Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Dec 14:8:129.
doi: 10.1038/s41378-022-00468-x. eCollection 2022.

Wafer-level vapor cells filled with laser-actuated hermetic seals for integrated atomic devices

Affiliations

Wafer-level vapor cells filled with laser-actuated hermetic seals for integrated atomic devices

Vincent Maurice et al. Microsyst Nanoeng. .

Abstract

Atomic devices such as atomic clocks and optically-pumped magnetometers rely on the interrogation of atoms contained in a cell whose inner content has to meet high standards of purity and accuracy. Glass-blowing techniques and craftsmanship have evolved over many decades to achieve such standards in macroscopic vapor cells. With the emergence of chip-scale atomic devices, the need for miniaturization and mass fabrication has led to the adoption of microfabrication techniques to make millimeter-scale vapor cells. However, many shortcomings remain and no process has been able to match the quality and versatility of glass-blown cells. Here, we introduce a novel approach to structure, fill and seal microfabricated vapor cells inspired from the century-old approach of glass-blowing, through opening and closing single-use zero-leak microfabricated valves. These valves are actuated exclusively by laser, and operate in the same way as the "make-seals" and "break-seals" found in the filling apparatus of traditional cells. Such structures are employed to fill cesium vapor cells at the wafer-level. The make-seal structure consists of a glass membrane that can be locally heated and deflected to seal a microchannel. The break-seal is obtained by breaching a silicon wall between cavities. This new approach allows adapting processes previously restricted to glass-blown cells. It can also be extended to vacuum microelectronics and vacuum-packaging of micro-electro-mechanical systems (MEMS) devices.

Keywords: Applied optics; Physics.

PubMed Disclaimer

Conflict of interest statement

Competing interestsThe authors declare no competing interests.

Figures

Fig. 1
Fig. 1. Methods for filling alkali vapor cells.
a Conventional alkali vapor cell filling based on glass-blowing. Empty glass cells are fused to a manifold providing access to an alkali metal source, gas cylinders and a high-vacuum pump. Sealing is achieved by heating a thin glass capillary, which shrinks due to the pressure difference. Once the channel is fully closed, the cell can be pulled away and pinched-off from the manifold. b Concept wafer layout integrating laser-actuated make-seal and break-seal structures. Multiple cells are arranged in an array and connected through a common channel providing access to a single alkali metal source (e.g. a dispenser pill). Cells also feature gas reservoirs initially separated from the main cell chamber. Integrated laser-actuated make-seal and break-seal structures allow connecting or isolating the different chambers and channel. After filling and sealing, the individual cells can be released by saw-dicing
Fig. 2
Fig. 2. Test wafer of make-seal (MS) structures.
a Make-seal demonstrator diagram. b Top view of a cell wafer with 4-cell clusters having different membrane diameters, all connected to a square dispenser cavity. c Individual cell released by saw-dicing, the channel with 40 µm-diameter mouth is visible under the glass membrane. The latter has been locally heated by laser to deflect it towards the channel mouth and seal it. d Cross-section of a sealed glass membrane obtained by saw-dicing, revealing the sealed vertical channel as well as the stack of 5 different substrates made of borosilicate glass and silicon. The white scale bars are 1 mm long
Fig. 3
Fig. 3. Spectroscopic measurements of Cs-vapor cells equipped with make-seal structures.
D1 absorption lines contrast and full width at half maximum (FWHM) for a 6-cell cluster (a) and a 4-cell cluster (b) preceding and following the make-seal actuation. Despite a slight and temporary change of the contrast, the values are not affected by laser sealing. Note that the measurement gaps correspond to MS actuation campaigns of other cell clusters during which the wafer was removed from the spectroscopy bench. c Evolutions of absorption contrast and FWHM during 30 weeks for a dozen of sealed and individually diced cells, including MS-CP shown in Fig. 2c. Note that the time origin (t = 0 days) corresponds to the date of activation of the dispensers. The graph therefore starts at t = 240 days when the cells were diced. Average values of contrast and linewidths are slightly larger in c than in a and b due to better thermalization of the individual cells compared to the wafer
Fig. 4
Fig. 4. Test wafer of break-seal structures.
a Break-seal demonstrator diagram. b Top-view of a cell wafer with different reservoir patterns. c Saw-diced individual cell, where the inner walls separating the reservoirs from the main cavity are 50 µm thick. The scale bar is 1 mm long. d Connection between cavities generated through femtosecond laser ablation of inner Si walls. The cell here has been monitored for almost 6 months, during which 3 reservoirs have been sequentially opened as shown in Fig. 5a
Fig. 5
Fig. 5. Spectroscopic measurements of cesium vapor cells equipped with break-seal structures.
a The cell shown in Fig. 4d has been monitored with linear absorption spectroscopy for almost 6 months, during which 3 reservoirs have been sequentially opened. This is well visible in the evolution of contrast and FWHM of absorption lines recorded during sequential activation of break-seals, where contrast is decreased and FWHM increased stepwise. b The same cell (BS-EL) has afterwards been measured on a coherent population trapping (CPT) clock bench for nearly 1 month. The clock frequency drift is measured at the level of 0.06 Hz/day, i.e. 6.5 × 10−12 fractional frequency stability at one day, despite a remaining adjacent gas reservoir filled with 186 Torr of neon
Fig. 6
Fig. 6. Fabrication flow-charts.
Flow-charts of the am make-seal and nw break-seal demonstrators. The processes are applied on 4-inch wafers, which are afterwards saw-diced into individual cells once the make-seal is actuated and the cells separated from the dispenser cavity. Step m: Cell wafer axonometric view enlightening the channel network structured in the cell lid. Note that fabrication of a cell with both seals is straightforward and solely requires switching the make-seal cell body with the break-seal one

References

    1. Bothwell T, et al. Resolving the gravitational redshift across a millimetre-scale atomic sample. Nature. 2022;602:420–424. doi: 10.1038/s41586-021-04349-7. - DOI - PubMed
    1. Boto E, et al. Moving magnetoencephalography towards real-world applications with a wearable system. Nature. 2018;555:657–661. doi: 10.1038/nature26147. - DOI - PMC - PubMed
    1. Wade CG, et al. Real-time near-field terahertz imaging with atomic optical fluorescence. Nat. Photonics. 2017;11:40–43. doi: 10.1038/nphoton.2016.214. - DOI
    1. Degen C, Reinhard F, Cappellaro P. Quantum sensing. Rev. Mod. Phys. 2017;89:035002. doi: 10.1103/RevModPhys.89.035002. - DOI
    1. Knappe S, et al. A microfabricated atomic clock. Appl. Phys. Lett. 2004;85:1460–1462. doi: 10.1063/1.1787942. - DOI