How macroscopic can a quantum state be and still retain its entanglement with a companion microscopic system? In recent years considerable experimental effort has been devoted to the observation of micro–macro entangled states. Writing in Nature Physics, Natalia Bruno et al.1 and Alexander Lvovsky et al.2 describe two remarkable experiments that report micro–macro optical entanglement, with the macro subsystem involving as many as 108 photons.

One of the fundamental questions of quantum physics is where the boundary between the classical and quantum worlds lies. Whereas microscopic systems obey the quantum superposition principle, the macroscopic realm is characterized by mutually exclusive realities. Is it then possible to establish a bridge between these two regimes? Such a situation is realised in Erwin Schrödinger's paradigmatic thought-experiment in which a cat (as an example of a macroscopic system) is locked in a box with a decaying atom (the microscopic counterpart). The quest for micro–macro entanglement reflects the recent growing interest in observing quantum phenomena in larger physical systems — quantum superpositions of optical fields in a cavity, trapped ions, fullerene molecules, ultracold atoms and superconducting circuits.

A direct approach to the generation of micro–macro entanglement is to start from a micro–micro quantum state and then amplify one of the two subsystems. This quantum amplification of entanglement was pioneered3 in 1998 and was later pursued in various experiments4. However, the amplification process is not a trivial operation in the quantum domain, where the no-cloning theorem implies that the amplified field is not a simple copy of the initial micro system3. By exploiting parametric amplification from a nonlinear crystal, micro–macro systems have been experimentally generated, but the demonstration of entanglement was not conclusive4,5,6. In fact, this would require the measurement of the macroscopic part of the superposition with single-photon resolution5,6,7, or adopting high-efficiency homodyne measurements7,3 — and both scenarios are technically very challenging.

Along the lines of a recent proposal8, Bruno et al.1 and Lvovsky et al.2 designed and implemented a clever approach that exploits linear optics instead of non-linear amplification, leading to reliable state generation and measurement. The two experiments exploit the concept of entanglement between one photon and the vacuum9: a single photon is sent into a beamsplitter, the output state being the coherent superposition of a state where the optical mode a is in vacuum and mode b carries one photon, and the complementary case, where a is populated (Fig. 1). This embodies the initial micro–micro entangled state. To take the quantum micro system a all the way to the macroscopic regime, both experiments adopt a smart trick: mode a is allowed to interfere with an intense light beam on a highly unbalanced beamsplitter1,2. This process — which effectively displaces the state of mode a in its phase space — is described by a local unitary transformation, thus fully preserving the entanglement of the initial state. However, now the overall state involves a micro–macro system.

Figure 1: Experimental set-up for the observation of micro–macro entanglement.
figure 1

A micro–micro entangled state is generated by sending a single photon into a beamsplitter. The micro subsystem a is combined with a strong light beam on an unbalanced beamsplitter to transfer the state to the macro regime. Finally the macro part is displaced back to the (micro) single-photon level to verify the presence of entanglement.

Several questions pop up: how do we detect this hybrid entanglement? Does the sensitivity of the micro–macro state to the environmental coupling increase with the size of the macro system? Is this a 'proper' micro–macro entangled state?

To verify the entanglement of the final state the macro-system was displaced back to the single photon level (by another local unitary analogous to the first displacement operation). There the existence of micro–micro quantum correlations provides evidence for the existence of entanglement in the micro–macro state.

The experiments may seem 'easy' to perform, but the larger the size of the macroscopic optical field, the higher the sensitivity to phase fluctuations. This is the behaviour one would expect from a macroscopic system with quantum features. To achieve a highly phase-stable set-up, Bruno et al.1 built a proper stabilization system, whereas in order to avoid instabilities Lvovsky et al.2 exploited beams with orthogonal polarization propagating along the same optical mode.

The final question relates to the effective 'size' of the quantum macroscopicity — does the macro field corresponding to the displacement of zero or one photon essentially behave like a cat which can be either alive or dead, as in Schrödinger's original paradox? To this end, different criteria have been proposed and the most suitable one is still debated within the community10.

The increasing sensitivity of the system to environmental decoherence can be used to support the macroscopic character of the quantum state. However, to be even more faithful to Schrödinger's paradigm, one needs to generate two macroscopic states that can be classically discriminated. This condition can be translated into the single-shot state distinguishability using detectors with no microscopic resolution. This is fulfilled for the implemented micro-macro states1,2, albeit with some intrinsic error probability in distinguishing between the two states. Whereas the discrimination efficiency is predicted to be 90% (ref. 2), due to technical imperfections Lvovsky et al. achieved a discrimination efficiency of about 68%.

The two experiments will offer new perspectives and will certainly inspire the design of other physical platforms for similar studies. A relevant challenge is to conceive a viable method for the direct measurement of micro–macro entanglement without displacing the macroscopic quantum state back to the single-photon level. Other intriguing directions may involve the coupling of optical fields to atomic or mechanical systems through resonant reflection or radiation pressure mechanisms to transfer the superposition of the optical fields to a superposition of massive objects. Finally, one could apply the displacement operations on both subsystems to obtain a macro–macro entanglement.