First-hand information for everyone
This paper studies what happens when a moving quantum wavepacket hits a region that removes particles. The authors build a one-dimensional “synthetic” lattice from internal states of a rubidium Bose–Einstein condensate. By tuning how strongly sites in one half of the chain lose atoms, they observe three behaviors: nearly lossless ballistic passage, strong absorption, and strong reflection caused by the quantum Zeno effect.
The experimental lattice uses the hyperfine ground states of 87Rb as the sites of a chain. The two subchains are the F=2 manifold (non-dissipative, left side) and the F=1 manifold (dissipative, right side). Radio-frequency fields set the hopping inside each subchain with an average tunneling ¯J ≈ 0.98 ΩRF (ΩRF/(2π)=1.968(4) kHz, so ¯J≈2π×1.929(4) kHz). A microwave field controls the tunneling across the interface (J0). Loss on the right side is created by optically coupling F=1 to an excited state (5P3/2, F′=0) that decays with lifetime 26.3 ns; atoms that decay pick up momentum and leave the measured population. Experiments start with about NT≈10^5 atoms prepared on the leftmost site |j=−4⟩ and track site populations over hundreds of microseconds using Stern–Gerlach separation and time-of-flight imaging.
At small loss rate γ the packet just travels through both halves and returns nearly unchanged. As γ grows to about the size of the tunneling rate (γ/¯J ≈ 1) absorption becomes large. If γ is made much larger than ¯J, absorption drops and the wavepacket is mostly reflected back into the non-dissipative side. The authors explain this as an impedance-matching condition: maximal absorption happens when tunneling into the lossy region and the loss rate are balanced. Extremely large loss acts like an imaginary potential step that reflects the packet, an effect related to the quantum Zeno effect. The paper quantifies these observations: the single-pass absorption proxy peaks near γmax/¯J = 1.34(5), and maximum absorption also occurs for an interface tunneling J0,max/¯J = 1.32(7).
The work matters because it demonstrates controlled dissipation as a tool for steering quantum transport. The synthetic-lattice method gives site-by-site control of tunneling and loss that is hard to achieve in ordinary spatial lattices. The experiments agree well with numerical simulations and with an idealized semi-infinite model, indicating that the basic balance between tunneling and dissipation governs absorption across the observed regimes.
There are important caveats. The lattice here is finite and has a small energy offset at the interface, which causes some reflection even in nominally optimal cases. The absorption is measured with a specific single-pass proxy at a chosen time (t = 7.2/¯J), so reported numbers are context dependent. Also, very strong loss does not improve absorption but instead suppresses it through enhanced reflection (the quantum Zeno effect). The results are presented with experimental uncertainties on the key ratios and with numerical comparisons; they establish principles rather than immediate practical devices.