Directly pumped 10 GHz microcomb modules from low-power diode lasers
Myoung-Gyun Suh
1
, Christine Y. Wang
2
, Cort Johnson
2
, and Kerry Vahala
1
∗
1
T. J. Watson Laboratory of Applied Physics, California Institute of Technology, Pasadena, California 91125, USA.
2
Charles Stark Draper Laboratories, Cambridge, MA, 02139, USA
∗
Corresponding author: vahala@caltech.edu
Soliton microcombs offer the prospect of ad-
vanced optical metrology and timing systems
in compact form factors. In these applications,
pumping of microcombs directly from a semicon-
ductor laser without amplification or triggering
components is desirable for reduced power op-
eration and to simplify system design. At the
same time, low repetition rate microcombs are
required in many comb applications for interface
to detectors and electronics, but their increased
mode volume makes them challenging to pump at
low power. Here, 10 GHz repetition rate soliton
microcombs are directly pumped by low-power
(
<
20 mW) diode lasers. High-Q silica microres-
onators are used for this low power operation and
are packaged into fiber-connectorized modules
that feature temperature control for improved
long-term frequency stability.
Optical frequency combs (OFCs) have revolution-
ized a wide range of applications
1
and in recent
years a miniature frequency comb or microcomb
has been demonstrated using compact (often chip-
based) microresonators
2
.
These devices use the op-
tical Kerr effect to induce soliton mode locking from
a continuous-wave optical pump
3–7
.
Soliton micro-
combs provide a pathway towards miniaturization of
OFC systems for application to time-standards
8
, op-
tical frequency synthesis
9
, precision spectroscopy
10,11
,
telecommunications
12,13
, LIDAR
14,15
, imaging
16
, and
astronomy
17,18
. Pumping power of microcombs is an im-
portant design consideration and severely impacts gen-
eration of low repetition rate solitons. Such devices are
required for microcomb interface to detectors and elec-
tronics, but are particularly challenging to operate at low
pumping power. This can be understood by considering
the Kerr parametric oscillation threshold power which is
given by,
5,19
P
th
=
nω
0
A
eff
8
n
2
1
ηQ
2
T
1
f
FSR
where
n
is refractive index,
ω
0
is the optical frequency,
A
eff
is effective mode area,
n
2
is the Kerr coefficient,
η
=
Q
T
/Q
E
is the waveguide-to-resonator loading factor,
Q
T
(
Q
E
) is the total (external) quality factor, and
f
FSR
is the free spectral range (FSR) or the repetition rate.
The operating power of all microcombs is a multiple of
this threshold power, and the expression makes clear that
decreased repetition rate (smaller FSR) increases thresh-
Diode Laser
TEC for
Temp. Control
Pump
Soliton Microcomb
λ
λ
Optical Fiber
5 cm
Servo Control
Silicon Substrate
Silica Disk
Microresonator
< 3D Cross-sectional view >
FIG. 1:
Schematics of experimental setup and
microcomb module.
Single wavelength diode laser
directly pumps a silica microresonator packaged into a
fiber-connectorized module to generate a 10 GHz soli-
ton microcomb. Pump laser is polarization controlled
before optical coupling via fiber taper. Output soliton
power is tapped and used as an error signal for servo
control of laser
20
. Device temperature is controlled
using a thermoelectric cooler (TEC). Upper left in-
set: photograph of 10 GHz soliton module with fiber
pigtails. Lower right inset: schematic cross-sectional
view of silica wedge disk resonator. Optical modes are
guided at the perimeter of the wedged silica disk.
old and in turn comb power. Physically, this happens
because the pumping volume of the resonator is larger
for smaller FSR devices.
The impact of lower FSRs on microcomb power can
be offset by leveraging the favorable inverse quadratic
dependence of power on optical Q factor in the thresh-
old equation as has been demonstrated using crystalline
and silica-based microcombs
3,5,21,22
. Recently, such de-
signs have been applied to realize directly-pumped mi-
crocombs with low-repetition-rate (silica: 22 GHz,
∼
40
mW
23
, and crystalline: 12.5 GHz,
>
100 mW pump
24
),
wherein a resonator is directly connected to a pump laser
without the need for intermediate amplification or soli-
ton triggering hardware. It is also worth noting that
improving optical Q factors in silicon nitride resonators
have enabled recent demonstrations of
∼
100 GHz soliton
rates with
∼
10 mW optical pumping
25
as well as
<
1 W
Distribution Statement ”A” (Approved for Public Release, Distribution Unlimited)
arXiv:1901.08126v1 [physics.optics] 23 Jan 2019
b
a
1545
1550
1555
1560
1565
1570
Power (dBm)
Pump laser
P
pump
~ 19 mW
RBW
1 kHz
Power (20 dB / div)
MHz + 10.02 GHz
-5
0
5
SNR
> 70 dB
Wavelength (nm)
1550
1551
Power
(1 0 dB / div)
10.02 GHz
sech
2
fitting
(~ 370 fs)
-10
-20
-30
-40
-50
-60
-70
FIG. 2:
Soliton spectrum.
(a) Optical spectrum of
10 GHz soliton microcomb at 19 mW pump power.
The squared hyperbolic secant envelope (dashed green
curve) gives a soliton pulse width of 370 fs. Pump laser
line at 1556 nm is filtered out by a fiber bragg grating.
Inset: Electrical spectrum showing the soliton repeti-
tion rate of 10.02 GHz. Signal-to-noise (SNR) is greater
than 70 dB for a resolution bandwidth (RBW) of 1
kHz. (b) Zoom-in of the optical spectrum over a 1 nm
span.
electrical pumping
26
; and
∼
200 GHz soliton rates with
∼
100 mW electrical pumping from a battery
27
.
In this letter, direct pumping of 10 GHz repetition
rate soliton microcombs from low-power (
<
20 mW)
diode lasers is reported. In the experiment, the high-Q
whispering-gallery-mode silica resonators
28
featured an
∼
6.5 mm diameter corresponding to a 10.02 GHz soliton
comb line spacing. Single soliton generation was possi-
ble at
∼
15 mW pump power using the devices. Typ-
ical unloaded quality factors and parametric oscillation
threshold for this design are reported elsewhere
21
. The
resonators are specifically designed for operation in the
final frequency division stage of an optical atomic clock
8
and a robust fiber-taper coupling inside a compact mod-
ule (30 mm x 94 mm x 15 mm) was implemented (Fig.1)
that included fiber pigtails and a thermoelectric cooling
(TEC) element. Two modules were fabricated at the Cal-
ifornia Institute of Technology (Pasadena, CA, USA) and
shipped to the Draper Laboratories (Cambridge, MA,
USA) for testing relating to the clock experiment.
In the experimental setup at Draper Laboratories, the
packaged devices are directly pumped by a diode laser
b
a
0
500
1000
Time (second)
0
500
1000
Time (second)
Relative Freq. (
MH
z)
100
-100
0
1
10
100
10
5
10
6
10
7
10
8
Allan Deviation (Hz)
Averaging Time (s)
w/o
w/ temperature feedback
w/o t emperature f
eedback
w/ temperature feedback
FIG. 3:
Temperature feedback control of soliton
offset frequency.
(a) Relative frequency drift of a
comb line relative to an HCN reference laser without
(left) and with (right) temperature feedback control.
The data is acquired with 1 second gate time over 1000
seconds using a frequency counter. (b) Allan deviation
plot of the data in panel (a) showing 300 kHz deviation
at 100 seconds averaging.
(RIO PLANEX Series 1556 nm laser). The diode laser
has a short-term linewidth of 1 kHz and a maximum
output power of approximately 23 mW at 150 mA bias
current. Laser frequency can be swept at a speed of 100
GHz/s by tuning the bias current. This enables rapid
tuning of the pump into the soliton existence detun-
ing range for soliton triggering
23
. After the transmitted
pump power is filtered using a fiber bragg grating (FBG)
the soliton power is tapped via a 90/10 fiber splitter.
The 10 percent port of soliton power is photodetected
and used as an error signal for servo control to maintain
the soliton state
20
. Generation of single soliton states
using under 20 mW pump power was possible in both
devices. Fig. 2a shows the optical spectrum of the gen-
erated soliton from one device at 19 mW pump power.
The squared hyperbolic secant envelope (dashed green
curve) indicates the soliton pulse width is 370 fs. The
electrical spectrum (Fig. 2a inset) shows the soliton rep-
etition rate of 10.02 GHz with
>
70dB signal-to-noise.
The zoomed-in optical spectrum in Fig. 2b shows the
equidistantly-spaced comb lines.
To improve the long term stability of the system, the
device temperature is controlled via a TEC element in-
stalled inside the module. In an independent measure-
ment (performed at Caltech) to test the performance of
the temperature feedback loop, one of the comb lines was
heterodyned with an HCN reference laser (1559 nm Clar-
Distribution Statement ”A” (Approved for Public Release, Distribution Unlimited)
ity laser manufactured by Wavelength References, Inc.)
and the beat frequency was measured using a frequency
counter. Figure 3a shows that the temperature feedback
can suppress the beat frequency drift due to the envi-
ronmental temperature changes. With the temperature
feedback, an absolute frequency instability of 300 kHz at
100 second averaging time (Fig. 3b) was measured. Con-
sidering that the HCN reference laser shows an absolute
frequency instability of
∼
200 kHz at 100 seconds aver-
aging, we can conclude that an absolute stability of the
comb offset frequency better than 200 kHz is achieved at
100 seconds averaging with the temperature feedback.
In summary, 10 GHz soliton microcomb devices were
fabricated, packaged into fiber connectorized modules,
transported across the North American continent and
then successfully operated by direct pumping from
low-power (
<
20 mW) diode lasers. This low direct
pumping power represents a record for this low comb
repetition rate, and its X-band rate also enables use of
low-bandwidth detectors and simplifies the electronic
interface with the microcomb system (here a clock).
These devices can be potentially integrated with SiN
waveguides
22
for more compact systems. The modules
also feature a built-in TEC control that was able
to stabilize the comb offset frequency in an open lab
environment better than 200 kHz at 100 s averaging time.
Funding Information
The authors acknowledge research funding from the
Defense Advanced Research Projects Agency (DARPA)
Atomic Clocks with Enhanced Stability (ACES) program
(HR0011-16-C-0118). The views, opinions and/or find-
ings expressed are those of the authors and should not be
interpreted as representing the official views or policies
of the Department of Defense or the U.S. Government.
This work is also supported by the Kavli Nanoscience
Institute
1
Diddams, S. A. The evolving optical frequency comb.
JOSA B
27
, B51–B62 (2010).
2
Kippenberg, T. J., Gaeta, A. L., Lipson, M. & Gorodetsky,
M. L. Dissipative kerr solitons in optical microresonators.
Science
361
, eaan8083 (2018).
3
Herr, T.
et al.
Temporal solitons in optical microres-
onators.
Nature Photonics
8
, 145–152 (2014).
4
Brasch, V.
et al.
Photonic chip–based optical frequency
comb using soliton cherenkov radiation.
Science
351
, 357–
360 (2016).
5
Yi, X., Yang, Q.-F., Yang, K. Y., Suh, M.-G. & Vahala,
K. Soliton frequency comb at microwave rates in a high-q
silica microresonator.
Optica
2
, 1078–1085 (2015).
6
Joshi, C.
et al.
Thermally controlled comb generation and
soliton modelocking in microresonators.
Optics Letters
41
,
2565–2568 (2016).
7
Wang, P.-H.
et al.
Intracavity characterization of micro-
comb generation in the single-soliton regime.
Optics Ex-
press
24
, 10890–10897 (2016).
8
Newman, Z.
et al.
Photonic integration of an optical
atomic clock.
arXiv preprint arXiv:1811.00616
(2018).
9
Spencer, D. T.
et al.
An optical-frequency synthesizer us-
ing integrated photonics.
Nature
557
, 81–85 (2018).
10
Suh, M.-G., Yang, Q.-F., Yang, K. Y., Yi, X. & Vahala,
K. J. Microresonator soliton dual-comb spectroscopy.
Sci-
ence
354
, 600–603 (2016).
11
Dutt, A.
et al.
On-chip dual-comb source for spectroscopy.
Science Advances
4
, e1701858 (2018).
12
Marin-Palomo, P.
et al.
Microresonator-based solitons for
massively parallel coherent optical communications.
Na-
ture
546
, 274–279 (2017).
13
Mazur, M.
et al.
Enabling high spectral efficiency coherent
superchannel transmission with soliton microcombs.
arXiv
preprint arXiv:1812.11046
(2018).
14
Suh, M.-G. & Vahala, K. J. Soliton microcomb range mea-
surement.
Science
359
, 884–887 (2018).
15
Trocha, P.
et al.
Ultrafast optical ranging using microres-
onator soliton frequency combs.
Science
359
, 887–891
(2018).
16
Bao, C., Suh, M.-G. & Vahala, K. Microresonator soli-
ton dual-comb imaging.
arXiv preprint arXiv:1809.09766
(2018).
17
Suh, M.-G.
et al.
Searching for exoplanets using a mi-
croresonator astrocomb.
Nature Photonics
13
, 25 (2019).
18
Obrzud, E.
et al.
A microphotonic astrocomb.
Nature
Photonics
13
, 31 (2019).
19
Kippenberg, T., Spillane, S. & Vahala, K.
Kerr-
nonlinearity optical parametric oscillation in an ultrahigh-
q toroid microcavity.
Physical Review Letters
93
, 083904
(2004).
20
Yi, X., Yang, Q.-F., Youl, K. & Vahala, K. Active capture
and stabilization of temporal solitons in microresonators.
Optics Letters
41
, 2037–2040 (2016).
21
Suh, M.-G. & Vahala, K. Gigahertz-repetition-rate soliton
microcombs.
Optica
5
, 65–66 (2018).
22
Yang, K. Y.
et al.
Bridging ultrahigh-q devices and pho-
tonic circuits.
Nature Photonics
12
, 297 (2018).
23
Volet, N.
et al.
Micro-resonator soliton generated directly
with a diode laser.
Laser & Photonics Reviews
12
, 1700307
(2018).
24
Pavlov, N.
et al.
Narrow-linewidth lasing and soliton kerr
microcombs with ordinary laser diodes.
Nature Photonics
12
, 694 (2018).
25
Liu, J.
et al.
Ultralow-power chip-based soliton micro-
combs for photonic integration.
Optica
5
, 1347–1353
(2018).
26
Raja, A. S.
et al.
Electrically pumped photonic integrated
soliton microcomb.
arXiv preprint arXiv:1810.03909
(2018).
27
Stern, B., Ji, X., Okawachi, Y., Gaeta, A. L. & Lipson,
M. Battery-operated integrated frequency comb generator.
Nature
562
, 401 (2018).
28
Lee, H.
et al.
Chemically etched ultrahigh-q wedge-
resonator on a silicon chip.
Nature Photonics
6
, 369–373
(2012).
Distribution Statement ”A” (Approved for Public Release, Distribution Unlimited)