The use of quantum technology-based sensors has been rapidly growing recently, as they can be used in various sensing applications that require highly sensitive detection of physical quantities. In particular, in order to implement a quantum sensor, it is known that the application of various quantum platforms depends on their purpose [1]. A cold atomic system is a versatile quantum sensor for measuring time, acceleration, and magnetic field [2, 3]. In addition, the cold atomic system has been spotlighted as a candidate for quantum navigation systems related to inertial navigation [4, 5]. The utility of the latest diamond sensor is surprising. It is drawing much attention for its potential in electric and magnetic field measurement as well as temperature sensing [6–9]. Photon-based quantum sensors that can be obtained by manipulating the state of photons, such as entanglement or squeezing, have also been reported [10, 11]. A photonic platform is a promising technology for implementing quantum imaging or quantum radar [12, 13].

Atomic vapor cells are widely used as convenient platforms because of their relatively small size, simple structure, and ease of fabrication compared to the aforementioned quantum sensor platforms. In the early 1990s, the famous electromagnetically induced transparency (EIT) effect was reported from atomic vapor and produced related applications in atomic vapors [14–18]. Moon et al. [16] suggested atomic clocks based on coherent population trapping obtained by an injection-locked laser system and atomic vapor. Phillips et al. [17] were the first to propose the storage of light in atomic vapor, which could be useful for realizing quantum memory. Bae and Moon [18] investigated continuous control of light group velocity in a Rb vapor cell. Budker and Romalis [19] demonstrated the optical magnetometer, which is sensitive to external magnetic fields based on the interaction of light and atoms in a vapor cell. In order to expand the applicability of the atomic vapor cell, vapor cells containing various buffer gases, an atomic vapor in hollow-core fiber, and a chip-scale vapor cell for integrated photonics have been developed [20–22].

Based on the atomic vapor cell, quantum electrometers are rapidly emerging due to a strong need for precision measurement in broadband telecommunication and THz imaging [23–27]. The traditional electric field measurement technology not only depends on the size of the antenna aperture, but also suffers from scattering or distortion of the electric field caused by metallic parts. On the other hand, the use of a glass vapor cell is a better solution that is free from these problems in electric field measurement [24]. Early studies mainly focused on measuring Autler-Townes (AT) splitting frequency after preparing Rydberg atom-based EIT. Sedlacek et al. [25] reported that a quantum electrometer corresponding to Rydberg-EIT is a powerful candidate for detecting the microwave. Vector microwave electrometry using Rydberg-EIT was also reported [26]. In 2014, the Holloway group at NIST announced that electric field measurements could be made using an atom-based probe capable of SI-traceable, self-calibrated measurements corresponding to a broadly tunable Rydberg receiver [27]. In particular, studies on the fiber-coupled atomic vapor cell used to flexibly guide light have been performed to minimize the interference effect caused by external scattering of the input radio-frequency (RF) [28].

Although the atomic vapor cell is the core of realizing a quantum sensor, such as a quantum electrometer, it depends mainly on the manufacturer, and its limitations are determined. Here, we introduce the Rb vapor cell assembly process in a lab fabrication system. This work is a necessary technology that must be used to manufacture a vapor cell with a proper structure that minimizes electric field distortion in absolute electric field measurement [29]. To prepare the Rydberg atoms, a vapor cell containing natural Rb atoms was made by a lab fabrication system. The fabricated atomic vapor cell was filled with Rb atoms to avoid collisional dephasing [30]. The system was built up with a vacuum chamber and gas line. After cleaning the stem cell and baking the system, we performed injection of natural Rb atoms. Finally, we achieved a high-purity Rb vapor cell by sealing and welding the Rb-contained stem cell. Inspection of the fabricated vapor cell was performed by achieving saturated absorption spectroscopy (SAS) spectra. By using probe and coupling lasers resonant on the ladder-type EIT transition, we clearly observed AT splitting while the RF source incidents into the lab-fabricated atomic vapor.

Two types of cuvettes, cylindrical and cubic Pyrex (Highborn group) with 50 mm stems, were prepared to contain natural Rb atoms. The outer diameter and length of the cylindrical cuvette were 13 mm and 20 mm, respectively. The thickness of the cylindrical cuvette corresponds to 1 mm. The cubic cuvette was 10 mm long, 10 mm wide, and 10 mm high. The thickness of the cubic cuvette was 0.75 mm.

If the internal diameter of the stem on the vapor cell is relatively small or the length of the stem is relatively long, we notice that it is difficult to move Rb atoms into the cuvette. In this case, it takes a long time to collect the Rb atoms inside the cuvette. In contrast, if the internal diameter of the stem is relatively large, the tip-off process may not be smooth after the completion of collecting Rb atoms. If the length of the stem is relatively short, the Wilson seal may be damaged by heat. Therefore, a proper type of vapor cell suitable for the structure of the chamber to be used is required.

Figure 1 shows 1(a) a simplified schematic and 1(b) photograph of the facility configuration of the chamber system for Rb vapor cell fabrication. The chamber system is composed of a 1.33-inch 6-way cube chamber, a pinch-off tube, three 1/2-inch diaphragm valves, two 1/4-inch diaphragm valves, and one needle valve. A Wilson seal adapter was installed below the 6-way cube chamber to allow attachment of the stem of the vapor cell. An additional flexible resistance foil heater was installed on the Wilson seal adapter. The foil heater effectively minimizes the temperature difference in the movement path of rubidium atoms when rubidium is collected. Due to the reduced temperature difference, few atoms are generated near the Wilson seal adapter and the vapor cell can collect the rubidium more efficiently. A copper pinch-off tube containing a natural Rb break-seal ampoule (99.75% metals basis; Alfa Aesar, MA, USA) was mounted on the right side of the 6-way cube chamber, and a diaphragm valve was installed to control the movement of Rb atoms. A nitrogen gas supply line for purging nitrogen inside the chamber was built on the left side of the 6-way cube chamber. In addition, a vacuum gauge was mounted on the nitrogen gas supply line to measure the degree of vacuum inside the chamber. In order to create a high vacuum inside the chamber, flexible bellows were used in the cube chamber and the nitrogen gas supply line. Those were connected to a pump station consisting of a combination of a diaphragm pump and a turbo pump.

Figure 1. The chamber system for Rb vapor cell fabrication. (a) Simplified schematic, and (b) photograph of facility configuration of the chamber system.

We describe the fabrication process for assembling the Rb vapor cell with cleaning, trapping, and sealing steps, as shown in Fig. 2. Before the vapor cell was mounted in the chamber, a cleaning operation was performed to remove impurities inside the vapor cell. The cleaning operation with a solution of distilled water, ethanol, and methanol was sequentially repeated five times or more. After cleaning, baking was performed at 550 ℃ for 24 hours using a hot plate to evaporate the remaining residues inside and outside the vapor cell.

Figure 2. The fabrication process for assembling the Rb vapor cell with cleaning, trapping, and sealing steps.

After the cleaning steps, the vapor cell was mounted on a Wilson seal adapter. When installing the vapor cell, the purging process is carried out by filling nitrogen gas with positive pressure to prevent contamination inside the chamber. After the installation of the cell was completed, a heating mantle was installed around the vapor cell to raise the temperature about 450 ℃ while making the ultra-high vacuum to completely remove the remaining impurities. The pressure of the pump station is 4.7 × 10⁻⁹ Torr during the baking process. After confirming a sufficient degree of vacuum to collect Rb atoms through the vacuum baking process, the Rb ampoule was placed in a pinch-off copper tube. The copper tube with the ampoule was made to a certain amount of positive pressure through nitrogen purging. If the surrounding environment has more negative pressure than the ampoule, glass and Rb may be scattered around when the ampoule is broken. To break the Rb ampoule, the pinch-off copper tube was compressed with appropriate force. Note that the ampoule is made of thin Pyrex, and the inside of the break-seal ampoule is already sealed in a vacuum, so no contamination of the chamber occurred during the ampoule breaking process. To prevent glass fragments from rapidly scattering into the chamber, a curvature of approximately 20 degrees was applied to the connection part of the pinch-off copper tube where the Rb ampoule is mounted.

After breaking the ampoule, the vacuum operation was performed again. When the chamber system reached a suitable degree of vacuum, the temperatures of the copper pinch-off tube and the chambers part were raised to produce Rb atom vapor. By the vapor pressure of Rb atoms according to temperature referred in [31, 32], the ampoule part and chamber part were set to an appropriate temperature for cold traps to collect the Rb vapor. Specifically, in our system, the temperature of the copper pinch-off tube and the 6-way cube chamber were maintained at 90 ℃ and 100 ℃ through the PID controller, respectively. The vapor cell for collecting Rb atoms was kept at a low temperature using an ice-filled beaker and chiller. In the process of activating rubidium and collecting rubidium vapor in the cell, the vacuum pressure of the pump station was about 1.4 × 10⁻⁸ Torr.

After approximately 48 hours, it was confirmed that Rb atoms were collected in the vapor cell, as shown in Fig. 3. When heat was applied to the stem and the main body of the vapor cell during the tip-off process, we carefully collected Rb atoms on the bottom of the vapor cell to avoid evaporation of the atoms gathered in the vapor cell, as shown in Fig. 3. After collecting the Rb atoms inside the vapor cell, heat was applied to the stem using a gas torch to continue the tip-off process.

Figure 3. Rb atoms collected on the bottom of the vapor cells.

To inspect the quality of the lab-fabricated atomic vapor cell, we compared the hyperfine structure of the D₂ transition of the commercial vapor cell with the fabricated vapor cells. The SAS spectra of Rb atoms with one commercial vapor cell and two fabricated Rb vapor cells are shown in Fig. 4. Upper (black solid line), middle (red solid line with square), and lower (blue solid line with circle) spectra represent SAS signals resulting from the commercial, cubic-shaped and cylinder-shaped Rb vapor cells, respectively. In Fig. 4, we observed F_g = 2 → F_e = 3 transition and F_g = 3 → F_e = 4 transition for the D₂ line of ⁸⁷Rb and ⁸⁵Rb atoms, respectively. The other dominant peaks are crossover peaks of the D₂ line of ⁸⁷Rb and ⁸⁵Rb atoms. All SAS spectra from the vapor cells were achieved at room temperature. The amplitudes of the SAS spectra produced by the fabricated vapor cell were much smaller than the commercial vapor cell. It means that the optical depth difference of the vapor cells used and the strength of the SAS spectra are quite different. However, considering the length of each cell, these results are very reasonable. From the results, we confirmed that our lab-fabricated atomic vapor cells are in good agreement with a commercial one.

Figure 4. Comparison of hyperfine structure of D₂ transition between commercial vapor cell and the fabricated vapor cells.

In this chapter, we will introduce the atomic electrometer based on the lab-fabricated Rb vapor cell. Rb atoms are excited to their highly excited Rydberg state using EIT in a ladder-type scheme. In this experiment, the principal quantum number n was 28, estimated by measuring the resonant RF field. Figure 5 represents an experimental schematic of a ladder-type EIT corresponding to the highly excited Rydberg states. We used one external cavity diode laser (ECDL) near 780 nm as a probe laser and another ECDL near 480 nm as a coupling laser. Dichroic mirrors such as DM1 and DM2 are used for distinguishing the probe and coupling lasers. The frequency of the probe laser was investigated by observing the resonant transition of the Rb vapor cell, and the wavelength of the coupling laser was monitored by using a wavemeter (WM). To avoid feedback damage, isolators (ISOs) were placed in front of both lasers.

Figure 5. Experimental schematic of Rydberg-electromagnetically induced transparency (EIT).

Finally, we could measure the AT splitting spectra with the incident RF field generated by the open-ended waveguide, as shown in Fig. 6. Figure 6(a) shows an energy diagram of the ⁸⁵Rb. Rydberg-EIT and AT splitting spectra were observed while the probe laser was scanning the 5S_1/2 → 5P_3/2 transitions of the ⁸⁵Rb atom, as shown in Fig. 6(b). The central peak and separated peaks represent the Rydberg-EIT and AT splitting spectra, respectively. The RF field used is resonant on the 28D_5/2 → 29P_3/2 transition, while the open-ended waveguide RF source is tuned to 100 GHz. AT splitting spectra as a function of incident RF field power was measured in steps of 1 dBm. With increasing RF power, we also observed an increment of the frequency gap between wing-like peaks.

Figure 6. AT splitting spectra with incident RF field that generated from the open-ended waveguide; (a) Energy level diagram of Rydberg-EIT with AT splitting, and (b) AT splitting spectra resonant on the 28D_5/2 → 29P_3/2 transition as a function of RF field power. Note that Ω_RF in (a) and Λ_EIT in (b) represent RF Rabi frequency and linewidth of Rydberg-EIT. (AT: Autler-Townes; RF: radio-frequency; EIT: electromagnetically induced transparency).

In this paper, we introduced an atomic vapor cell fabrication process and compared fabricated vapor cells with a commercial one. We did not observe any difference between the vapor cells with the measurement of F_g = 2 → F_e = 3 transitions and F_g = 3 → F_e = 4 transitions for the D₂ line of ⁸⁷Rb and ⁸⁵Rb atoms, respectively, and we could confirm that the lab-assembled atomic vapor cell was excellently fabricated. Using this atomic vapor cell, we prepared a highly excited Rydberg atom resonant on the 5S_1/2 → 5P_3/2 → 28D_5/2 transitions of the ⁸⁵Rb atom. With the incident RF field produced by the open-ended waveguide, we could clearly observe an AT splitting signal that was changed from the EIT signal. The spectra of AT splitting on the strength of the input electric field was also measured as a function of incident RF power. We believe that this work can be used to develop a quantum electrometer based on the atomic vapor cell.

The authors declare no conflict of interest.

Data underlying the results presented in this paper are not publicly available at the time of publication, but may be obtained from the authors upon reasonable request.

Institute of Information & communications Technology Planning & Evaluation (IITP) funded by the Korea government (MIST) (No. 2021-0-00890); Development of ultra-sensitive electric field detection technology based on non-metals.