Thomas Udem

Laser Spectroscopy

Max Planck Institute of Quantum Optics

Hans-Kopfermann-Str. 1

85748 Garching

Tel. +49 89 32905 282

thomas.udem[at]mpq.mpg.de

Research Website

Description

Research focus: precision laser spectrscopy of hydrogen and hydrogen-like systems

Our group is conducting high resolution laser spectroscopy on simple atomic systems such as atomic hydrogen or hydrogen like helium ions. The primary goal is to compare experimental results with calculations in the framework of quantum electrodynamics (QED). The latter includes constants that cannot be determined theoretically but have to be fixed by measurements. To test whether QED is consistent with observations more measurements than constants are required. Within the CODATA least squares adjustment, two constants, the Rydberg constant and the proton charge radius are obtained from 14 high precision measurements in atomic hydrogen. The values obtained in this way are in disagreement with the measured Lamb shift in muonic hydrogen. We are working on shedding light on this so called proton radius puzzle by improving previous values of experimental transition frequencies in atomic hydrogen. In addition we are working towards laser spectroscopy of hydrogen like helium ions and helium like lithium ions. The latter is coming up for high precision tests of QED as calculations of two electron systems are constantly improving.


Continuous wave laser spectroscopy

Udem_Hydrogen_spectrometer
The only metrologically relevant transition in atomic hydrogen that comes with a very narrow natural linewidth occurs between the 1S ground and the metastable 2S state. Its transition frequency has been measured in our lab with a relative uncertainty of a few parts in 10^15. Any of the other metrologically relevant transitions have natural line widths of the order of MHz. At least one of these broad transitions has to be used in order to fix the values of the Rydberg constant and the proton charge radius. Another one is required to test QED. Therefore the only path to improve QED tests on hydrogen is to precisely understand and model the large line width transitions.

For this reason we have been re-measuring the 2S-4P transition frequency using the previous 1S-2S apparatus that is now employed as a source of cold, laser excited 2S atoms. The result is as accurate as all previous hydrogen data combined and confirms the muonic value of the proton charge radius. Hence it is in significant contradiction with the previous hydrogen data. The cause of this discrepancy is as of yet unknown. One may speculate about yet unknown systematic errors of the measurements, errors in QED calculations or even new physics.


Direct frequency comb spectroscopy

Udem_comb
Another route followed by our group is to determine the 1S-3S transition frequency with improved accuracy. For this experiment we use pulsed excitation with a frequency comb. This has the advantage that the excitation volume within the atomic hydrogen beam is small and can be easily shielded against perturbing external fields. Using the comb structure results in an observational line width that is only limited by the natural line width of the 1S-3S transition. With this experiment we have been able to update the values for the Rydberg constant and the proton charge radius as well as improving experimental verification of the predictions of QED.


XUV direct frequency comb spectroscopy

XUV_frequency_comb_system
To extend comb spectroscopy to even shorter wavelengths we started a project to perform spectroscopy on a single trapped helium ion. This system is as simple as hydrogen but eliminates its main experimental problem, the thermal motion of the atoms through the laser beam. In addition, the scaling with large powers of the nuclear charge makes the helium ion 10 times more sensitive to high order QED contributions. Its nucleus, the alpha particle, is much better understood than the proton. In addition the experience with muonic hydrogen urges that systems different from regular atomic hydrogen need to investigated. The helium ion is the perfect candidate because its energy levels can be computed as accurately as atomic hydrogen and muonic hydrogen. For this project we received an ERC Advanced Grant in 2017.

Publications

Towards trapping of hydrogen atoms for computable optical clock applications

O. Amit, D. Taray, V. Wirthl, V. Weis, M. W. Syed, A. Ozawa, J. Weitenberg, S. G. Karshenboim, J. T. M. Walraven, L. Maisenbacher, R. Pohl, Z. Burkley, F. Schmid, T. W. Hansch, D. C. Yost, T. Udem

Physical Review A 112 (3), 33101 (2025).

Show Abstract

"Because of its simple structure, the hydrogen atom is often used as a testbed for quantum electrodynamics. Spectroscopy of trapped atomic samples can greatly improve the accuracy of these tests. Trapping atomic hydrogen in an optical dipole trap or an optical lattice has never been achieved. Only trapping in magnetic fields that lead to large Zeeman shifts has been demonstrated. Standard techniques of atomic physics are difficult to apply to atomic hydrogen. The small mass of the atom and the large photon energy of the 1S-2P cooling transition significantly complicate Doppler cooling. This proposal introduces a photon recoil-assisted loading scheme that uses these properties to our advantage to load atomic hydrogen into an optical dipole trap without laser cooling. The magic wavelength (515 nm) for the 1S-2S clock transition (1.3-Hz natural linewidth) is easily accessible with current laser technology. Since the 1S-2S clock transition can be driven Doppler free, we do not require a very low temperature. Besides improving spectroscopy for fundamental science, such a system can also be used as a ""computable"" atomic clock that may one day justify the redefinition of the SI second in terms of the Rydberg constant."

10.1103/3bnr-q23f

Astigmatic thermal lensing due to surface bulging in Yb:KYW laser crystals

M. Bieringer, J. Weitenberg, T. Udem, A. Ozawa

Optical Materials Express 14 (11), 2527-2541 (2024).

Show Abstract

One of the sources of thermal lensing in crystals is surface bulging that results from an inhomogeneous temperature distribution. We investigate a thermal lens caused by surface bulging in an end-pumped Yb:KYW crystal, which serves as a gain medium in lasers and optical amplifiers. The surface profile of the pumped crystal is measured using a Fizeau-type interferometer and compared with a numerical simulation using a finite element method. The study reveals that due to anisotropic thermal expansion, the surface shape of the Yb:KYW crystal is anisotropic and the profile of the expansion is transversely displaced with respect to the pump beam profile which generates the temperature distribution. The observed surface bulging gives rise to aberrations and deflection of the transmitted beam. It was found that the surface bulging introduces astigmatism that is significantly larger than previously estimated [Appl. Opt. 56, 3857 (2017)]. Our results allow the evaluation of the bulging contribution with improved accuracy. We show that these effects can be significant in certain designs of amplifiers and lasers.

10.1364/ome.530949

Proposal for a computable optical Clock

O. Amit, V. Wirthl, D. Taray, V. Weis, S. Sarkar, M.-C. Kim, M. W. Syed, J. Moreno, F. Egli, J. Weitenberg, M. Thariq, T. Strobl, M. Bieringer, L. Hein, S. Karshenboim, R. Pohl, D. C. Yost, A. Ozawa, T. W. Haensch, T. Udem

9th Symposium on Frequency Standards and Metrology 2889, (2023).

Show Abstract

With the recent update of the SI system, all but one of the units are now based on defining the values of some fundamental constants. This development began in 1983 when the speed of light was assigned an exact fixed value. The advantage of this method is that it separates the definition from the realization, allowing new realizations to be introduced as technology advances without further redefinition. In addition, it allows unit realizations that are adapted to the scale of their intended use. Because of these advantages, we expect that one day also the last remaining object in the current SI system, the caesium atom, will also disappear. The purpose of this proposal is to outline possible paths for realizations of a future SI second based on the definition of the value of the Rydberg constant. Hydrogen and hydrogen{like systems would be the obvious candidates. The emphasis here is on the development of optical clock systems that circumvent difficulties associated with the short wavelength lasers otherwise required for cooling and driving the clock transition. The proposed clock systems based on atomic hydrogen and hydrogen-like He+, should be no more complex than current optical lattice clocks.

10.1088/1742-6596/2889/1/012027

Improved active fiber-based retroreflector with intensity stabilization and a polarization monitor for the near UV (vol 29, pg 7024, 2021)

V. Wirthl, L. Maisenbacher, J. Weitenberg, A. Hertlein, A. Grinin, A. Matveev, R. Pohl, T. W. Hänsch, T. Udem

Optics Express 30 (5), 7340-7341 (2022).

Show Abstract

In Sec. 6 (polarization monitor) of our recent publication [Opt. Express 29(5), 7024 (2021)], we assumed a small value of delta. This is however incorrect. The correct approximation for small beta leads to the updated Eqs. (10)-(11), resulting in a corrected Fig. 12. (C) 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement

DOI: 10.1364/oe.454374

Improved active fiber-based retroreflector with intensity stabilization and a polarization monitor for the near UV

V. Wirthl, L. Maisenbacher, J. Weitenberg, A. Hertlein, A. Grinin, A. Matveev, R. Pohl, T. W. Hänsch, T. Udem

Optics Express 29 (5), 7024-7048 (2021).

Show Abstract

We present an improved active fiber-based retroreflector (AFR) providing high-quality wavefront-retracing anti-parallel laser beams in the near UV. We use our improved AFR for first-order Doppler-shift suppression in precision spectroscopy of atomic hydrogen, but our setup can be adapted to other applications where wavefront-retracing beams with defined laser polarization are important. We demonstrate how weak aberrations produced by the fiber collimator may remain unobserved in the intensity of the collimated beam but limit the performance of the AFR. Our general results on characterizing these aberrations with a caustic measurement can be applied to any system where a collimated high-quality laser beam is required. Extending the collimator design process by wave optics propagation tools, we achieved a four-lens collimator for the wavelength range 380-486 nm with the beam quality factor of M-2 similar or equal to 1.02, limited only by the not exactly Gaussian beam profile from the single-mode fiber. Furthermore, we implemented precise fiber-collimator alignment and improved the collimation control by combining a precision motor with a piezo actuator. Moreover, we stabilized the intensity of the wavefront-retracing beams and added in-situ monitoring of polarization from polarimetry of the retroreflected light. (C) 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

DOI: 10.1364/oe.417455

Two-photon frequency comb spectroscopy of atomic hydrogen

A. Grinin, A. Matveev, D. C. Yost, L. Maisenbacher, V. Wirthl, R. Pohl, T. W. Hänsch, T. Udem

Science 370 (6520), 1061-+ (2020).

Show Abstract

We have performed two-photon ultraviolet direct frequency comb spectroscopy on the 1S-3S transition in atomic hydrogen to illuminate the so-called proton radius puzzle and to demonstrate the potential of this method. The proton radius puzzle is a significant discrepancy between data obtained with muonic hydrogen and regular atomic hydrogen that could not be explained within the framework of quantum electrodynamics. By combining our result [f(1s-3s) = 2,922,743,278,665.79(72) kilohertz] with a previous measurement of the 1S-2S transition frequency, we obtained new values for the Rydberg constant [R-infinity = 10,973,731.568226(38) per meter] and the proton charge radius [r(p) = 0.8482(38) femtometers]. This result favors the muonic value over the world-average data as presented by the most recent published CODATA 2014 adjustment.

DOI: 10.1126/science.abc7776

Scroll to top