Precision Measurement Of The Clock-transition Spectrum In A Cold Ytterbium Atomic Clock

For about60years, the atomic frequency standards have played a crucial role in the precision measurements, fundamental physical researches and technical applications. Tremendous progress on the laser cooling and trapping techniques has enabled the flexible manipulation of atoms or ions. Since year2000, it is also possible to make a direct and reliable frequency measurement from the optical domain to the microwave domain via the optical frequency comb. Moreover, the ultrastable laser with an extremely narrow linewidth, which acts as a local oscillator, has been successfully achieved. All these possibilities allow us to build optical frequency standards with a lower fractional frequency uncertainty and higher stability. Maybe in the future the second, one of the International System of Units that is currently realized by the microwave caesium frequency standards, will be redefined by an optical atomic transition.Ytterbium is regarded as one of the promising candidates for the optical clocks. At present, two optical clocks based on cold ytterbium atoms are being developed in the State Key Laboratory of Precision Spectroscopy, East China Normal University. We have been working on the realization of ytterbium optical clocks for years and we have accomplished the first-stage cooling, second-stage cooling and preliminary optical lattice confinement of ytterbium atoms, which have been detailedly described in the previous works of our group.This thesis presents some improvements in each experimental part and emphasizes the precision measurement of the clock-transition spectrum in a cold ytterbium optical clock. In the first cooling stage, the399nm laser is remotely controlled via a Digilock110module. For laser frequency stabilization utilizing the modulation transfer spectroscopy method, we compare two connection schemes to fully exploit the Digilockl10, and finally obtain an error signal with a high signal to noise ratio. In the second cooling stage, a low noise556nm laser is achieved by locking the frequency to a high finesse Fabry Perot (FP) cavity with the Pound-Drever-Hall (PDH) technique. The laser linewidth is narrowed to about3kHz. We optimize the parameters for the556nm magneto-optical trap (MOT) phase, such as the556nm laser frequency, laser intensity and the B-field. The typical temperature of the171Yb atomic cloud is measured to be about20μK. By careful alignment, we construct a so-called "(1,1,1)" light configuration for the optical lattice on three dimensions. Experiments on trapping ytterbium atoms in various optical lattices are presented. The ultracold171Yb atoms are visibly confined in the one-, two-, and three-dimensional optical lattices operating at the Stark-free wavelength, respectively. For the one-dimensional lattice, the temperature, number and lifetime of cold171Yb atoms are measured. After optimization, the one-dimensional optical lattice with171Yb is readily used for the clock laser interrogation.Precision measurement of the clock-transition spectrum is performed using the electron shelving technique, which cancels out the shot-to-shot atom number fluctuations. The clock-transition spectrum, taken by scanning a double-passed acousto-optic modulator (AOM) for a total range of a few hundred kHz, consist of a carrier, red sideband and blue sideband spectral structure. The anharmonicity of the lattice results in the coupling between the longitudinal and transverse degrees of freedom. So the sideband structures appear to be broad and asymmetric:a sharp edge with its back to the carrier and a slope towards the carrier. A typical sideband spectral shape is modeled to know about the efficient lattice depth and the temperature of in-lattice atoms. For different lattice depths, the longitudinal oscillation frequency is demonstrated experimentally to be proportional to the square root of the lattice beam power, which is consistent with the harmonic approximation of the lattice. As the carrier favors the purely electronic transition when the lattice is at the magic wavelength, generally it has nothing to do with the nonzero motional excitation and transverse confinement. Maybe the main problem is the line pulling induced by the tail of the sideband spectrum. In fact, the carrier spectrum with a linewidth of several kHz can already be used to lock the clock laser. The linewidth broadening mainly results from the residual magnetic field around the lattice region so that the degeneration of four characteristic Zeeman sub-level transitions are removed. If we zoom in the carrier, all four transitions are separated from each other. After the power broadening effect of the578nm clock laser is eliminated, the residual B-field can be well compensated with our compensation coils. With an external applied B-field perpendicular to the clock laser propagation direction, a linear-polarized clock laser pulse allow only "π" transitions. The spectral lines feature a linewidth of about16Hz, which approaches the Fourier limit for our experimental parameters.This thesis presents some analysis as well. The differential light shifts induced by lattice beams can be cancelled out by operating at the magic wavelength. Here magic wavelengths of two ytterbium transition lines are calculated. The Stark free wavelength of6s21S0-6s6p3P1, which has not been reported yet, may be helpful in accurately mapping the ytterbium photoassciation spectrum using this556-nm intermediate line. For the ytterbium optical clock, I will give the detailed evaluation of the blackbody radiation shift. Isotope separation by laser deflecting an atomic beam is analyzed theoretically, which will give a guideline for simply obtaining pure isotopes of various elements. Meanwhile, the concept of laser deflection can be applied to the blackbody radiation suppression, of which I will just outline the physics here.