Optical spectroscopy of Bose-Einstein condensates at finite temperature
R. M. F. Andersen, L. N. Stokholm, I. Zebergs, N. R. Neubert, A. S. Chatterley, N. Kjærgaard, J. J. Arlt · 2026
Read Optical spectroscopy of Bose-Einstein condensates at finite temperature by R. M. F. Andersen, L. N. Stokholm, I. Zebergs, N. R. Neubert, A. S. Chatterley, N. Kjærgaard, J. J. Arlt (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-24568.txt.
Introduction
Andersen and colleagues read the temperature of an ultracold atom cloud from the colours it scatters. Bose-Einstein condensates, matter waves formed when atoms crowd into one quantum state, coexist with a thermal component at finite temperature. Telling the two apart in place has been difficult. Their 2026 measurements use optical spectroscopy on rubidium-87 gases, with a dark field setup sensitive to single photons and a frequency agile laser sweeping the spectrum.
Distinct spectral features from the thermal and condensed components appear directly in the frequency domain. A model of light travel through the uneven cloud interprets the spectra. Temperature and atom number are extracted, benchmarked against standard time of flight absorption imaging. The spectral parameters prove sensitive to small thermal fractions in nearly pure condensates, exactly where imaging struggles. Dark field operation beats bright field clearly. Andersen's team establishes the method as a mapping tool for ultracold systems.
Catalog record
| CREATOR | R. M. F. Andersen, L. N. Stokholm, I. Zebergs, N. R. Neubert, A. S. Chatterley, N. Kjærgaard, J. J. Arlt |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 42 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.24568
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Optical spectroscopy of Bose-Einstein condensates at finite temperature R. M. F. Andersen, L. N. Stokholm, I. Zebergs, N. R. Neubert, A. S. Chatterley, N. Kjærgaard, J. J. Arlt Abstract We report on optical spectroscopic measurements of ultracold and partially condensed 87 Rb gases, which show distinct spectral features due to the thermal and the Bose condensed components in the frequency domain. These features are detected in-situ by using a dark-field configuration with single-photon sensitivity and a frequency-agile laser system. To interpret the observed spectra, we develop a model for light propagation through an inhomogeneous atomic cloud. This model enables the extraction of temperature and atom number, which we benchmark against conventional time-of-flight absorption imaging. The spectroscopically obtained cloud parameters show enhanced sensitivity to small thermal fractions in nearly pure condensates. We further compare the spectroscopy in dark-field and bright-field configurations, demonstrating the superior performance of the former. Our results reveal previously unexplored spectral structure in optically dense ultracold gases and establish spectroscopy as a tool for characterizing ultracold systems at very low temperatures. I Introduction The realization of Bose–Einstein condensation (BEC) in dilute atomic gases has had a significant impact on modern physics, providing an unprecedented platform for exploring quantum phenomena on macroscopic scales 32 ; 33 . The high degree of experimental control available in ultracold atomic systems, including external potentials, dimensionality, and interactions, has enabled their application in quantum simulation, fundamental physics, and in particular for the investigation of many-body physics 5 ; 34 . Thus, the continued development of accurate probing techniques for BECs is essential for understanding these strongly correlated quantum systems. The detection of ultracold atomic clouds is typically achieved by optical methods, based on the interaction of the constituent atoms with light. These methods can broadly be divided into resonant and dispersive techniques. These techniques have been reviewed in detail recently 40 ; 39 . Resonant imaging techniques rely on detecting photons resonantly scattered by the atomic sample, with absorption and fluorescence imaging being the most widely used approaches. In both cases, the atom number is inferred from the number of photons scattered. Fluorescence imaging, which directly collects this scattered light, has seen widespread use across atomic physics experiments. Absorption imaging, in particular, has become a cornerstone of ultracold atomic physics, including investigations of Bose–Einstein condensation 20 ; 36 . Notably, it can also be implemented in a dark-field configuration, where spatial filtering in the Fourier plane suppresses the unscattered probe beam, enhancing sensitivity to weak signals 29 ; 37 . Dispersive imaging techniques, in contrast, measure the phase shift imprinted on off-resonant probe light by the spatially varying refractive index of the atomic cloud 3 ; 8 . Since they do not rely on photon absorption, these methods can be made minimally destructive, enabling repeated probing of the same ultracold sample. The resulting signal depends sensitively on the density-dependent refractive index of the gas 12 . In practice, such techniques are typically operated at a fixed, non-zero detuning, necessitating a degree of spectroscopic understanding of the system—even when a complete spectral characterization is not explicitly performed. At low temperatures and large condensate fractions, it becomes increasingly difficult to extract the thermal atom number and temperature using imaging techniques, since the small thermal component is typically obscured by the BEC. Nonetheless, absorption imaging can be used to characterize very cold finite temperature BECs 22 , which can be exploited in low-gravity environments 11 . Alternative techniques include using embedded impurities such as atoms 16 or magnons 28 , and Bragg scattering 13 . Although optical spectroscopy is the standard tool for probing atomic systems, the complex spectral signatures that arise in partially condensed ultracold gases have received relatively limited attention. The most important previous application was the spectroscopic detection of the condensation of atomic hydrogen 10 . Moreover, optical spectroscopy has been applied in cold atomic gases to investigate dipolar interactions 17 ; 19 ; 30 or electromagnetically induced transparency 15 ; 35 ; 31 . This paper examines optical-frequency spectroscopy of partially condensed ultracold atomic clouds at the single-photon level. We observe rich spectral features that arise from the large in-situ optical densities of the samples. The spectra are investigated using a dark-field (DF) configuration, where the unscattered light is suppressed, as well as a bright-field (BF) setup, where the transmitted light is directly measured. The DF method is sensitive to both absorption and phase-shift, resulting in enhanced sensitivity to the system parameters. The spectroscopic measurements are compared with a theoretical model, which is also used to extract the thermal atom number, condensed atom number, and temperature. These results are benchmarked against absorption imaging to validate the method. An advantage of using spectroscopy is that it can be measured in-situ, eliminating the need to model expansion dynamics, which is especially difficult in the high condensate fraction regime 2 . The non-destructive nature of the spectroscopic method also allows for repeated measurements of the same cloud. Furthermore, the spatial resolution requirements for obtaining spectroscopic information are less stringent than for imaging methods. This manuscript is organized as follows. Section II introduces the experimental techniques and theoretical framework necessary to measure the spectra and extract cloud parameters. It also discusses the novel spectral features caused by high in-situ optical densities of ultracold clouds. Section III compares the extracted cloud parameters from fitted spectra to the established TOF imaging method and illustrates the advantages of applying the spectroscopic method for very cold partially condensed clouds. Section IV summarizes the results and discusses its applicability in future experiments. II Spectroscopy of partially condensed atomic clouds II.1 Experimental setup The experiments are performed on clouds of 87 Rb atoms prepared in the \lvert F=2,m_{\text{F}}=2\rangle state and confined in a crossed optical dipole trap (cODT) at a wavelength of 1064 nm with approximately 80\text{\,}\mathrm{\SIUnitSymbolMicro m} beam waists. Using standard evaporative cooling techniques, BECs containing up to 6.5\cdot 10^{5} atoms can be produced. The condensate fraction is controlled by adjusting the final dipole trap power. A detailed description of the preparation sequence can be found in 1 ; 38 . Figure 1 : Schematic of the detection setup. Collimated probe light (orange) is sent through a cloud of trapped atoms, and some of the light is scattered (blue). At the focus plane of the probe beam, an optional small dark-field obstacle can be placed to block only the unscattered light. Additional filtering is achieved by a pinhole at an intermediate image plane. The image is either relayed onto a camera or focused onto a single-photon detector (SPD). To perform spectroscopy, \sigma^{+} -polarized probe light drives the cycling transition between the \lvert 2,2\rangle and the \lvert 3,3\rangle excited state. The probe laser is beat-locked to a reference laser using an optical phase-lock loop (OPLL) 4 , enabling agile frequency detuning on a 9\text{\,}\mathrm{GHz} frequency range with a maximal sweep rate of 90\text{\,}\mathrm{MHz}\text{/}\mathrm{ms} . A peak probe field intensity of \approx 10^{-6}I_{\mathrm{sat}} , is used, where I_{\mathrm{sat}} is the saturation intensity. Thus, on average, at most one photon is present in the cloud at any moment. The spatial mode of the probe field is much larger than the cloud. The transmitted probe light is relayed through an imaging system onto a single-photon