[en] We report on the precision measurement of aluminum atoms ${}^2{P}_{1/2}-<!--\; ???\; -->{}^2{S}_{1/2}$ transition at 394 nm and ${}^2{P}_{3/2}-<!--\; ???\; -->{}^2{S}_{1/2}$ transition at 396 nm in a hollow-cathode lamp. Both absorption and saturated absorption spectra are measured. From the absorption spectra, the Doppler linewidth is estimated to be 2.65 GHz. The saturated absorption spectra are analyzed based on a velocity-changing collisions model. With a frequency comb calibrated wavemeter, the frequencies of ${}^2{P}_{1/2}(F^{\mathrm{\prime <!--\; ???\; -->}}=3)-<!--\; ???\; -->{}^2{S}_{1/2}(F=2)$ transition and ${}^2{P}_{3/2}(F^{\mathrm{\prime <!--\; ???\; -->}}=4)-<!--\; ???\; -->{}^2{S}_{1/2}(F=3)$ transition are measured to be 759.905 204(1) THz and 756.547 133(3) THz, respectively. The hyperfine coupling constants of aluminum atoms are determined, and are compared with previously reported measurement results and theoretical calculation results. Reasonable agreement is found for the magnetic dipole constant (A constant), while the electric quadrupole constant (B constant) has a large deviation. (paper)

[en] We report construction of an iodine-stabilized laser frequency standard at 532 nm based on modulation transfer spectroscopy (MTS) technology with good reproducibility. A frequency stability of 2.5 × 10⁻¹⁴ at 1 s averaging time is achieved, and the frequency reproducibility has a relative uncertainty of 3.5 × 10⁻¹³, demonstrating the great stability of our setup. The systematic uncertainty of the iodine-stabilized laser frequency standard is evaluated, especially the contribution of the residual amplitude modulation (RAM). The contribution of the RAM in MTS cannot be evaluated directly. To solve this problem, we theoretically deduce the MTS signal with RAM under large modulation depth, and prove that the non-symmetric shape of the MTS signal is directly related to the MTS effect. The non-symmetric shape factor r can be calibrated with a frequency comb, and in real experiments, this r value can be obtained by least-squares fitting of the MTS signal, from which we can infer the RAMinduced frequency shift. The full frequency uncertainty is evaluated to be 5.3 kHz (corresponding to a relative frequency uncertainty of 9.4 × 10⁻¹²). The corrected transition frequency has a difference from the BIPM-recommended value of 2 kHz, which is within 1 σ uncertainty, proving the validity of our evaluation. (paper)