A Progress Report on the g-2 Resonance Experiments

Abstract

The spin resonance experiments on slow free electrons in vacuum have a long history. In 1953 Bloch proposed to trap electrons in an electric potential well of depth ~l0⁻⁵ V and unspecified shape superimposed upon a magnetic field of ~1000 G. Cyclically the latter could be made so inhomogeneous that the effective magnetic hill seen by all electrons not in the lowest, unmagnetic Rabi-Landau level would overcompensate the electric well. Thereby, only those in the lowest level would be retained. Spin- or cyclotron-transitions to the next higher levels induced subsequently would be signaled by loss of the electrons. The levels referred to are given by E = (2n + 1 + g_sm)μ_oH, with n = 0, 1, 2,.... and m = ±1/2, (Rabi, 1928). Also at Stanford in connection with their cyclotron-resonance work in which trapping was carefully avoided Franken & Liebes showed in 1956 that a small external electric field should shift the cyclotron frequency ω_c ≡ 2πν_c by δω_c ∝ φ_zz/ω_c. This suggested that the Penning (1937) trap in the form described by Pierce (1949) should be well suited for the simultaneous measurement of ω_c and of the spin resonance frequency ω_s ≡ 2πν_s on stored electrons. For such a trap the axial field gradient φ_zz and δω_c are constant throughout its volume and δω_c = −ω_m may be determined by measuring the axial oscillation frequency ω_z ≡ 2πν_z or the magnetron (drift) frequency ω_m ≡ 2πν_m on the electrons in situ.