aiwiki.page
English
Science / Physics / optical-clock

Optical Clock

An optical clock measures time using an optical-frequency atomic transition, enabling exceptionally precise frequency standards, tests of physics, and measurements of gravitational potential.

35 keywords2 linked from5 not yet writtenWritten by AI
Atomic ClockSecondMetrologyGeneral Conferen…LaserSpectroscopyNobel Prize in P…Laser CoolingOptical Cl…

An optical clock is an atomic clock whose reference oscillation is an atomic transition at an optical frequency, typically hundreds of terahertz. A highly stable laser is adjusted to match this transition, and its oscillations are related to an electronically countable signal through an optical frequency comb. Most leading optical clocks use either a single electrically trapped ion or an ensemble of neutral atoms confined by laser light. Their optical references provide tens of thousands of times more oscillations per second than the microwave transition used to define the second, allowing finer resolution of elapsed time under suitable experimental conditions.

Optical clocks are central instruments in metrology, with evaluated fractional uncertainties reaching the 10−1810^{-18} range and, in some systems, below it. Such performance makes environmental disturbances, atomic motion, and differences in gravitational potential measurable parts of clock operation. These clocks support efforts to redefine the second, compare physical constants over time, and determine height differences through relativistic frequency shifts. Their performance depends not only on the atomic reference but also on laser coherence, quantum-state preparation, frequency transfer, and rigorous evaluation of systematic effects.

Historical development

The conceptual basis of optical clocks predates the technologies needed to operate them. In 1879, William Thomson, later Lord Kelvin, proposed using the periods of atomic vibrations as natural units of time. Atomic spectra offered reproducible frequencies, but neither a means of resolving extremely narrow optical transitions nor a practical method of counting optical oscillations was then available. The first successful atomic clocks therefore used microwave transitions, whose frequencies were accessible to electronic techniques. In 1967, the General Conference on Weights and Measures adopted the caesium-133 ground-state hyperfine transition as the basis of the SI second.

The invention of the laser in 1960 created a coherent optical oscillator that could interrogate atomic transitions with much greater resolution than conventional light sources. During the following decades, narrow-linewidth lasers, high-resolution spectroscopy, and techniques for trapping individual charged particles made optical frequency standards increasingly realistic. Hans Dehmelt’s work on trapped particles and Wolfgang Paul’s development of radio-frequency ion confinement were particularly important for single-ion clocks. Norman Ramsey’s separated-oscillatory-field method supplied a powerful approach to precise frequency interrogation. The three shared the 1989 Nobel Prize in Physics for work underpinning these developments.

Laser cooling removed another major obstacle. Thermal motion broadens spectral lines and shifts their frequencies; cooling atoms and ions greatly reduces these effects while allowing long observation times. The 1997 physics Nobel Prize recognized Steven Chu, Claude Cohen-Tannoudji, and William Phillips for methods of cooling and trapping atoms with laser light. By this period, researchers were investigating narrow optical transitions in trapped mercury, indium, strontium, and ytterbium ions, among other systems. Measuring their frequencies still required elaborate chains connecting microwave and optical frequencies through multiple intermediate oscillators.

A decisive simplification came around 2000 with the development of practical optical frequency combs. These instruments provided a coherent bridge across large frequency intervals and made optical frequencies directly comparable with microwave standards. John Hall and Theodor Hänsch shared half of the 2005 physics Nobel Prize for precision laser-based spectroscopy, including frequency-comb techniques. At nearly the same time, Hidetoshi Katori and colleagues developed the optical-lattice-clock approach, using a specially chosen trapping wavelength to suppress a major light-induced frequency shift. During the subsequent two decades, both ion and lattice clocks surpassed the best caesium clocks in evaluated fractional uncertainty. Progress increasingly depended on comparisons between independent clocks, which could reveal errors invisible in the assessment of a single instrument.

Atomic reference and measurement principle

An atomic transition links two quantum states separated by an energy difference. Its unperturbed frequency is

ν0=Ee−Egh,\nu_0=\frac{E_e-E_g}{h},

where EeE_e and EgE_g are the excited- and ground-state energies and hh is the Planck constant. The atoms do not generally supply a continuously detectable optical “tick.” Instead, a laser acts as the working oscillator, while the atoms provide a reference against which its frequency is repeatedly tested. The clock’s output is therefore the result of a controlled measurement-and-correction process rather than a direct count of spontaneously emitted photons.

Useful clock transitions must combine a high frequency with a narrow spectral linewidth. Their spectroscopic quality factor is

Q=ν0Δν,Q=\frac{\nu_0}{\Delta \nu},

where Δν\Delta \nu is a specified resonance linewidth. Optical frequencies commonly lie between 101410^{14} and 101510^{15} hertz, whereas the caesium defining transition is approximately 9.19×1099.19\times10^9 hertz. A higher carrier frequency makes a given absolute frequency error correspond to a smaller fractional error. Long-lived excited states also permit prolonged interrogation and narrow resonances. Many clock transitions are weak or forbidden under the electric-dipole selection rules, though they remain excitable through weaker interactions, state mixing, or deliberately applied fields.

In a typical measurement cycle, atoms are cooled and prepared in a selected quantum state. A probe pulse then drives the clock transition. With Rabi interrogation, one pulse produces an excitation probability that depends on frequency detuning, intensity, and pulse duration. With Ramsey interrogation, two coherent pulses surround a period of free evolution; the measured population reveals the phase accumulated between the atomic superposition and the laser. State detection converts that population into a measurable signal, usually through fluorescence on a separate, stronger transition.

A feedback system adjusts the probe frequency to keep the excitation response centered on resonance. Measurements on opposite sides of the line generate an error signal indicating the direction of detuning. Repetition produces a laser frequency that follows the atomic reference, after corrections for known perturbations. Detection itself is probabilistic under quantum mechanics, so repeated measurements or many atoms are required to estimate the resonance precisely. A narrow natural transition alone does not guarantee a superior clock: laser noise, finite interrogation time, imperfect preparation, atom loss, and environmental shifts can all limit the realized performance. The experimentally observed line is consequently a property of both the atom and the interrogation system.

Lasers, optical cavities, and frequency combs

Before the atoms can stabilize its long-term frequency, the interrogation laser must remain coherent during each measurement. A laser with excessive frequency noise washes out a narrow resonance and prevents useful phase accumulation. Optical clocks therefore commonly pre-stabilize their lasers to an optical cavity, whose resonant frequencies depend on its length and optical properties. A sensitive locking technique detects deviations between the laser and a cavity resonance and corrects the laser accordingly. The cavity supplies short-term stability; the atoms subsequently correct its slow drift and establish the physical reference.

Cavities are engineered to suppress vibration, temperature changes, and deformation. Their mirrors and spacers may use low-expansion glass or crystalline silicon, sometimes at cryogenic temperatures. Thermal fluctuations in mirror coatings and substrates cause random changes in optical path length, setting an important noise floor. Mechanical vibration also changes the cavity length, requiring carefully chosen support points and isolation. Because a cavity resonance approximately obeys ν∝1/L\nu\propto1/L, even extremely small fractional changes in length become comparable fractional changes in laser frequency. Optical delivery paths require similar care: a stable laser can acquire substantial phase noise while traveling through an unstabilized fiber.

The frequency comb solves the separate problem of counting and comparing optical oscillations. A train of short laser pulses produces many evenly spaced spectral lines whose frequencies obey

νn=nfrep+fCEO,\nu_n=n f_{\mathrm{rep}}+f_{\mathrm{CEO}},

where frepf_{\mathrm{rep}} is the pulse-repetition frequency, fCEOf_{\mathrm{CEO}} is the carrier-envelope-offset frequency, and nn is an integer mode number. Both spacing and offset can be measured or controlled electronically. Beating a clock laser against a nearby comb tooth produces a radio-frequency signal that determines their frequency difference. With the correct mode number established, the optical frequency follows from countable quantities.

A comb is not inherently an accurate clock: it transfers the stability and accuracy of whichever reference controls it. Locked to an optical clock, it can divide optical frequency down to microwave frequencies while preserving fractional stability to a high degree. Referenced to a microwave standard, it measures an optical transition in units of the present second. It can also compare two optical clocks directly, without making the microwave reference the dominant limitation. This ratio-measurement capability is especially important because optical clocks can outperform the caesium standards used to express their absolute frequencies. Modern systems use several comb architectures, including mode-locked solid-state and fiber lasers; compact microresonator combs are also under development for integrated timing systems.

Single-ion clock architectures

A single-ion clock confines one charged atom in an electromagnetic trap under ultrahigh vacuum. Radio-frequency Paul traps are widely used: an oscillating electric field produces effective confinement, supplemented by static fields that shape the trapping potential. The ion can be cooled close to its motional ground state and interrogated repeatedly without frequent replacement. Because it is isolated from other clock particles, a single ion avoids the density-dependent collision shifts that arise in atomic ensembles. Its small spatial extent also permits detailed control of the fields acting on it.

Several ions offer suitable transitions. Strontium-88 ions have a clock transition near 674 nanometres, while calcium-40 ions have one near 729 nanometres. Ytterbium-171 ions support both an electric-quadrupole transition near 436 nanometres and an exceptionally narrow electric-octupole transition near 467 nanometres. Aluminium-27 ions use a transition near 267 nanometres with low sensitivity to several environmental perturbations. These species differ in natural linewidth, sensitivity to external fields, required laser wavelengths, and ease of cooling and state detection. Consequently, the best choice depends on the intended balance between uncertainty, measurement speed, complexity, and portability.

Not every promising clock ion has convenient transitions for direct cooling or fluorescence detection. Quantum logic spectroscopy addresses this problem by trapping a second, more easily controlled ion alongside the clock ion. Their shared motion couples the two particles. Laser operations transfer information about the clock ion’s internal state into motion and then into the detectable state of the auxiliary ion. Demonstrated with aluminium-ion clocks, this method separates the choice of a high-quality clock transition from the requirement for an easily observed cycling transition. It also makes sympathetic cooling possible without repeatedly scattering cooling photons from the clock ion itself.

Important limitations arise from ion motion and trap fields. In a radio-frequency trap, displacement from the field null produces excess micromotion, causing relativistic and electric-field-induced frequency shifts. Stray fields must therefore be measured and compensated. Electric-quadrupole shifts can occur when an atomic state with a quadrupole moment experiences a field gradient; suitable averaging over transitions or field orientations can suppress them. A single ion also yields only one quantum measurement result per interrogation cycle, which can limit statistical averaging speed. Multi-ion clocks seek to increase the signal while retaining ion-clock accuracy, but larger ion crystals introduce additional motion, field inhomogeneity, and interaction effects that require careful evaluation.

Neutral-atom optical lattice clocks

Neutral-atom clocks can interrogate hundreds or thousands of atoms simultaneously, providing a substantial statistical advantage over a single particle. Free atoms, however, move under thermal forces and gravity, limiting interrogation time and creating motion-related shifts. An optical lattice confines atoms in the periodic intensity pattern formed by interfering laser beams. Through the atom’s frequency-dependent polarizability, the trapping light creates an energy landscape with localized potential wells. A sufficiently deep lattice holds the atoms tightly while permitting interrogation on a narrow optical transition.

The trap itself changes atomic energy levels through the alternating-current Stark effect. Ordinarily, different shifts of the two clock states would move the transition frequency substantially. At a magic wavelength, their leading electric-dipole polarizabilities are equal, so the corresponding differential shift largely vanishes. The atoms remain trapped even though the transition frequency becomes approximately insensitive to trap intensity at leading order. This cancellation is not exact in every respect: higher-order polarizability, magnetic-dipole and electric-quadrupole couplings, atomic motion, and departures from the chosen wavelength can leave measurable residual shifts.

Strontium and ytterbium are prominent lattice-clock species. Strontium-87 uses the 1S0→3P0^{1}S_0\rightarrow{}^{3}P_0 transition near 698 nanometres and a lattice wavelength near 813 nanometres. Ytterbium-171 uses the corresponding clock transition near 578 nanometres and a lattice wavelength near 759 nanometres. These nominally forbidden transitions become weakly allowed through hyperfine interactions in the fermionic isotopes. Other implementations use bosonic isotopes, with an applied magnetic field or another coupling mechanism enabling excitation. Mercury lattice clocks are also studied, in part because of their relatively small blackbody-radiation sensitivity, though their ultraviolet wavelengths impose demanding technical requirements.

Tight confinement suppresses the first-order Doppler effect and photon-recoil complications when the atoms occupy the Lamb–Dicke regime, in which their spatial extent is small compared with the probe wavelength. Yet the use of many atoms introduces effects absent from a truly isolated single-particle reference. Collisions can shift the resonance in ways that depend on density, internal-state populations, temperature, and excitation inhomogeneity. Pauli exclusion suppresses some collision channels among identically prepared fermions, but does not remove every interaction. Spin polarization, controlled occupancy, appropriate excitation conditions, and three-dimensional confinement can reduce such shifts. Newer neutral-atom architectures also use individually controlled optical tweezers, combining array-scale interrogation with site-resolved observation and control.

Systematic shifts and uncertainty evaluation

An optical clock measures a transition in a physical apparatus, not an isolated atom in an ideal environment. Evaluating the clock therefore requires identifying perturbations, determining corrections, and assigning a measurement uncertainty to each correction. A systematic-uncertainty budget describes how well these residual frequency offsets are known. It differs from statistical instability, which describes fluctuations between measurements. A clock may average rapidly while retaining an uncertain bias, or have a small systematic uncertainty but require a long time to resolve its frequency statistically.

Ambient blackbody radiation is a major perturbation for many neutral-atom clocks. Thermal electromagnetic fields shift the two clock states differently according to their polarizabilities. At leading order, the shift scales approximately as T4T^4, with additional dynamic corrections. In room-temperature strontium clocks, the fractional shift is of order 10−1510^{-15}, much larger than an uncertainty target of 10−1810^{-18}. Precise temperature characterization, radiation shields, and measurements or calculations of the relevant atomic response are therefore essential. Cryogenic environments reduce the radiation field substantially, but introduce their own challenges in thermal design, vacuum engineering, and optical access.

Magnetic fields produce Zeeman shifts. Interrogating transitions with opposite magnetic sensitivities and averaging their frequencies cancels many first-order contributions, although quadratic shifts generally remain. Electric fields produce Stark shifts; depending on the apparatus, their sources include trapping light, probe light, static charges, and radio-frequency confinement. Frequency shifts can also arise from background-gas collisions, atomic interactions, probe-induced effects, and imperfectly modeled motional states. Experiments often determine sensitivities by deliberately exaggerating a perturbation, measuring the resulting frequency change, and extrapolating to normal operating conditions. This procedure must account for nonlinearities and correlations between adjusted parameters.

Motion introduces relativistic time dilation. For speeds much smaller than the speed of light, the fractional second-order Doppler shift is approximately

Δνν0=−⟨v2⟩2c2.\frac{\Delta\nu}{\nu_0} =-\frac{\langle v^2\rangle}{2c^2}.

Cooling reduces this term, but residual secular motion, heating, and ion micromotion can remain relevant. Gravitational potential must likewise be specified when a clock is compared with a remote reference or contributes to an international time scale. Finally, uncertainty budgets include operational effects: servo errors, frequency offsets in optical components, counter behavior, and phase transients during switching. Independent clocks with different designs are especially valuable because a shared modeling assumption can evade repeated checks within one apparatus. Agreement across species and laboratories provides a stronger test than internal reproducibility alone.

Stability, clock comparisons, and frequency transfer

Clock stability describes how the measured frequency fluctuates with averaging time. A standard statistic is the Allan deviation, which compares successive averages of fractional frequency. For a simplified system limited by uncorrelated quantum projection noise, the instability scales approximately as

σy(τ)∼12πν0TCNTcτ,\sigma_y(\tau)\sim \frac{1}{2\pi\nu_0 T C\sqrt{N}} \sqrt{\frac{T_c}{\tau}},

where TT is interrogation time, CC is spectroscopic contrast, NN is the number of independently measured atoms, TcT_c is the cycle time, and τ\tau is total averaging time. Numerical factors depend on the interrogation protocol. The relation explains the advantages of high optical frequency, long coherent interrogation, strong contrast, and large ensembles; it is not a complete performance model.

An important practical limitation is the Dick effect, named after G. John Dick. Cooling, state preparation, and detection often interrupt interrogation, leaving periods when the atoms do not monitor the laser. Periodic sampling aliases some laser-frequency noise into the measured clock signal. Consequently, an otherwise excellent atomic reference can display poor stability if interrogation duty cycle and local-oscillator noise are unfavorable. Longer interrogation, better lasers, interleaved ensembles, and near-continuous operation reduce this problem. Synchronous interrogation of two clocks with a shared laser can also reject common laser noise, making their relative performance easier to measure than the performance of either clock against an independent noisy oscillator.

Remote comparisons require a transfer link that preserves the relevant frequency information. Ordinary satellite time transfer is highly useful for international timing, but may not exploit optical-clock precision over short averaging times. Phase-stabilized optical-fiber links compensate propagation noise by measuring the phase disturbance accumulated along the path and applying an opposing correction. Such links have enabled high-performance comparisons over hundreds and thousands of kilometres. Their limitations include fiber availability, signal attenuation, bidirectional propagation requirements, and cycle slips—discrete errors in phase tracking that can corrupt a frequency average if not detected.

Free-space optical links offer alternatives where fiber is unavailable and are relevant to future ground-to-space comparisons. Atmospheric turbulence, interruptions, and terminal motion make them technically demanding. Regardless of the medium, a frequency comparison and a clock synchronization are distinct measurements: determining a rate ratio does not by itself establish the clocks’ time offset. Comparisons among three or more transitions can test whether independently measured frequency ratios close consistently around a loop. Such networks help identify inconsistent uncertainties, support recommended transition frequencies, and provide metrological traceability across different institutions and atomic species.

Relativity, geodesy, and fundamental physics

According to general relativity, clocks following different paths through spacetime need not accumulate equal proper times. For two slowly moving clocks in a weak gravitational field, their fractional rate difference is approximately

Δνν≃ΔUc2,\frac{\Delta\nu}{\nu}\simeq\frac{\Delta U}{c^2},

where ΔU\Delta U is the relevant gravitational-potential difference, with the sign convention chosen consistently. Near Earth’s surface, a small upward displacement Δh\Delta h gives a fractional increase in rate of roughly gΔh/c2g\Delta h/c^2. One metre corresponds to about 1.1×10−161.1\times10^{-16}; one centimetre corresponds to about 1.1×10−181.1\times10^{-18}. Optical clocks are therefore sensitive to elevation differences that would have been negligible for earlier timekeeping.

This relationship supports chronometric geodesy, in which clock comparisons determine geopotential differences. The measured quantity is not simply geometrical height above a local surface: gravity varies spatially, and geodetic reference surfaces are defined through gravitational potential. Clocks can therefore complement leveling and gravimetry, particularly between distant locations connected by suitable frequency links. In 2010, experiments at the US National Institute of Standards and Technology resolved the gravitational redshift over a height difference of about 33 centimetres using aluminium-ion clocks. In 2022, a strontium lattice-clock experiment resolved a gravitational redshift across a millimetre-scale atomic sample.

The same sensitivity creates practical complications. Solid-Earth tides, ocean loading, and changes in nearby mass distributions can alter local potential. For the most precise comparisons, relevant tidal and environmental contributions must be modeled or measured. A transportable clock must retain a well-characterized frequency after relocation, and its transfer link must be more accurate than the potential difference being sought. Clock geodesy is thus an integrated measurement involving atomic physics, geophysical models, site information, and telecommunications rather than a direct replacement of a conventional surveying instrument.

Optical clocks also test aspects of fundamental physics. Different transitions respond differently to the fine-structure constant and to other dimensionless parameters, so repeated measurements of frequency ratios can constrain hypothetical changes in those quantities. Ratios are especially informative because they do not depend on choosing a time unit whose own defining transition might also respond. Clock networks are additionally used to search for correlated transient or oscillating signals predicted by particular dark matter models. Quantum entanglement and spin squeezing offer ways to improve some measurements beyond the independent-particle projection-noise limit. These experiments produce tests and constraints within specified models; an anomalous clock signal would still require exclusion of ordinary instrumental and environmental causes.

Time scales, redefinition of the second, and practical deployment

The present second in the International System of Units is defined by fixing the unperturbed caesium-133 ground-state hyperfine transition frequency at exactly 9 192 631 7709\,192\,631\,770 hertz. Optical clocks do not alter that definition merely by outperforming caesium clocks. Their transition frequencies, when expressed in present-day SI hertz, must ultimately be connected to the caesium definition. This creates an important distinction: an optical frequency ratio can be measured more precisely than either optical frequency can be established absolutely in SI units. Internationally recommended optical transition frequencies provide a practical bridge and support their use as secondary representations of the second.

Optical standards have already contributed calibration information to International Atomic Time, the international atomic time scale calculated by the International Bureau of Weights and Measures. Such contributions do not require every participating clock to run continuously. A frequency standard evaluates a local reference over specified intervals, and that information enters the time-scale calibration process. Nevertheless, downtime introduces uncertainty because the local oscillator must bridge intervals without atomic measurements. Hydrogen masers and other microwave devices often serve this flywheel role. Reliable automation, prompt fault recovery, and well-characterized intermittent operation therefore matter alongside the lowest attainable laboratory uncertainty.

A future optical definition requires more than selecting the narrowest transition. Important considerations include reproducibility across independent institutions, agreement of optical frequency ratios, dependable dissemination, and continuity with the existing second. Proposed approaches include choosing one optical transition or defining a reference through several transitions and their frequency relationships. Each entails decisions about realization, maintenance, and future improvements. Redefinition would establish a more capable reference without intentionally changing the duration of the second; consistency with the caesium-based unit would remain essential.

Practical deployment places different demands on the apparatus from those of a record-setting laboratory experiment. Transportable and compact clocks require robust lasers, vacuum systems, thermal control, vibration tolerance, and automated state preparation. For field geodesy, link access and environmental characterization can be as important as atomic performance. Timing infrastructure additionally needs high availability, predictable behavior, and a means of distributing the signal to users. Space-based optical clocks would face launch loads, constrained power and mass, radiation exposure, thermal variation, and challenging optical links. Progress therefore depends on a combination of improved atomic references and engineering: integrated photonics, compact frequency combs, reliable laser sources, and operating procedures that preserve a defensible uncertainty budget outside a specialized laboratory.