Atomic Thermometers

Summary

The Atomic Thermometers program aims to establish primary, quantum-based sensors of both radiative and thermodynamic temperature in the range 250 K – 400 K. Atoms may absorb or emit blackbody radiation (BBR) to change their quantum state; the temperature dependence of these transitions is calculable by ab inito methods. By monitoring the populations in one or more quantum states, the radiative temperature can be inferred.

Description

Approach 1

Compact Blackbody Radiation Atomic Sensor (CoBRAS)

The Compact Blackbody Radiation Atomic Sensor (CoBRAS) uses a thermal vapor of atoms excited by a single laser to detect BBR. From the optically excited state, atomic population is transferred to other, nearby excited states by a combination of spontaneous decay and BBR-stimulated transitions. Each of the nearby states emits fluorescence at one or more unique wavelengths. The fluorescence intensity ratio of two transition wavelengths depends on the temperature-dependent steady-state populations of the various excited states. Monitoring the fluorescence intensity ratios gives a temperature dependent signal. Monitoring the same fluorescence intensity ratio when using the laser to excite to a different optically excited state changes the fluorescence intensity ratio by a calculable amount that only depends on atomic properties. This allows for the total collection efficiency of the fluorescence detector to be calibrated to the atom, resulting in a primary thermometer.

Atomic vapor cells are intrinsically compact containers of identical quantum systems (i.e. atoms). By constructing the vapor cell to be either transparent or emissive at the sensed blackbody radiation wavelength, the measurement may be sensitive to the blackbody radiation of the environment surrounding the vapor cell (non-contact thermometer) or to the blackbody radiation emitted by the vapor cell itself (contact thermometer). This makes CoBRAS a promising platform for practical thermometry, such as platinum resistance thermometer replacement.

Using a rubidium vapor cell, we have demonstrated a CoBRAS with a statistical uncertainty as low as 0.1% in one second. We resolve temperature with a precision of 0.04 % in the range 308 K to 344 K when averaging for several seconds. When calibrating to a transfer standard, the total demonstrated temperature accuracy of 1.0% was limited by thermal gradients present in our measurement apparatus, which can be reduced by roughly a factor of 100 in an improved enclosure. When acting as a self-calibrated thermometer, the total uncertainty is 1.5% owing to theoretical uncertainty in atom transition dipole matrix elements.

a) Simplified level diagram for the CoBRAS. A laser at 359 nm excites atoms from the ground 5S state to the 7P3/2 state. Atoms can then be further excited by blackbody radiation into states including the 8S or 6D (blue lines) or decay via spontaneous emission to lower energy states including the 7S and 5D states (green lines). Fluorescence from 6D, 7S, and 5D to 5P3/2 (red arrows) is monitored by PMTs with the appropriate filter wavelengths indicated. (b, top) The ratio of PMT signals recorded with the 630 nm filter to the 740 nm filter vs. temperature. Blue points indicate the data; solid black curve the thoeretical prediction with one fitted parameter. Gray shaded region indicates the uncertainty in the theory due to atomic physics parameters. (b, bottom) fractional residual of between the data and prediction. Blue band shows the mean and root-mean-squared uncertainty. (c) The ratio of PMT signals recorded with the 762 nm filter to the 740 nm filter (top) and the fractional residuals between the data and prediction (bottom) vs. temperature. Same symbols as is (b).

Credit: NIST

Approach 2

Cold Atom Thermometer (CAT)

In the Cold Atom Thermometer, atoms are excited to a Rydberg energy level via a two-photon excitation. From there, blackbody radiation drives transitions to other, nearby Rydberg states. The populations of the various Rydberg states are then measured as a function of interaction time with the BBR using the selective field ionization (SFI) technique. The rate of population transfer is determined solely by the BBR temperature and fixed properties of the atomic transitions, resulting in a primary thermometer.

Collisions between atoms excited to the Rydberg state and ground-state atoms can cause the nearly-free electron of the Rydberg atom to change state or be completely ionized. To suppress these effects, we laser cool atoms to temperatures of ∼ 0.5 mK in a magneto-optical trap. Furthermore, the cold atomic temperature reduces the spread in the time-of-flight of the ionized atoms for improved state resolution in the SFI signal compared to a thermal vapor.

Using a MOT of rubidium atoms, we have demonstrated a CAT with a statistical uncertainty as low as 9% in one second. We resolve temperature with a precision of 0.5 % in the range 297 K to 338 K when averaging for a few minutes. The total demonstrated temperature accuracy of 0.8% was limited by overlap in the time-of-flight of the SFI signal due to adjacent states. Improved electronic readout techniques should be able to reduce this systematic by many orders of magnitude.

(a) Relevant level diagram of 85Rb for the cold atom thermometer (CAT) experiment. Atoms in the 5S1/2 ground state are pumped using a 780 nm laser into 5P3/2 intermediate state and further excited using a 480 nm laser into the 32S1/2 state. Blackbody radiation (BBR) then transfers some of the atoms into nearby P states, including 30P, 31P, and 32P. (b) Relative population in various states as a function of BBR interaction time t at T = 296 K. The theoretical curves are calculated by a rate equation model (2.3), using decay rates from ARC [8]. (c, left panel) Ratio of the 32S+31P peak to the 32P peak R as a function of blackbody interaction time. Data are shown as points, theoretical prediction with two fit parameters shown as solid curves. (c, right panel) The atomic-determined temperatures (points) together with the contact thermometry determined temperatures (horizontal lines with corresponding uncertainty bars).

Credit: NIST

Opportunities

If you are interested in joining our team as a postdoc, guest researcher, collaborator, or student volunteer, send us an email.