High-Precision Measurement of Parity Nonconserving Optical Rotation in Atomic Lead.

This dissertation describes a new measurement of parity nonconserving optical rotation near the ^3P_0 to ^3P_1 magnetic dipole absorption line in lead at 1.279 mum. This measurement has an accuracy of 1%, the best in atomic PNC to date. We measure the quantity {cal R} equiv {cal I}m ({cal E} _{pnc}/{cal M}), where {cal M} is the magnetic -dipole amplitude of the absorption line and { cal E}_{pnc} is the electric-dipole amplitude coupled into the same line by the PNC interaction within the lead atom. The interference between the two multipoles produces a rotation of the plane of polarized light in the vapor that is proportional to the refractive index of the absorption line. Calibration is accomplished by separate measurements of the Faraday rotation associated with the absorption line, using a known magnetic field parallel to the light beam. An important improvement allowed the windowless lead vapor tube to be moved in and out of the optical path without any change of optical elements, permitting spurious background rotations to be measured separately and subtracted off, and reducing the scatter of {cal R} by more than an order of magnitude. The data from the absorption and Faraday rotation traces are fitted to theoretical functions to obtain spectral widths, absorption length, and frequency axis parameters. Next, these parameters are used to fit the parity rotation data. Extensive analysis of the statistics and correlations of the parameters provides strong checks against systematic error. In particular, data are taken for absorption depths of the lead vapor from 10 to 60, all yielding the same value of {cal R}. Our result is {cal R} = (-9.86 +/- 0.04 +/- 0.11) times 10^{-8}. The first error is statistical, caused by imperfect subtraction of the drifting background, and the second error is systematic, caused by lineshape uncertainties and calibration error. Our value is consistent with the atomic PNC calculations for lead, which give {cal R} = (-10.7 +/- 0.8) times 10^{-8}. Difficulties in the atomic theory of lead presently limit the extent to which our result tests the standard model of electroweak interactions. The amplitude of the nuclear spin-dependent PNC rotation is found to be less than 2 times 10^{-2} of the nuclear spin-independent rotation.