| The Lamb shift of the 1S hydrogen ground state is measured to 6ppm by comparing the frequencies of the hydrogen 1S-2S and 2S-4P transitions. The 1S-2S transition is observed with high resolution using Doppler-free two-photon absorption in a cell containing flowing atomic hydrogen. The required 243-nm light is generated by frequency-doubling the output of a 486-nm ring dye laser, and its intensity is enhanced in an optical cavity. Transitions to the 2S state are detected by monitoring the collisionally-induced Lyman-{dollar}alpha{dollar} fluorescence. In a separate apparatus, the single-photon 486-nm 2S-4P{dollar}sb{lcub}1/2, 3/2{rcub}{dollar} transitions are excited in a collimated H(2S) beam with a laser beam propagating at right angles to the atomic beam. The resonance is observed as a depletion of the H(2S) population. The residual Doppler shift is reduced by retro-reflecting the laser beam, and any remaining shift is measured by changing the angle at which the laser and atomic beams cross. Each transition frequency is calibrated by heterodyning the scanning laser with a second laser locked to a molecular tellurium Doppler-free transition half-way between the two optical hydrogen transition frequencies. The 1S Lamb shift is extracted from the 5 GHz difference frequency of the two 486-nm hydrogen transitions, giving L(1S) = 8172.819(51) MHz. This is the most precise measurement of a Lamb shift in any system. This value agrees with the theoretical value L(1S) = 8172.79(3) MHz, which uses the larger, most recent value of the proton charge radius. It is discrepant with the theoretical value L(1S) = 8172.64(3) MHz, which uses the smaller, older value of the proton charge radius. |
