2025 Volume 34 Issue 6
Article Contents

Shiyao Shao, Qing Li, Lihua Zhang, Bang Liu, Zhengyuan Zhang, Qifeng Wang, Jun Zhang, Yu Ma, Tianyu Han, Hanchao Chen, Jiadou Nan, Yiming Yin, Dongyang Zhu, Yajun Wang, Dongsheng Ding, Baosen Shi. A Rb–Cs dual-species magneto-optical trap[J]. Chinese Physics B, 2025, 34(6): 063702. doi: 10.1088/1674-1056/adc190
Citation: Shiyao Shao, Qing Li, Lihua Zhang, Bang Liu, Zhengyuan Zhang, Qifeng Wang, Jun Zhang, Yu Ma, Tianyu Han, Hanchao Chen, Jiadou Nan, Yiming Yin, Dongyang Zhu, Yajun Wang, Dongsheng Ding, Baosen Shi. A Rb–Cs dual-species magneto-optical trap[J]. Chinese Physics B, 2025, 34(6): 063702. doi: 10.1088/1674-1056/adc190

A Rb–Cs dual-species magneto-optical trap

  • Received Date: 14/01/2025
    Accepted Date: 15/03/2025
    Available Online: 01/06/2025
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(5)

Article Metrics

Article views(723) PDF downloads(7) Cited by(0)

Access History

A Rb–Cs dual-species magneto-optical trap

Abstract: We describe a three-dimensional (3D) magneto-optical trap (MOT) capable of simultaneously capturing 85Rb and 133Cs atoms. Unlike conventional setups, our system utilizes two separate laser systems that are combined before entering the vacuum chamber, enabling the simultaneous trapping of two different atomic species. We trapped 85Rb and 133Cs atoms using relatively low total power: 8 mW cooling and 4 mW repump for 85Rb, and 7.5 mW cooling and 1.5 mW repump for 133Cs. The number of trapped atoms was 1.6 × 108 for 85Rb and 1.4 × 108 for 133Cs. The optical depths were 3.71 for 85Rb and 3.45 for 133Cs. The temperature of trapped atoms was ∼200 μK for 85Rb and ∼200 μK for 133Cs. Our 3D MOT setup allows full horizontal optical access to the trapped atomic ensembles without spatial interference from the trapping or repump laser beams. Our vacuum system is also quite simple, avoiding much of the complexity typically encountered in similar dual-species systems. However, the red detuning of the cooling laser used for atomic trapping in our system is relatively small, leaving room for further optimization. This system offers a versatile platform for exploring complex phenomena in ultracold atom physics, such as Rydberg molecule formation and interspecies interactions.

1.   Introduction
  • The magneto-optical trap (MOT)[1] is a key tool in the field of cold atom physics and the most common laser cooling method. Since its introduction in 1987,[1] the MOT has been widely used in various fundamental and applied research areas. By combining laser cooling and magnetic trapping techniques, the MOT can cool atoms to micro-Kelvin temperatures or even lower. These ultracold temperatures significantly reduce the thermal motion of atoms, enabling precise control and measurement of quantum states. As a result, the MOT has become an essential foundational tool in cutting-edge fields such as quantum optics,[2] cold molecule research,[3] atomic clock[4] and precision spectroscopy.[5] Its primary advantage lies in its ability to achieve high-density, low-temperature atomic trapping with relatively low technical complexity, providing a stable source of cold atoms for conducting complex experiments.

    Rubidium (Rb) and cesium (Cs), as the most commonly used species in cold atom experiments, are widely employed in studies of cold atomic gases due to their excellent laser cooling properties and rich hyperfine energy level structures.[6] Although single-species MOT systems have been extensively optimized,[7–10] the scope of research has expanded, and this singularity has increasingly become a bottleneck in the study of multicomponent systems. The high sealing of the vacuum system makes it extremely difficult to introduce new atomic species into an established system. A single-species MOT cannot facilitate the study of interactions between different species. In many emerging fields, such as quantum information processing and cold atom quantum simulation,[11–13] the need to simultaneously cool and manipulate multiple atomic species has become more prominent. For example, in ultracold heteronuclear molecule synthesis, simultaneous cooling of multiple atomic species is a prerequisite. Since the first synthesis of KRb diatomic molecules through magnetic association and stimulated Raman adiabatic passage in 2008,[3] heteronuclear molecules such as RbCs, NaK, and NaRb have been successively synthesized.[14–17] Significant progress was made in 2022 with the first quantum-coherent synthesis of triatomic molecules in a laboratory setting.[18] A system capable of cooling multiple atomic species not only enables these complex experiments but also provides new perspectives for studying interatomic interactions, multicomponent effects in mixed cold atomic gases, and dynamical behaviors in diatomic gases.[19–22]

    However, traditional MOT systems are typically designed for specific atomic species from the outset, and the high vacuum environment makes it extremely difficult to add or change atomic species later. To address this issue, we introduced both Rb and Cs atomic sources at the design stage. In principle, the system can simultaneously capture three atomic species: 85Rb, 87Rb, and 133Cs. This design overcomes the limitations of single-species MOT systems, providing a more flexible experimental platform for multicomponent research. Due to the significant frequency difference between the cooling light for Rb and Cs atoms, we designed independent laser systems for each species and achieved precise control to enable efficient cooling and trapping. This technical foundation is a critical guarantee for conducting mixed-species atomic experiments.

    In this paper, we review and describe a dual-species 3D magneto-optical trap (MOT) apparatus. This system enables simultaneous trapping of 85Rb and 133Cs atoms at the same position within a vacuum chamber and achieves excellent performance with smaller red detuning of the cooling light. We arranged the trapping beams at a 45° angle to the vertical axis. This configuration ensures complete optical access in the horizontal direction and avoids interference between the cooling and repump beams. Furthermore, with both cooling and repump beams controlled by acousto-optic modulators (AOMs), the MOT operates at a high repetition rate. In this paper, the experimental cycle is set to 10 ms, corresponding to a MOT repetition rate of 100 Hz. In our experiments, we utilized low laser power conditions — 8 mW for the trapping light and 4 mW for the repump light for 85Rb, and 7.5 mW for the trapping light and 1.5 mW for the repump light for 133Cs, with all values representing the total power. Under these conditions, we successfully trapped 1.6 × 10885Rb atoms and 1.4 × 108133Cs atoms, achieving optical depths (OD) of 3.71 and 3.45 for 85Rb and 133Cs, respectively. Additionally, we observed suppression of 133Cs atom absorption due to 85Rb atom clouds, further confirming the high spatial overlap between the two atomic species. In recent years, similar dual-species systems have been applied to many cutting-edge studies, such as the synthesis of heteronuclear Rydberg molecules[23] and the construction of quantum gates.[12,24]

2.   The 3D MOT setup
  • A typical MOT apparatus comprises an ultra-high vacuum (UHV) cell, magnetic fields, and laser beams. The vacuum system we used is shown in Fig. 1(a) The rectangular UHV science cell we used is made from 3 mm thick borosilicate glass and has internal dimensions of 20 × 20 × 60 mm. A cylindrical graded index section increases the overall length of the cell from the flange to ∼110 mm. We install two rubidium (Rb) dispensers (SAES,RB/NF/7/25 FT10+10) to provide controllable Rb vapor pressure. During the MOT operation, we only run one dispenser and the other one is reserved for backup. A constant voltage current source is used to power the dispensers, and the Rb atom vapor pressure is controlled by adjusting the current. The naturally occurring rubidium is composed of 85Rb (72.2%) and 87Rb (27.8%). In this work, we describe the 3D MOT for 85Rb atoms, but the system can also be used to cool and trap 87Rb atoms by changing the MOT laser frequencies. For the Cs source, we place metallic Cs in a copper tube connected to a valve, which is installed on one side of the multi-port vacuum chamber. A heating tape is wrapped around the copper tube, powered by another constant voltage current source. The Cs vapor pressure is adjusted by changing the valve size and the current supplied to the heating tape. The experiment needs the pressure inside the science cell to be below 10−5 Pa. In our system, we first use a mechanical pump to lower the pressure to 10−4 Pa. Then, we start the compound pump, which has an ion pump and a getter pump. Once the compound pump is operational, we strategically seal the mechanical pump’s interface, allowing its complete removal and optimizing the experimental setup’s spatial efficiency. As a result, our vacuum system becomes very streamlined, avoiding the complexities typically associated with similar dual-atom systems. With the help of both pumps, the final pressure in our science cell is below 2.1 × 10−9 Pa.

    As illustrated in Fig. 1(b), the magneto-optical configuration comprises a spatially varying magnetic field produced from a magnetic coil and six trapping beams. To produce the magnetic field, we utilized MAG-3000 magnetic field coils from ColdQuantaTM, which consist of six independently powered coils integrated into a single structure mounted outside the glass vacuum cell. In this system, we only power the two opposing coils in reverse to create the spatial magnetic field. The two powered coils have 120 turns each, measuring 60 mm in length and 35 mm in width, with a spacing of 45 mm between them. In the experiment, the coils are powered at 1.58 A, resulting in an axial magnetic field gradient of approximately 22 Gauss/cm at the center of the trapping area, which meets the experimental requirements. By varying the current, we can adjust the spatial magnetic field gradient to satisfy different experimental conditions. Only the two opposing coils among the six independently powered coils are used to construct the magneto-optical trap for capturing atoms, while the remaining coils can be utilized for compensating the magnetic field and other operations, providing a significant operational redundancy.

    The separated laser systems are used for cooling and trapping Rb and Cs, as shown in Fig. 2. The relevant energy level diagrams for the cooling processes of the two species are shown in Figs. 1(c) and 1(d). Both Rb and Cs require separate lasers to provide cooling and repump light due to the relatively large hyperfine splitting in their ground states. All lasers in our system are frequency-locked using saturated absorption spectroscopy. For Rb, the cooling laser is red detuned by 6.5 MHz from the transition |S1/2, F = 3〉 → |P3/2, F′ = 4〉, and the repump laser is on resonance with the transition |S1/2, F = 2〉 → |P3/2, F′ = 3〉. For Cs, the cooling laser is red detuned by 5.5 MHz from the transition |S1/2, F = 4〉 → |P3/2, F′ = 5〉, and the repump laser is on resonance with the transition |S1/2, F = 3〉 → |P3/2, F′ = 4〉. For the Rb probe light, we choose to split a portion from the cooling light path. The final probe light frequency corresponds to the resonance of the transition |S1/2, F = 3〉 → |P3/2, F′ = 4〉. The probe light for Cs is also split from the cooling light, following the same procedure as for Rb. The final frequency corresponds to the |S1/2, F = 4〉 → |P3/2, F′ = 5〉 resonance. All lasers are coupled into optical fibers after passing through the AOM and transmitted to the vicinity of the MOT. The three fibers for Rb and Cs are connected to a beam expander located near the MOT to complete the beam expansion. The circular laser beam is expanded to about 15 mm in 1/e diameter. After passing through a λ/2 wave plate and a λ/4 wave plate, the cooling light for both Rb and Cs is combined using a dichroic mirror (DM). The combined beams pass through a broadband λ/4 wave plate before entering the science cell. One pair of counterpropagating trapping laser beams (1 and 2) is aligned along the y axis. Unlike conventional 3D MOT setups, where the trapping beams are orthogonal to the cell surfaces, in our system, the four trapping beams (labeled 3, 4, 5, and 6) are aligned at 45° angles to both the x and z axes, as shown in Fig. 1(b). This configuration allows full horizontal optical access to the trapped atomic ensembles without spatial interference from the trapping or repump laser beams. During atom trapping, we keep the broadband λ/4 wave plate stationary and adjust the wave plates after the beam expanders for the Rb and Cs cooling light. This ensures that the cooling light for both Rb and Cs entering the science cell through the broadband λ/4 wave plate is in the appropriate circular polarization state, which is required for trapping the atoms. The polarizations (σ+ and σ−) of the six trapping beams are shown in Fig. 1(b). In the two-species 3D MOT, we simultaneously trap two types of atoms, using a cooling laser power of only 8 mW for 85Rb and only 7.5 mW for 133Cs, along with a repump laser power of 4 mW for 85Rb and 1.5 mW for 133Cs, with all values representing the total power.

3.   Optical depth and measurement
  • Figures 3(a) and 3(c) show the CCD fluorescence image of the trapped atoms in the 3D MOT. The diameters of the trapped Rb and Cs atomic clouds are 0.78 mm and 0.40 mm, respectively. The experimental system employs a design where two atomic capture laser beams are co-injected into the science cell, sharing a common set of magnetic field coils. This configuration allows for the simultaneous capture of 85Rb and 133Cs atoms at the same spatial location within the cell. Compared to typical systems, the red detuning of the cooling light in this system is significantly smaller.[25–28] Generally, the red detuning of the capture laser is approximately 2Γ–3 Γ (where Γ is the natural linewidth of the cooling transition). In this system, the red detuning for capturing 85Rb atoms is 1.05 Γ, while for 133Cs atoms, it is 1.07 Γ. In general, since the scattering force from the cooling laser interacting with the atoms is inversely proportional to detuning, F∝Γ2(Γ2δ)2IIsat, a smaller red detuning leads to a stronger scattering force, which may cause heating effects. Additionally, a reduction in red detuning can also decrease the photon scattering rate, thus prolonging the time required for cooling.[29] Doppler cooling relies on the compensation of the Doppler shift in the atomic motion. When the red detuning decreases, the velocity distribution of atoms that can absorb the cooling light and slow down narrows, potentially reducing the number of atoms affected by the cooling.[30] Overall, a relatively small red detuning is not advantageous for atomic trapping and may reduce the system’s cooling and trapping efficiency, indicating that there is still room for optimization in this system. After the atomic trapping is completed, we take OD measurements to characterize the 3D MOT properties. Optical depth (OD), as an important physical quantity used to quantify the degree of attenuation of light as it propagates through a medium, helps in understanding and predicting the behavior of light in different media and has wide-ranging applications. The OD measurement scheme is shown in Fig. 4, where (a) is the optical setup, and (b) is the timing. The two probe beams are transmitted via optical fibers to the vicinity of the MOT. After passing through a half-wave plate and a polarizing beam splitter (PBS), the two probe beams are combined using a dichroic mirror (DM). The combined beam is then focused onto the trapped atomic cloud using an achromatic lens with a focal length of 100 mm. After passing through the atomic cloud, the probe light is collimated again using another achromatic lens with the same focal length (100 mm). The beams are then separated by the dichroic mirror and individually coupled into optical fibers. Finally, we measure the probe absorption spectrum with a photomultiplier tube (PMT). The measurement is taken periodically. At each period of T = 10 ms, we set the MOT trapping time tMOT = 9 ms and the measurement duty (including state preparation and OD measurement) time tduty = 0.55 ms. At each cycle, after the cooling laser is switched off, we keep the repump laser on for additional Δt = 0.5 ms to optically pump all the atoms into the state |S1/2, F = 3〉 for 85Rb and |S1/2, F = 4〉 for 133Cs, which are preferable for the OD scheme. After the atoms are prepared in the suitable state, we switch on the probe lasers for absorption measurement inside the duty window. All the laser powers are controlled by AOMs.

    The transmission of the weak probe laser after passing through the medium is given by[31]

    where L is the medium length and k(ωp)=(ωp/c)1+χ≃(ωp/c)(1+χ/2) is the complex wave number in the medium with the speed of light in vacuum c. The linear susceptibility (complex) is given as[32]

    For Rb, α0 is the on-resonance absorption coefficient of the transition |S1/2, F = 3〉 → |P3/2, F′ = 4〉, k0 = ωr/c, ωr is the resonance frequency of the transition |S1/2, F = 3〉 → |P3/2, F′ = 4〉, Δωp = ωp–ωr is the probe detuning. The atomic-lifetime-determined electric dipole relaxation rate of |S1/2, F = 3〉 → |P3/2, F′ = 4〉 is γ13. The atomic optical depth OD = α0L and γ13 are obtained as the best fitting parameters of Eq. (1) to the measurement data. After substituting the OD and related values into Eq. (1) and simplifying, L will be eliminated. During the process of fitting the function to obtain the OD, the value of L does not affect the result. For Cs, the relevant parameters can be replaced with those specific to Cs atoms during the calculations.

    Figures 3(b) and 3(d) present the typical absorption measurement results for the two-species 3D magneto-optical trap MOT. The probe lasers are maximally absorbed at resonance, as shown in Figs. 3(b) and 3(d). The solid curves represent the best fit of Eq. (1) to the experimental data, yielding an optical depth (OD) of 3.71 and γ13 = 2π × 1.8 MHz for 85Rb, and an OD of 3.45 and γ13 = 2π × 1.2 MHz for 133Cs. Using the method developed in Refs. [33,34], the total number of atoms in the MOT is 1.6 × 108 for 85Rb and 1.4 × 108 for 133Cs. By using the time-of-flight (TOF) method,[35] we measured the temperature of the trapped 85Rb atomic cloud to be ∼ 200 K, and the temperature of the trapped 133Cs atomic cloud to be ∼ 200 μK.

    Figure 5 illustrates the inhibitory effect of trapped Rb atoms on the absorption of trapped Cs atoms when both species are simultaneously captured. Initially, with both Rb and Cs trapping light activated, the OD of the Cs atomic cloud was measured while both species were present in the MOT. Subsequently, the Rb trapping light was turned off, releasing the Rb atomic cloud, and the OD of the Cs atomic cloud was measured again with only Cs atoms trapped. From Fig. 5, it is evident that the OD of the Cs atomic cloud increased by at least 20% (from 1.69 to 2.04) in the absence of the Rb atomic cloud, indicating a clear inhibitory effect of the Rb atoms on the absorption of Cs atoms. This effect is attributed to inelastic scattering between the Cs and Rb atoms, which are trapped at the same spatial location, interfering with the loading of Cs atoms and resulting in a reduction of the OD for Cs.[25] This phenomenon also indicates a significant spatial overlap between the two types of atoms captured in this system.

4.   Conclusion
  • In summary, we have described the apparatus of 3D MOT of 85Rb and 133Cs atoms and its OD measurements. Our vacuum system is simple and avoids much of the complexity usually encountered in many similar two-species systems. In our six-trapping-beam optical setup, four of the trapping beams are aligned at 45° angles to the vertical axis. This trapping-repump beam configuration opens full optical access along the longitudinal axis and it is important for quantum optics experiments. Our system requires relatively low power for atom trapping. With only 8 mW of total cooling light and 4 mW of total repump light, we trapped 1.6 × 10885Rb atoms with an optical depth of 3.71, and the temperature was ∼ 200 μK. Using just 7.5 mW of total cooling light and 1.5 mW of total repump light, we trapped 1.4 × 108133Cs atoms with an OD of 3.45, and the temperature was ∼ 200 μK. The observed suppression of 133Cs atom absorption due to the presence of trapped 85Rb atoms indicates a high degree of spatial overlap between the two species in the trapping state. However, the red detuning of the cooling laser in this system is relatively small compared to typical systems, indicating that there is still room for optimization in the system. In this work, we describe the 3D MOT for 85Rb and 133Cs atoms, but the system can also be used to trap 87Rb atoms by changing the MOT laser frequencies.

    Future enhancements will focus on refining control schemes, improving trapping efficiency, and incorporating Rydberg state excitation. These upgrades will enable studies of long-range interactions[36,37] and heteronuclear Rydberg molecules,[38,39] contributing to the integration of Rydberg physics with ultracold atom research. With its low laser power requirements and capability to simultaneously trap Rb and Cs atoms, our system offers a nearly optimal cold atom source for advanced quantum optics research. Our systems will support research in cutting-edge fields such as quantum computing,[40] quantum simulation[41,42] and precision measurement.[43]

Figure (5)  Reference (43)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return