Nonclassical Light from Large Ensembles of Trapped Ions

The vast majority of physical objects we are dealing with are almost exclusively made of atoms. Because of their discrete level structure, single atoms have proved to be emitters of light, which is incompatible with the classical description of electromagnetic waves. We demonstrate this incompatibility for atomic fluorescence when scaling up the size of the source ensemble, which consists of trapped atomic ions, by several orders of magnitude. The presented measurements of nonclassical statistics on light unconditionally emitted from ensembles containing up to more than a thousand ions promise further scalability to much larger emitter numbers. The methodology can be applied to a broad range of experimental platforms focusing on the bare nonclassical character of single isolated emitters.

Figure 1

The simplified scheme of the experimental setup and laser pulse sequence for generation and detection of nonclassical light. (a) The energy level scheme of the ${}^{40}{\mathrm{Ca}}^{+}$ ion including employed transitions with their respective wavelengths. (b) An ensemble of ions is trapped in the linear Paul trap, and the 397 nm fluorescence emitted along the applied magnetic field direction ($B$ field) is collected by a lens objective and separated from the 866 nm light by an optical interference filter (IF). The fluorescence is then directed towards the nonclassicality analyzing setup comprising a single beam-splitter (BS), a pair of avalanche photodiodes (APDs) and the time-tagging module. The trapped ion crystals can be imaged on an electron multiplying CCD (EMCCD) camera for the purposes of ion number estimation and optimization of trapping stability. (c) The generation of analyzed light from the trapped ion ensemble in the pulsed regime begins by the Doppler cooling period, in which both lasers are switched on. It is followed by the optical pumping stage, where the populations are shuffled to the metastable $3D_{3/2}$ state using only the 397 nm laser. The analyzed fluorescence at the $4P_{1/2}\leftrightarrow 4S_{1/2}$ transition is then generated by fast depopulation of the $3D_{3/2}$ state using the 866 nm laser pulse.

Figure 2

The measured violation of nonclassicality as a function of the number of trapped single-photon emitters and the same measurements realized just with the laser light. All measured values of the witness $d$ for trapped ion ensembles in pulsed (red circles) and continuous (yellow circles) regimes are provably in the nonclassical region given by $d>0$. The green triangles correspond to the numerical simulation of the witness value in the pulsed regime; details of the simulation can be found in the Supplemental Material, Sec. C [24]. Estimated mean photon numbers per single emission pulse $\overline{N}$ in the pulsed regime are shown on the top axis. $d$ for the 397 nm laser light scattered from the trap electrodes has been measured using the same detection scheme in both excitation regimes. The laser intensity has been set to reach the detection count rates corresponding to measurement on light from a given trapped ion ensemble. The dark blue and grey squares are measurements on the laser light in the pulsed and continuous regimes, respectively. All measurements corresponding to a given ion number are grouped in arrowlike shaded regions. Error bars correspond to one standard deviation.