supported by the National Key R&D Program of China (Grant No. 2017YFA0304502) and the National Natural Science Foundation of China (Grant Nos. 11634008, 11674203, 11574187, and 61227902).
supported by the National Key R&D Program of China (Grant No. 2017YFA0304502) and the National Natural Science Foundation of China (Grant Nos. 11634008, 11674203, 11574187, and 61227902).
† Corresponding author. E-mail:
supported by the National Key R&D Program of China (Grant No. 2017YFA0304502) and the National Natural Science Foundation of China (Grant Nos. 11634008, 11674203, 11574187, and 61227902).
Atomic spin relaxation in a vapor cell, which can be characterized by the magnetic resonance linewidth (MRL), is an important parameter that eventually determines the sensitivity of an atomic magnetometer. In this paper, we have extensively studied how the pump intensity affects the spin relaxation. The experiment is performed with a cesium vapor cell, and the influence of the pump intensity on MRL is measured at room temperature at zero-field resonance. A simple model with five atomic levels of a Λ-like configuration is discussed theoretically, which can be used to represent the experimental process approximately, and the experimental results can be explained to some extent. Both the experimental and the theoretical results show a nonlinear broadening of the MRL when the pump intensity is increasing. The work helps to understand the mechanism of pump induced atomic spin relaxation in the atomic magnetometers.
Atomic magnetometry[1,2] has been extensively investigated in the past decades. The sensitivity of atomic magnetometry is related to the spin relaxation and it can be measured by observing the change of atomic polarization in a magnetic field.[3] In this measurement, optical pumping is necessary to generate the atomic polarization[4] through the absorption–emission cycle of atom–light interaction,[5,6] and such optical pumping process has been considered as one of the fundamentals of magnetic field measurement.[7,8] The atomic spin relaxation time T2 in the vapor cell is a very important parameter which directly determines the sensitivity of an atomic magnetometer.[9–11] The spin relaxation in the vapor cell is affected by many factors, such as diffusion of atoms,[12,13] collision between atoms,[14–16] optical pumping, etc. The relaxation caused by diffusion and collision can be suppressed by some advanced techniques, such as atom-relaxation coating of the vapor cell,[17] optimizing the filling of buffer gas,[18] controlling the temperature of the vapor cell,[19] reducing the magnetic field gradient,[20] etc. The relaxation caused by optical pumping can be controlled by optimizing the pump intensity,[21] which is inevitable and cannot be eliminated completely.
The atomic spin relaxation time T2 is characterized by the magnetic resonance linewidth (MRL) of the light-induced magnetic resonance Δ ω (Δ ω = 1/T2).[10,22] By observing the atomic polarization in different magnetic field around zero-point (so called zero-field resonance),[23] or by tuning the modulation frequency of the pump beam around the magnetic resonance frequency,[24–26] the magnetic resonance spectrum (MRS) can be measured. For a linear pump, the magnetic resonance frequency is equal to twice of the Larmor frequency (2 ωL).[27,28] The magnetic resonance process can be treated by Bloch equations and the MRS can be figured out which is normally a Lorentzian shape,[29] from which the MRL (the half width at half maximum of MRS) is obtained.[30]
Usually, the MRL can simply be treated as linear broadening[22,31] with the increase of the pump intensity. However this treatment is actually incomplete. In 1999, the nonlinear light narrowing was investigated theoretically and experimentally,[21] and the nonlinear broadening was also analyzed later[32] based on a simplified ideal atomic system. By solving the Liouville equation with an unmodulated pump beam, the nonlinear broadening was obtained. Recently, Han et al. finished the experimental testing of pump induced nonlinear broadening in a heated rubidium vapor cell, in which the circularly polarized pump beam was amplitude modulated.[33]
In this paper, we have investigated the influence of the pump intensity on spin relaxation at room temperature. The experiment is performed with a cesium vapor cell with buffer gas, and an unmodulated linear polarized pump beam locked to the D1 line of cesium is used. The MRL is measured with the probe beam tuned on resonance to the D2 transition. The MRL is measured at zero-field resonance, and the results show that the relation between MRL and pump intensity is nonlinear broadening. To the best of our knowledge, this is the first time this phenomenon is observed experimentally. We have also used a simplified five-level atomic system model with one excited state and four ground states to represent our experimental configuration, and by solving the Liouville equation,[34,35] the experimental result can be explained approximately.
The experiment for testing the influence of the pump intensity on atomic spin relaxation is carried out at room temperature (∼ 25 °C), and the experimental setup is shown in Fig.
The dependence of MRL on the pump intensity is shown in Fig.
In this part, we are trying to give an explanation to the above observed results. As shown in Fig.
(I) Since all the Zeeman sublevels in excited state 62P1/2 F′ = 3 are involved in the optical pumping process, and all these Zeeman sublevels have the same spontaneous decay rate Γ0[6] (for cesium, Γ0 = 2π × 4.5 × 106 Hz), we can treat all these seven sublevels as one state |e⟩.
(II) The Zeeman sublevels in ground state 62S1/2 F = 4 are separated into two parts: the first part is the Zeeman sublevels involved in the optical pumping process, corresponding to −3 ⩽ mF ⩽ 3, and this part can be represented as state |2⟩. The second part is the rest of the Zeeman sublevels, corresponding to mF = 4 and mF = −4, which can be represented as states |1⟩ and |3⟩, respectively.
(III) The ground state 62S1/2 F = 3 is not involved in the optical pumping process, but some of the atoms still fall into this ground state through spontaneous emission process from the excited state 62P1/2 F = 3, so we can represent these Zeeman sublevels as one state |4⟩.
Since we have simplified our atomic level system, the spontaneous decay rate from excited state |e⟩ to each ground state |1⟩, |2⟩, and |3⟩ must be modified. We use η Γ0 and θ Γ0 to represent the recalculated spontaneous decay rates from |e⟩ to balanced states (|1⟩ and |3⟩) and state |4⟩, respectively. η and θ can be estimated by using the dipole matrix elements[36]
As shown in Fig.
The second and third terms in Eq. (
A probe beam is used to detect the component of the atomic polarization in 62S1/2 F = 4 along the Y direction, which can be expressed by an operator
The dependence of
We have investigated the influence of the pump intensity on atomic spin relaxation. The spin relaxation is characterized by the MRL. The experiment is performed in a cesium vapor cell with 20 Torr helium as a buffer gas at room temperature. The nonlinear broadening of MRL along with the pump intensity increasing is observed. A simplified five-level atomic configuration is analyzed which can represent our experiment system to a certain extent, and the dependence of the MRL on pump intensity is obtained theoretically. The nonlinear relation between MRL and pump intensity is expected, which explains our experimental results to some extent. The reported results provide a better understanding of the atom–field interaction in atomic magnetometers.
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] | |
[30] | |
[31] | |
[32] | |
[33] | |
[34] | |
[35] | |
[36] | |
[37] | |
[38] |