HOW SCIENCE WORKS · PHYSICS
AC Stark Shift: How Oscillating Light Moves Atomic Energy Levels
Light can alter an atom’s measured transition frequency even when it is not resonantly absorbed. The oscillating electric field mixes quantum states and produces an intensity- and detuning-dependent energy shift.
See what changes
Read the measurement logic
Find where the model breaks
Read in order for the full explanation, or use a route to enter at the question that matters most.
Scientific note: The equations below describe idealised regimes. Real measurements require calibration, uncertainty analysis and tests of competing mechanisms.
Chapter index
Core idea
Evidence and calculation
Limits and transfer
CHAPTER 1 OF 8
The observation: a resonance moves when light is applied
A probe records one transition frequency with the perturbing beam off and another with it on. The shift can reverse direction when the light is tuned from one side of a resonance to the other.
This is the AC Stark shift, also called the light shift. It is the oscillating-field counterpart of the static Stark effect and follows from coherent coupling, not merely from sample heating.
CHAPTER 2 OF 8
The mechanism: virtual admixture of states
An off-resonant field couples two states without efficiently completing the transition. Their amplitudes mix slightly, changing the eigenenergies of the driven system.
The shift depends on the state’s polarizability, field frequency, intensity and polarisation. A transition moves according to the difference between the shifts of its upper and lower levels.
← Previous · Chapter index · Next →
CHAPTER 3 OF 8
The model: Rabi frequency and detuning
For a far-detuned two-level system, a characteristic angular-frequency shift has magnitude |δ| ≈ |Ω|²/(4|Δ|), where Ω is the Rabi frequency and Δ the detuning. The sign depends on the state and detuning convention.
The approximation requires |Δ| much larger than Ω and the linewidth. Nearby levels, counter-rotating terms and tensor polarizabilities can require a multilevel calculation.
← Previous · Chapter index · Next →
CHAPTER 4 OF 8
A worked estimate
Take Ω/2π = 10 MHz and Δ/2π = 100 MHz. The far-detuned estimate gives |δ|/2π ≈ 10²/(4 × 100) MHz = 0.25 MHz.
Doubling field amplitude doubles Ω and makes the shift four times larger; doubling intensity doubles Ω² and therefore doubles the shift in this regime.
| Change | Ideal far-detuned prediction | Diagnostic |
|---|---|---|
| Intensity ×2 | Shift ×2 | Tests Ω² scaling |
| |Detuning| ×2 | Shift ÷2 | Tests inverse-detuning scaling |
| Detuning sign reversed | Shift direction changes | Separates coherent shift from heating |
| Beam made less uniform | Less inhomogeneous broadening | Checks spatial averaging |
← Previous · Chapter index · Next →
CHAPTER 5 OF 8
How the shift is measured
A 2016 precision study in rubidium vapour measured the resonant frequency while tuning a perturbing laser across a broad range around the D1 line.
A convincing dataset maps shift versus intensity and detuning, reverses or removes the perturbing field, and monitors temperature, power and magnetic-field drift.
← Previous · Chapter index · Next →
CHAPTER 6 OF 8
How alternatives are excluded
Heating can also move or broaden a line, while laser-frequency drift can imitate a slow shift. Beam inhomogeneity produces different shifts across an ensemble and can widen the spectrum.
Interleaved light-on/light-off measurements, independent thermometry, intensity calibration and extrapolation to zero power separate these effects.
← Previous · Chapter index · Next →
CHAPTER 7 OF 8
Where the simple formula fails
Near resonance, absorption and dressed-state splitting become important; the perturbation is no longer a small correction. Multilevel atoms can have scalar, vector and tensor shifts.
Strong fields may also cause ionisation, power broadening or decoherence. A successful model must predict both the line centre and collateral changes such as width and contrast.
← Previous · Chapter index · Next →
CHAPTER 8 OF 8
Why light shifts matter
Optical traps rely on controlled light shifts, while atomic clocks must measure and cancel unwanted differential shifts. Qubit readout and control also use dispersive shifts.
The methodological lesson is that a measuring or trapping field can alter the quantity being measured. Precision requires modelling that backaction rather than treating illumination as passive.