise measurements of the hydrogen spectral series (Balmer, Lyman, Paschen) but no theoretical account of why those particular wavelengths, and no others, appeared. Bohr's model — later shown by Schrödinger's wave mechanics to be an approximation valid for one-electron systems, but conceptually exactly right about quantization — reproduced the empirical Rydberg formula essentially perfectly and gave physicists their first mechanistic picture of why atomic and molecular spectra are lines, not continuous bands. The extension of the same quantization logic to molecular vibration and rotation came somewhat later, built on the full machinery of quantum mechanics developed by Heisenberg, Schrödinger, and Born in the mid-1920s, and applied specifically to molecular spectroscopy by researchers such as Robert Mulliken (electronic structure and molecular orbital theory) whose systematic notation and interpretation of vibronic and rotational fine structure in molecular spectra remains the standard framework used in every modern spectroscopy textbook.

Origins · 1 minNiels Bohr and how this was discovered

Niels Bohr

1885–1962 · Copenhagen, working within the emerging quantum theory of matter, corresponding closely with Rutherford's nuclear model and Planck's quantum hypothesis

History

- 1885 — Balmer publishes the empirical formula fitting the visible hydrogen spectral lines, with no physical explanation.

- 1900 — Planck introduces energy quantization to explain blackbody radiation,

E = h\nu

- 1913 — Bohr proposes quantized electron orbits in the hydrogen atom, explaining the Balmer/Rydberg formula mechanistically.

- 1926 — Schrödinger's wave equation supersedes the Bohr model, correctly generalizing quantized energy levels to all bound quantum systems, including molecular vibration and rotation.

- 1930s — Systematic rotational-vibrational (rovibrational) spectroscopy of diatomic and small polyatomic gases establishes the rigid-rotor/harmonic-oscillator framework as the standard interpretive model still taught today.

Why do electronic transitions need so much more energy than vibrational or rotational transitions?

Electronic transitions rearrange the distribution of electrons around nuclei against strong Coulombic forces, energy scales comparable to full chemical bond strengths (electron volts, hundreds of kJ/mol). Vibrational transitions only stretch or bend an already-formed bond slightly, governed by a much softer "spring constant" restoring force, typically tens of kJ/mol at most. Rotational transitions cost the least because they involve no distortion of bond length or electron distribution at all — merely a change in how fast the whole rigid molecular framework tumbles, governed only by its moment of inertia, which is a comparatively weak quantum-mechanical constraint. This hierarchy — electronic ≫ vibrational ≫ rotational — is why the three types of transitions appear in completely separate spectral regions: UV-visible, infrared, and microwave respectively.

If energy levels are discrete, why do real absorption spectra show broad peaks instead of infinitely sharp lines?

Real spectral lines are broadened by several physical mechanisms layered on top of the underlying discreteness. Natural linewidth arises from the Heisenberg uncertainty principle applied to a finite excited-state lifetime. Doppler broadening comes from the thermal motion of gas-phase molecules shifting the apparent absorption frequency. In condensed phases (liquids, solids), collisional and solvent interactions perturb energy levels moment to moment, smearing a sharp line into a broad envelope, and rotational fine structure that would be resolved in the gas phase is often washed out entirely in solution spectra. The underlying levels remain perfectly discrete; only their experimental manifestation is blurred by these physical broadening mechanisms.

How does the harmonic oscillator model break down for real chemical bonds?

A real bond cannot stretch indefinitely without breaking, so its true potential energy curve (the Morse potential) rises steeply at short internuclear distance, flattens toward a dissociation limit at large distance, and only resembles a symmetric parabola near the potential minimum. The harmonic oscillator approximation, giving perfectly evenly spaced levels $E_v = h\nu_0(v+\tfrac12)$ , is accurate only for the lowest few vibrational quantum numbers. At higher $v$ , real vibrational levels converge and crowd together as the bond approaches dissociation, and overtone/combination bands appear at frequencies slightly different from simple integer multiples of the fundamental — a real, measurable anharmonicity correction routinely reported in high-resolution vibrational spectroscopy.

Why is the rotational constant B so much smaller than typical vibrational or electronic energy gaps?

$B = h^2/(8\pi^2 I)$ , and molecular moments of inertia $I = \mu r^2$ are comparatively large numbers built from atomic masses (roughly $10^{-27}$ – $10^{-26}$ kg) and bond lengths on the order of $10^{-10}$ m. Squaring the bond length and multiplying by reduced mass gives $I$ around $10^{-46}$ – $10^{-45}$ kg·m $^2$ , and dividing $h^2$ (an extremely small number, since $h$ itself is $6.6\times10^{-34}$ J·s) by $8\pi^2 I$ yields an energy on the order of $10^{-23}$ – $10^{-22}$ J — orders of magnitude below typical vibrational quanta ( $10^{-20}$ J) or electronic gaps ( $10^{-19}$ – $10^{-18}$ J). This is a direct numerical consequence of nuclei being far more massive than the electrons responsible for bonding and electronic structure.

Do all molecules show rotational fine structure in their electronic and vibrational spectra?

Only in the gas phase, and only when instrumental resolution is high enough to resolve individual rotational lines — this is routinely seen for small, rigid gas-phase molecules like HCl, CO, or I $_2$ vapor, whose rotational spacings ( $B \sim$ a few cm $^{-1}$ or less) are well within reach of modern high-resolution spectrometers. In solution or condensed phases, frequent molecular collisions interrupt free rotation on a timescale far shorter than the rotational period, so no discrete rotational structure survives — only a smooth, broadened vibrational or electronic envelope is observed, which is why routine solution-phase UV-Vis and IR spectra look like broad humps rather than combs of sharp lines.

What is the practical difference between reading an IR spectrum in wavenumbers versus reading a UV-Vis spectrum in wavelength?

Wavenumber ( $\tilde\nu = 1/\lambda$ , units cm $^{-1}$ ) is directly proportional to energy ( $\Delta E = hc\tilde\nu$ ), so equal spacings in wavenumber correspond to equal spacings in energy — convenient for vibrational spectroscopy, where force constants and quantum spacings are naturally expressed in energy terms. Wavelength (nm) is not linearly related to energy ( $\Delta E = hc/\lambda$ , an inverse relationship), so UV-Vis spectra reported in nanometers require converting through the reciprocal relationship to compare energies directly; this convention split is historical but consistently applied across the two techniques, and forgetting the reciprocal (rather than linear) relationship between wavelength and energy is a common source of unit-conversion errors.

How does this quantized-level framework connect to why atoms and molecules absorb only specific colors of light?

Every color a molecule can absorb corresponds to a photon whose energy exactly matches some allowed $\Delta E$ between two quantized states — no other photon energy is resonant with the system, so it passes through largely unabsorbed. This is precisely why a copper(II) sulfate solution looks blue (it absorbs orange/red light, corresponding to a specific d-d electronic energy gap) while a chlorophyll solution looks green (it absorbs red and blue light, leaving green transmitted). The specific set of allowed electronic energy gaps — set by the identity of the metal, its oxidation state, the ligands present, and the overall molecular symmetry — is what gives every compound its characteristic color and absorption spectrum, the central subject of UV-Vis Chromophores and Electronic Transitions.

Energy Levels and Molecular Transitions

Where this lives today

The physical picture

The core ideas

Worked examples

Limiting cases, scope & common traps

The chemist behind it

Before this

Where this leads