Direct determination of atomic number density in MEMS vapor cells via single-pass absorption spectroscopy (SPAS)

Micro-electro-mechanical systems (MEMS)-based chip-scale alkali vapor cells are the essential components in emerging quantum technologies, including compact atomic clocks, chip-scale magnetometers, and miniature quantum opto-electronic systems. The sensitivity of MEMS-based atomic quantum technology devices depends on the atomic number density. Thus, it is important to have an accurate estimate of the atomic number density in chip-scale alkali vapor cells to optimize light-matter interactions and design efficient quantum sensing systems. Here, we present a direct and quantitatively validated method for determining the rubidium (Rb) number density in warm alkali vapor using single-pass absorption spectroscopy (SPAS). The absolute transmission spectra are measured and modeled using the 780.24nm as well as the 420.29nm transition in a Rb-filled MEMS vapor cell. The atomic number density measurements and the model were also validated using a commercial vapor cell of length 100mm. The theoretical model employs a density-matrix formalism within the Lindblad framework and incorporates directly measurable experimental parameters, such as laser beam power, diameter, and cell temperature. The model explicitly accounts for optical pumping, Doppler broadening, and transit-time broadening effects and exhibits quantitative agreement ($> 99%$) with experimental spectra over a broad range of temperatures (293–343K), laser intensities ($\sim0.2, I_{\mathrm{sat}}$ to $\sim2, I_{\mathrm{sat}}$), and cell lengths (2–100~mm). The extracted densities from the MEMS cell closely follow the empirical vapor-pressure model by Alcock et al. The demonstrated methodology provides a practical, well-controlled method for determining the atomic number density in alkali vapor cells relevant to the characterization and development of compact alkali-vapor-based devices for quantum sensing and metrology.

💡 Research Summary

This paper presents a rigorously validated technique for directly determining the rubidium (Rb) atomic number density in warm alkali vapor cells, with a focus on micro‑electro‑mechanical systems (MEMS) chip‑scale devices. The authors employ single‑pass absorption spectroscopy (SPAS) on both the D₂ transition at 780.24 nm and the higher‑lying transition at 420.29 nm. Absolute transmission spectra are recorded for a MEMS cell (2 mm path length) and a conventional 100 mm reference cell over a temperature range of 293–343 K, laser intensities from ≈0.2 Isat to ≈2 Isat, and various beam diameters.

The theoretical framework is built on a density‑matrix master equation solved within the Lindblad formalism. Each stable isotope (⁸⁵Rb, ⁸⁷Rb) is modeled as a four‑level system comprising two hyperfine ground states and three excited states, allowing explicit treatment of optical pumping, spontaneous emission, pure dephasing, and transit‑time decay. The Hamiltonian includes the relevant Rabi frequencies and detunings, and the rotating‑wave approximation is applied to obtain a time‑independent representation. The Lindblad super‑operator incorporates decay channels with experimentally known branching ratios and dephasing rates.

In steady state (∂ρ/∂t = 0) the off‑diagonal density‑matrix elements ρij provide the optical coherences, from which the complex electric susceptibility χ(Δ,T) is calculated as χ = –2 N(T) D²/(ħε₀Ω) ρij. Here N(T) is the only free parameter; all other quantities—cell length L, temperature T, laser power P, beam waist w—are measured independently. Doppler broadening is introduced by integrating over the Maxwell‑Boltzmann velocity distribution, while transit‑time broadening is modeled as Γtt ≈ v̄/w, with v̄ the mean thermal speed.

Experimental spectra are fitted by adjusting N(T) to minimize the mean‑square deviation between measured and simulated transmission. Across the full parameter space the model reproduces the data with > 99 % correlation. The extracted densities follow the empirical vapor‑pressure relation of Alcock, Itkin, and Horrigan (log₁₀ Pvap = 2.881 + a – b/T) and the ideal‑gas law N = Pvap/(kBT) to within experimental uncertainty.

Key advantages of this approach are: (1) it requires no external calibration beyond absolute transmission; (2) it simultaneously accounts for optical pumping, Doppler, and transit‑time effects, yielding high‑precision density values even for short optical paths; (3) it is computationally efficient because the four‑level model captures the essential physics of both isotopes without excessive matrix size; (4) it is readily extensible to other alkali species or molecular systems. Limitations include sensitivity to laser linewidth and beam profile uniformity, and potential deviations at temperatures above 350 K where cell materials may introduce non‑ideal pressure behavior.

The authors suggest future extensions such as incorporating multi‑step excitation schemes (e.g., 5 S→5 P→5 D), exploiting quantum‑interference effects for sub‑Doppler resolution, and integrating real‑time temperature feedback to maintain density stability in field‑deployed devices. By providing a straightforward, accurate, and broadly applicable method for measuring atomic number density, this work supports the design, optimization, and metrological validation of compact quantum sensors, atomic clocks, and integrated photonic‑atomic platforms.