## Abstract

Superconducting quantum systems (artificial atoms) have been recently successfully used to demonstrate on-chip effects of quantum optics with single atoms in the microwave range. In particular, a well-known effect of four wave mixing could reveal a series of features beyond classical physics, when a non-linear medium is scaled down to a single quantum scatterer. Here we demonstrate the phenomenon of quantum wave mixing (QWM) on a single superconducting artificial atom. In the QWM, the spectrum of elastically scattered radiation is a direct map of the interacting superposed and coherent photonic states. Moreover, the artificial atom visualises photon-state statistics, distinguishing coherent, one- and two-photon superposed states with the finite (quantised) number of peaks in the quantum regime. Our results may give a new insight into nonlinear quantum effects in microwave optics with artificial atoms.

## Introduction

In
systems with superconducting quantum circuits—artificial atoms—strongly
coupled to harmonic oscillators, many amazing phenomena of on-chip
quantum optics have been recently demonstrated establishing the
direction of circuit quantum electrodynamics^{1,2,3}, particularly, in such systems one is able to resolve photon number states in harmonic oscillators^{4}, manipulate with individual photons^{5,6,7}, generate photon (Fock) states^{8} and arbitrary quantum states of light^{9}, demonstrate the lasing effect from a single artificial atom^{10}, study nonlinear effects^{11, 12}. The artificial atoms can also be coupled to open space^{13}(microwave transmission lines) and also reveal many interesting effects such as resonance fluorescence of continuous waves^{14, 15}, elastic and inelastic scattering of single-frequency electromagnetic waves^{16, 17}, amplification^{18}, single-photon reflection and routing^{19}, non-reciprocal transport of microwaves^{20}, coupling of distant artificial atoms by exchanging virtual photons^{21}, superradiance of coupled artificial atoms^{22}.
All these effects require strong coupling to propagating waves and
therefore are hard to demonstrate in quantum optics with natural atoms
due to low-spatial mode matching of propagating light.

In our
work, we focus on the effect of wave mixing. Particularly, the four wave
mixing is a textbook optical effect manifesting itself in a pair of
frequency side peaks from two driving tones on a classical
Kerr-nonlinearity^{23, 24}.
Ultimate scaling down of the nonlinear medium to a single artificial
atom, strongly interacting with the incident waves, results in time
resolution of instant multi-photon interactions and reveals effects
beyond classical physics. Here, we demonstrate the physical phenomenon
of quantum wave mixing (QWM) on a superconducting artificial atom in the
open one-dimensional (1D) space (coplanar transmission line on-chip).
We show two regimes of QWM comprising different degrees of
‘quantumness’: the first and most remarkable one is QWM with
nonclassical superposed states, which are mapped into a finite number of
frequency peaks. In another regime, we investigate the different orders
of wave mixing of classical coherent waves on the artificial atom. The
dynamics of the peaks exhibits a series of Bessel-function Rabi
oscillations, different from the usually observed harmonic ones, with
orders determined by the number of interacting photons. Therefore, the
device utilising QWM visualises photon-state statistics of classical and
non-classical photonic states in the open space. The spectra are
fingerprints of interacting photonic states, where the number of peaks
due to the atomic emission always exceeds by one the number of
absorption peaks. Below, we summarise several specific findings of this
work: (1) demonstration of the wave mixing on a single quantum system;
(2) in the quantum regime of mixing, the peak pattern and the number of
the observed peaks is a map of coherent and superposed photonic states,
where the number of peaks *N*_{peaks} is related to the number of interacting photons *N*_{ph} as *N*_{peaks} = 2*N*_{ph}
+ 1. Namely, the one-photon state (in two-level atoms) results in
precisely three emission peaks; the two-photon state (in three-level
atoms) results in five emission peaks; and the classical coherent
states, consisting of infinite number of photons, produce a spectrum
with an infinite number of peaks; (3) Bessel function Rabi oscillations
are observed and the order of the Bessel functions depends on the peak
position and is determined by the number of interacting photons.

## Results

### Coherent and zero-one photon superposed state

To
evaluate the system, we consider electromagnetic waves propagating in a
1D transmission line with an embedded two-level artificial atom^{15} (see also Supplementary Methods, Supplementary Fig. 1) shown in Fig. 1a.
In this work, we are interested in photon statistics, which will be
revealed by QWM, therefore, we will consider our system in the photon
basis. The coherent wave in the photon (Fock) basis$\u2223N\u27e9$ is presented as

and consists of an infinite number of photonic states. A two-level atom with ground and excited states $\u2223g\u27e9$ and $\u2223e\u27e9$ driven by the field can be prepared in superposed state $\Psi =\mathrm{cos}\frac{\theta}{2}\u2223g\u27e9+\mathrm{sin}\frac{\theta}{2}\u2223e\u27e9$ and, if coupled to the external photonic modes, transfers the excitation to the mode, creating zero-one photon superposed state

where $\beta =\mathrm{tan}\frac{\theta}{2}$ (Supplementary Note 1). The superposed state comprises coherence, however $\u2223\beta \u27e9$ state is different from classical coherent state $\u2223\alpha \u27e9$, consisting of an infinite number of Fock states. The energy exchange process is described by the operator ${b}^{-}{b}^{+}\u2223g\u27e9\u27e8g\u2223+{b}^{+}\u2223g\u27e9\u27e8e\u2223$, which maps the atomic to photonic states, where ${b}^{+}=\u22231\u27e9\u27e80\u2223$ and ${b}^{-}=\u22230\u27e9\u27e81\u2223$ are creation/annihilation operators of the zero-one photon state. The operator is a result of a half-period oscillation in the evolution of the atom coupled to the quantised photonic mode and we keep only relevant for the discussed case (an excited atom and an empty photonic mode) terms (Supplementary Note 1).

We discuss and demonstrate experimentally an elastic scattering of two waves with frequencies *ω*_{−} = *ω*_{0} − *δω* and *ω*_{+} = *ω*_{0} + *δω*, where *δω* is a small detuning, on a two-level artificial atom with energy splitting $\hslash {\omega}_{0}$. The scattering,
taking place on a single artificial atom, allows us to resolve instant
multi-photon interactions and statistics of the processes. Dealing with
the final photonic states, the system Hamiltonian is convenient to
present as the one, which couples the input and output fields

using creation and annihilation operators ${a}_{\pm}^{\u2020}$ (*a*_{±}) of photon states ${\u2223N\u27e9}_{\pm}$ (*N* is an integer number) and ${b}_{\pm}^{+}$ and ${b}_{\pm}^{-}$ are creation/annihilation operators of single-photon output states at frequencies *ω*_{±}. Here $\hslash g$ is the field-atom coupling energy. Operators ${b}_{\pm}^{+}$ and ${b}_{\pm}^{-}$ also describe the atomic excitation/relaxation, using substitutions ${b}_{\pm}^{+}\leftrightarrow {e}^{\mp i\phi}\u2223e\u27e9\u27e8g\u2223$ and ${b}_{\pm}^{-}\leftrightarrow {e}^{\pm i\phi}\u2223g\u27e9\u27e8e\u2223$, where *φ* = *δωt* is a slowly varying phase (Supplementary Note 2). The phase rotation results in the frequency shift according to *ω*_{±}*t* = *ω*_{0}*t* ± *δωt* and more generally for ${b}_{m}^{\pm}$ (with integer *m*) the varied phase *mδφ* results in the frequency shift *ω*_{
m
} = *ω*_{0} + *mδω*.

The system evolution over the time interval [*t*, *t*′] (*t*′ = *t* + Δ*t* and $\delta \omega \Delta t\ll 1$) described by the operator *U*(*t*, *t*′) = exp(−*iH*Δ*t*/$\hslash $) can be presented as a series expansion of different order atom–photon interaction processes ${a}_{\pm}^{\u2020}{b}_{\pm}^{-}$ and ${a}_{\pm}{b}_{\pm}^{+}$—sequential absorption-emission accompanied by atomic excitations/relaxations (Supplementary Note 2). Operators *b* describe the atomic states (instant interaction of the photons in the atom) and, therefore, satisfy the following identities: ${b}_{p}^{-}{b}_{m}^{+}={\u22230\u27e9}_{m-p}\u27e80\u2223$, ${b}_{j}^{\pm}{b}_{p}^{\mp}{b}_{m}^{\pm}={b}_{j-p+m}^{\pm}$, ${b}_{p}^{\pm}{b}_{m}^{\pm}=0$. The excited atom eventually relaxes producing zero-one superposied photon field ${\u2223\beta \u27e9}_{m}$ at frequency *ω*_{
m
} = *ω*_{0} + *mδω* according to ${b}_{m}^{+}\u22230\u27e9={\u22231\u27e9}_{m}$. We repeat the evolution and average the emission on the time interval *t*>*δω*^{−1}
and observe narrow emission lines. In the general case, the atom in a
superposed state generates coherent electromagnetic waves of amplitude

at frequency *ω*_{
m
}, where *Γ*_{1} is the atomic relaxation rate and *μ* is the atomic dipole moment^{15, 17}.

### Elastic scattering and Bessel function Rabi oscillations

To
study QWM, we couple the single artificial atom (a superconducting loop
with four Josephson junctions) to a transmission line via a capacitance
(Supplementary Methods). The atom relaxes with the photon emission rate found to be *Γ*_{1}/2*π*
≈ 20 MHz. The coupling is strong, which means that any non-radiative
atom relaxation is suppressed and almost all photons from the atom are
emitted into the line. The sample is held in a dilution refrigerator
with base temperature 15 mK. We apply periodically two simultaneous
microwave pulses with equal amplitudes at frequencies *ω*_{−} and *ω*_{+}, length Δ*t* = 2 ns and period *T*_{r} = 100 ns (much longer than the atomic relaxation time ${\Gamma}_{1}^{-1}\approx 8$ ns). A typical emission power spectrum integrated over many periods (bandwidth is 1 kHz) is shown in Fig. 2a. The pattern is symmetric with many narrow peaks (as narrow as the excitation microwaves), which appeared at frequencies *ω*_{0} ± (2*k* + 1)*δω*, where *k* ≥ 0 is an integer number. We linearly change driving amplitude (Rabi frequency)* Ω*,
which is defined from the measurement of harmonic Rabi oscillations
under single-frequency excitation. The dynamics of several side peaks
versus linearly changed *Ω*Δ*t* (here we vary *Ω*, however, equivalently Δ*t* can be varied) is shown on plots of Fig. 2b. Note that the peaks exhibit anharmonic oscillations well fitted by the corresponding 2*k* + 1-order Bessel functions of the first kind. The first maxima are delayed with the peak order, appearing at *Ω*Δ*t* ∝ *k* + 1. Note also that detuning *δω* should be within tens of megahertz (≤Γ_{1}). However, in this work, we use *δω*/2*π* = 10 kHz to be able to quickly span over several *δω* of the spectrum analyser (SA) with the narrow bandwidth.

Figure 1b
examplifies the third-order process (known as the four-wave mixing in
the case of two side peaks), resulting in the creation of the right
hand-side peak at *ω*_{3} = 2*ω*_{+} − *ω*_{−}. The process consists of the absorption of two photons of frequency *ω*_{+} and the emission of one photon at *ω*_{−}. More generally, the 2*k* + 1-order peak at frequency *ω*_{2k+1} = (*k* + 1)*ω*_{+} − *kω*_{−} (≡*ω*_{0} + (2*k* + 1)*δω*) is described by the multi-photon process ${\left({a}_{+}{a}_{-}^{\u2020}\right)}^{k}{a}_{+}{b}_{2k+1}^{+}$, which involves the absorption of *k* + 1 photons from *ω*_{+} and the emission of *k* photons at *ω*_{−}; and the excited atom eventually generates a photon at *ω*_{2k+1}. The symmetric left hand-side peaks at *ω*_{0} − (2*k* + 1)*δω* are described by a similar processes with swapped indexes (+ ↔ −). The peak amplitudes from Eq. (4) are described by expectation values of *b*-operators, which at frequency *ω*_{2k+1} can be written in the form of $\u27e8{b}_{2k+1}^{+}\u27e9={D}_{2k+1}\u27e8{\left({a}_{+}{a}_{-}^{\u2020}\right)}^{k}{a}_{+}\u27e9$. The prefactor *D*_{2k+1} depends on the driving conditions and can be calculated summing up all virtual photon processes (e.g., ${a}_{+}^{\u2020}{a}_{+}$, ${a}_{-}^{\u2020}{a}_{-}$, etc.) not changing frequencies (Supplementary Note 2). For instance, the creation of a photon at 2*ω*_{+} − *ω*_{−} is described by $\u27e8{b}_{3}^{+}\u27e9={D}_{3}\u27e8{a}_{+}{a}_{-}^{\u2020}{a}_{+}\u27e9$.

As the number of required photons increases with *k*, the emission maximum takes longer time to appear (Fig. 2b). To derive the dependence observed in our experiment, we consider the case with initial state $\Psi =\u22230\u27e9\otimes \left({\u2223\alpha \u27e9}_{-}+{\u2223\alpha \u27e9}_{+}\right)$ and $\alpha \gg 1$. We find then that the peaks exhibit Rabi oscillations described by $\u27e8{b}_{2k+1}\u27e9={\left(-1\right)}^{k}\u22152\times {J}_{2k+1}\left(2\Omega \Delta t\right)$ (Supplementary Note 2, Eq. (29)) and the mean number of generated photons per cycle in 2*k* + 1-mode is

The symmetric multi-peak
pattern in the spectrum is a map of an infinite number of interacting
classical coherent states. The dependence from the parameter 2*Ω*Δ*t* observed in our experiment can also be derived using a semiclassical approach, where the driving field is given by *Ω**e*^{iδωt} + *Ω**e*^{−iδωt} = 2*Ω*cos *δωt*. As shown in Supplementary Note 2,
a classical description can be mathematically more straightforward and
leads to the same result, but fails to provide a qualitative picture of
QWM discussed below. The Bessel function dependencies have been earlier
observed in multi-photon processes, however in frequency domain^{25,26,27}.

### QWM and dynamics of non-classical photonic states

Next,
we demonstrate one of the most interesting results: QWM with
non-classical photonic states. We further develop the two-pulse
technique separating the excitation pulses in time. Breaking
time-symmetry in the evolution of the quantum system should result in
asymmetric spectra and the observation of series of spectacular quantum
phenomena. The upper panel in Fig. 3a demonstrates such a spectrum, when the pulse at frequency *ω*_{+} is applied after a pulse at *ω*_{−}. Notably, the spectrum is asymmetric and contains only one side peak at frequency 2*ω*_{+} − *ω*_{−}. There is no any signature of other peaks, which is in striking contrast with Fig. 2a. Reversing the pulse sequence mirror reflects the pattern revealing the single side peak at 2*ω*_{−} − *ω*_{+} (not shown here).

The quantitative explanation of the process is provided on the left panel of Fig. 1c. The first pulse prepares superposed zero-one photon state ${\u2223\beta \u27e9}_{-}$ in the atom, which contains not more than one photon (*N*_{ph} = 1). Therefore, only a single-positive side peak 2*ω*_{+} − *ω*_{−} due to the emission of the *ω*_{−}-photon, described by ${a}_{+}{a}_{-}^{\u2020}{a}_{+}$, is allowed. See Supplementary Note 3 for details.

To
prove that there are no signatures of other peaks, except for the
observed three peaks, we vary the peak amplitudes and compare the
classical and QWM regimes with the same conditions. Figure 3b
demonstrates the side peak power dependencies in different mixing
regimes: classical (two simultaneous pulses) (left panels) and quantum
(two consecutive pulses) (right panels). The two cases reveal a very
similar behaviour of the right hand-side four-wave mixing peak at 2*ω*_{+} − *ω*_{−},
however the other peaks appear only in the classical wave mixing,
proving the absence of other peaks in the mixing with the quantum state.

The
asymmetry of the output mixed signals, in principle, can be
demonstrated in purely classical systems. It can be achieved in several
ways, e.g., with destructive interference, phase-sensitive
detection/amplification^{28},
filtering. All these effects are not applicable to our system of two
mixed waves on a single point-like scatterer in the open (wide frequency
band) space. What is more important than the asymmetry is that the
whole pattern consists of only three peaks without any signature of
others.

This demonstrates another remarkable property of our device: it probes photonic states, distinguishing the coherent, $\u2223\alpha \u27e9$, and superposed states with the finite number of the photon states. Moreover, the single peak at *ω*_{3} shows that the probed state was $\u2223\beta \u27e9$ with *N*_{ph} = 1. This statement can be generalised for an arbitrary state. According to the picture in Fig. 1c,
adding a photon increases the number of peaks from the left- and
right-hand side by one, resulting in the total number of peaks *N*_{peaks} = 2*N*_{ph} + 1.

### Probing the two-photon superposed state

To
have a deeper insight into the state-sensing properties and to
demonstrate QWM with different photon statistics, we extended our
experiment to deal with two-photon states (*N*_{ph} = 2). The two lowest transitions in our system can be tuned by adjusting external magnetic fields to be equal to $\hslash $*ω*_{0}, though higher transitions are off-resonant ($\ne \hslash {\omega}_{0}$, See Supplementary Fig. 2). In the three-level atom, the microwave pulse at *ω*_{−} creates the superposed two-photon state

where $C=\sqrt{1+{\u2223{\gamma}_{1}\u2223}^{2}+{\u2223{\gamma}_{2}\u2223}^{2}}$. The plot in Fig. 4 shows the modified spectrum. As expected, the spectrum reveals only peaks at frequencies consisting of one or two photons of *ω*_{−}. The frequencies are *ω*_{3} = 2*ω*_{+} − *ω*_{−}, *ω*_{−3} = 2*ω*_{−} − *ω*_{+}, and *ω*_{5} = 3*ω*_{+} − 2*ω*_{−} corresponding, for instance, to processes ${a}_{+}{a}_{-}^{\u2020}{a}_{+}{c}_{3}^{+}$, ${a}_{-}{a}_{-}{a}_{+}^{\u2020}{c}_{-3}^{+}$ and ${a}_{+}{a}_{-}^{\u2020}{a}_{-}^{\u2020}{a}_{+}{a}_{+}{c}_{5}^{+}$, where ${c}_{m}^{+}$ and ${c}_{m}^{-}$ are creation and annihilation operators defined on the two-photon space ($\u2223n\u27e9$, where *n*
takes 0, 1 or 2). The intuitive picture of the two-photon state mixing
is shown on the central and right-hand side panels of Fig. 1c. The two photon state (*N*_{ph}
= 2) results in the five peaks. This additionally confirms that the atom
resolves the two-photon state. See Supplementary Note 4 for the details.

The
QWM can be also understood as a transformation of the quantum states
into quantised frequencies similar to the Fourier transformation. The
summarised two-dimensional plots with *N*_{ph} are presented in Fig. 5. The mixing with quantum states is particularly revealed in the asymmetry. Note that for arbitrary *N*_{ph} coherent states, the spectrum asymmetry will remain, giving *N*_{ph} and *N*_{ph}−1 peaks at the emission and absorption sides.

According
to our understanding, QWM has not been demonstrated in systems other
than superconducting quantum ones due to the following reasons. First,
the effect requires a single quantum system because individual
interaction processes have to be separated in time^{29}
and it will be washed out in multiple scattering on an atomic ensemble
in matter. Next, although photon counters easily detect single photons,
in the visible optical range, it might be more difficult to detect
amplitudes and phases of weak power waves^{30, 31}. On the other hand, microwave techniques allow one to amplify and measure weak coherent emission from a single quantum system^{17, 32},
due to strong coupling of the single artificial atom; the confinement
of the radiation in the transmission line; and due to an extremely high
phase stability of microwave sources. The radiation can be selectively
detected by either SAs or vector network analysers with narrow frequency
bandwidths, efficiently rejecting the background noise.

In summary, we have demonstrated QWM—an interesting phenomenon of quantum optics. We explore different regimes of QWM and prove that the superposed and coherent states of light are mapped into a quantised spectrum of narrow peaks. The number of peaks is determined by the number of interacting photons. QWM could serve as a powerful tool for building new types of on-chip quantum electronics.

### Data availability

Relevant data is available from A.Yu.D. upon request.

## Additional Information

**Publisher's note:** Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

## References

- 1.
Clarke, J. & Wilhelm, F. K. Superconducting quantum bits.

*Nature***453**, 1031–1042 (2008). - 2.
Wallraff, A. et al. Strong coupling of a single photon to a superconducting qubit using circuit quantum electrodynamics.

*Nature***431**, 162–167 (2004). - 3.
You, J. Q. & Nori, F. Atomic physics and quantum optics using superconducting circuits.

*Nature***474**, 589597 (2011). - 4.
Schuster, D. I. et al. Resolving photon number states in a superconducting circuit.

*Nature***445**, 515–518 (2007). - 5.
Peng, Z., De Graaf, S., Tsai, J. & Astafiev, O. Tuneable on-demand single-photon source in the microwave range.

*Nat. Commun.***7**, 12588 (2016). - 6.
Houck, A. A. et al. Generating single microwave photons in a circuit.

*Nature***449**, 328–331 (2007). - 7.
Lang, C. et al. Correlations, indistinguishability and entanglement in Hong-Ou-Mandel experiments at microwave frequencies.

*Nat. Phys.***9**, 345–348 (2013). - 8.
Hofheinz, M. et al. Generation of Fock states in a superconducting quantum circuit.

*Nature***454**, 310–314 (2008). - 9.
Hofheinz, M. et al. Synthesizing arbitrary quantum states in a superconducting resonator.

*Nature***459**, 546–549 (2009). - 10.
Astafiev, O. et al. Single artificial-atom lasing.

*Nature***449**, 588–590 (2007). - 11.
Hoi, I.-C. et al. Giant cross-Kerr effect for propagating microwaves induced by an artificial atom.

*Phys. Rev. Lett.***111**, 053601 (2013). - 12.
Kirchmair, G. et al. Observation of quantum state collapse and revival due to the single-photon Kerr effect.

*Nature***495**, 205 (2013). - 13.
Roy, D., Wilson, C. M. & Firstenberg, O. Colloquium: strongly interacting photons in one-dimensional continuum.

*Rev. Mod. Phys.***89**, 021001 (2017). - 14.
Hoi, I.-C. et al. Microwave quantum optics with an artifficial atom in one-dimensional open space.

*New J. Phys.***15**, 025011 (2013). - 15.
Astafiev, O. et al. Resonance fluorescence of a single artificial atom.

*Science***327**, 840–843 (2010). - 16.
Toyli, D. M. et al. Resonance fluorescence from an artificial atom in squeezed vacuum.

*Phys. Rev. X***6**, 031004 (2016). - 17.
Abdumalikov, A. A. Jr, Astafiev, O. V., Pashkin, Y. A., Nakamura, Y. & Tsai, J. Dynamics of coherent and incoherent emission from an artificial atom in a 1D space.

*Phys. Rev. Lett.***107**, 043604 (2011). - 18.
Astafiev, O. V. et al. Ultimate on-chip quantum amplifier.

*Phys. Rev. Lett.***104**, 183603 (2010). - 19.
Hoi, I.-C. et al. Demonstration of a singlephoton router in the microwave regime.

*Phys. Rev. Lett.***107**, 073601 (2011). - 20.
Fang, Y.-L. L. & Baranger, H. U. Multiple emitters in a waveguide: nonreciprocity and correlated photons at perfect elastic transmission.

*Phys. Rev. A***96**, 013842 (2017). - 21.
van Loo, A. F. et al. Photon-mediated interactions between distant artificial atoms.

*Science***342**, 1494–1496 (2013). - 22.
Mlynek, J., Abdumalikov, A. A., Eichler, C. & Wallraff, A. Observation of Dicke superradiance for two artificial atoms in a cavity with high decay rate.

*Nat. Commun.***5**, 5186 (2014). - 23.
Boyd, R. W.

*Nonlinear Optics*(Academic press, New York, 2003). - 24.
Scully, M. O. & Zubairy, M.

*Quantum Optics*(Cambridge University Press, Cambridge, 1997). - 25.
Oliver, W. D. et al. Mach-Zehnder interferometry in a strongly driven superconducting qubit.

*Science***310**, 1653–1657 (2005). - 26.
Sillanpää, M., Lehtinen, T., Paila, A., Makhlin, Y. & Hakonen, P. Continuous-time monitoring of Landau-Zener interference in a cooper-pair box.

*Phys. Rev. Lett.***96**, 187002 (2006). - 27.
Neilinger, P. et al. Landau-Zener-Stückelberg-Majorana lasing in circuit quantum electrodynamics.

*Phys. Rev. B***94**, 094519 (2016). - 28.
Schackert, F., Roy, A., Hatridge, M., Devoret, M. H. & Stone, A. D. Three-wave mixing with three incoming waves: signal-idler coherent attenuation and gain enhancement in a parametric amplifier.

*Phys. Rev. Lett.***111**, 073903 (2013). - 29.
Maser, A., Gmeiner, B., Utikal, T., Götzinger, S. & Sandoghdar, V. Few-photon coherent nonlinear optics with a single molecule.

*Nat. Photon.***10**, 450–453 (2016). - 30.
Lvovsky, A. I. & Raymer, M. G. Continuous-variable optical quantum-state tomography.

*Rev. Mod. Phys.***81**, 299–332 (2009). - 31.
Ip, E., Lau, A. P. T., Barros, D. J. & Kahn, J. M. Coherent detection in optical fiber systems.

*Opt. Express***16**, 753–791 (2008). - 32.
Shen, J.-T. & Fan, S. Coherent single photon transport in a one-dimensional waveguide coupled with superconducting quantum bits.

*Phys. Rev. Lett.***95**, 213001 (2005).

## Acknowledgements

We acknowledge Russian Science Foundation (grant N 16-12-00070) for supporting the work. We thank A. Semenov and E. Ilichev for useful discussions.

## Author information

### Affiliations

#### Laboratory of Artificial Quantum Systems, Moscow Institute of Physics and Technology, Dolgoprudny, 141700, Russia

- A. Yu. Dmitriev
- , R. Shaikhaidarov
- , V. N. Antonov
- & O. V. Astafiev

#### Department of Physics, Royal Holloway, University of London, Egham, Surrey, TW20 0EX, UK

- R. Shaikhaidarov
- , V. N. Antonov
- , T. Hönigl-Decrinis
- & O. V. Astafiev

#### National Physical Laboratory, Teddington, TW11 0LW, UK

- T. Hönigl-Decrinis
- & O. V. Astafiev

#### Institute of Solid State Physics, Russian Academy of Sciences, Chernogolovka, 142432, Russian Federation

- A. Yu. Dmitriev

### Authors

### Search for A. Yu. Dmitriev in:

### Search for R. Shaikhaidarov in:

### Search for V. N. Antonov in:

### Search for T. Hönigl-Decrinis in:

### Search for O. V. Astafiev in:

### Contributions

O.V.A. planned and designed the experiment, R.S., A.Yu.D. and T.H.-D. fabricated the sample and built the set-up for measurements. A.Yu.D., R.S. and T.H.-D. measured the raw data. A.Yu.D., V.N.A. and O.V.A. made calculations, analysed and processed the data and wrote the manuscript, with important contributions from all the authors.

### Competing interests

The authors declare no competing financial interests.

### Corresponding authors

Correspondence to A. Yu. Dmitriev or O. V. Astafiev.