Experimental Test of Macroscopic Realism Problem Using a Superconducting Flux Qubit

1. Introduction

NTT and the University of Illinois have conducted an experimental test on the breaking of macroscopic realism^*1 in a superconducting circuit using a flux qubit^*2.

The idea that an observed object exists even before it is observed is called realism. This is commonly accepted in the macroscopic world in which we live. In contrast, in the microscopic world described by quantum mechanics^*3, realism is broken^*4. This means that a state cannot be determined before it is observed and is in fact determined by the observation. If the macroscopic world is also described by quantum mechanics, realism must be broken there. The question is, can realism be broken in the macroscopic world, or is there a limitation to the application of quantum mechanics in the macroscopic world? This is one of the biggest unsolved problems since the discovery of quantum mechanics.

We tested this using a superconducting flux qubit that had 10¹² electrons flowing around it per second and found that realism was broken in this supercurrent. We demonstrated that quantum mechanics is applicable on a macroscopic scale in this supercurrent. This contributes significantly to the field of fundamental physics and shows that a superconducting flux qubit works as a real quantum device.

This work was published in the online version of Nature Communications on the 4th of November, 2016 (GMT) [1].

2. Background

Even a single electron passing through a double slit exhibits interference fringes^*5. Interference fringes do not appear if the electron passes through only one slit. This means that superposition is realized between the left slit passed state and the right slit passed state. These interference fringes disappear when we obtain information (by observation) revealing the slit that the electron passed through. This means that realism, as it is commonly accepted in the macroscopic world, does not hold in the microscopic world described by quantum mechanics. We can expect realism to be broken in the macroscopic world if this world also obeys quantum mechanics. The purpose of our work is to demonstrate the breaking of macroscopic realism, and thus imply that our commonly held belief is wrong (Fig. 1).

Fig. 1. Breaking of macroscopic realism.

3. Experiment

NTT Basic Research Laboratories carried out a test experiment on breaking macroscopic realism using a superconducting flux qubit made of aluminum and demonstrated breaking of the realism of a superconducting flux qubit current state by 84 times the standard deviation. This work demonstrated that quantum mechanics can be applied on a macroscopic scale with a large supercurrent and is not simply limited to a microscopic scale.

The experiment was performed by cooling a superconducting qubit to 10 mK. This enables us to prepare qubits in the lowest energy state (ground state: –1) because there is no thermal excitation in this temperature region. We then read out the qubit state after performing two operations (qubit state rotations) on the superconducting flux qubit. We compared the results obtained with and without observation during these two qubit operations.

We prepared the flux qubit in a coherent superposition of its ground (–1) and excited (+1) states with a first state operation and measured the difference between the results with and without observation (main experiment^*6). If realism is true, we can expect the readout result to be independent of the observation because the state is determined before the observation. On the other hand, if realism is broken by quantum superposition, the qubit state will become +1 or –1 as a result of the observation. In our case, differences appeared in the readout results because of the observation.

Next, we carried out control experiments to confirm that the observation did not disturb the qubit state (control experiment^*7). We prepared +1 or –1 states by using the first state operation and measured the difference between the results caused by the existence of the observation. By confirming that this difference is small, we can confirm the non-invasiveness of the observation, which means that our (+1, –1) control state was not changed by the observation. The difference from the result of our main experiment was well beyond that of the control experiments, meaning that we successfully demonstrated realism breaking on macroscopic objects such as a current state.

4. Technical features

(1) Measurement method mathematically equivalent to Leggett-Garg inequality^*8

To confirm the breaking of realism on a physical system, we first need to demonstrate the breaking of the condition called the Leggett-Garg inequality. This inequality is always satisfied if realism is correct; however, it can be broken in systems that do not obey realism, for instance, quantum mechanics. It is difficult to show the breaking of the Leggett-Garg inequality directly because we have to measure the qubit state at least three times using high fidelity detection while the qubit maintains its quantum characteristics. These are hard conditions to achieve experimentally. We used an alternative method that is mathematically equivalent to the Leggett-Garg inequality and that avoids this detection difficulty.

(2) Superconducting flux qubit and state observation

To show the breaking of macroscopic realism, we need to detect small disturbances caused by observing a macroscopic system. A superconducting flux qubit has two states corresponding to +1 and –1. In each state, a few hundred nanoamperes of superconducting circulating current flow clockwise and anti-clockwise, respectively. These currents correspond to 10¹² electrons flowing around the loop per second, and this is a sufficiently macroscopic value to be detected with a current meter. It is necessary to observe the qubit current state without disturbing it. However, quantum non-demolition measurement^*9, which can achieve this, has proved difficult to realize. We achieved fast and highly accurate quantum non-demolition measurement of a superconducting flux qubit current state using a Josephson bifurcation amplifier^*10.

5. Future plans

We aim to demonstrate measurements with much less observation disturbance and to increase the macroscopic nature of the system by using a large supercurrent qubit or an ensemble of superconducting flux qubits.

Reference

For Inquiries

Public Relations,

NTT Science and Core Technology Laboratory Group

Email: a-info@lab.ntt.co.jp