A Rydberg atom-based mixer: Measuring the
phase of a radio frequency wave
Cite as: Appl. Phys. Lett. 114, 114101 (2019); https://doi.org/10.1063/1.5088821
Submitted: 14 January 2019 • Accepted: 04 March 2019 • Published Online: 18 March 2019
Matthew T. Simons, Abdulaziz H. Haddab, Joshua A. Gordon, et al.
ARTICLES YOU MAY BE INTERESTED IN
Determining the angle-of-arrival of a radio-frequency source with a Rydberg atom-based
sensor
Applied Physics Letters 118, 114001 (2021); https://doi.org/10.1063/5.0045601
A “real-time” guitar recording using Rydberg atoms and electromagnetically induced
transparency: Quantum physics meets music
AIP Advances 9, 065110 (2019); https://doi.org/10.1063/1.5099036
Electric field metrology for SI traceability: Systematic measurement uncertainties in
electromagnetically induced transparency in atomic vapor
Journal of Applied Physics 121, 233106 (2017); https://doi.org/10.1063/1.4984201
Appl. Phys. Lett. 114, 114101 (2019); https://doi.org/10.1063/1.5088821 114, 114101
© 2019 Author(s).
Applied Physics Letters ARTICLE scitation.org/journal/apl
A Rydberg atom-based mixer: Measuring
the phase of a radio frequency wave
Cite as: Appl. Phys. Lett. 114, 114101 (2019); doi: 10.1063/1.5088821
Submitted: 14 January 2019 . Accepted: 4 March 2019 .
Published Online: 18 March 2019
Matthew T. Simons,1,a) Abdulaziz H. Haddab,2 Joshua A. Gordon,3 and Christopher L. Holloway3
AFFILIATIONS
1
Department of Physics, University of Colorado, Boulder, Colorado 80302, USA
2
Department of Electrical Engineering, University of Colorado, Boulder, Colorado 80302, USA
3
National Institute of Standards and Technology (NIST), RF Technology Division, U.S. Department of Commerce,
Boulder Laboratories, Boulder, Colorado 80305, USA
a)
Electronic mail: [email protected].
ABSTRACT
Rydberg atoms have been shown to be very useful in performing absolute measurements of the magnitude of a radio frequency (RF) field
using electromagnetically induced transparency. However, there has been less success in using Rydberg atoms for the measurement of the
phase of an RF field. Measuring the phase of a RF field is a necessary component for many important applications, including antenna
metrology, communications, and radar. We demonstrate a scheme for measuring the phase of an RF field by using Rydberg atoms as a mixer
to down-convert an RF field at 20 GHz to an intermediate frequency on the order of kHz. The phase of the intermediate frequency corre-
sponds directly to the phase of the RF field. We use this approach to measure the phase shift on an electromagnetic wave from a horn
antenna as the antenna is placed at different distances from the Rydberg atom sensor. The atom-based RF phase measurements allow us to
measure the propagation constant of the RF wave to within 0.1% of the theoretical value.
Published under license by AIP Publishing. https://doi.org/10.1063/1.5088821
Recently, great progress has been achieved in the development of The motivation of this work is to advance quantum-based E-field
a spectroscopic approach for radio-frequency (RF) electric (E) field sensor technology. To this end, we demonstrate the ability of Rydberg
measurements.1–7 This approach utilizes electromagnetically induced atoms to measure the phase of an RF field. Inclusion of phase mea-
transparency (EIT) and Autler-Townes (AT) splitting to create a direct surements allows the Rydberg atom system to fully characterize an RF
International System of Units (SI) traceable, self-calibrated measure- E-field in a single quantum-based sensor. This technique relies on
ment of an RF E-field amplitude. Rydberg atoms are very sensitive to using the atom system as a Rydberg atom-based mixer. Applying an
RF fields due to their large dipole moments over a broad frequency RF field that is on-resonance with the Rydberg transition acts as a local
range, from MHz to THz.5,8,9 Thus far, Rydberg atoms have been used oscillator (“LO” in Fig. 1), which causes the EIT/AT effect in the
to extend the frequency range and the sensitivity of RF electric field Rydberg atoms to demodulate a second, co-polarized RF field (labeled
amplitude3,10 and polarization measurements,7 but have not been used “SIG” in Fig. 1). The difference frequency, or intermediate frequency
for RF phase measurements. This Rydberg-atom based sensor can act (IF), is detected by optically probing the Rydberg atoms (Fig. 1). The
as a compact receiver/antenna for detecting modulated signals, which phase of the IF signal corresponds directly to the relative phase
allows for communication applications. The first demonstration of between the LO and RF signals. This technique can be applied over
this was presented in Ref. 11, and later by others.12–18 This has led to the same frequency range as E-field amplitude measurements, result-
the new term “atom-radio.”15,16 The work in Refs. 11–18 have demon- ing in an atom-based sensor that can measure amplitude, polarization,
strated the detection and reception of amplitude modulated (AM) and and now phase for fields from 500 MHz to 1 THz in a single setup.
frequency modulated (FM) signals, where the AM and FM detection Rydberg atoms can detect and measure RF E-fields through the
are based on the measurement of the amplitude of the carrier. To use of two related optical phenomena. EIT occurs when a laser field
detect and receive BPSK/QPSK/QAM digital communication signals (probe field) that is on-resonance with an atomic transition and nor-
requires the ability to measure the phase of an RF field. Other applica- mally be absorbed by an atomic vapor is instead transmitted in the
tions such as near-field antenna metrology and radar also require a presence of a second laser field (coupling field), which is tuned to
phase measurement. another atomic transition. This occurs over a very narrow frequency
Appl. Phys. Lett. 114, 114101 (2019); doi: 10.1063/1.5088821 114, 114101-1
Published under license by AIP Publishing
Applied Physics Letters ARTICLE scitation.org/journal/apl
the EIT/AT splitting. By setting the probe laser frequency to the center
of the EIT peak, the probe laser transmission is modulated by the
interference between the two RF fields. If one of the RF fields is
detuned from the other, the beat note can be observed in time in the
transmission. This effect is demonstrated by placing two standard gain
horn antennas 12 cm from the Cs vapor cell (Fig. 2). One antenna is
fed by a signal generator set to fLO ¼ 19.626 000 GHz to act as the LO.
Another signal generator feeds the second antenna to provide the sec-
ond RF signal, fSIG. The two signal generators are connected to a
rubidium (Rb) clock with 10 MHz signal. Note that this synchroniza-
tion signal need not be physically connected if each signal generator is
FIG. 1. Block diagram of the Rydberg atom “mixer.” The Rydberg atoms separate
the difference frequency (IF) from two RF signals (LO and SIG). This demodulated
connected to atomic clocks that were previously synchronized, allow-
signal is carried in the probe laser. ing the LO and Rydberg atom systems to be physically separated from
the SIG signal.
The total electric field in the vapor cell (~
E atoms ) is the sum of the
range. When a third electromagnetic field, tuned to another linked
LO and SIG fields (E1 and E2). For two co-polarized fields, this is the
atomic transition, is present, the original transparency region is split into scalar total Eatoms ¼ E1 þ E2. We define the LO and SIG fields as E1
two regions separated in frequency (AT-splitting). The frequency separa- ¼ ELO cosðxLO t þ /LO Þ and E2 ¼ ESIG cos ðxSIG t þ /SIG Þ. Assuming
tion is directly related to the strength of the third RF field (typically an that the difference between the two frequencies Dx ¼ xLO xSIG is
RF field). More details of this process can be found in Refs. 1–6. much less than the average frequency x ¼ ðxLO þ xSIG Þ=2, such that
In this work, the atomic vapor cell is filled with cesium (133Cs) Dx x, then we can express the total field Eatoms as
atoms. Figure 2 shows a diagram of the experimental setup. The probe
laser is tuned to a wavelength of 852 nm to excite 133Cs from the Eatoms ¼ E1 þ E2 ; (1)
ground state to the first excited state (6S1=2 ! 6P3=2 ). The coupling pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 2
¼ cos ðxLO þ /LO Þ ELO þ ESIG þ 2ELO ESIG cos ðDxt þ D/Þ ;
laser is counter-propagated through the probe laser and tuned to
511.148 nm, coupling the excited state to the Rydberg state 34D5/2, (2)
and creating EIT in the probe laser, increasing the transmission where xLO ¼ 2pfLO ; xSIG ¼ 2pfSIG , and D/ ¼ /LO /SIG .
through the vapor cell. The probe laser has a full-width at half- Equation (2) can be written to separate the term on resonance with
maximum (FWHM) of 425 lm and a power of 49 lW. The coupling the Rydberg transition Eres ðxLO Þ from the low frequency term
laser has a FWHM of 620 lm and a power of 60.6 mW. Applying an Emod ðDxÞ as Eatoms ¼ Eres Emod , where
RF field of 19.626 GHz causes a transition of 34D5=2 ! 35P3=2 , pro-
ducing AT-splitting. Eres ¼ cos ðxLO þ /LO Þ; (3)
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
With the probe laser frequency fixed at the first excited state tran- 2 2 þ 2E E
Emod ¼ ELO þ ESIG LO SIG cos ðDx t þ D/Þ: (4)
sition frequency, the transmission through the vapor cell is reduced.
The interference of two RF fields can then be detected by the effect on The resonant term Eres causes an Autler-Townes effect as described
above, reducing the peak of the EIT line. The low frequency term Emod
modulates the amplitude of the resonant effect.
The probe transmission is a function of absorption of the atomic
vapor19
2pL
Tprobe / exp Im½v ; (5)
kp
where L is the length of the cell, kp is the wavelength of the probe laser,
and v is the susceptibility. The susceptibility is a function of the E-field
strength jEatoms j at xLO, thus the probe transmission is a function of
the magnitude of the E-field on resonance with the Rydberg transition.
The magnitude of the resonant term jEres j ¼ 1, so jEatoms j ¼ jEres
Emod j ¼ jEmod j. For a SIG field that is small compared to the LO field
(ESIG ELO ), the magnitude of the field can be written as
jEatoms j u ELO þ ESIG cos ðDx t þ D/Þ: (6)
FIG. 2. A probe and a coupling laser are overlapped through a Cs vapor cell using The atoms demodulate the high-frequency xLO field and the probe
two dichroic mirrors. Two RF fields (E1 and E2) are applied using two horn anten- transmission as a function of time is given by
nas fed by two signal generators (LO and SIG). The SIG horn is placed on a micro-
meter translation stage. Transmission of the probe through the 133Cs vapor is Tprobe / jEatoms j u ELO þ ESIG cos ðDx t þ D/Þ: (7)
monitored with a balanced photodetector. Part of the probe beam is split and sent
to the second port of the balanced photodetector without sampling the Cs (not The term Eres causes AT-splitting through the resonant interaction
pictured). which reduces the probe transmission (Tprobe) below the EIT peak.
Appl. Phys. Lett. 114, 114101 (2019); doi: 10.1063/1.5088821 114, 114101-2
Published under license by AIP Publishing
Applied Physics Letters ARTICLE scitation.org/journal/apl
This reduction is then modulated in time by the low-frequency beat frequency of SIG is detuned from the frequency of SIG (the different
note, Emod. In effect, the probe laser transmission varies in time by the frequency ranges were plotted on different scales for illustration; the
difference frequency Dx, with a phase given by the difference between magnitude of the low-frequency noise gives some indication of
the LO and SIG phases D/. the change in scale). The detuning is limited by the response rate of
The observation of a signal at Df ¼ Dx/2p in the detected probe the EIT/AT effect. The minimum time for the EIT/AT to develop is
transmission indicates the presence of an RF field at either fSIG ¼ fLO on the order of 0.1 ls,12,13,17,19 which limits the modulation frequen-
6 Df. This is demonstrated in Fig. 3, where a discrete Fourier trans- cies that are observable to Df ⱗ 10 MHz.
form (DFT) of the probe signal is plotted as the SIG is varied from From Eq. (7), we see that if we measure the phase (D/) of the
fSIG ¼ 19.626 001 GHz to 19.628 500 GHz. Each subfigure shows a dif- intermediate frequency signal in the probe transmission, we directly
ferent frequency range. The Df signal gets weaker the further the measure the phase between LO and SIG. A direct method for varying
the phase of an RF field is to physically translate the source. We dem-
onstrate the ability of the Rydberg atoms to detect the RF phase by
translating the SIG horn linearly away from the atoms (in the z-direc-
tion), as shown in Fig. 2. The SIG horn is translated by a motorized
linear stage, with a step size of Dd ¼ kRF/8 ¼ 1.91 mm. The LO fre-
quency is set to fLO ¼ 19.626 000 GHz, and the SIG is set to
fSIG ¼ 19.626 030 GHz, for a difference in frequency of Df ¼ 30 kHz. A
function generator is set to 30 kHz for a reference signal to serve as the
trigger for the oscilloscope. The two signal generators and the function
generator are connected to a Rb clock 10 MHz signal. The LO horn
antenna generates an electric field of jEj ¼ 2:97 V=m in the vapor cell,
and the SIG horn a field of jEj ¼ 1:22 V=m.
By translating the SIG horn antenna while leaving the LO horn
antenna fixed, we shift the relative phase between the SIG and LO
fields in the vapor cell. We measure the change in phase (D/) of our
detected signal as a function of the distance traveled by the RF horn
antenna Dd, giving us the propagation constant20 b
D/
b¼ : (8)
Dd
Four sample traces corresponding to different distances of the SIG
horn relative to the Rydberg atoms are shown in Fig. 4. We track the
phase of the signal relative to the first trace as the SIG horn is moved,
keeping track of the accumulated phase. Figure 5 shows the measured
FIG. 3. DFT of the probe transmission for different values of Df, keeping the LO fre-
quency fixed. Each plot (a)–(c) has a different y-scale. (a) Values of Df from bottom
to top of 1 kHz, 3 kHz, 5 kHz, 7 kHz, 9 kHz, and 11 kHz, (b) values of 10 kHz, 50
kHz, 90 kHz, 130 kHz, 170 kHz, and 210 kHz, and (c) values of 500 kHz, 1 MHz, FIG. 4. Signal detected on a photodetector for different SIG horn distances Dd.
1.5 MHz, 2 MHz, and 2.5 MHz. The IF frequency is Df ¼ 30 kHz.
Appl. Phys. Lett. 114, 114101 (2019); doi: 10.1063/1.5088821 114, 114101-3
Published under license by AIP Publishing
Applied Physics Letters ARTICLE scitation.org/journal/apl
opens up a wide range of possible applications for Rydberg atom-
based sensors. The Rydberg atom-based mixer can be used to open up
several different avenues for quantum-based RF field detection, includ-
ing communication applications, near-field antenna measurements,
and radar. It is not necessary for the incident RF source to be physi-
cally connected to the LO source as long as the phases can be synchro-
nized, for instance, through the use of atomic clocks. This would be
necessary to use Rydberg atom receivers in high-frequency communi-
cations. Rydberg atoms have been shown to measure RF E-field ampli-
tude1–6 and polarization.7 This work shows that Rydberg atoms can
also be used to measure the RF phase, previously a missing link for
complete E-field metrology with Rydberg atom-based quantum
sensors.
REFERENCES
1
C. Holloway, J. Gordon, S. Jefferts, A. Schwarzkopf, D. Anderson, S. Miller, N.
Thaicharoen, and G. Raithel, IEEE Trans. Antennas Propag. 62, 6169 (2014).
2
M. Tanasittikosol, J. D. Pritchard, D. Maxwell, A. Gauguet, K. J. Weatherill, R.
M. Potvliege, and C. S. Adams, J. Phys. B: At., Mol. Opt. Phys. 44, 184020
FIG. 5. Measured phase shift vs. distance SIG horn was moved. Error bars are the (2011).
3
standard deviation from 5 measurements. Line shows predicted phase shift, with a J. Sedlacek, A. Schwettmann, H. Kubler, R. Low, T. Pfau, and J. Shaffer, Nat.
slope from Eq. (9) of 409.7 rad/m. The fit to the data gives a measured slope of Phys. 8, 819 (2012).
4
409.3 rad/m. J. A. Gordon, C. L. Holloway, A. Schwarzkopf, D. A. Anderson, S. Miller, N.
Thaicharoen, and G. Raithel, Appl. Phys. Lett. 105, 024104 (2014).
5
C. L. Holloway, M. T. Simons, J. A. Gordon, P. F. Wilson, C. M. Cooke, D. A.
phase difference vs. distance. This measurement was repeated 5 times Anderson, and G. Raithel, IEEE Trans. Electromagn. Compat. 59, 717 (2017).
6
and averaged. The fit to the data in Fig. 5 gives a b ¼ 409.3 rad/m. The H. Fan, S. Kumar, J. Sedlacek, H. K€ubler, S. Karimkashi, and J. Shaffer, J. Phys.
theoretical value for the propagation constant for a wave propagating B: At. Mol. Opt. Phys. 48, 202001 (2015).
7
in free space is given by20 b ¼ xRF/c. In this experimental setup J. A. Sedlacek, A. Schwettmann, H. K€ ubler, and J. P. Shaffer, Phys. Rev. Lett.
111, 063001 (2013).
(Fig. 2), the horn antenna for the SIG is at an angle of h 5 from the 8
T. F. Gallagher, “Cambridge monographs on atomic, molecular and chemical
orthogonal of the two laser beams. As such, the propagation constant physics,” in Rydberg Atoms (Cambridge University Press, 1994).
measured in this setup must be corrected20 by cos ðhÞ, giving 9
A. K. Mohapatra, T. R. Jackson, and C. S. Adams, Phys. Rev. Lett. 98, 113003
(2007).
xSIG rad 10
M. T. Simons, J. A. Gordon, C. L. Holloway, D. A. Anderson, S. A. Miller, and
b¼ cos ðhÞ ¼ 409:7 : (9)
c m G. Raithel, Appl. Phys. Lett. 108, 174101 (2016).
11
D. Stack, B. Rodenburg, S. Pappas, W. Su, M. S. John, P. Kunz, M. Simons, J.
After correcting for the angle, the measured propagation constant is Gordon, and C. Holloway, in APS DAMOP, 2017.
within 0.1% of the theoretical value. This shows that the phase of the 12
D. Meyer, K. Cox, F. Fatemi, and P. Kunz, Appl. Phys. Lett. 112, 211108
SIG field was measured relative to the phase of LO. (2018).
13
This work demonstrates a Rydberg atom-based RF mixer and the K. C. Cox, D. H. Meyer, F. K. Fatemi, and P. D. Kunz, Phys. Rev. Lett. 121,
ability of Rydberg atoms to directly measure the phase of an RF field. 110502 (2018).
14
Z. Song, W. Zhang, H. Liu, X. Liu, H. Zou, J. Zhang, and J. Qu, preprint
The Rydberg atoms demodulate the RF field and transfer the IF fre-
arXiv:1808.10839 (2018).
quency directly to an optical field. The presence of a second RF field 15
D. Anderson, R. E. Sapiro, and G. Raithel, preprint arXiv:1808.08589 (2018).
(analogous to an LO) creates a beat note that the Rydberg atoms can 16
MIT Technology Review, https://www.technologyreview.com/s/611977/get-
demodulate. The phase of the demodulated signal in the optical field is ready-for-atomicradio/ for “Get ready for atom radio” (2018).
17
directly related to the phase of the incident RF field. The uncertainties C. L. Holloway, M. T. Simons, A. H. Haddab, J. A. Gordon, and S. D. Voran,
of this type of phase measurement technique are currently being inves- “A multiple-band Rydberg-atom based receiver/antenna: AM/FM stereo
tigated. However, the results in this paper show that the phase can be reception,” IEEE Antennas Propag. Mag. (submitted).
18
M. T. Simons, A. H. Haddab, R. Horansky, and C. L. Holloway, “Atom-based
determined to at least 0.1% when compared to theoretical values. The receiver: A pseudo-random bit sequence in the presence of noise,” IEEE
ability to measure the phase allows a single quantum sensor to fully Antennas Propag. Lett. (submitted).
characterize an RF field, at frequencies from MHz to THz. This allows 19
Y.-q. Li and M. Xiao, Opt. Lett. 20, 1489 (1995).
20
atomic receivers to be competitive with current technologies and C. A. Balanis, Advanced Engineering Electromagnetics (Wiley, 1989), Chap. 4.
Appl. Phys. Lett. 114, 114101 (2019); doi: 10.1063/1.5088821 114, 114101-4
Published under license by AIP Publishing