Atomic Vapor Cell with Wall-Integrated Electrodes

• The paper introduces an innovative design that integrates silicon ring electrodes directly into a glass vapor cell, achieving high field uniformity for precise DC and microwave measurements.
• Methodologies include anodic bonding and FEM/HFSS simulations, which are validated by Stark spectroscopy and Autler–Townes splitting to accurately map electric fields.
• Applications extend to compact quantum sensors, atomic clocks, and ion or plasma traps, with scaling insights that enhance field homogeneity and device integration.

An atomic vapor cell with wall-integrated electrodes is a hybrid optical-electronic platform in which electrode structures, typically highly doped silicon rings, are directly embedded within the walls of a glass vapor cell. This device architecture enables precise, local application and measurement of both static (DC) and oscillatory (AC/microwave) electric fields inside the cell using alkali vapor (rubidium, Rb) as a sensitive electromagnetic probe. The integration of electrodes into the vapor cell wall contrasts with traditional external electrode arrangements, permitting higher field uniformity, compactness, and new avenues for quantum metrology and field sensing (Ma et al., 2021).

1. Device Architecture and Fabrication

The atomic vapor cell utilizes a borosilicate (Pyrex) glass vacuum tube, into which eight highly doped, double-side polished silicon (Si) ring electrodes are anodically bonded. Each silicon ring is approximately 0.5 mm thick (matching typical wafer thickness), has an axial width near 2 mm, and an inner diameter closely matching the glass tube bore (approximately 15 mm). The rings are aligned with center-to-center spacing of 6 mm along the longitudinal axis. Glass tubes undergo chemical-mechanical polishing to achieve optical flatness prior to electrode integration.

Anodic bonding is performed at approximately 300 °C, with 600–900 V applied between a Si ring bus and glass surfaces via a custom jig. Following electrode integration, the cell is evacuated, filled with natural-abundance Rb vapor, and flame-sealed. Thin, gold-plated wires are bonded around the outer rim of each Si ring, providing electrical access for external DC biasing of individual electrodes. This configuration is summarized in the table below.

Component	Dimension/Parameter	Function
Si ring	0.5 mm thick, 2 mm wide	Internal electrode
Ring spacing	6 mm center-to-center	Field region definition
Tube inner bore	≃15 mm	Beam/field propagation path
Gold wires	—	External connection

The anodic bonding process enables the monolithic assembly of glass and silicon, preserving vacuum integrity and optical access for spectroscopic interrogation (Ma et al., 2021).

2. DC Electric Field Sensing via Stark Spectroscopy

Rubidium atoms within the vapor cell permit all-optical electric field measurements through Rydberg-EIT-based Stark spectroscopy. In a two-photon ladder configuration, a 780 nm “probe” excites the 5S$_{1/2}$ → 5P$_{3/2}$ transition, while a 480 nm “coupler” drives 5P$_{3/2}$ → 32S$_{1/2}$. In the presence of a static electric field $E_{\text{DC}}$, the 32S$_{1/2}$ Rydberg level experiences a quadratic Stark shift:

$\Delta E = -\frac{1}{2} \alpha E_{\text{DC}}^2$

$$\Delta \nu = -\frac{ \alpha E_{\text{DC}}^2 }{ 2 h }$$

where $\alpha = 2.214$ MHz/(V/cm)$^2$ is the 32S$_{3/2}$0 DC polarizability, and $_{3/2}$1 is Planck’s constant.

Applying voltage to a single Si ring yields a new, red-shifted EIT spectral feature; at 10 V bias, the peak separation of $_{3/2}$2 MHz corresponds to $_{3/2}$3 V/cm (10% relative uncertainty). Finite-element method (FEM) modeling, assuming charge-free glass, overestimates the field by approximately a factor of three. Introducing grounded “sheath” regions adjacent to the biased ring (mimicking surface charge accumulation) attenuates the modeled field by ≃3×, bringing simulated lineshapes into quantitative agreement with experiment without ad hoc scaling.

3. Microwave Field Mapping via Autler–Townes Splitting

The device architecture enables injection and mapping of AC (microwave) fields using Autler–Townes splitting of Rydberg transitions as a local field probe. A horn antenna delivers 18.149805 GHz radiation resonant with the 49D$_{3/2}$4 → 50P$_{3/2}$5 transition; the coupler beam addresses 5P$_{3/2}$6 → 49D$_{3/2}$7. The observed Autler–Townes splitting $_{3/2}$8 corresponds to a microwave Rabi frequency $_{3/2}$9, with $_{3/2}$0 C$_{3/2}$1m the dipole matrix element.

In practice, the total measured splitting is described by

$_{3/2}$2

with a best fit $_{3/2}$3 MHz/$_{3/2}$4, $_{3/2}$5 MHz (residual splitting from DC/magnetic fields). The corrected splitting is

$_{3/2}$6

Field amplitudes inferred from this method for horn powers $_{3/2}$7 to $_{3/2}$8 dBm at polarization angle $_{3/2}$9 span $_{1/2}$0 to $_{1/2}$1 V/m, consistent with theoretical expectations. Polarization dependence follows $_{1/2}$2 at large angles, but shows enhanced leakage for $_{1/2}$3 and remains at $_{1/2}$4 of $_{1/2}$5 for $_{1/2}$6, indicating incomplete polarization rejection due to the cell’s dielectric heterogeneity.

4. Theoretical Modeling and Validation

Both DC and microwave fields in the cell are rigorously modeled and compared with experiment. DC fields are calculated using FEM, including conductive Si electrodes (resistivity $_{1/2}$7–$_{1/2}$8 Ω$_{1/2}$9cm), borosilicate glass ($E_{\text{DC}}$0), and surface-charge-mimicking grounded sheaths. This modeling reproduces the observed field attenuation (factor of ≃3×) when surface charges are considered. AC field distributions are simulated using Ansys HFSS, incorporating the experimental horn-cell geometry. For polarization $E_{\text{DC}}$1, the time-averaged field along the axis is ≃60% of the value without a cell; for $E_{\text{DC}}$2, ≃25%. Simulations reveal microwave field maxima between rings and minima near ring planes, and predict residual ellipticity (minima $E_{\text{DC}}$35–10% of peak), consistent with the measured Autler-Townes signals.

5. Applications and Scaling Considerations

Wall-integrated electrodes enable unprecedented control and measurement of DC and RF fields in vapor cells. Applications include:

• Miniature DC and radiofrequency (RF) quadrupole traps (Paul, Penning, cusp) for ions, electrons, or plasmas, where in situ field control is achieved via embedded electrodes with nonperturbing Rydberg-EIT-based field sensing.
• Compact, SI-traceable microwave and RF sensors embedded in atomic receivers, utilizing RF dressing or Autler–Townes spectroscopy for vector and amplitude microwave field measurements.
• Integrated photonic–atomic devices such as chip-scale atomic clocks, magnetometers, and phase sensors, wherein electrode geometry can be precisely tailored. The electrode spacing ($E_{\text{DC}}$4) sets the length scale for field homogeneity, while ring width ($E_{\text{DC}}$5) and thickness ($E_{\text{DC}}$6) impact field enhancement and capacitance.

Scaling relationships indicate that halving $E_{\text{DC}}$7 increases field uniformity by a factor of ≃4 for a fixed spatial region, at the cost of increased capacitance and bonding voltage demands.

6. Significance and Outlook

The demonstration of an anodically bonded vapor cell with silicon ring electrodes, together with quantitative characterization of internal DC and AC fields using Rydberg atom spectroscopy, establishes a versatile platform for quantum metrology and charged-particle control. The integrated approach allows both precise field delivery and high-fidelity, spatially selective in situ diagnostics. Modeling and experiment are in close agreement for both static and oscillating fields, supporting the reliability of the platform for advanced sensing and microtrap applications (Ma et al., 2021).